high-side current sensor...

J

John Larkin

Guest
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
<klauskvik@hotmail.com> wrote:

On 12/01/2022 03.44, jlarkin@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
langwadt@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd <whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8, jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the output currents
from nine low-V regulators, or is it input to the regulators, and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to include initial
charging of filter capacitors? How big are those capacitors? Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the high-side? A secondary
low-current power supply grounded at the high rail might be economic, if
you want to have nine sensors and nine cutoffs all near the +48V source.
There will be eight user-programmable power supplies, each running off
+48, all off a common +48 bus fed from a kilowatt bulk power supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to protect
things if they do it.

there is quite a few high side load switches with short circuit protection
and a combined current monitor / error output , but I think most of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address. There\'s
no need to actually remove 48v power from the baby boards. The master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

--

If a man will begin with certainties, he shall end with doubts,
but if he will be content to begin with doubts he shall end in certainties.
Francis Bacon
 
L

Lasse Langwadt Christensen

Guest
tirsdag den 18. januar 2022 kl. 20.52.58 UTC+1 skrev John Larkin:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klau...@hotmail.com> wrote:

On 12/01/2022 03.44, jla...@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd <whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8, jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the output currents
from nine low-V regulators, or is it input to the regulators, and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to include initial
charging of filter capacitors? How big are those capacitors? Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the high-side? A secondary
low-current power supply grounded at the high rail might be economic, if
you want to have nine sensors and nine cutoffs all near the +48V source.
There will be eight user-programmable power supplies, each running off
+48, all off a common +48 bus fed from a kilowatt bulk power supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to protect
things if they do it.

there is quite a few high side load switches with short circuit protection
and a combined current monitor / error output , but I think most of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address. There\'s
no need to actually remove 48v power from the baby boards. The master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the FPGA/Micro

Add a resistor with known drop, to calibrate
It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

or is is just a matter of supply and demand and Digikey is doing you a favor
by discouraging hoarding, so if you really need a part you can get it
 
J

Jasen Betts

Guest
On 2022-01-18, John Larkin <jlarkin@highland_atwork_technology.com> wrote:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klauskvik@hotmail.com> wrote:

On 12/01/2022 03.44, jlarkin@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
langwadt@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd <whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8, jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the output currents
from nine low-V regulators, or is it input to the regulators, and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to include initial
charging of filter capacitors? How big are those capacitors? Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the high-side? A secondary
low-current power supply grounded at the high rail might be economic, if
you want to have nine sensors and nine cutoffs all near the +48V source.
There will be eight user-programmable power supplies, each running off
+48, all off a common +48 bus fed from a kilowatt bulk power supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to protect
things if they do it.

there is quite a few high side load switches with short circuit protection
and a combined current monitor / error output , but I think most of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address. There\'s
no need to actually remove 48v power from the baby boards. The master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

Or at-least an agent for a scalper. They\'re listing Rochester Electronics
as supplier on some lines (other parts).

sfaik digikey mostly sells to designers, people making large numbers
of devices find their parts elsewhere.

--
Jasen.
 
K

Klaus Vestergaard Kragelund

Guest
On 19/01/2022 09.27, Jasen Betts wrote:
On 2022-01-18, John Larkin <jlarkin@highland_atwork_technology.com> wrote:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klauskvik@hotmail.com> wrote:

On 12/01/2022 03.44, jlarkin@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
langwadt@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd <whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8, jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the output currents
from nine low-V regulators, or is it input to the regulators, and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to include initial
charging of filter capacitors? How big are those capacitors? Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the high-side? A secondary
low-current power supply grounded at the high rail might be economic, if
you want to have nine sensors and nine cutoffs all near the +48V source.
There will be eight user-programmable power supplies, each running off
+48, all off a common +48 bus fed from a kilowatt bulk power supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to protect
things if they do it.

there is quite a few high side load switches with short circuit protection
and a combined current monitor / error output , but I think most of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address. There\'s
no need to actually remove 48v power from the baby boards. The master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

Or at-least an agent for a scalper. They\'re listing Rochester Electronics
as supplier on some lines (other parts).

sfaik digikey mostly sells to designers, people making large numbers
of devices find their parts elsewhere.

Yes, nobody buys in scale from Digikey

80cents is expensive in my world, but then again I also come from a
business where qty products sold was counted in millions
 
S

server

Guest
On Wed, 19 Jan 2022 13:22:38 +0100, Klaus Vestergaard Kragelund
<klauskvik@hotmail.com> wrote:

On 19/01/2022 09.27, Jasen Betts wrote:
On 2022-01-18, John Larkin <jlarkin@highland_atwork_technology.com> wrote:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klauskvik@hotmail.com> wrote:

On 12/01/2022 03.44, jlarkin@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
langwadt@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd <whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8, jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the output currents
from nine low-V regulators, or is it input to the regulators, and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to include initial
charging of filter capacitors? How big are those capacitors? Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the high-side? A secondary
low-current power supply grounded at the high rail might be economic, if
you want to have nine sensors and nine cutoffs all near the +48V source.
There will be eight user-programmable power supplies, each running off
+48, all off a common +48 bus fed from a kilowatt bulk power supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to protect
things if they do it.

there is quite a few high side load switches with short circuit protection
and a combined current monitor / error output , but I think most of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address. There\'s
no need to actually remove 48v power from the baby boards. The master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

Or at-least an agent for a scalper. They\'re listing Rochester Electronics
as supplier on some lines (other parts).

sfaik digikey mostly sells to designers, people making large numbers
of devices find their parts elsewhere.

Yes, nobody buys in scale from Digikey

80cents is expensive in my world, but then again I also come from a
business where qty products sold was counted in millions

We sell modest volumes of fairly expensive stuff. Parts cost is
roughly 12% of sales price, and people (purchasing, engineering,
manufacturing, testing) are a much bigger cost. Lately, availability
dominates cost.

PCBs and metal cases, both custom, are big costs.

We buy a fair amount from Mouser and Digikey, expecially things like
resistors that cost a fraction of a cent each.

A sales guy, from Arrow or Avnet or one of the biggies, said that
\"distributors\" don\'t stock parts much any more, but Mouser and Digikey
do. It looks to me that Digikey at least is trending that way too;
it\'s hard to tell right now.





--

I yam what I yam - Popeye
 
P

Phil Hobbs

Guest
Klaus Vestergaard Kragelund wrote:
On 19/01/2022 09.27, Jasen Betts wrote:
On 2022-01-18, John Larkin <jlarkin@highland_atwork_technology.com
wrote:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klauskvik@hotmail.com> wrote:

On 12/01/2022 03.44, jlarkin@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
langwadt@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev
jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd
whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8,
jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight
plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the
output currents
from nine low-V regulators, or is it input to the regulators,
and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to
include initial
charging of filter capacitors? How big are those capacitors?
Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the
high-side? A secondary
low-current power supply grounded at the high rail might be
economic, if
you want to have nine sensors and nine cutoffs all near the
+48V source.
There will be eight user-programmable power supplies, each
running off
+48, all off a common +48 bus fed from a kilowatt bulk power
supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to
protect
things if they do it.

there is quite a few high side load switches with short circuit
protection
and a combined current monitor / error output , but I think most
of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address.
There\'s
no need to actually remove 48v power from the baby boards. The
master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider
on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the
FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

Or at-least an agent for a scalper.  They\'re listing Rochester
Electronics
as supplier on some lines (other parts).

sfaik digikey mostly sells to designers, people making large numbers
of devices find their parts elsewhere.

Yes, nobody buys in scale from Digikey

Somebody must--from the Thief River Falls Reginal Airport webpage:

\"From FAA-published information gathered in 2012, the Airport
experiences 31,200 aircraft operations per year (takeoffs and landings)
of which 30,000 operations are attributed to general aviation
operations, 1,250 from air carrier commercial service, 2,600 from other
commercial/cargo service, and 100 military operations. Nearly 5,000
passengers (2,500 enplanement boardings) are processed annually from
scheduled airline flights. The Airport has 26 based aircraft, including
3 multi-piston and 3 jet turbine airplanes.

So that\'s 3750 commercial takeoffs and landings, and a total of 2500
passengers. (The webmaster is apparently a bit arithmetically
challenged, but oh well.)

Since there\'s not a lot in Thief River Falls except Digikey, one gathers
that _somebody_ must be buying a lot of stuff from them.

80cents is expensive in my world, but then again I also come from a
business where qty products sold was counted in millions

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
S

server

Guest
On Wed, 19 Jan 2022 10:59:04 -0500, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

Klaus Vestergaard Kragelund wrote:
On 19/01/2022 09.27, Jasen Betts wrote:
On 2022-01-18, John Larkin <jlarkin@highland_atwork_technology.com
wrote:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klauskvik@hotmail.com> wrote:

On 12/01/2022 03.44, jlarkin@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
langwadt@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev
jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd
whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8,
jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight
plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the
output currents
from nine low-V regulators, or is it input to the regulators,
and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to
include initial
charging of filter capacitors? How big are those capacitors?
Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the
high-side? A secondary
low-current power supply grounded at the high rail might be
economic, if
you want to have nine sensors and nine cutoffs all near the
+48V source.
There will be eight user-programmable power supplies, each
running off
+48, all off a common +48 bus fed from a kilowatt bulk power
supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to
protect
things if they do it.

there is quite a few high side load switches with short circuit
protection
and a combined current monitor / error output , but I think most
of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address.
There\'s
no need to actually remove 48v power from the baby boards. The
master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider
on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the
FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

Or at-least an agent for a scalper.  They\'re listing Rochester
Electronics
as supplier on some lines (other parts).

sfaik digikey mostly sells to designers, people making large numbers
of devices find their parts elsewhere.

Yes, nobody buys in scale from Digikey

Somebody must--from the Thief River Falls Reginal Airport webpage:

\"From FAA-published information gathered in 2012, the Airport
experiences 31,200 aircraft operations per year (takeoffs and landings)
of which 30,000 operations are attributed to general aviation
operations, 1,250 from air carrier commercial service, 2,600 from other
commercial/cargo service, and 100 military operations. Nearly 5,000
passengers (2,500 enplanement boardings) are processed annually from
scheduled airline flights. The Airport has 26 based aircraft, including
3 multi-piston and 3 jet turbine airplanes.

So that\'s 3750 commercial takeoffs and landings, and a total of 2500
passengers. (The webmaster is apparently a bit arithmetically
challenged, but oh well.)

Since there\'s not a lot in Thief River Falls except Digikey, one gathers
that _somebody_ must be buying a lot of stuff from them.


80cents is expensive in my world, but then again I also come from a
business where qty products sold was counted in millions

Cheers

Phil Hobbs

We lazy engineers often specify a Digikey part for an initial buy, and
our excellent purchasing lady buys quantities of the same thing
somewhere else, at a much better price.

Digikey has/had a good multi-vendor search engine. It seems to make
less sense lately. A category might for example have 25,000 items, and
selecting a couple of very obvious qualifiers narrows it down to a
few, or none.

Selecting for example a Vcc range gives a huge list of confusing,
overlapping choices.

Or specifying something explicit gets me non-qualifying parts.

Lots of broken links too.



--

I yam what I yam - Popeye
 
P

Phil Hobbs

Guest
jlarkin@highlandsniptechnology.com wrote:
On Wed, 19 Jan 2022 10:59:04 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

Klaus Vestergaard Kragelund wrote:
On 19/01/2022 09.27, Jasen Betts wrote:
On 2022-01-18, John Larkin <jlarkin@highland_atwork_technology.com
wrote:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klauskvik@hotmail.com> wrote:

On 12/01/2022 03.44, jlarkin@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
langwadt@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev
jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd
whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8,
jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight
plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the
output currents
from nine low-V regulators, or is it input to the regulators,
and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to
include initial
charging of filter capacitors? How big are those capacitors?
Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the
high-side? A secondary
low-current power supply grounded at the high rail might be
economic, if
you want to have nine sensors and nine cutoffs all near the
+48V source.
There will be eight user-programmable power supplies, each
running off
+48, all off a common +48 bus fed from a kilowatt bulk power
supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to
protect
things if they do it.

there is quite a few high side load switches with short circuit
protection
and a combined current monitor / error output , but I think most
of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address.
There\'s
no need to actually remove 48v power from the baby boards. The
master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider
on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the
FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

Or at-least an agent for a scalper.  They\'re listing Rochester
Electronics
as supplier on some lines (other parts).

sfaik digikey mostly sells to designers, people making large numbers
of devices find their parts elsewhere.

Yes, nobody buys in scale from Digikey

Somebody must--from the Thief River Falls Reginal Airport webpage:

\"From FAA-published information gathered in 2012, the Airport
experiences 31,200 aircraft operations per year (takeoffs and landings)
of which 30,000 operations are attributed to general aviation
operations, 1,250 from air carrier commercial service, 2,600 from other
commercial/cargo service, and 100 military operations. Nearly 5,000
passengers (2,500 enplanement boardings) are processed annually from
scheduled airline flights. The Airport has 26 based aircraft, including
3 multi-piston and 3 jet turbine airplanes.

So that\'s 3750 commercial takeoffs and landings, and a total of 2500
passengers. (The webmaster is apparently a bit arithmetically
challenged, but oh well.)

Since there\'s not a lot in Thief River Falls except Digikey, one gathers
that _somebody_ must be buying a lot of stuff from them.

80cents is expensive in my world, but then again I also come from a
business where qty products sold was counted in millions



We lazy engineers often specify a Digikey part for an initial buy, and
our excellent purchasing lady buys quantities of the same thing
somewhere else, at a much better price.

Never mess with purchasing ladies--it\'s almost as stupid as insulting
somebody who\'s preparing your food. ;)

Digikey has/had a good multi-vendor search engine. It seems to make
less sense lately. A category might for example have 25,000 items, and
selecting a couple of very obvious qualifiers narrows it down to a
few, or none.

Selecting for example a Vcc range gives a huge list of confusing,
overlapping choices.

Yeah, their VCC, slew rate, GBW, and 3 dB bandwidth columns are so
broken. They could usefully add a column for total supply voltage--that
would help a good bit all by itself.h

Or specifying something explicit gets me non-qualifying parts.

Lots of broken links too.

Especially Samsung MLCC characteristic sheets. They keep breaking them
for some reason--you\'d think DK would just give up and mirror them.

Cheers

Phil Hobbs


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
J

John Larkin

Guest
On Wed, 19 Jan 2022 13:39:46 -0500, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

jlarkin@highlandsniptechnology.com wrote:
On Wed, 19 Jan 2022 10:59:04 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

Klaus Vestergaard Kragelund wrote:
On 19/01/2022 09.27, Jasen Betts wrote:
On 2022-01-18, John Larkin <jlarkin@highland_atwork_technology.com
wrote:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klauskvik@hotmail.com> wrote:

On 12/01/2022 03.44, jlarkin@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
langwadt@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev
jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd
whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8,
jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight
plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the
output currents
from nine low-V regulators, or is it input to the regulators,
and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to
include initial
charging of filter capacitors? How big are those capacitors?
Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the
high-side? A secondary
low-current power supply grounded at the high rail might be
economic, if
you want to have nine sensors and nine cutoffs all near the
+48V source.
There will be eight user-programmable power supplies, each
running off
+48, all off a common +48 bus fed from a kilowatt bulk power
supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to
protect
things if they do it.

there is quite a few high side load switches with short circuit
protection
and a combined current monitor / error output , but I think most
of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address.
There\'s
no need to actually remove 48v power from the baby boards. The
master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider
on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the
FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

Or at-least an agent for a scalper.  They\'re listing Rochester
Electronics
as supplier on some lines (other parts).

sfaik digikey mostly sells to designers, people making large numbers
of devices find their parts elsewhere.

Yes, nobody buys in scale from Digikey

Somebody must--from the Thief River Falls Reginal Airport webpage:

\"From FAA-published information gathered in 2012, the Airport
experiences 31,200 aircraft operations per year (takeoffs and landings)
of which 30,000 operations are attributed to general aviation
operations, 1,250 from air carrier commercial service, 2,600 from other
commercial/cargo service, and 100 military operations. Nearly 5,000
passengers (2,500 enplanement boardings) are processed annually from
scheduled airline flights. The Airport has 26 based aircraft, including
3 multi-piston and 3 jet turbine airplanes.

So that\'s 3750 commercial takeoffs and landings, and a total of 2500
passengers. (The webmaster is apparently a bit arithmetically
challenged, but oh well.)

Since there\'s not a lot in Thief River Falls except Digikey, one gathers
that _somebody_ must be buying a lot of stuff from them.

80cents is expensive in my world, but then again I also come from a
business where qty products sold was counted in millions



We lazy engineers often specify a Digikey part for an initial buy, and
our excellent purchasing lady buys quantities of the same thing
somewhere else, at a much better price.

Never mess with purchasing ladies--it\'s almost as stupid as insulting
somebody who\'s preparing your food. ;)

Digikey has/had a good multi-vendor search engine. It seems to make
less sense lately. A category might for example have 25,000 items, and
selecting a couple of very obvious qualifiers narrows it down to a
few, or none.

Selecting for example a Vcc range gives a huge list of confusing,
overlapping choices.

Yeah, their VCC, slew rate, GBW, and 3 dB bandwidth columns are so
broken. They could usefully add a column for total supply voltage--that
would help a good bit all by itself.h


Or specifying something explicit gets me non-qualifying parts.

Lots of broken links too.

Especially Samsung MLCC characteristic sheets. They keep breaking them
for some reason--you\'d think DK would just give up and mirror them.

Cheers

Phil Hobbs

Sometimes they link to a vendor site that requires registration and a
password and all that nonsense, just to see a data sheet.

\"The easiest thing in the world is not to sell.\"

- Melvin Goldstein



--

If a man will begin with certainties, he shall end with doubts,
but if he will be content to begin with doubts he shall end in certainties.
Francis Bacon
 
T

three_jeeps

Guest
On Wednesday, January 19, 2022 at 10:47:43 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Wed, 19 Jan 2022 13:22:38 +0100, Klaus Vestergaard Kragelund
klau...@hotmail.com> wrote:

On 19/01/2022 09.27, Jasen Betts wrote:
On 2022-01-18, John Larkin <jlarkin@highland_atwork_technology.com> wrote:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klau...@hotmail.com> wrote:

On 12/01/2022 03.44, jla...@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd <whi...@gmail..com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8, jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the output currents
from nine low-V regulators, or is it input to the regulators, and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to include initial
charging of filter capacitors? How big are those capacitors? Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the high-side? A secondary
low-current power supply grounded at the high rail might be economic, if
you want to have nine sensors and nine cutoffs all near the +48V source.
There will be eight user-programmable power supplies, each running off
+48, all off a common +48 bus fed from a kilowatt bulk power supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to protect
things if they do it.

there is quite a few high side load switches with short circuit protection
and a combined current monitor / error output , but I think most of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address. There\'s
no need to actually remove 48v power from the baby boards. The master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

Or at-least an agent for a scalper. They\'re listing Rochester Electronics
as supplier on some lines (other parts).

sfaik digikey mostly sells to designers, people making large numbers
of devices find their parts elsewhere.

Yes, nobody buys in scale from Digikey

80cents is expensive in my world, but then again I also come from a
business where qty products sold was counted in millions
We sell modest volumes of fairly expensive stuff. Parts cost is
roughly 12% of sales price, and people (purchasing, engineering,
manufacturing, testing) are a much bigger cost. Lately, availability
dominates cost.

PCBs and metal cases, both custom, are big costs.

We buy a fair amount from Mouser and Digikey, expecially things like
resistors that cost a fraction of a cent each.

A sales guy, from Arrow or Avnet or one of the biggies, said that
\"distributors\" don\'t stock parts much any more, but Mouser and Digikey
do. It looks to me that Digikey at least is trending that way too;
it\'s hard to tell right now.
--

I yam what I yam - Popeye

Interesting, so what is a rule of thumb, from your perspective, to go from parts cost to the manufactured product?
Years ago, a colleague told me that the multiplier was 7-10X, e.g. if there were $10 worth of electronic parts, the cost of the manufactured product would be ~ $70-$100. That was internal costs. Add more for profit. This was for an all-electronic widget. Am sure things would be different if it were electro-mechanical. Lots of other factors could obviously impact this rule of thumb.

Another colleague of mine worked at the Zenith TV electronics engineering department. When given a subsection of a TV receiver to redesign for a new product, they had two over-riding guidelines: 1) minimize parts count, 2) If the cost of parts for the new design increased cost about $0.05 as compared to the old design, he had to go through a justification review. This was back in the 80\'s so I am sure things have changed a bit. It was, for me, just interesting to hear about this.
j
 
L

Lasse Langwadt Christensen

Guest
onsdag den 19. januar 2022 kl. 20.39.07 UTC+1 skrev three_jeeps:
On Wednesday, January 19, 2022 at 10:47:43 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Wed, 19 Jan 2022 13:22:38 +0100, Klaus Vestergaard Kragelund
klau...@hotmail.com> wrote:

On 19/01/2022 09.27, Jasen Betts wrote:
On 2022-01-18, John Larkin <jlarkin@highland_atwork_technology.com> wrote:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klau...@hotmail.com> wrote:

On 12/01/2022 03.44, jla...@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd <whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8, jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the output currents
from nine low-V regulators, or is it input to the regulators, and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to include initial
charging of filter capacitors? How big are those capacitors? Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the high-side? A secondary
low-current power supply grounded at the high rail might be economic, if
you want to have nine sensors and nine cutoffs all near the +48V source.
There will be eight user-programmable power supplies, each running off
+48, all off a common +48 bus fed from a kilowatt bulk power supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to protect
things if they do it.

there is quite a few high side load switches with short circuit protection
and a combined current monitor / error output , but I think most of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address. There\'s
no need to actually remove 48v power from the baby boards. The master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

Or at-least an agent for a scalper. They\'re listing Rochester Electronics
as supplier on some lines (other parts).

sfaik digikey mostly sells to designers, people making large numbers
of devices find their parts elsewhere.

Yes, nobody buys in scale from Digikey

80cents is expensive in my world, but then again I also come from a
business where qty products sold was counted in millions
We sell modest volumes of fairly expensive stuff. Parts cost is
roughly 12% of sales price, and people (purchasing, engineering,
manufacturing, testing) are a much bigger cost. Lately, availability
dominates cost.

PCBs and metal cases, both custom, are big costs.

We buy a fair amount from Mouser and Digikey, expecially things like
resistors that cost a fraction of a cent each.

A sales guy, from Arrow or Avnet or one of the biggies, said that
\"distributors\" don\'t stock parts much any more, but Mouser and Digikey
do. It looks to me that Digikey at least is trending that way too;
it\'s hard to tell right now.
--

I yam what I yam - Popeye
Interesting, so what is a rule of thumb, from your perspective, to go from parts cost to the manufactured product?
Years ago, a colleague told me that the multiplier was 7-10X, e.g. if there were $10 worth of electronic parts, the cost of the manufactured product would be ~ $70-$100. That was internal costs. Add more for profit. This was for an all-electronic widget. Am sure things would be different if it were electro-mechanical. Lots of other factors could obviously impact this rule of thumb.

Another colleague of mine worked at the Zenith TV electronics engineering department. When given a subsection of a TV receiver to redesign for a new product, they had two over-riding guidelines: 1) minimize parts count, 2) If the cost of parts for the new design increased cost about $0.05 as compared to the old design, he had to go through a justification review. This was back in the 80\'s so I am sure things have changed a bit. It was, for me, just interesting to hear about this.
j

I\'ve heard of stories of gadgets made in volume so that the monthly cost of something like a few extra resistors could pay an engineers monthly salary
 
K

Klaus Kragelund

Guest
19.01.22 20:47, Lasse Langwadt Christensen wrote:
onsdag den 19. januar 2022 kl. 20.39.07 UTC+1 skrev three_jeeps:
On Wednesday, January 19, 2022 at 10:47:43 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Wed, 19 Jan 2022 13:22:38 +0100, Klaus Vestergaard Kragelund
klau...@hotmail.com> wrote:

On 19/01/2022 09.27, Jasen Betts wrote:
On 2022-01-18, John Larkin <jlarkin@highland_atwork_technology.com> wrote:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klau...@hotmail.com> wrote:

On 12/01/2022 03.44, jla...@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd <whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8, jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the output currents
from nine low-V regulators, or is it input to the regulators, and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to include initial
charging of filter capacitors? How big are those capacitors? Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the high-side? A secondary
low-current power supply grounded at the high rail might be economic, if
you want to have nine sensors and nine cutoffs all near the +48V source.
There will be eight user-programmable power supplies, each running off
+48, all off a common +48 bus fed from a kilowatt bulk power supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to protect
things if they do it.

there is quite a few high side load switches with short circuit protection
and a combined current monitor / error output , but I think most of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address. There\'s
no need to actually remove 48v power from the baby boards. The master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

Or at-least an agent for a scalper. They\'re listing Rochester Electronics
as supplier on some lines (other parts).

sfaik digikey mostly sells to designers, people making large numbers
of devices find their parts elsewhere.

Yes, nobody buys in scale from Digikey

80cents is expensive in my world, but then again I also come from a
business where qty products sold was counted in millions
We sell modest volumes of fairly expensive stuff. Parts cost is
roughly 12% of sales price, and people (purchasing, engineering,
manufacturing, testing) are a much bigger cost. Lately, availability
dominates cost.

PCBs and metal cases, both custom, are big costs.

We buy a fair amount from Mouser and Digikey, expecially things like
resistors that cost a fraction of a cent each.

A sales guy, from Arrow or Avnet or one of the biggies, said that
\"distributors\" don\'t stock parts much any more, but Mouser and Digikey
do. It looks to me that Digikey at least is trending that way too;
it\'s hard to tell right now.
--

I yam what I yam - Popeye
Interesting, so what is a rule of thumb, from your perspective, to go from parts cost to the manufactured product?
Years ago, a colleague told me that the multiplier was 7-10X, e.g. if there were $10 worth of electronic parts, the cost of the manufactured product would be ~ $70-$100. That was internal costs. Add more for profit. This was for an all-electronic widget. Am sure things would be different if it were electro-mechanical. Lots of other factors could obviously impact this rule of thumb.

Another colleague of mine worked at the Zenith TV electronics engineering department. When given a subsection of a TV receiver to redesign for a new product, they had two over-riding guidelines: 1) minimize parts count, 2) If the cost of parts for the new design increased cost about $0.05 as compared to the old design, he had to go through a justification review. This was back in the 80\'s so I am sure things have changed a bit. It was, for me, just interesting to hear about this.
j

I\'ve heard of stories of gadgets made in volume so that the monthly cost of something like a few extra resistors could pay an engineers monthly salary


I was the lead innovator on a new product with release in a couple of months

Quantity of 500.000 pcs per month, so a 2 cent cost reduction could pay an engineers salary, not counting all the other overhead


--
Klaus
 
S

server

Guest
On Wed, 19 Jan 2022 11:39:02 -0800 (PST), three_jeeps
<jjhudak@gmail.com> wrote:

On Wednesday, January 19, 2022 at 10:47:43 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Wed, 19 Jan 2022 13:22:38 +0100, Klaus Vestergaard Kragelund
klau...@hotmail.com> wrote:

On 19/01/2022 09.27, Jasen Betts wrote:
On 2022-01-18, John Larkin <jlarkin@highland_atwork_technology.com> wrote:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klau...@hotmail.com> wrote:

On 12/01/2022 03.44, jla...@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd <whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8, jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the output currents
from nine low-V regulators, or is it input to the regulators, and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to include initial
charging of filter capacitors? How big are those capacitors? Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the high-side? A secondary
low-current power supply grounded at the high rail might be economic, if
you want to have nine sensors and nine cutoffs all near the +48V source.
There will be eight user-programmable power supplies, each running off
+48, all off a common +48 bus fed from a kilowatt bulk power supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to protect
things if they do it.

there is quite a few high side load switches with short circuit protection
and a combined current monitor / error output , but I think most of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address. There\'s
no need to actually remove 48v power from the baby boards. The master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

Or at-least an agent for a scalper. They\'re listing Rochester Electronics
as supplier on some lines (other parts).

sfaik digikey mostly sells to designers, people making large numbers
of devices find their parts elsewhere.

Yes, nobody buys in scale from Digikey

80cents is expensive in my world, but then again I also come from a
business where qty products sold was counted in millions
We sell modest volumes of fairly expensive stuff. Parts cost is
roughly 12% of sales price, and people (purchasing, engineering,
manufacturing, testing) are a much bigger cost. Lately, availability
dominates cost.

PCBs and metal cases, both custom, are big costs.

We buy a fair amount from Mouser and Digikey, expecially things like
resistors that cost a fraction of a cent each.

A sales guy, from Arrow or Avnet or one of the biggies, said that
\"distributors\" don\'t stock parts much any more, but Mouser and Digikey
do. It looks to me that Digikey at least is trending that way too;
it\'s hard to tell right now.
--

I yam what I yam - Popeye

Interesting, so what is a rule of thumb, from your perspective, to go from parts cost to the manufactured product?

We define \"direct cost\" as unburdened parts cost plus labor. A selling
price of 3x DC is our lower limit. 6s is better, sometimes 10x is
possible where the value to the customer is high and there is no
competition.

We used to sell NMR gradient amplifiers to Varian/Agilent at maybe 4x
DC. They marked our price up 6x to their customers.

I wonder what a high-end iphone costs Apple.

Years ago, a colleague told me that the multiplier was 7-10X, e.g. if there were $10 worth of electronic parts, the cost of the manufactured product would be ~ $70-$100. That was internal costs. Add more for profit. This was for an all-electronic widget. Am sure things would be different if it were electro-mechanical. Lots of other factors could obviously impact this rule of thumb.

Another colleague of mine worked at the Zenith TV electronics engineering department. When given a subsection of a TV receiver to redesign for a new product, they had two over-riding guidelines: 1) minimize parts count, 2) If the cost of parts for the new design increased cost about $0.05 as compared to the old design, he had to go through a justification review. This was back in the 80\'s so I am sure things have changed a bit. It was, for me, just interesting to hear about this.
j

I get the impression that in high volume, contract manufacturers make
a few per cent over their net costs. They often have contracts that
let the customer audit their actual costs and pay them a few per cent
more.

One of our customers proposed to do that to us. We declined.



--

I yam what I yam - Popeye
 
J

John S

Guest
On 1/19/2022 1:36 PM, John Larkin wrote:
On Wed, 19 Jan 2022 13:39:46 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

jlarkin@highlandsniptechnology.com wrote:
On Wed, 19 Jan 2022 10:59:04 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

Klaus Vestergaard Kragelund wrote:
On 19/01/2022 09.27, Jasen Betts wrote:
On 2022-01-18, John Larkin <jlarkin@highland_atwork_technology.com
wrote:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klauskvik@hotmail.com> wrote:

On 12/01/2022 03.44, jlarkin@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
langwadt@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev
jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd
whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8,
jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight
plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the
output currents
from nine low-V regulators, or is it input to the regulators,
and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to
include initial
charging of filter capacitors? How big are those capacitors?
Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the
high-side? A secondary
low-current power supply grounded at the high rail might be
economic, if
you want to have nine sensors and nine cutoffs all near the
+48V source.
There will be eight user-programmable power supplies, each
running off
+48, all off a common +48 bus fed from a kilowatt bulk power
supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to
protect
things if they do it.

there is quite a few high side load switches with short circuit
protection
and a combined current monitor / error output , but I think most
of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address.
There\'s
no need to actually remove 48v power from the baby boards. The
master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider
on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the
FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

Or at-least an agent for a scalper.  They\'re listing Rochester
Electronics
as supplier on some lines (other parts).

sfaik digikey mostly sells to designers, people making large numbers
of devices find their parts elsewhere.

Yes, nobody buys in scale from Digikey

Somebody must--from the Thief River Falls Reginal Airport webpage:

\"From FAA-published information gathered in 2012, the Airport
experiences 31,200 aircraft operations per year (takeoffs and landings)
of which 30,000 operations are attributed to general aviation
operations, 1,250 from air carrier commercial service, 2,600 from other
commercial/cargo service, and 100 military operations. Nearly 5,000
passengers (2,500 enplanement boardings) are processed annually from
scheduled airline flights. The Airport has 26 based aircraft, including
3 multi-piston and 3 jet turbine airplanes.

So that\'s 3750 commercial takeoffs and landings, and a total of 2500
passengers. (The webmaster is apparently a bit arithmetically
challenged, but oh well.)

Since there\'s not a lot in Thief River Falls except Digikey, one gathers
that _somebody_ must be buying a lot of stuff from them.

80cents is expensive in my world, but then again I also come from a
business where qty products sold was counted in millions



We lazy engineers often specify a Digikey part for an initial buy, and
our excellent purchasing lady buys quantities of the same thing
somewhere else, at a much better price.

Never mess with purchasing ladies--it\'s almost as stupid as insulting
somebody who\'s preparing your food. ;)

Digikey has/had a good multi-vendor search engine. It seems to make
less sense lately. A category might for example have 25,000 items, and
selecting a couple of very obvious qualifiers narrows it down to a
few, or none.

Selecting for example a Vcc range gives a huge list of confusing,
overlapping choices.

Yeah, their VCC, slew rate, GBW, and 3 dB bandwidth columns are so
broken. They could usefully add a column for total supply voltage--that
would help a good bit all by itself.h


Or specifying something explicit gets me non-qualifying parts.

Lots of broken links too.

Especially Samsung MLCC characteristic sheets. They keep breaking them
for some reason--you\'d think DK would just give up and mirror them.

Cheers

Phil Hobbs

Sometimes they link to a vendor site that requires registration and a
password and all that nonsense, just to see a data sheet.

Yeah, that scorches me, too.


\"The easiest thing in the world is not to sell.\"

- Melvin Goldstein

Mel got that from me.
 
L

Lasse Langwadt Christensen

Guest
torsdag den 20. januar 2022 kl. 03.54.12 UTC+1 skrev jla...@highlandsniptechnology.com:
On Wed, 19 Jan 2022 11:39:02 -0800 (PST), three_jeeps
jjh...@gmail.com> wrote:

On Wednesday, January 19, 2022 at 10:47:43 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Wed, 19 Jan 2022 13:22:38 +0100, Klaus Vestergaard Kragelund
klau...@hotmail.com> wrote:

On 19/01/2022 09.27, Jasen Betts wrote:
On 2022-01-18, John Larkin <jlarkin@highland_atwork_technology.com> wrote:
On Mon, 17 Jan 2022 23:59:27 +0100, Klaus Vestergaard Kragelund
klau...@hotmail.com> wrote:

On 12/01/2022 03.44, jla...@highlandsniptechnology.com wrote:
On Tue, 11 Jan 2022 14:44:57 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 23.31.35 UTC+1 skrev John Larkin:
On Tue, 11 Jan 2022 12:12:37 -0800 (PST), Lasse Langwadt Christensen
lang...@fonz.dk> wrote:

tirsdag den 11. januar 2022 kl. 18.09.07 UTC+1 skrev jla...@highlandsniptechnology.com:
On Mon, 10 Jan 2022 22:34:08 -0800 (PST), whit3rd <whi...@gmail.com
wrote:
On Monday, January 10, 2022 at 7:26:19 PM UTC-8, jla...@highlandsniptechnology.com wrote:
We\'ll have a 48 volt, 20 amp power supply that feeds eight plug-in
modular load boards. We want to measure all 9 currents ...

So, is this regulated voltages, nine of \'em, and you want the output currents
from nine low-V regulators, or is it input to the regulators, and is that input somewhere
in the vicinity of 48VDC? Are these nine currents going to include initial
charging of filter capacitors? How big are those capacitors? Are some of the
regulators switchers?

As for \'intelligent cutoffs\', do you want to do that at the high-side? A secondary
low-current power supply grounded at the high rail might be economic, if
you want to have nine sensors and nine cutoffs all near the +48V source.
There will be eight user-programmable power supplies, each running off
+48, all off a common +48 bus fed from a kilowatt bulk power supply.
Users can potentially install modules and program and load the
supplies such as to cave in the main source, which would be really
ugly. That\'s against the rules in the manual, but we need to protect
things if they do it.

there is quite a few high side load switches with short circuit protection
and a combined current monitor / error output , but I think most of them
are limited to a ~24V supply
The c code can shut down supply modules as needed. Each power supply
module has its own FPGA that the main controller can address. There\'s
no need to actually remove 48v power from the baby boards. The master
controller just needs to know the currents and have some reasonable
rules.

check.

depending on how good the ADC is, maybe just a resistivity divider on the 48V and after each shunt is good enough?





Looks like we can use the XADC in the Zynq, the main controller FPGA.
That\'s a 1-volt full-scale, pretty terrible ADC. The voltage drop
across a shunt might be 50 mV with a common-mode of 48 volts. So we
need some sort of high-side amp.

Maybe use the idea with the resistive divider, but then don\'t use the
ADC, use a slope converter for increased precision.
That can be done with one cheap comparator, and a loop from the FPGA/Micro

Add a resistor with known drop, to calibrate

It\'s fun to design circuits, until some clown sells an IC that does it
all for you cheap. Looks like we\'ll use INA280, which is available and
costs 80 cents, according to TI.

Digikey wants $3. Digikey has become a scalper.

Or at-least an agent for a scalper. They\'re listing Rochester Electronics
as supplier on some lines (other parts).

sfaik digikey mostly sells to designers, people making large numbers
of devices find their parts elsewhere.

Yes, nobody buys in scale from Digikey

80cents is expensive in my world, but then again I also come from a
business where qty products sold was counted in millions
We sell modest volumes of fairly expensive stuff. Parts cost is
roughly 12% of sales price, and people (purchasing, engineering,
manufacturing, testing) are a much bigger cost. Lately, availability
dominates cost.

PCBs and metal cases, both custom, are big costs.

We buy a fair amount from Mouser and Digikey, expecially things like
resistors that cost a fraction of a cent each.

A sales guy, from Arrow or Avnet or one of the biggies, said that
\"distributors\" don\'t stock parts much any more, but Mouser and Digikey
do. It looks to me that Digikey at least is trending that way too;
it\'s hard to tell right now.
--

I yam what I yam - Popeye

Interesting, so what is a rule of thumb, from your perspective, to go from parts cost to the manufactured product?
We define \"direct cost\" as unburdened parts cost plus labor. A selling
price of 3x DC is our lower limit. 6s is better, sometimes 10x is
possible where the value to the customer is high and there is no
competition.

We used to sell NMR gradient amplifiers to Varian/Agilent at maybe 4x
DC. They marked our price up 6x to their customers.

I wonder what a high-end iphone costs Apple.

afaict the estimations are that the BOM for a $1000 phone is ~$500
 

Welcome to EDABoard.com

Sponsor

Top