Switching Regulator Noise...

R

Rickster C

Guest
Power source supplies 12 volts @ up to 10 Amps surge for a motor. A boost switcher converts this voltage to a level needed to charge the 12V battery (the data sheet shows a circuit for that, LT3796). A steering circuit selects the DC input when above 10V or the battery when the input is below 10 volts. A limiter/switch limits the voltage to the motor to 12V (manufacturer warning) and acts as a cutoff when over current.

The rest of the electronics needs 5 and 3.3 volt supplies with one 5V supply providing up to 300 mA mostly for the LCD back light LED current. The rest is maybe 50 mA combined. Some of this is sensitive analog circuitry. One device we are measuring has a nominal signal of 10 mV. The sensors are all slow, so power supply noise can and will be filtered in analog before the ADC and in digital in the delta-sigma ADC.

The switching charger circuit and the motor controller will be on a separate board at the back of the chassis, perhaps two feet/500 mm from the front of the chassis where the front panel, battery and controller board are. The motor is mostly in the back.

Some are concerned about the power wasted in the linear regulators for 5 and 3.3 volts and want them to be switchers. Rather than making them switchers, I\'m thinking add a 7V switcher on the rear board to supply a voltage for linear regulators on the front controller board. Then the wasted heat that the controller board has to dissipate is much lower.

I\'d really like to see a display used that doesn\'t consume so much current, but that ain\'t go\'na happen. The rest of the electronics is pretty low power even when using power from the 12V rail.

There is some irony in the fact that one engineer is opposed to driving the linears directly from 12V because the higher temps will lower the reliability of the linear parts while some mechanical parts of the machine are designed for a lifetime of weeks. I\'m pretty sure running a linear at 80 °C instead of 50°C is not going to make it crap out in a few weeks.

Does adding a 7V switcher to the PSU board in the rear of the chassis sound like a reasonable approach to reducing the power dissipated in the linear regulators on the front board?

What would you do?

--

Rick C.

- Get 1,000 miles of free Supercharging
- Tesla referral code - https://ts.la/richard11209
 
Just throw a AOZ1282 at it already.

Tim

--
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Design
Website: https://www.seventransistorlabs.com/

\"Rickster C\" <gnuarm.deletethisbit@gmail.com> wrote in message
news:2c716b80-f451-42d4-bd10-2c1c079881e0o@googlegroups.com...
Power source supplies 12 volts @ up to 10 Amps surge for a motor. A boost
switcher converts this voltage to a level needed to charge the 12V battery
(the data sheet shows a circuit for that, LT3796). A steering circuit
selects the DC input when above 10V or the battery when the input is below
10 volts. A limiter/switch limits the voltage to the motor to 12V
(manufacturer warning) and acts as a cutoff when over current.

The rest of the electronics needs 5 and 3.3 volt supplies with one 5V supply
providing up to 300 mA mostly for the LCD back light LED current. The rest
is maybe 50 mA combined. Some of this is sensitive analog circuitry. One
device we are measuring has a nominal signal of 10 mV. The sensors are all
slow, so power supply noise can and will be filtered in analog before the
ADC and in digital in the delta-sigma ADC.

The switching charger circuit and the motor controller will be on a separate
board at the back of the chassis, perhaps two feet/500 mm from the front of
the chassis where the front panel, battery and controller board are. The
motor is mostly in the back.

Some are concerned about the power wasted in the linear regulators for 5 and
3.3 volts and want them to be switchers. Rather than making them switchers,
I\'m thinking add a 7V switcher on the rear board to supply a voltage for
linear regulators on the front controller board. Then the wasted heat that
the controller board has to dissipate is much lower.

I\'d really like to see a display used that doesn\'t consume so much current,
but that ain\'t go\'na happen. The rest of the electronics is pretty low
power even when using power from the 12V rail.

There is some irony in the fact that one engineer is opposed to driving the
linears directly from 12V because the higher temps will lower the
reliability of the linear parts while some mechanical parts of the machine
are designed for a lifetime of weeks. I\'m pretty sure running a linear at
80 °C instead of 50°C is not going to make it crap out in a few weeks.

Does adding a 7V switcher to the PSU board in the rear of the chassis sound
like a reasonable approach to reducing the power dissipated in the linear
regulators on the front board?

What would you do?

--

Rick C.

- Get 1,000 miles of free Supercharging
- Tesla referral code - https://ts.la/richard11209
 
On 10/11/2020 05:00, Rickster C wrote:

The rest of the electronics needs 5 and 3.3 volt supplies with one 5V
supply providing up to 300 mA mostly for the LCD back light LED
current. The rest is maybe 50 mA combined. Some of this is
sensitive analog circuitry. One device we are measuring has a
nominal signal of 10 mV. The sensors are all slow, so power supply
noise can and will be filtered in analog before the ADC and in
digital in the delta-sigma ADC.

It depends how much your power budget is when compared to the motor.

If the motor is really running at around 100W then 3W wasted in the
electronics is not all that important. If the motor runs at 10W then you
could get 30% more runtime with mains offline before the battery dies.

Maybe power the display with a 5W dropper resistor so that the linear
regulators only have to deal with the clean analogue supply <100mA.

How bright is this thing? 300mA is insanely bright with modern bulk LED
efficiency even spread over quite a large area.

> What would you do?

The supply voltages you wanted are all common on consumer PC PSUs.

--
Regards,
Martin Brown
 
On Tue, 10 Nov 2020 01:14:56 -0600, \"Tim Williams\"
<tiwill@seventransistorlabs.com> wrote:

Just throw a AOZ1282 at it already.

Tim

I like to use the little 3-lead potted things that are switcher
drop-ins for 78xx regulators. That\'s sure easy. Most of them can be
used to convert pos to neg supplies, too.

SRH05 accepts up to 72 volts in.




--
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Design
Website: https://www.seventransistorlabs.com/

\"Rickster C\" <gnuarm.deletethisbit@gmail.com> wrote in message
news:2c716b80-f451-42d4-bd10-2c1c079881e0o@googlegroups.com...
Power source supplies 12 volts @ up to 10 Amps surge for a motor. A boost
switcher converts this voltage to a level needed to charge the 12V battery
(the data sheet shows a circuit for that, LT3796). A steering circuit
selects the DC input when above 10V or the battery when the input is below
10 volts. A limiter/switch limits the voltage to the motor to 12V
(manufacturer warning) and acts as a cutoff when over current.

The rest of the electronics needs 5 and 3.3 volt supplies with one 5V supply
providing up to 300 mA mostly for the LCD back light LED current. The rest
is maybe 50 mA combined. Some of this is sensitive analog circuitry. One
device we are measuring has a nominal signal of 10 mV. The sensors are all
slow, so power supply noise can and will be filtered in analog before the
ADC and in digital in the delta-sigma ADC.

The switching charger circuit and the motor controller will be on a separate
board at the back of the chassis, perhaps two feet/500 mm from the front of
the chassis where the front panel, battery and controller board are. The
motor is mostly in the back.

Some are concerned about the power wasted in the linear regulators for 5 and
3.3 volts and want them to be switchers. Rather than making them switchers,
I\'m thinking add a 7V switcher on the rear board to supply a voltage for
linear regulators on the front controller board. Then the wasted heat that
the controller board has to dissipate is much lower.

I\'d really like to see a display used that doesn\'t consume so much current,
but that ain\'t go\'na happen. The rest of the electronics is pretty low
power even when using power from the 12V rail.

There is some irony in the fact that one engineer is opposed to driving the
linears directly from 12V because the higher temps will lower the
reliability of the linear parts while some mechanical parts of the machine
are designed for a lifetime of weeks. I\'m pretty sure running a linear at
80 °C instead of 50°C is not going to make it crap out in a few weeks.

Does adding a 7V switcher to the PSU board in the rear of the chassis sound
like a reasonable approach to reducing the power dissipated in the linear
regulators on the front board?

What would you do?
 
On Tuesday, November 10, 2020 at 2:04:54 PM UTC-5, Martin Brown wrote:
On 10/11/2020 05:00, Rickster C wrote:

The rest of the electronics needs 5 and 3.3 volt supplies with one 5V
supply providing up to 300 mA mostly for the LCD back light LED
current. The rest is maybe 50 mA combined. Some of this is
sensitive analog circuitry. One device we are measuring has a
nominal signal of 10 mV. The sensors are all slow, so power supply
noise can and will be filtered in analog before the ADC and in
digital in the delta-sigma ADC.

It depends how much your power budget is when compared to the motor.

If the motor is really running at around 100W then 3W wasted in the
electronics is not all that important. If the motor runs at 10W then you
could get 30% more runtime with mains offline before the battery dies.

Maybe power the display with a 5W dropper resistor so that the linear
regulators only have to deal with the clean analogue supply <100mA.

That\'s actually a good idea. The resistor deals with the brunt of the dissipation near the motor where it will already be warm and not in the linear regulators.

The motor in operation runs at approximately 5 amps @ 12V so 50-60 watts. Perhaps not fully that because the high current is maybe half the cycle. So the wasted power is no where near 30%, but could possibly more like 5%-10%.

How bright is this thing? 300mA is insanely bright with modern bulk LED
efficiency even spread over quite a large area.

Don\'t know. I don\'t have a setup and the people who have them are not so good at figuring out how to make measurements. The data sheet lists 65 to 156 mA per and we are using two. There\'s a small, low power FPGA for some 5-10-20 mA (don\'t have it on a board yet) and an ST ARM processor for a few more mA. Some sensors for single digit mA each and not much else I can think of.


What would you do?

The supply voltages you wanted are all common on consumer PC PSUs.

Yeah, no, this is not a PC. I think the size alone would preclude that. I don\'t know what approval requirements they are going to put on this device if/when a company wants to build them, but using an external power brick like on a laptop is not a bad idea. We just need to get the team leader to give up on the crap barrel power connector... which ain\'t gonna happen.

So many bad compromises, so little time.

--

Rick C.

+ Get 1,000 miles of free Supercharging
+ Tesla referral code - https://ts.la/richard11209
 

Welcome to EDABoard.com

Sponsor

Back
Top