EDAboard.com | EDAboard.de | EDAboard.co.uk | WTWH Media

bias tees

Ask a question - edaboard.com

elektroda.net NewsGroups Forum Index - Electronics Design - bias tees

Goto page Previous  1, 2, 3, 4  Next

John Larkin
Guest

Tue Feb 05, 2019 6:45 pm   



On Tue, 5 Feb 2019 08:28:40 -0800 (PST),
bloggs.fredbloggs.fred_at_gmail.com wrote:

Quote:
On Monday, February 4, 2019 at 5:28:11 PM UTC-5, John Larkin wrote:
https://www.amazon.com/gp/product/B07D1GMXMD/ref=ppx_yo_dt_b_asin_title_o02__o00_s00?ie=UTF8&psc=1

We needed some bias tees so I tried Amazon. Prime, extra $4 for
overnight delivery.

Amazon is amazing.

We TDRd them, and they really are pretty good 6 GHz tees.

Gotta open one up and see what's inside.

You'll ruin it if you open it.

http://www.microwavejournal.com/articles/2809-a-broadband-microwave-choke

These things have been around 20 years.


And the patents have expired! The Piconics prices were outrageous.

I think there's too much emphasis on SRF, as if a part becomes
instantly useless above SRF.

A 47uF 1206 cap might have 1 nH of ESL, hence an SRF around 700 KHz.
But above 700 KHz, it's still only a nanohenry. It doesn't become
useless above 700K.

Same with inductors. A fraction of a pF is still a high impedance.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

Clifford Heath
Guest

Tue Feb 05, 2019 11:45 pm   



On 6/2/19 2:14 am, John Larkin wrote:
Quote:
On Tue, 5 Feb 2019 08:18:56 -0600, "Tim Williams"
tiwill_at_seventransistorlabs.com> wrote:

"Clifford Heath" <no.spam_at_please.net> wrote in message
news:R_46E.269058$0z4.266603_at_fx03.iad...
These are cool, ferrite filled like the conicals but higher current.

https://www.coilcraft.com/4310lc.cfm

I don't understand how they can have "Flat bandwidth with high impedance
to 6 GHz" but an SRF of 235MHz. Surely if they turn capacitive above
235MHz they're still capacitive at 6GHz?

Well, think about it. If there's more than 1/4 wave of wire there, then it
must fall over at some frequency of interest, and the fields will be very
different along the coil, not just a simple gradient. The first resonance
is good, it's a parallel resonance, high Z. The second resonance, and every
other one thereafter, is bad. The only way to keep impedance high, is to
absorb those resonances in a loss element.

If you use a tapped coil, you can add taps at strategic points, to dampen
across the antinodes of each resonance of interest.

If you use a solenoid or toroidial coil, a lossy material inside -- and
outside, mind, because the fields are external at these frequencies -- can
dampen that.

This keeps impedance flat, and a flat impedance must necessarily be lossy.
The loss sink is the core material.

Note that you can't simply taper the coil (coreless), or do a progressive
wound solenoid or whatever. There _must_ be a loss sink somewhere,
otherwise you're just making resonators in different shapes!

If you use a conical coil, you're really making a conical spiral antenna,
which has circular polarization and good directivity. Direct that energy
into a wad of ferrite, and it's gone forever -- rather than reflecting and
causing peaks and dips.

I don't think there's really anything very special that you can do with a
conical. The self-similar shape may make it easier to get a flat impedance,
but that impedance still isn't going to be very high, on account of how much
of the spiral is inducting versus how much is absorbing.

Tim

A conical inductor is the distributed equivalent of the old trick of
putting progressively larger damped inductors in series.


Thanks Tim, John, very enlightening. I'd seen that pattern on the PSPL
bias T photos but didn't understand the significance. The series RC
loading on successive inductors depends on the SRF of each inductor.

Clifford Heath.

John Larkin
Guest

Wed Feb 06, 2019 12:45 am   



On Wed, 6 Feb 2019 09:15:41 +1100, Clifford Heath <no.spam_at_please.net>
wrote:

Quote:
On 6/2/19 2:14 am, John Larkin wrote:
On Tue, 5 Feb 2019 08:18:56 -0600, "Tim Williams"
tiwill_at_seventransistorlabs.com> wrote:

"Clifford Heath" <no.spam_at_please.net> wrote in message
news:R_46E.269058$0z4.266603_at_fx03.iad...
These are cool, ferrite filled like the conicals but higher current.

https://www.coilcraft.com/4310lc.cfm

I don't understand how they can have "Flat bandwidth with high impedance
to 6 GHz" but an SRF of 235MHz. Surely if they turn capacitive above
235MHz they're still capacitive at 6GHz?

Well, think about it. If there's more than 1/4 wave of wire there, then it
must fall over at some frequency of interest, and the fields will be very
different along the coil, not just a simple gradient. The first resonance
is good, it's a parallel resonance, high Z. The second resonance, and every
other one thereafter, is bad. The only way to keep impedance high, is to
absorb those resonances in a loss element.

If you use a tapped coil, you can add taps at strategic points, to dampen
across the antinodes of each resonance of interest.

If you use a solenoid or toroidial coil, a lossy material inside -- and
outside, mind, because the fields are external at these frequencies -- can
dampen that.

This keeps impedance flat, and a flat impedance must necessarily be lossy.
The loss sink is the core material.

Note that you can't simply taper the coil (coreless), or do a progressive
wound solenoid or whatever. There _must_ be a loss sink somewhere,
otherwise you're just making resonators in different shapes!

If you use a conical coil, you're really making a conical spiral antenna,
which has circular polarization and good directivity. Direct that energy
into a wad of ferrite, and it's gone forever -- rather than reflecting and
causing peaks and dips.

I don't think there's really anything very special that you can do with a
conical. The self-similar shape may make it easier to get a flat impedance,
but that impedance still isn't going to be very high, on account of how much
of the spiral is inducting versus how much is absorbing.

Tim

A conical inductor is the distributed equivalent of the old trick of
putting progressively larger damped inductors in series.

Thanks Tim, John, very enlightening. I'd seen that pattern on the PSPL
bias T photos but didn't understand the significance. The series RC
loading on successive inductors depends on the SRF of each inductor.

Clifford Heath.


We've done that, a series string of inductors. The fast end is usually
a small ferrite bead, then a bigger bead, then a real inductor or two
with shunt resistors to kill the Qs.

Lately, we buy iron-powder filled inductors, solenoid or conical.

It's a challenge to make a truly wideband, stable current source.



--

John Larkin Highland Technology, Inc

lunatic fringe electronics

George Herold
Guest

Wed Feb 06, 2019 2:45 am   



On Tuesday, February 5, 2019 at 6:07:54 PM UTC-5, John Larkin wrote:
Quote:
On Wed, 6 Feb 2019 09:15:41 +1100, Clifford Heath <no.spam_at_please.net
wrote:

On 6/2/19 2:14 am, John Larkin wrote:
On Tue, 5 Feb 2019 08:18:56 -0600, "Tim Williams"
tiwill_at_seventransistorlabs.com> wrote:

"Clifford Heath" <no.spam_at_please.net> wrote in message
news:R_46E.269058$0z4.266603_at_fx03.iad...
These are cool, ferrite filled like the conicals but higher current.

https://www.coilcraft.com/4310lc.cfm

I don't understand how they can have "Flat bandwidth with high impedance
to 6 GHz" but an SRF of 235MHz. Surely if they turn capacitive above
235MHz they're still capacitive at 6GHz?

Well, think about it. If there's more than 1/4 wave of wire there, then it
must fall over at some frequency of interest, and the fields will be very
different along the coil, not just a simple gradient. The first resonance
is good, it's a parallel resonance, high Z. The second resonance, and every
other one thereafter, is bad. The only way to keep impedance high, is to
absorb those resonances in a loss element.

If you use a tapped coil, you can add taps at strategic points, to dampen
across the antinodes of each resonance of interest.

If you use a solenoid or toroidial coil, a lossy material inside -- and
outside, mind, because the fields are external at these frequencies -- can
dampen that.

This keeps impedance flat, and a flat impedance must necessarily be lossy.
The loss sink is the core material.

Note that you can't simply taper the coil (coreless), or do a progressive
wound solenoid or whatever. There _must_ be a loss sink somewhere,
otherwise you're just making resonators in different shapes!

If you use a conical coil, you're really making a conical spiral antenna,
which has circular polarization and good directivity. Direct that energy
into a wad of ferrite, and it's gone forever -- rather than reflecting and
causing peaks and dips.

I don't think there's really anything very special that you can do with a
conical. The self-similar shape may make it easier to get a flat impedance,
but that impedance still isn't going to be very high, on account of how much
of the spiral is inducting versus how much is absorbing.

Tim

A conical inductor is the distributed equivalent of the old trick of
putting progressively larger damped inductors in series.

Thanks Tim, John, very enlightening. I'd seen that pattern on the PSPL
bias T photos but didn't understand the significance. The series RC
loading on successive inductors depends on the SRF of each inductor.

Clifford Heath.


We've done that, a series string of inductors. The fast end is usually
a small ferrite bead, then a bigger bead, then a real inductor or two
with shunt resistors to kill the Qs.

Lately, we buy iron-powder filled inductors, solenoid or conical.

It's a challenge to make a truly wideband, stable current source.

Huh, can I ask a stupid question? (too late :^)
I made a bias tee, I didn't think too much about the inductor*.
(I picked one that fit in my form factor.)
Upstream is a bigger inductor and resistor, all fed by a
current source.

Why should I care about the inductor so much?

(Oh, it sucks the RF from my laser diode?)

George H.
Quote:


--

John Larkin Highland Technology, Inc

lunatic fringe electronics


Jeff Liebermann
Guest

Wed Feb 06, 2019 5:45 am   



On Mon, 04 Feb 2019 14:28:02 -0800, John Larkin
<jjlarkin_at_highland_snip_technology.com> wrote:

Quote:

https://www.amazon.com/gp/product/B07D1GMXMD/ref=ppx_yo_dt_b_asin_title_o02__o00_s00?ie=UTF8&psc=1

We needed some bias tees so I tried Amazon. Prime, extra $4 for
overnight delivery.

Amazon is amazing.

We TDRd them, and they really are pretty good 6 GHz tees.

Gotta open one up and see what's inside.


There are a bunch of photos of the guts found with Google Images.
<https://www.google.com/search?tbm=isch&q=bias+Tee&chips=bias+tee+lna,online_chips:6000+mhz>

These all claim 10 to 6000MHz and appear NOT to use conical inductors:
<https://www.ebay.com/itm/Bias-Tee-Wideband-1-6000-MHz-for-HAM-radio-RTL-SDR-LNA-Low-Noise-Amplifier-/273629284113>
<https://www.ebay.com/itm/New-10MHz-6000MHz-6GHz-Wideband-RF-Feeder-RF-Isolator-Bias-Feeder-Tee-50VDC-/172438957382>
<https://www.ebay.com/itm/Bias-Tee-10-6000-MHz-6GHz-For-HAM-radio-RTL-SDR-LNA-Low-Noise-Amplifier-/192724549249>
<https://www.ebay.com/itm/Bias-Tee-10MHz-6000MHz-6GHz-for-HAM-radio-RTL-SDR-LNA-Low-Noise-Amplifier-/223312879280>
etc...

My guess(tm) is that when you open up the package, you'll find an
ordinary solenoid wound chip inductor. I tried to find a Bias-Tee
that uses a conical inductor and failed. I did find some conical
inductors hiding inside an SMD package:
<http://www.piconics.com/smt/>
<https://www.globalsources.com/gsol/I/Chip-inductor/p/sm/1156321206.htm>
However, it's possible that I didn't recognize it on the photos
because those have an upper frequency limit of 15 to 65GHz and a 6GHz
conical inductor would be larger and probably be more obvious.




--
Jeff Liebermann jeffl_at_cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

Jeff Liebermann
Guest

Wed Feb 06, 2019 6:45 am   



On Mon, 04 Feb 2019 16:26:01 -0800, John Larkin
<jjlarkin_at_highland_snip_technology.com> wrote:

Quote:
On 02/04/2019 05:28 PM, John Larkin wrote:
https://www.amazon.com/gp/product/B07D1GMXMD/ref=ppx_yo_dt_b_asin_title_o02__o00_s00?ie=UTF8&psc=1
We TDRd them, and they really are pretty good 6 GHz tees.
Gotta open one up and see what's inside.

https://www.dropbox.com/s/hz4wj3fltsetqqj/Amazon_Tee.jpg?dl=0


Ugh. Using the 1mm dia SMA center pin as a measurement reference, the
device seems to have about 3mm of exposed and ungrounded center wire
on the RF port and about 2mm on the RF+DC port. Those appear as
inductors and can radiate of appear as a mismatch loss. The self
inductance is 0.215 nH/mm. At 6GHz, that would be
XL = 2Pi F L = 2 * 3.14 * 6*10^9 * 0.215*10^-9
= 8.1 ohms/mm
So, the RF port would have 16 ohms of added reactance to the 50 ohm
input impedance, and the RF+DC port would add 24 ohms. That's NOT a
great impedance match. The PCB should butt right up to the SMA
connectors leaving as little exposed center conductor as possible. At
least they used the correct SMA connector type and milled the package
to the correct wall thickness. However, the PCB was cut too small.

Did it really look good on a TDR?


--
Jeff Liebermann jeffl_at_cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

John Larkin
Guest

Wed Feb 06, 2019 6:45 am   



On Tue, 05 Feb 2019 20:34:56 -0800, Jeff Liebermann <jeffl_at_cruzio.com>
wrote:

Quote:
On Mon, 04 Feb 2019 14:28:02 -0800, John Larkin
jjlarkin_at_highland_snip_technology.com> wrote:


https://www.amazon.com/gp/product/B07D1GMXMD/ref=ppx_yo_dt_b_asin_title_o02__o00_s00?ie=UTF8&psc=1

We needed some bias tees so I tried Amazon. Prime, extra $4 for
overnight delivery.

Amazon is amazing.

We TDRd them, and they really are pretty good 6 GHz tees.

Gotta open one up and see what's inside.

There are a bunch of photos of the guts found with Google Images.
https://www.google.com/search?tbm=isch&q=bias+Tee&chips=bias+tee+lna,online_chips:6000+mhz

These all claim 10 to 6000MHz and appear NOT to use conical inductors:
https://www.ebay.com/itm/Bias-Tee-Wideband-1-6000-MHz-for-HAM-radio-RTL-SDR-LNA-Low-Noise-Amplifier-/273629284113
https://www.ebay.com/itm/New-10MHz-6000MHz-6GHz-Wideband-RF-Feeder-RF-Isolator-Bias-Feeder-Tee-50VDC-/172438957382
https://www.ebay.com/itm/Bias-Tee-10-6000-MHz-6GHz-For-HAM-radio-RTL-SDR-LNA-Low-Noise-Amplifier-/192724549249
https://www.ebay.com/itm/Bias-Tee-10MHz-6000MHz-6GHz-for-HAM-radio-RTL-SDR-LNA-Low-Noise-Amplifier-/223312879280
etc...


I wonder what's special about 6 GHz. Most tees spec 6 GHz.


Quote:
My guess(tm) is that when you open up the package, you'll find an
ordinary solenoid wound chip inductor. I tried to find a Bias-Tee
that uses a conical inductor and failed. I did find some conical
inductors hiding inside an SMD package:
http://www.piconics.com/smt/
https://www.globalsources.com/gsol/I/Chip-inductor/p/sm/1156321206.htm
However, it's possible that I didn't recognize it on the photos
because those have an upper frequency limit of 15 to 65GHz and a 6GHz
conical inductor would be larger and probably be more obvious.


Mine seems to use a small inductor or, likely, a ferrite bead.

https://www.dropbox.com/s/hz4wj3fltsetqqj/Amazon_Tee.jpg?dl=0

It has about a 60 ps rise time, which corresponds to about 6 GHz.

This one is 12 GHz:

https://www.minicircuits.com/WebStore/dashboard.html?model=TCBT-123%2B

$9.95.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

Jeff Liebermann
Guest

Wed Feb 06, 2019 7:45 am   



On Tue, 05 Feb 2019 20:58:52 -0800, John Larkin
<jjlarkin_at_highlandtechnology.com> wrote:


>I wonder what's special about 6 GHz. Most tees spec 6 GHz.

Dunno. My guess is that 6GHz is where the BNC connector craps out.
SMA will go much higher in frequency, but most of Bias-Tee devices I
played with many years ago used BNC.
<http://www.interfacebus.com/RF_Connector_Frequency_Range.html>
Tradition must be maintained even if nobody can remember why.

Quote:
Mine seems to use a small inductor or, likely, a ferrite bead.
https://www.dropbox.com/s/hz4wj3fltsetqqj/Amazon_Tee.jpg?dl=0


Probably a solenoid wound inductor (or straight through wire) wrapped
in a ferrite bead. That would allow placement on the PCB without an
air gap and reduce any tendency for the inductor to turn into an
antenna.

Same Bias-Tee for only $3.21 from Hong Kong. Just add a can and you
saved about $25:
<https://www.ebay.com/itm/Bias-Tee-10-6000-MHz-6GHz-For-HAM-radio-RTL-SDR-LNA-Low-Noise-Amplifier-/192724549249>
(Sorry, but I'm pathologically cheap).

>It has about a 60 ps rise time, which corresponds to about 6 GHz.

0.35 / rise_time(nsec) = bandwidth(GHz)
0.35 / 0.060nsec = 5.83GHz.
Yep, close enough.

I thought you were also looking at reflections along the 50 ohm
transmission line with the TDR. Methinks you'll find impedance bumps
at the SMA to PCB transition distance.

Also, the PCB should look like the amp posted by Gerhard Hoffmann
which has zero exposed SMA center pin:
<https://www.flickr.com/photos/137684711_at_N07/33112324418/in/dateposted-public/>
Note the holes drilled into the aluminum block to allow some clearance
for the corners of the PCB. One could radius or chamfer the corners
of the PCB, but drilling can be automated and is easier.

Quote:


It's difficult to be certain from the data sheet photo, but I don't
see a conical inductor.


--
Jeff Liebermann jeffl_at_cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

Gerhard Hoffmann
Guest

Wed Feb 06, 2019 7:45 am   



Am 06.02.19 um 05:58 schrieb John Larkin:

> I wonder what's special about 6 GHz. Most tees spec 6 GHz.

That is meant to suggest that they have a HP8753 opt 006 network
analyzer. Without opt 006 it only goes to 3 GHz; the option is
a frequency doubler.

The 8753 was user friendly in comparison to the really fast ones,
and good price/performance. Very popular.
The S-parameters of a lot of RF transistors end at 6 GHz for this.

I have just opened a Mini circuits ZFBT-4R2G-FT.
< https://ww2.minicircuits.com/pdfs/ZFBT-4R2G-FT+.pdf >

The first impression was OMG! everything was wrapped in some
teflon tape / goo and I was thinking: Whatever is inside, I can
do that better. So I cut it open, which was more work than expected.

There are 2 toroids close to the stripline and 3 larger two-hole cores.
2 of them have a lot of windings, trifilar or worse and all in series.
There are also some resistors and a lot of red glue to fix everything.
So, there seems more to it than meets the eye, but only one 1210 cap.

Tonight, I'll take some photos.

Quote:
Mine seems to use a small inductor or, likely, a ferrite bead.

https://www.dropbox.com/s/hz4wj3fltsetqqj/Amazon_Tee.jpg?dl=0

It has about a 60 ps rise time, which corresponds to about 6 GHz.

This one is 12 GHz:

https://www.minicircuits.com/WebStore/dashboard.html?model=TCBT-123%2B

$9.95.


That SMD housing is a good thing. The Piconics I have must be soldered
with their ultrathin wires, and glued to the board or they'll rip loose.

Now back to this f"§$%&/ ADA-4898-2 opamp that thinks it must oscillate
for no reason. Unity gain stable. Pah!

regards,
Gerhard

John Larkin
Guest

Wed Feb 06, 2019 4:45 pm   



On Wed, 6 Feb 2019 06:51:17 +0100, Gerhard Hoffmann
<ghf_at_hoffmann-hochfrequenz.de> wrote:

Quote:
Am 06.02.19 um 05:58 schrieb John Larkin:

I wonder what's special about 6 GHz. Most tees spec 6 GHz.

That is meant to suggest that they have a HP8753 opt 006 network
analyzer. Without opt 006 it only goes to 3 GHz; the option is
a frequency doubler.

The 8753 was user friendly in comparison to the really fast ones,
and good price/performance. Very popular.
The S-parameters of a lot of RF transistors end at 6 GHz for this.

I have just opened a Mini circuits ZFBT-4R2G-FT.
https://ww2.minicircuits.com/pdfs/ZFBT-4R2G-FT+.pdf

The first impression was OMG! everything was wrapped in some
teflon tape / goo and I was thinking: Whatever is inside, I can
do that better. So I cut it open, which was more work than expected.

There are 2 toroids close to the stripline and 3 larger two-hole cores.
2 of them have a lot of windings, trifilar or worse and all in series.
There are also some resistors and a lot of red glue to fix everything.
So, there seems more to it than meets the eye, but only one 1210 cap.

Tonight, I'll take some photos.


Please!

Quote:

Mine seems to use a small inductor or, likely, a ferrite bead.

https://www.dropbox.com/s/hz4wj3fltsetqqj/Amazon_Tee.jpg?dl=0

It has about a 60 ps rise time, which corresponds to about 6 GHz.

This one is 12 GHz:

https://www.minicircuits.com/WebStore/dashboard.html?model=TCBT-123%2B

$9.95.

That SMD housing is a good thing. The Piconics I have must be soldered
with their ultrathin wires, and glued to the board or they'll rip loose.

Now back to this f"$%&/ ADA-4898-2 opamp that thinks it must oscillate
for no reason. Unity gain stable. Pah!

regards,
Gerhard


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

John Larkin
Guest

Wed Feb 06, 2019 4:45 pm   



On Tue, 05 Feb 2019 21:45:10 -0800, Jeff Liebermann <jeffl_at_cruzio.com>
wrote:

Quote:
On Tue, 05 Feb 2019 20:58:52 -0800, John Larkin
jjlarkin_at_highlandtechnology.com> wrote:


I wonder what's special about 6 GHz. Most tees spec 6 GHz.

Dunno. My guess is that 6GHz is where the BNC connector craps out.
SMA will go much higher in frequency, but most of Bias-Tee devices I
played with many years ago used BNC.
http://www.interfacebus.com/RF_Connector_Frequency_Range.html
Tradition must be maintained even if nobody can remember why.

Mine seems to use a small inductor or, likely, a ferrite bead.
https://www.dropbox.com/s/hz4wj3fltsetqqj/Amazon_Tee.jpg?dl=0

Probably a solenoid wound inductor (or straight through wire) wrapped
in a ferrite bead. That would allow placement on the PCB without an
air gap and reduce any tendency for the inductor to turn into an
antenna.

Same Bias-Tee for only $3.21 from Hong Kong. Just add a can and you
saved about $25:
https://www.ebay.com/itm/Bias-Tee-10-6000-MHz-6GHz-For-HAM-radio-RTL-SDR-LNA-Low-Noise-Amplifier-/192724549249
(Sorry, but I'm pathologically cheap).


I had an engineer waiting to test a part, so I got two at $25 each,
overnight, in a can with SMAs. Engineers are pathologically expensive.

Quote:

It has about a 60 ps rise time, which corresponds to about 6 GHz.

0.35 / rise_time(nsec) = bandwidth(GHz)
0.35 / 0.060nsec = 5.83GHz.
Yep, close enough.


I just figured that 6x6 = 36.

Quote:

I thought you were also looking at reflections along the 50 ohm
transmission line with the TDR. Methinks you'll find impedance bumps
at the SMA to PCB transition distance.


When I get to work (later... this is Slacker Wednesday) I'll post the
TDR pics.

Quote:

Also, the PCB should look like the amp posted by Gerhard Hoffmann
which has zero exposed SMA center pin:
https://www.flickr.com/photos/137684711_at_N07/33112324418/in/dateposted-public/
Note the holes drilled into the aluminum block to allow some clearance
for the corners of the PCB. One could radius or chamfer the corners
of the PCB, but drilling can be automated and is easier.

This one is 12 GHz:
https://www.minicircuits.com/WebStore/dashboard.html?model=TCBT-123%2B
$9.95.

It's difficult to be certain from the data sheet photo, but I don't
see a conical inductor.


I've seen one in the flesh. It's some wire wrapped around a ferrite
cylinder.



--

John Larkin Highland Technology, Inc

lunatic fringe electronics


Guest

Wed Feb 06, 2019 5:45 pm   



On Monday, February 4, 2019 at 5:28:11 PM UTC-5, John Larkin wrote:
Quote:
https://www.amazon.com/gp/product/B07D1GMXMD/ref=ppx_yo_dt_b_asin_title_o02__o00_s00?ie=UTF8&psc=1

We needed some bias tees so I tried Amazon. Prime, extra $4 for
overnight delivery.

Amazon is amazing.

We TDRd them, and they really are pretty good 6 GHz tees.

Gotta open one up and see what's inside.


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com


You might have given me an idea. I have been pondering throwing something together and I might be able to make use of a couple of those.

In case you're curious, it is a cable/remote extender. We got a cable box in one room and a set top box in another. I want to take some RG6 to send the signal from one room to the other, and make it bidirectional. Not at the same time, but with a flip of the switch it turns around. One position it sends the set top box through the cable and the remote signal the other way. Flip the switch then it sends the cable output one way and switched the IR emitters and receivers. I was just going to use a resistor but using those will keep the cable signal integrity better. Actually it's not that critical, it'll just be channel 4 but this is much neater if nothing else.

Of course that will have to wait until I again have too much time on my hands.

John Larkin
Guest

Wed Feb 06, 2019 5:45 pm   



On Tue, 05 Feb 2019 20:57:35 -0800, Jeff Liebermann <jeffl_at_cruzio.com>
wrote:

Quote:
On Mon, 04 Feb 2019 16:26:01 -0800, John Larkin
jjlarkin_at_highland_snip_technology.com> wrote:

On 02/04/2019 05:28 PM, John Larkin wrote:
https://www.amazon.com/gp/product/B07D1GMXMD/ref=ppx_yo_dt_b_asin_title_o02__o00_s00?ie=UTF8&psc=1
We TDRd them, and they really are pretty good 6 GHz tees.
Gotta open one up and see what's inside.

https://www.dropbox.com/s/hz4wj3fltsetqqj/Amazon_Tee.jpg?dl=0

Ugh. Using the 1mm dia SMA center pin as a measurement reference, the
device seems to have about 3mm of exposed and ungrounded center wire
on the RF port and about 2mm on the RF+DC port. Those appear as
inductors and can radiate of appear as a mismatch loss. The self
inductance is 0.215 nH/mm. At 6GHz, that would be
XL = 2Pi F L = 2 * 3.14 * 6*10^9 * 0.215*10^-9
= 8.1 ohms/mm
So, the RF port would have 16 ohms of added reactance to the 50 ohm
input impedance, and the RF+DC port would add 24 ohms. That's NOT a
great impedance match. The PCB should butt right up to the SMA
connectors leaving as little exposed center conductor as possible. At
least they used the correct SMA connector type and milled the package
to the correct wall thickness. However, the PCB was cut too small.


Well, it is China Export. [1]

I use cheap edge-launch SMAs that have a big center pin that we solder
onto a microstrip. It's a pretty good connection for 30 ps edges. One
can neck down the trace at the pin end and make it even better. We
used ATLC to tweak that.

https://www.dropbox.com/s/1zb71vy9g576c6y/E-field.jpg?dl=0

>Did it really look good on a TDR?

I'll post later.

[1] Japan had a similar quality reputation once, but they got that
under control.

https://en.wikipedia.org/wiki/W._Edwards_Deming



--

John Larkin Highland Technology, Inc

lunatic fringe electronics

John Larkin
Guest

Wed Feb 06, 2019 5:45 pm   



On Wed, 6 Feb 2019 08:40:24 -0800 (PST), jurb6006_at_gmail.com wrote:

Quote:
On Monday, February 4, 2019 at 5:28:11 PM UTC-5, John Larkin wrote:
https://www.amazon.com/gp/product/B07D1GMXMD/ref=ppx_yo_dt_b_asin_title_o02__o00_s00?ie=UTF8&psc=1

We needed some bias tees so I tried Amazon. Prime, extra $4 for
overnight delivery.

Amazon is amazing.

We TDRd them, and they really are pretty good 6 GHz tees.

Gotta open one up and see what's inside.


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

You might have given me an idea. I have been pondering throwing something together and I might be able to make use of a couple of those.

In case you're curious, it is a cable/remote extender. We got a cable box in one room and a set top box in another. I want to take some RG6 to send the signal from one room to the other, and make it bidirectional. Not at the same time, but with a flip of the switch it turns around. One position it sends the set top box through the cable and the remote signal the other way. Flip the switch then it sends the cable output one way and switched the IR emitters and receivers. I was just going to use a resistor but using those will keep the cable signal integrity better. Actually it's not that critical, it'll just be channel 4 but this is much neater if nothing else.

Of course that will have to wait until I again have too much time on my hands.


Use relays.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

Jeff Liebermann
Guest

Wed Feb 06, 2019 8:45 pm   



On Wed, 6 Feb 2019 08:40:24 -0800 (PST), jurb6006_at_gmail.com wrote:

Quote:
I have been pondering throwing something together and I
might be able to make use of a couple of those.

In case you're curious, it is a cable/remote extender. We
got a cable box in one room and a set top box in another.
I want to take some RG6 to send the signal from one room
to the other, and make it bidirectional. Not at the same
time, but with a flip of the switch it turns around. One
position it sends the set top box through the cable and
the remote signal the other way. Flip the switch then it
sends the cable output one way and switched the IR emitters
and receivers. I was just going to use a resistor but using
those will keep the cable signal integrity better. Actually
it's not that critical, it'll just be channel 4 but this
is much neater if nothing else.


Yech. More like a flip of 2 switches. However, you don't need a
switch. See "RF directional coupler" or "RF hybrid coupler". One at
each end will give you a bi-directional connection without flipping
switches. If you don't like hybrid couplers, which potentially can
have intermodulation (mixing) problems, look into upconverters or RF
modulators that upconvert the IR remote carrier (about 40KHz) to RF
frequencies using baseband in one direction and cramming some RF
between two cable channels for the other direction.

If the cable frequencies on the coax are different from whatever the
remote control sends, you can do much the same thing with an "RF
splitter" or "diplexer".

Look into "MoCA" (multimedia over coax) which I think will do whatever
you're contemplating using ethernet.
<http://www.mocalliance.org>
<http://www.mocainyourhouse.com>

Last resort... run two coax cables, one for each direction. IR to RF
converter perhaps. Maybe an IR to fiber optic cable converter. Maybe
use CAT5 instead of coax:
<https://www.amazon.com/Remote-Over-Cat5-Extender-Repeater/dp/B0191D1R9O>

Quote:
Of course that will have to wait until I again have too much
time on my hands.


Perhaps wash your hands occasionally?

--
Jeff Liebermann jeffl_at_cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

Goto page Previous  1, 2, 3, 4  Next

elektroda.net NewsGroups Forum Index - Electronics Design - bias tees

Ask a question - edaboard.com

Arabic version Bulgarian version Catalan version Czech version Danish version German version Greek version English version Spanish version Finnish version French version Hindi version Croatian version Indonesian version Italian version Hebrew version Japanese version Korean version Lithuanian version Latvian version Dutch version Norwegian version Polish version Portuguese version Romanian version Russian version Slovak version Slovenian version Serbian version Swedish version Tagalog version Ukrainian version Vietnamese version Chinese version Turkish version
EDAboard.com map