EMRFD Message Archive 12045

Message Date From Subject
12045 2015-12-19 09:35:36 Ken Chase Termination Insensitive Amp (TIA)

Hi All


Our homebrew group is working on Pete Juliano's LBS and I would like to use the TIA as an RF Preamp. I was wondering if it was possible to very the gain of the TIA. I would have a var. pot. on the front panel if possible.





12046 2015-12-19 09:41:37 Ashhar Farhan Re: Termination Insensitive Amp (TIA)
of course it is. the paper on termination insensitive amplifiers has all the details.

- f

12047 2015-12-19 10:00:04 zx97lite Re: Termination Insensitive Amp (TIA)
Figure 5
12049 2015-12-19 12:07:56 w7zoi Re: Termination Insensitive Amp (TIA)
Hi Ken, and group,

I think that Farhan missed the point of your quest.  I believe you wanted to have an amplifier with gain that could be varied with a front panel pot.    I’m not familiar with the Juliano “LBS” (?) design that you mention or the scheme he might use to realize a termination insensitive amplifier.   You will have to communicate with him about that.

The termination insensitive amplifiers that Bob Kopski and I presented in an on-line article were two stage designs with resistor feedback in the first stage.   That input stage is a bipolar transistor amplifier with two forms of resistor negative feedback:  Emitter degeneration (Rd) and parallel feedback in the form of a collector to base resistor (Rf.)    The amplifier has a well defined input impedance match when the product (Rd)x(Rf) is a constant.     If Rd is increased and Rf is decreased, amplifier gain will drop, but input Z will remain constant so long as the product is constant.

The first stage amplifier input impedance is insensitive to output load impedance when the first stage collector load is a resistor within the overall circuit rather than a direct connection to an external load.  The second amplifier stage in our design isolates the load from the first stage collector.

Negative feedback amplifiers were popular at one time and new circuit variations appeared monthly in the trade journals.  I recall one article, although I must apologize for not remembering the source or author, where the Rd and Rf elements were replaced with PIN diodes.   The diodes had an RF resistance that was inversely proportional to the DC current.    These diodes were AC coupled to the feedback amplifier, but had DC flowing in them that was controlled by an op-amp (or similar) circuit that kept the product of the two diode currents constant.    An approximation to this is to maintain the sum of the diode currents as a constant.

EMRFD Fig 6.46 (p6.18) shows a circuit with five PIN diodes.   A differential amplifier feeds a constant current to the diodes.  Altering bias on the diff-amp distributes the current between the diodes to control overall circuit attenuation while maintaining a nearly constant input impedance.   I believe that this circuit appeared in an early Ulrich Rohde article.   A similar scheme might be used to realize a feedback amplifier with gain that could be controlled with a pot.    There are a lot of details left for you to implement, but that’s the fun of the game.

Good luck with the experiments.

73, Wes

12051 2015-12-19 15:56:17 Ken Chase Re: Termination Insensitive Amp (TIA)
Hi Wes

I was hoping you would jump in on this.

Pete's LBS is a stage by stage build of a 40m SSB Tcvr. It's a project to learn how to build and learn how the stages work. Pete uses common easily obtainable parts. The 2N3904 is the transisitor used throughout the build. He uses 3 small signal amps. Two for boosting the SI5351 VFO/BFO generator and one for an optional preamp. As it turns out the amps are not needed on the SI5351.

I have tried several of the feedback amps in EMRFD, trying to use var. pot in the emitter degen. Also tried a var. attenuator. I thought there might be a way to vary control the gain on the TIA. I do see that when the Rd emitter degen is changed so does the Rf feedback resistor.

I thought of going to the pin diode route, but trying to stay with the theme of common parts.

73 Ken VA3ABN

12053 2015-12-20 06:52:03 biastee Re: Termination Insensitive Amp (TIA)
Why continue to recommend the old PIN diode attenuator topology in EMRFD fig. 6.46 when a better & simpler alternative is available in the Waugh 4-diode attenuator? The latter requires fewer components and has better linearity. Google for "AN1048 A Low-Cost Surface Mount PIN Diode PI Attenuator" or read this: www.microwaves101.com/encyclopedias/waugh-attenuator.

  73, Chin-leong Lim, 9W2LC.

---In emrfd@yahoogroups.com, wrote :

EMRFD Fig 6.46 (p6.18) shows a circuit with five PIN diodes.   A differential amplifier feeds a constant current to the diodes.  Altering bias on the diff-amp distributes the current between the diodes to control overall circuit attenuation while maintaining a nearly constant input impedance.   I believe that this circuit appeared in an early Ulrich Rohde article.   (snipped)

12054 2015-12-20 09:03:37 Ashhar Farhan Re: Termination Insensitive Amp (TIA)
whoops! my bad.
how much gain variation are you looking for? much depends upon this. you could for instance, place a pot between two TIAs (as they are now called).
- f

12055 2015-12-20 10:25:24 w7zoi Re: Termination Insensitive Amp (TIA)
That attenuator in EMRFD Fig 6.46 was not a recommended circuit.   Rather, it was a circuit that used a PIN diode and, of greater significance, was driven with a differential amplifier.    A similar (and that does not mean identical) circuit might be applied to a feedback amplifier.

There are many possible routes to follow for this circuit.   The particular circuit that I mentioned is in many ways similar to the circuit that Waugh described back in 1992. 

12056 2015-12-20 14:16:09 kb1gmx Re: Termination Insensitive Amp (TIA)
TIAs are nice but don't AGC directly very well.  I have however put a 
attenuator (twin T flavor) before and after a TIA with good result.  The circuit 
used from the book with 1n1007s figure 6-46 and got a 50DB range.  Using two
one before and after the amp netted a range of -1 to -91db gain control.
With two TIAs age,tia, agc, TIA, I got a working gain from +38DB to -59db
with smooth gain control, low noise, and no issues with overload.  Stability 
was excellent.   FYI at HF 4007s work about the best of the "poor mans 
pin diodes".

By using switched TIAs (using relays) It was possible to have the amp
in the RX path with variable attenuators and TX path with fixed 
attenuators for stable TX gain.  Made optimizing the path gains 
easy while keeping the active device count low.  Relays were 
cheap and have no ESD issues like CMOS switches.  


12059 2015-12-20 18:35:47 Ken Chase Re: Termination Insensitive Amp (TIA)
Thanks for everyone's contribution.

Pete's LBS is a club group build and we'd like to keep it simple, true to the project. For some it might be too much. I thought of the PIN diode, but it steps outside of the "common" parts.

I will have to try and figure something out.



12060 2015-12-20 20:12:09 kb1gmx Re: Termination Insensitive Amp (TIA)
Depening on the what you do it and how I've found 1n4007 both common and very usable as:

Pin diode 
Varicap diode

They are not microwave diodes by any stretch but behave well and predictably.

That's the trick of common parts.  Understanding them and utilizing them.


12062 2015-12-21 09:55:23 Ken Chase Re: Termination Insensitive Amp (TIA)
Hi Allison

I would like a way to adjust the gain of Wes' Termination Insensitive Amp. I would like to use only one stage, not cascading more than one. If is necessary to use more than a varaible pot, then try to keep to common parts. Nothing exotic.

Many of the group members building Pete's LBS are learning, so need to keep it simple. If it was only me I would have go "exotic" or multiple stages, whatever it took to vary the gain.



12063 2015-12-21 11:42:14 kb1gmx Re: Termination Insensitive Amp (TIA)
Simple just use one.

I used two as I needed more gain and the range of AGC.

In that case I was not building a gain limited radio. it was a test fixture to see if 
cascaded stages and controls were a workable format.  So I used gain as I desired.
Also one stage of AGC would likely do fine.  I've done that before and even if 
the agc range is limited so was the gain and it did much to make my ears happier.

One flaw is to try and make a 1uV signal and a 500uv (S9++) signal sound the same.
a simple AGC using 1n4007 as control diodes with only 25-30db of range is often 
adequate. I refer back to the chapter on AGC. 

I'd go off and say build as Pete suggests and once working then introduce age 
as a further step.  It keeps the initial build simple and assures success as all the basic 
parts are now working rather than many parts that maybe don't.

While LBS is a good project you have to be careful to not introduce changes 
and features not well understood and then great a troubleshooting nightmare for
less experienced.  

12071 2015-12-21 18:49:49 Ken Chase Re: Termination Insensitive Amp (TIA)
Hi Allison

Thanks for the input. I did have a quick look at the agc circuits and I may have found something usable.

Unfortunately Pete's amp distorts the signal and impedances are not 50 ohms.