Circuit conundrum... - UK Vintage Radio Repair and Restoration Discussion Forum

Ian - G4JQT · 26th Feb 2023, 11:53 am

Can anyone explain what's going on here please?

This is the output circuit of a wobbulator I made decades ago and works really well.

I had reason to do some checks and found the following odd results. At the base of the transistor there is about 15mV p-p, but at the emitter it rises to about 200 mV p-p. It's an emitter-follower with no gain so how can this be? I expect I put the 470k in series with the base to get the appropriate the output.

Experimenting today, I replaced the transistor with a FET with the same results.

How is gain possible using a emitter or source follower? Is it perhaps something to do with impedance matching? All voltages measured on a 'scope.

Thank you...

Attachment 273937

ms660 · 26th Feb 2023, 11:57 am

'scope probe/cable loading?

Lawrence.

trh01uk · 26th Feb 2023, 12:06 pm

I suspect its something to do with how you are making the measurements, which you haven't described. At the base, the source impedance is at the least the value of the series resistor, 470k as shown. Your scope probe/voltmeter probe/whatever will have at least a few pF of capacitance - and depending on the frequency that will be enough to load down the base circuit considerably, and give you a false "low" reading for voltage.

The emitter/source follower does its job admirably and lowers the source impedance to give you a "true" reading of the voltage.

Ian - G4JQT · 26th Feb 2023, 12:18 pm

Yes, possibly something to do with the scope/probe, although using a x20 probe and a Signent scope.

Bazz4CQJ · 26th Feb 2023, 12:49 pm

The FET circuit has come up in discussions on the Forum quite a number of times. It's been used in all kinds of published designs (I think the Haigue Wobbulator has something similar), but knowledgeable members using techniques like LT Spice say that a FET will exhibit negative resistance in such a circuit with consequent problems, though I freely admit that negative resistance is way above my pay-grade.

B

Neutrino · 26th Feb 2023, 4:34 pm

It is usually best to use 10x scope probes to minimise loading on the circuit compared to 1x scope probes.

If you have a dual channel scope check the base and emitter voltages simultaneously and see if the emitter voltage changes when you remove the scope probe from the base.

David

Cruisin Marine · 26th Feb 2023, 5:15 pm

Quote:

Originally Posted by Bazz4CQJ

The FET circuit has come up in discussions on the Forum quite a number of times. It's been used in all kinds of published designs (I think the Haigue Wobbulator has something similar), but knowledgeable members using techniques like LT Spice say that a FET will exhibit negative resistance in such a circuit with consequent problems, though I freely admit that negative resistance is way above my pay-grade.

B

Negative (Oscillator) resistance is not as complex as you may imagine, this video explains it well and why for example the series resistor is changed to make aged crystal oscillators run in a stable way in many Transceivers.
This is well explained in this video.
https://www.youtube.com/watch?v=scc17Z8126c

Bazz4CQJ · 26th Feb 2023, 5:50 pm

Quote:

Originally Posted by Cruisin Marine

Negative (Oscillator) resistance is not as complex as you may imagine, this video explains it well and why for example the series resistor is changed to make aged crystal oscillators run in a stable way in many Transceivers.
This is well explained in this video.
https://www.youtube.com/watch?v=scc17Z8126c

Thanks for that. I watched that video, but then spotted this video https://www.youtube.com/watch?v=KDVYL0I1o_8 with it's reference to tunnel diodes and found that very useful.

However, I'm still not sure about what the implications of negative resistance in the FET source follower circuit are. I think that circuit is often proposed to create a high impedance buffer ahead of the following circuit (for example, in an RF probe). What are the issues that arise?

B

trh01uk · 26th Feb 2023, 6:11 pm

Quote:

Originally Posted by Ian - G4JQT

Yes, possibly something to do with the scope/probe, although using a x20 probe and a Signent scope.

I few quick sums confirm this. Since this is a wobbulator, its probably 455kHz (standard IF) frequency that you are measuring. A scope probe with a typical 10pF loading, will have a reactance of about 35k ohms at that frequency. When that is combined with the (minimum) 470k source impedance for the signal, we get a voltage divider effect of around 14:1. I think that explains what you observe.

A quick way to check is to move your scope probe to the point between the 10n coupling capacitor and the first 470k resistor. You will probably see something much closer to the 2V p-p that you note on the schematic.

Ian - G4JQT · 26th Feb 2023, 6:31 pm

Thanks for the responses chaps.

First my typo - yes it's X10 probe.

And yes 2v P-P between the input C and 470k.

So I think we've probably cracked it. The 460 kHz signal source Z is high, and even a x10/10pF probe will load it.

Seems all is well after all!

Thanks again.

Ian

Cruisin Marine · 26th Feb 2023, 7:47 pm

However, I'm still not sure about what the implications of negative resistance in the FET source follower circuit are. I think that circuit is often proposed to create a high impedance buffer ahead of the following circuit (for example, in an RF probe). What are the issues that arise?

Do they refer to the other definition of Negative resistance where if you increase the supply voltage the current drawn actually drops?

Bazz4CQJ · 26th Feb 2023, 8:21 pm

Quote:

Originally Posted by Cruisin Marine

However, I'm still not sure about what the implications of negative resistance in the FET source follower circuit are. I think that circuit is often proposed to create a high impedance buffer ahead of the following circuit (for example, in an RF probe). What are the issues that arise?

Do they refer to the other definition of Negative resistance where if you increase the supply voltage the current drawn actually drops?

The references to previous discussions go back to a lengthy thread dating back to (maybe) 2017? It was focussed on how to build a sensitive RF millivoltmeter, initiated by the late and lamented 'Skywave' (Al). Reference to a number of circuits published in ham radio magazines noted that a common feature was the use of the FET source follower as a high impedance buffer. Various people had constructed such circuits, but found very poor outcomes.

I could try find the thread if you wanted to look at it.

B

Cruisin Marine · 26th Feb 2023, 8:26 pm

Yes please, I missed it totally
Thanks

Bazz4CQJ · 27th Feb 2023, 12:55 pm

Here you go

https://www.vintage-radio.net/forum/...d.php?t=131879

8 pages

B

Bazz4CQJ · 27th Feb 2023, 1:50 pm

Actually, the above link is probably not the right thread.

As I recall, between 2016 and 2020, Al created 4 or 5 threads which centred around RF probes, RF Millivoltmeters and wideband low level RF amps. Even at the time, it was hard to recall what comments were in which threads, at least 2 of which were merged.

I think the discussion of negative resistance in FETs, and the LTSpice analysis was mainly lead by Jeremy (G0HZU_JMR).

There was a lot of really useful discussion in those threads. Al developed and posted a circuit with 3 or 4 transistors which he was pleased with as the basis for a RF millivoltmeter with a flat response up to ~150MHz.

Happy Days!

B

26th Feb 2023, 11:53 am	#1
Ian - G4JQT Octode Join Date: Jun 2004 Location: Reading/Fakenham, UK. Posts: 1,326	Circuit conundrum... Can anyone explain what's going on here please? This is the output circuit of a wobbulator I made decades ago and works really well. I had reason to do some checks and found the following odd results. At the base of the transistor there is about 15mV p-p, but at the emitter it rises to about 200 mV p-p. It's an emitter-follower with no gain so how can this be? I expect I put the 470k in series with the base to get the appropriate the output. Experimenting today, I replaced the transistor with a FET with the same results. How is gain possible using a emitter or source follower? Is it perhaps something to do with impedance matching? All voltages measured on a 'scope. Thank you... Attachment 273937 Attached Thumbnails

26th Feb 2023, 11:57 am	#2
ms660 Dekatron Join Date: Apr 2011 Location: Cornwall, UK. Posts: 13,454	Re: Circuit conundrum... 'scope probe/cable loading? Lawrence.

26th Feb 2023, 12:06 pm	#3
trh01uk Octode Join Date: Nov 2010 Location: Ross-on-Wye, Herefordshire, UK. Posts: 1,654	Re: Circuit conundrum... I suspect its something to do with how you are making the measurements, which you haven't described. At the base, the source impedance is at the least the value of the series resistor, 470k as shown. Your scope probe/voltmeter probe/whatever will have at least a few pF of capacitance - and depending on the frequency that will be enough to load down the base circuit considerably, and give you a false "low" reading for voltage. The emitter/source follower does its job admirably and lowers the source impedance to give you a "true" reading of the voltage.

26th Feb 2023, 12:18 pm	#4
Ian - G4JQT Octode Join Date: Jun 2004 Location: Reading/Fakenham, UK. Posts: 1,326	Re: Circuit conundrum... Yes, possibly something to do with the scope/probe, although using a x20 probe and a Signent scope.

26th Feb 2023, 12:49 pm	#5
Bazz4CQJ Dekatron Join Date: Apr 2009 Location: Oxfordshire, UK. Posts: 4,935	Re: Circuit conundrum... The FET circuit has come up in discussions on the Forum quite a number of times. It's been used in all kinds of published designs (I think the Haigue Wobbulator has something similar), but knowledgeable members using techniques like LT Spice say that a FET will exhibit negative resistance in such a circuit with consequent problems, though I freely admit that negative resistance is way above my pay-grade. B __________________ Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch.

26th Feb 2023, 4:34 pm	#6
Neutrino Pentode Join Date: Oct 2016 Location: Maldon, Essex, UK. Posts: 184	Re: Circuit conundrum... It is usually best to use 10x scope probes to minimise loading on the circuit compared to 1x scope probes. If you have a dual channel scope check the base and emitter voltages simultaneously and see if the emitter voltage changes when you remove the scope probe from the base. David

26th Feb 2023, 6:31 pm	#10
Ian - G4JQT Octode Join Date: Jun 2004 Location: Reading/Fakenham, UK. Posts: 1,326	Re: Circuit conundrum... Thanks for the responses chaps. First my typo - yes it's X10 probe. And yes 2v P-P between the input C and 470k. So I think we've probably cracked it. The 460 kHz signal source Z is high, and even a x10/10pF probe will load it. Seems all is well after all! Thanks again. Ian Last edited by Ian - G4JQT; 26th Feb 2023 at 6:42 pm.

26th Feb 2023, 7:47 pm	#11
Cruisin Marine Octode Join Date: Jul 2018 Location: Worthing, West Sussex, UK. Posts: 1,003	Re: Circuit conundrum... However, I'm still not sure about what the implications of negative resistance in the FET source follower circuit are. I think that circuit is often proposed to create a high impedance buffer ahead of the following circuit (for example, in an RF probe). What are the issues that arise? Do they refer to the other definition of Negative resistance where if you increase the supply voltage the current drawn actually drops? __________________ "Behind every crowd, there's a silver Moonshine"

26th Feb 2023, 8:26 pm	#13
Cruisin Marine Octode Join Date: Jul 2018 Location: Worthing, West Sussex, UK. Posts: 1,003	Re: Circuit conundrum... Yes please, I missed it totally Thanks __________________ "Behind every crowd, there's a silver Moonshine"

27th Feb 2023, 12:55 pm	#14
Bazz4CQJ Dekatron Join Date: Apr 2009 Location: Oxfordshire, UK. Posts: 4,935	Re: Circuit conundrum... Here you go https://www.vintage-radio.net/forum/...d.php?t=131879 8 pages B __________________ Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. Last edited by Bazz4CQJ; 27th Feb 2023 at 1:08 pm.

27th Feb 2023, 1:50 pm	#15
Bazz4CQJ Dekatron Join Date: Apr 2009 Location: Oxfordshire, UK. Posts: 4,935	Re: Circuit conundrum... Actually, the above link is probably not the right thread. As I recall, between 2016 and 2020, Al created 4 or 5 threads which centred around RF probes, RF Millivoltmeters and wideband low level RF amps. Even at the time, it was hard to recall what comments were in which threads, at least 2 of which were merged. I think the discussion of negative resistance in FETs, and the LTSpice analysis was mainly lead by Jeremy (G0HZU_JMR). There was a lot of really useful discussion in those threads. Al developed and posted a circuit with 3 or 4 transistors which he was pleased with as the basis for a RF millivoltmeter with a flat response up to ~150MHz. Happy Days! B __________________ Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch.