Valiant Audio Issue

[k7iou]

The other day my modulation was reported low well today it went haywire. I heard a sizzling sound so it's on the work bench. While on the bench I duplicated the issue. Negative feedback 0, PTT talking into mic I slowly increased the mic gain. Arcing started from the terminal strip for the modulation transformer to a 6146 modulation tube socket adjacent to the strip.
Transformer tested good with meg tester so no shorted windings. Tubes good in the modulation circuit. We believe it's going into oscillation.
Next step is to inject a signal into the microphone circuit and trace it with my oscilloscope.
Attached is my audio circuit.
Thoughts or suggestions?
Anyone seen this issue before?
Thanks
David
de k7iou

[WD5JKO]

David,

   Your modulator driver is a circuit I have not seen before. Did you design this, and if not where did it come from? I like several aspects of the circuit. It would be important to know if the Valiant had been modulating just fine before the failure.

    Well my take would be to see again where it appears to be sizzling. Perhaps the terminal strip connections are too close, or their is a conductive path now (carbon). Could you take some hi-resolution pictures in that area, and post them here? Somehow mark where the arc path is visible (mspaint).

   If you suspect there is feedback causing oscillation, then turn the feedback pot down, and to be sure clip the wire at the pot wiper. This is a high gain circuit, and improper power supply decoupling between the low level stages and the high level stages could cause some instability. This kind of instability is usually in the form of low frequency motor boating.

   Maybe all you need is to do some rewiring to replace a terminal strip. If you do maybe add more space between the points that are sizzling now.

Jim
Wd5JKO

[AB2EZ]

I don't see how the driver stage can be providing equal and opposite polarity audio waveforms to the 6146's.

Separately, the impedance to ground on each 6146 seems to be too high for the tubes to draw grid current (AB2) without introducing excessive distortion.

Stu

[k7iou]

Hi Jim, I purchased the rig several months ago and it was restored with this mod. He had the negative feedback disconnected. So I reconnected it. It was working great with my dynamic mic, repurposed phono tube preamp into Shure equalizer. I was using what he labeled Low Z input. The negative feedback smoothed out the raspiness on voice peaks.
A friend said it was arcing across the transformer terminal strips but I thought I saw it jump from the 6146 tube socket to terminal strip. I did notice the 6146 tube socket in question the metal base of the tube was noticeably hotter that the other. The tube tested ok. I tried another tube with same results, arcing.
So my dilemma is; is something failing? Ie transformer? Should I return the audio back to stock configuration?
I have to resolder the transformer to the terminal strip and will post a picture later.
After reading this article I wonder if this mod is too straining on the transformer?
http://www.w8ji.com/Johnson%20audio%20mods.htm
73 de k7iou

[WD5JKO]

I don't see how the drive stage can be providing equal and opposite polarity audio waveforms to the 6146's.

Separately, the impedance to ground on each 6146 seems to be too high for the tubes to draw grid current (AB2) without introducing excessive distortion.

Stu

  Stu, I made the same observations myself. That said, the cathode coupled phase inverter will have slightly different gain each side due to the less than perfect current source at the bottom (the 10K). If that resistor were replaced with a perfect CCS, the gain each side should be the same, so the plate load resistors would be the same. I see several high end audio circuits using a variation of this circuit, and some use a bipolar CCS at the bottom. As it is, staggering the plate load resistors as they did should work. I was thinking about a 2W 100K pot instead with the wiper going to B+, and adjust for a perfect balance.

   If you look up the 6146 audio specs, these put out a lot of power in ab1, so in the case of a Valiant, the little extra power gotten driving the grids into grid current (Ab2) is not justified.

Jim
Wd5JKO

[AB2EZ]

Jim

Good points!

I can now see how this would work.

Yes, the adjustable 2nd plate load resistor, to obtain balance, seems like a very good idea.

Now that you have pointed out the version with a CCS feeding the shared cathodes... I guess this is what is referred to as a "current mirror".

Thanks
Stu

[WD5JKO]

So my dilemma is; is something failing? Ie transformer? Should I return the audio back to stock configuration?
I have to resolder the transformer to the terminal strip and will post a picture later.
After reading this article I wonder if this mod is too straining on the transformer?
http://www.w8ji.com/Johnson%20audio%20mods.htm
73 de k7iou

I see I have been posted over....so lets see what the post was.. :-)

David,

   Going back to stock would be major at this point, and you'd have to procure an interstage transformer. Since it did work, and you were happy with it, I'd suggest studying what you got, and look into what happened first.

  People sometimes get mad at me, but I always use some kind of negative peak limiter. Think of an automobile analogy. Here the issue is a rear wheel drive street rod with rear leaf springs. Some folks put traction bars on each spring. Why bother because 99% of the time they do NOTHING? Well that 1% when you dump the clutch could take out the driveshaft when the rear end pinion angles upward putting a bad angle on the rear U-Joint. Limiting this tilting with the traction bars helps keep the car driving instead of FORD (Found ON Road Dead).

   When we beef up a modulator we are more prone to failure. Why? Same reason, and I won't go into all that here. Some of the bigger rigs put spark gaps across the mod transformer secondary, and sometimes we can play tricks with diodes and resistors to insure the modulator is always loaded; even when heavily over modulating. Perhaps what happened to you is something simple like dropping the microphone caused something to flash over. Once a gap is compromised, it will continue to arc.

Jim
Wd5JKO

[WD5JKO]

Now that you have pointed out the version with a CCS feeding the shared cathodes... I guess this is what is referred to as a "current mirror".

Here is a good example:
http://www.turneraudio.com.au/80W-AB1-amp-2014.html

The plate load resistors are identical here.

Notice the NFB loop connection on the bottom triode grid.

Jim
Wd5JKO

[Jim, W5JO]

That modification has been floating around for years.  I forget where it was originally published or who did it, but I have seen it or a slight variation in a couple of Valiants in years past.  At one time I had a copy of the original diagram with the information on it but I haven't seen it in years.

[Chuck...K1KW]

That is the classic long tail phase inverter, very common in many P-P amps.  I've used it to replace the driver transformer in Rangers, Valiants, and other rigs.  Works very well in a negative feedback loop due to its flat response.  Here is an article I found that explains it fairly well.  The two load resistors will wind up with unequal values for a balanced amplitude output from the inverted and non inverted outputs.

http://www.valvewizard.co.uk/acltp.html

As far as the arcing goes, is the terminal strip just bad and has carbonized due to arcing?  If so, replace it and add a small arc gap across the mod transformer.  The gap will be small, perhaps a 1/16" for those voltages.



Chuck

[k7iou]

Dug through my paperwork and found this mod. It is WA1HZK
Jim you were correct the primary's arced across each other and left a carbon track. I clecned them off with an eraser and alcohol, resoldered  the transformer in line, disconnected the negative feedback and tried the Valiant into a dummy load while slowly bringing up the mic gain. Seems ok.
Any idea on the 12bhy7 there is a 39k on one side and a 51k on the other?

[k7iou]

Well, the 3-100k resistors were tied in phase, we connected it to the other 6146 and using the oscilloscope we now could see the negative feedback going negative while the other was going positive. But when I reconnected the old clipper pot to the 2,2k on pin 7 of the 6C4 it went haywire and the modulation meter would peg full scale if you adjusted the clipper pot (NFB pot) up from 0 thru the 2 range. Disconnected both ends and back to normal and behaving. So something is out of wack with this circuit. Design or faulty component.
de k7iou

[WD5JKO]

This post is coming after a post that just came through. I hope I don't contradict anything.

Dave,

    The repair might not be a long term one. Time will tell. You might do as Chuck K1KW suggested and add a spark gap across the mod transformer secondary. Experiment with the gap, but start as he suggested. I'd curve the wires into a circle at the end so the gap will be like two spheres in close proximity.

    As far as those resistors are concerned, this was discussed earlier in this thread. If you want to refine things further, then you will need some test equipment. Do you have a dual trace scope with two 10X compensated probes? If so, you might inject a tone into the modulator, 1khz sine wave, adjust for 50% modulation. Then look at the 6146 grids with scope AC coupled to see if the AC drive levels are balanced.

   Another way is to have the lower of the two resistor values adjustable. If one is 51K, then make the other 25K with a 50K pot in series with it. With the pot centered then each 12BH7 plate has the same load. Vary from there.

   The actual modulator balance is dependent on the 6146 Gm matching as well as the mod transformer primary being equal turns either side of the CT. When you achieve balance, the 2nd harmonic should be at a null. If there was a lot of 2nd harmonic, you would see a dip in modulator plate current when balance is achieved.

   With an SDR receiver Pan Adapter this should be readily visible. Could also use a receiver CW filter and tune to the 2nd harmonic. I attach a image of one of my rigs. The 2nd harmonic shows a pretty deep null.

   Go ahead a roll back in the NFB if you want.

Jim
Wd5JKO

[WA5VGO]

Breakdown of the terminal strip that the modulation transformer terminates to is a well documented problem.

Darrell

[AB2EZ]

Focusing just on the feedback...

The 100k potentiometer might look, at first glance, to provide a wide range of adjustment for the feedback loop. But the resistance between the adjustable wiper and ground is just a few thousand ohms (2.2k ohms + the impedance looking into the cathode of the 6C4 stage... which is a little less than 1/the transconductance of the 6C4).

For simplicity, let's assume that the audio frequency impedance (resistance) looking into the cathode of the 6C4 is 1/(2500 micromhos)= 400 ohms

So, unless the wiper is almost all the way toward the grounded end of the potentiometer, the feedback ratio is between [400 ohms/300k ohms] and [400 ohms /400k ohms].

The voltage gain (looking into the cathode) of the 6C4 stage is 2500 micromhos x 15000 ohms = 37.5.

The voltage gain of (each phase) of the 12BH7 stage is in the ballpark of 0.5 x 39k ohms x 3100 micromhos = 60

The voltage gain of the 6146 is around 7000 micromhos x 10,000 ohms (or whatever the load is looking into the modulation transformer) = 70

Therefore, the total open loop voltage gain from the cathode of the 6C4 to the output of each 6146 is around 35.5 x 60 x 70 = 125,000 => around 100dB

It would seem (to me) that unless the wiper of the 100k ohm potentiometer is close to the grounded end, the product of the open loop gain and the feedback ratio is at least 125,000 x 400/400,000 = 125. I.e. 40dB of loop gain.

This seems high, from the perspective of maintaining stability at higher audio frequencies.

Therefore... to maintain stability, the wiper of the 100k ohm potentiometer probably needs to be close to the grounded end.

I think that a potentiometer with a smaller total resistance (maybe 5000 ohms) would do a better job of providing a useful feedback adjustment.

Stu
 

[AB2EZ]

When you moved the chain of 100k ohm resistors to the other 6146 plate... you switched the feedback from negative feedback to positive feedback. If you decide to reconnect the feedback loop... move the chain of 100k resistors back to the other 6146.

I.e. there are several phase inversions between the cathode input of the 6C4, and each of the 6146 outputs. It would take some careful thought to determine which 6146 output should be connected to the cathode input of the 6C4. Empirically, after make the switch... you picked the wrong one.

Well, the 3-100k resistors were tied in phase, we connected it to the other 6146 and using the oscilloscope we now could see the negative feedback going negative while the other was going positive. But when I reconnected the old clipper pot to the 2,2k on pin 7 of the 6C4 it went haywire and the modulation meter would peg full scale if you adjusted the clipper pot (NFB pot) up from 0 thru the 2 range. Disconnected both ends and back to normal and behaving. So something is out of wack with this circuit. Design or faulty component.
de k7iou

[k7iou]

Well, the 3-100k resistors were tied in phase, we connected it to the other 6146 and using the oscilloscope we now could see the negative feedback going negative while the other was going positive. But when I reconnected the old clipper pot to the 2,2k on pin 7 of the 6C4 it went haywire and the modulation meter would peg full scale if you adjusted the clipper pot (NFB pot) up from 0 thru the 2 range. Disconnected both ends and back to normal and behaving. So something is out of wack with this circuit. Design or faulty component.
de k7iou

When you moved the chain of 100k ohm resistors to the other 6146 plate... you switched the feedback from negative feedback to positive feedback. If you decide to reconnect the feedback loop... more the chain of 100k resistors back to the other 6146.

I.e. there are several phase inversions between the cathode input of the 6C4, and each of the 6146 outputs. It would take some careful thought to determine which 6146 output should be connected to the cathode input of the 6C4. Empirically, after make the switch... you picked the wrong one.

No they originally were in phase together. Now the waveform looks correct.



Mod. fix: please try to not put your response inside an existing quote. Makes it difficult to see who's saying what.

[k7iou]

hi Stu, I checked the pot after the haywire and it's a 350k pot.

Focusing just on the feedback...

The 100k potentiometer might look, at first glance, to provide a wide range of adjustment for the feedback loop. But the resistance between the adjustable wiper and ground is just a few thousand ohms (2.2k ohms + the impedance looking into the cathode of the 6C4 stage... which is a little less than 1/the transconductance of the 6C4).

For simplicity, let's assume that the audio frequency impedance (resistance) looking into the cathode of the 6C4 is 1/(2500 micromhos)= 400 ohms

So, unless the wiper is almost all the way toward the grounded end of the potentiometer, the feedback ratio is between [400 ohms/300k ohms] and [400 ohms /400k ohms].

The voltage gain (looking into the cathode) of the 6C4 stage is 2500 micromhos x 15000 ohms = 37.5.

The voltage gain of (each phase) of the 12BH7 stage is in the ballpark of 0.5 x 39k ohms x 3100 micromhos = 60

The voltage gain of the 6146 is around 7000 micromhos x 10,000 ohms (or whatever the load is looking into the modulation transformer) = 70

Therefore, the total open loop voltage gain from the cathode of the 6C4 to the output of each 6146 is around 35.5 x 60 x 70 = 125,000 => around 100dB

It would seem (to me) that unless the wiper of the 100k ohm potentiometer is close to the grounded end, the product of the open loop gain and the feedback ratio is at least 125,000 x 400/400,000 = 125. I.e. 40dB of loop gain.

This seems high, from the perspective of maintaining stability at higher audio frequencies.

Therefore... to maintain stability, the wiper of the 100k ohm potentiometer probably needs to be close to the grounded end.

I think that a potentiometer with a smaller total resistance (maybe 5000 ohms) would do a better job of providing a useful feedback adjustment.

Stu
 

[AB2EZ]

I'm referring to the pot that is in the feedback path (labeled "old clipper pot") in the schematic diagram that you posted.

Is that the pot you are referring to in your post (immediately above)

Stu

[AB2EZ]

I suspect that you do not completely understand how negative feedback works.

Stu

Well, the 3-100k resistors were tied in phase, we connected it to the other 6146 and using the oscilloscope we now could see the negative feedback going negative while the other was going positive. But when I reconnected the old clipper pot to the 2,2k on pin 7 of the 6C4 it went haywire and the modulation meter would peg full scale if you adjusted the clipper pot (NFB pot) up from 0 thru the 2 range. Disconnected both ends and back to normal and behaving. So something is out of wack with this circuit. Design or faulty component.
de k7iou

When you moved the chain of 100k ohm resistors to the other 6146 plate... you switched the feedback from negative feedback to positive feedback. If you decide to reconnect the feedback loop... more the chain of 100k resistors back to the other 6146.

I.e. there are several phase inversions between the cathode input of the 6C4, and each of the 6146 outputs. It would take some careful thought to determine which 6146 output should be connected to the cathode input of the 6C4. Empirically, after make the switch... you picked the wrong one.

No they originally were in phase together. Now the waveform looks correct.



Mod. fix: please try to not put your response inside an existing quote. Makes it difficult to see who's saying what.

[WD7F]

Turns out that when we compared the phase of the feedback with the signal on the cathode of the 6AU6, it appears that the feedback was in-phase.  So, we swapped it out.  Then everything looked good until we hooked it back in place and then it was really hosed.  Now, maybe I'm wrong, but to me, "negative" feedback means that the waveform coming from the plate of the 6146 will be out of phase from the mike audio on the cathode of the tube.  Didn't look like it was right.  We switched it and it was even worse.  So, for whatever reason, the feedback was disconnected altogether and it behaves normally.  That may explain why it was disconnected when Dave got the unit from the seller in the first place.  Now...here's the issue:  The audio is a bit raspy on peaks with Dave's voice without the feedback.  Why did it screw up?  Don't know.  But the good news is that the mod transformer and the audio chain is good.  I even figured out why one plate resistor of the phase splitter is a 51K and the other is 39K.  What do you think?

[k7iou]

How about someone posting a waveform showing negative feedback?
Thanks
David

[AB2EZ]

I can see how this might have confused you (or anyone else, including me).

The cathode of the 6C4 is a negative feedback point.

The grid-to-cathode voltage on the 6C4 is causing current to flow. This current flows through both the cathode resistor of the 6C4 and the plate resistor of the 6C4. The resulting cathode-to-ground voltage is in the direction that causes the grid-to-cathode voltage to decrease. I.e. the grid to cathode voltage being applied to the 6C4 is: [the grid-to-ground input voltage - the cathode-to-ground (feedback) voltage]

When the larger loop's feedback signal (from the output of one of the 6146s) is applied to the cathode of the 6C4, it must have the same polarity as the open loop 6C4 cathode-to-ground voltage... in order to produce negative feedback around the larger feedback loop.

Stu

Turns out that when we compared the phase of the feedback with the signal on the cathode of the 6AU6, it appears that the feedback was in-phase.  So, we swapped it out.  Then everything looked good until we hooked it back in place and then it was really hosed.  Now, maybe I'm wrong, but to me, "negative" feedback means that the waveform coming from the plate of the 6146 will be out of phase from the mike audio on the cathode of the tube.  Didn't look like it was right.  We switched it and it was even worse.  So, for whatever reason, the feedback was disconnected altogether and it behaves normally.  That may explain why it was disconnected when Dave got the unit from the seller in the first place.  Now...here's the issue:  The audio is a bit raspy on peaks with Dave's voice without the feedback.  Why did it screw up?  Don't know.  But the good news is that the mod transformer and the audio chain is good.  I even figured out why one plate resistor of the phase splitter is a 51K and the other is 39K.  What do you think?

[AB2EZ]

Separately, as I indicated in an earlier post... I think the feedback loop has much too much gain, unless the wiper of 100k ohm "the old clipper" potentiometer is almost all the way toward the grounded end. With too much gain around the loop (40db, in this case, unless the wiper is almost all the way toward the grounded end), there will me oscillation caused by phase shifts around the loop at higher audio frequencies. I.e. the loop gain will not decrease to below unity fast enough to prevent positive feedback (due to a 180 degree phase shift) at higher audio frequencies.

There are a couple of ways to fix this.

One way would be to replace this potentiometer with a 5k ohm potentiometer.

Another way to fix this would be to keep the existing "old clipper" potentiometer,  replace the 2.2k ohm resistor, between the wiper and the cathode of the 6C4, with a resistor whose value is roughly equal to the full resistance of the old clipper potentiometer. I.e. if the existing potentiometer has an end-to-end value of 100k ohms, then replace the 2.2k ohm resistor, between the wiper and the cathode of the 6C4, with a 100k ohm fixed resistor.

Turns out that when we compared the phase of the feedback with the signal on the cathode of the 6AU6, it appears that the feedback was in-phase.  So, we swapped it out.  Then everything looked good until we hooked it back in place and then it was really hosed.  Now, maybe I'm wrong, but to me, "negative" feedback means that the waveform coming from the plate of the 6146 will be out of phase from the mike audio on the cathode of the tube.  Didn't look like it was right.  We switched it and it was even worse.  So, for whatever reason, the feedback was disconnected altogether and it behaves normally.  That may explain why it was disconnected when Dave got the unit from the seller in the first place.  Now...here's the issue:  The audio is a bit raspy on peaks with Dave's voice without the feedback.  Why did it screw up?  Don't know.  But the good news is that the mod transformer and the audio chain is good.  I even figured out why one plate resistor of the phase splitter is a 51K and the other is 39K.  What do you think?

[WD7F]

What I posted about the phase of the feedback in relationship to the cathode to grid voltage of the 6C4 kept bothering me.  I worked on that in my head laying in bed last night and it became apparent that I must have been thinking wrong.  When I read Stu's comments, the light came on.  Yep, I was wrong.  I think we'll take another stab at it with a renewed understanding...except I have questions.

1.  Why not remove one of the 100K resistors to decrease the resistance to allow the correct feedback level not so close to ground on the wiper of the pot?

2.  What does 10db mean?  At what frequency is this to be measured and how would you measure it?  Is it linear across the audio range?  If so, when why do it? I'd guess not.

3.  Why are the plate resistors of the driver different?  Is it to make the positive portion of the modulation envelope larger?  Is it because you're hanging the feedback loop off of one of the plates and need to be rebalanced?

Questions, questions, so many questions and so little time.

John in Tucson
 

I can see how this might have confused you (or anyone else, including me).

The cathode of the 6C4 is a negative feedback point.

The grid-to-cathode voltage on the 6C4 is causing current to flow. This current flows through both the cathode resistor of the 6C4 and the plate resistor of the 6C4. The resulting cathode-to-ground voltage is in the direction that causes the grid-to-cathode voltage to decrease. I.e. the grid to cathode voltage being applied to the 6C4 is: [the grid-to-ground input voltage - the cathode-to-ground (feedback) voltage]

When the larger loop's feedback signal (from the output of one of the 6146s) is applied to the cathode of the 6C4, it must have the same polarity as the open loop 6C4 cathode-to-ground voltage... in order to produce negative feedback around the larger feedback loop.

Stu

Turns out that when we compared the phase of the feedback with the signal on the cathode of the 6AU6, it appears that the feedback was in-phase.  So, we swapped it out.  Then everything looked good until we hooked it back in place and then it was really hosed.  Now, maybe I'm wrong, but to me, "negative" feedback means that the waveform coming from the plate of the 6146 will be out of phase from the mike audio on the cathode of the tube.  Didn't look like it was right.  We switched it and it was even worse.  So, for whatever reason, the feedback was disconnected altogether and it behaves normally.  That may explain why it was disconnected when Dave got the unit from the seller in the first place.  Now...here's the issue:  The audio is a bit raspy on peaks with Dave's voice without the feedback.  Why did it screw up?  Don't know.  But the good news is that the mod transformer and the audio chain is good.  I even figured out why one plate resistor of the phase splitter is a 51K and the other is 39K.  What do you think?