The AM Forum

The Bias control (a pot) does not have the same effect on plate current it
used to have.

If I turn the pot to "zero", I get "zero" plate current. Advancing the pot
shows zero plate current until it suddenly jumps to 100 ma. Turning the pot
back does not affect the plate current until at a point it drops to zero.

I thought the pot had gone bad, but a test with an ohmmeter shows it
to be good.

I checked the bias voltage (at the pot wiper) with an Analog Meter and
it varies from ~-87 vdc on down with no hickups in the reading.

Any thoughts ??

73, Dick, W1KSZ

Dick

Check the connections from the meter to the shunt resistor that is used to measure the modulator plate current.

It could be a cold solder joint.

It could be oxidation (particularly on the positive side) of the associated meter switch contacts.

Check both the + side and the ground side of the meter connections in the "modulator" position. I.e. the switch lifts the ground side connection in the "plate" position, so you can't assume that the ground side connection of the meter is okay in the "modulator" position.


Stu

Dick,

Pull the modulator tubes and check the bias voltage at the grids to see what it shows.  Either check the modulator screen voltage VERY carefully or to be safe with the power off and HV discharged put the rig in the phone position and measure resistance from the screen terminals to the output of the HV filter choke.

I checked the resistance and it reads 12K. The Screen voltage reads +218 vdc
on my trusty $2 Harbor Freight DVM. I'll check the bias as soon as the HV cools
down.

73, Dick, W1KSZ

Could it possibly be a meter shunt value issue?

Al VE3AJM

The symptoms are non-linear, you can turn the pot 10 or 15 degrees
and nothing, then all of a sudden, plate current drops to zero from
around 150 ma or so. No adjustment between those two points.

73, Dick, W1KSZ

Before checking bias voltage check the mod cathode resistance to ground before turning power on.  It should read near zero.

If you don't have any spare 6146 tubes (any 6.3 variant of the family) for testing then it is safe to put 1 tube at a time in the modulator with no audio drive and see if that uncovers a bad tube IF the previous bias and resistance checks don't indicate an issue.

I agree that the problem could be a loose or corroded connection between the ground side of the cathode meter shunt resistor and chassis ground.

At low enough settings of the negative grid bias, this bad connection might arc through. If you make the grid bias negative enough it might open up again.

In between, it might be very non linear in terms of its I-V characteristic.

Stu

I found a spare set of 6146's (sitting in the Seneca I just got), I know they
work, I get 150 watts out on 6 from the Seneca.

Well, same problem ... so, the only thing left is the shunt resistor, buried
under some transformer. Guess I'll take a crack at removing it to see if
there's something wrong under there.

Thanks for all the replies, I'll post my findings on the shunt resistor.

73, Dick, W1KSZ

I finally got back to the Valiant after some delay.

I had to remove the Finals Bias Pot to get to a screw holding in
the choke blocking view of the Mod tube sockets. All looks well
in there.

Now it gets stranger, in fiddling around I noticed the Audio Gain
Pot will change the MOD current as I turn it. The higher I turn it
the lower the current ???

I thought it may be oscillating, but I listened on another receiver
and there is no sign of oscillation.

Anyone have any ideas on this one ??

Tnx, Dick, W1KSZ

Dick

This now seems like an oscillation problem, even though you haven't detected any oscillation. With all of the audio stages being AC coupled, it is hard to see how the audio gain pot would effect the resting cathode current in the modulator, except as a result of some kind of oscillation.

Try removing the tube in the audio chain that drives the modulator output tubes, and see if you can then properly adjust the modulator resting current.

Stu

Check the decoupling caps in the B+ chain that feed the audio circuits.

Fred

I remember somebody having some strange modulator issues when one of the 22 ohm resistors in the plate leads went up in value.  Take a look at them.

Put a scope on the RF out (go to dummy load, of course)

See if your scope either goes all fuzzy or you actually see an oscillation on the envelope.  If it acts normal while on the dummy, maybe you have an RF in the shack problem.  You might not hear a tone is the oscillation is ultrasonic (beyond your hearing of your receiver)

Dick

Further comments:

If there is an oscillation in the audio section, the audio oscillation frequency might be very high (ultrasonic). A high audio frequency oscillation would be greatly attenuated by the mod transformer, and by the presence of all of the .01uF capacitors (TVI filters) that are between the modulated B+ and ground in a stock Valiant. Therefore, you might not see or hear much modulation on the RF envelope of the transmitter's output.

Note that, looking at the schematic, there is no audio frequency bypass capacitor across the modulator bias supply (i.e. between the center tap of the secondary of the audio driver transformer and ground). The overall negative bias supply is bypassed; but the fraction of this overall negative supply voltage that goes to the modulator (via the voltage divider that includes the modulator bias adjustment pot) is not bypassed. Adding a 10uF capacitor from the center tap of the secondary of the audio driver transformer (or from the wiper of the modulator bias pot) to ground would be a good idea to try. [Note that the + side of the capacitor should be connected to ground.]

Note, also, that the clipper (being a non-linear function) could increase the tendency of the audio section to oscillate if it is not functioning properly.

I was also thinking about the screen voltage regulator (two VR tubes in series). If they were switching on and off, that would cause the screen voltage to jump up and down. However, that doesn't fit with the symptoms.

If there were a bad connection between the HV supply and the regulators (via the 12kohm dropping resistor) that might be causing the screen voltage to be turning on and off. There are a number of switch contacts in that path.

Stu

I pulled the 12AU7 Driver tube and I could adjust the Bias quite easily.

Thinking the tube might be bad, I changed it, no dice.

I then started probing with a scope;
Pin 6 - 6C4: No sign of oscillation
Pin 5 - 6C4: Shows signs of oscillation as soon as the Mod tubes draw
current
Pin 2 - 12AU7: Same as above.

Now, this rig was working OK, I made a couple of Q's with it and got
decent audio reports.

Also, it's been heavily modded in the Audio section, from comparing
schematics, it looks like the W6BM mods were done to it (after all
we are closer to the West Coast !!).

Open for suggestions.

Tnx, Dick, W1KSZ

That explains why your set is not working, it's all that foggy, smoggy West Coast air. ;D

Fred

Dick

Now try to see where the feedback is entering the audio chain.

Start by unplugging the microphone. See if that makes the oscillation stop.

If the oscillation is still there, remove the microphone preamplifier tube.

Stu

Here's what I tried and the results:

Connected Mike, still oscillated.

Pulled 12AU7 (Driver), Oscillation stopped.
Replaced 12AU7, still oscillated.

Pulled 6C4, Oscillation stopped.
Replaced 6C4, still oscillated.

Pulled 12AX7, still oscillated, OK, now it's got
to be downstream from the Mike Pre-Amp tube
(12AX7).

I replaced the Cathode By-Pass cap on the 12AU7,
still oscillated.

At this point I am stumped. Any recommendations ??

Tnx, Dick, W1KSZ

Mike

If there are mods, then I'm probably not looking at an accurate schematic (I'm looking at the stock Valiant schematic on the BAMA website)

If the clipper is still part of the circuit, try removing the 6AL5 clipper tube to see if that stops the oscillation.

Separately, has feedback from the modulation transformer's output to the 6C4 been added as an audio mod?

Stu

The rig has been modd'ed along the lines of the W6BM mod's. The Clipper
has been removed and the Feedback mod added.

I'll try breaking the feedback line to see what happens.

How important is this feedback mod ? Does it really improve the audio
that much ?

But, the rig did function earlier, got decent audio reports, so it's
something recent that happened.

I'll report back on my findings.

73, Dick, W1KSZ

Mike

You said that the transmitter was working okay before this problem arose.

Between the time it was working okay and the time the problem arose... did you by any chance remove the plate caps on the modulator tubes (maybe to change tubes) and then reinstall the plate caps on the opposite modulator tubes. Doing so would switch the polarity of the feedback from negative to positive

The feedback isn't critical, but it probably helps a little if it is implemented correctly

Stu

I broke the feedback loop and I can now adjust the Modulator bias
as before. Now to find out if that feedback really does make an
improvement, else I am ripping it out.

The only time I removed the Modulator tubes was after the problem
appeared. And the added feedback line to one of the plate caps
makes it near impossible to switch them.

So, possibly some component in the feedback loop went bad, no
problem if I just remove it all !!

73, Dick, W1KSZ

Does the feedback come back to the 6C4? If so, can you provide some details as to where it is connected.

Typically, one would use a capacitor to ground, at the point where the feedback is inserted... of suitable value to roll off the feedback ratio at higher audio frequencies.



Stu

Feedback comes back to the cathode of the 6C4. It consists of two
0.22 mfd 600 vdc caps in series, two 47K 2W in series, a 4.7K 5
Watt thence thru a 680 ohm 1/2 Watt to pin 7 of the 6C4. Pin 7
also has a 1.2K 1/2 Watt to ground.

Lots of parts, for what purpose ??

73, Dick, W1KSZ

Dick

(I apologize for calling you Mike in a few of my earlier posts)

The two capacitors in series are there to block the DC.

The two 47k ohm resistors in the resistor chain are part of the voltage divider. I'm not sure what the function is of the 4.7k ohm 5W resistor and the 680 ohm 1/2 watt resistor. I speculate that someone was trying to make these all add up to 100k ohms in order to literally match the recommended value.

The 1.2k ohm resistor from cathode to ground is in parallel with the impedance looking into the cathode of the tube (1/the transconductance)... which is roughly 1.2k ohms in parallel with 400 ohms = 300 ohms. This 300 ohms of resistance,  in conjunction with the 100k ohm resistor (above), is intended to form a 300 ohm : 100.3k ohm = .003:1 voltage divider (which is the feedback ratio, "beta", in the feedback formula)

If you want to keep the feedback, I would suggest that you try adding a 0.1uF capacitor from cathode to ground (in parallel with the existing 1.2kohm resistor). This will result in a 3dB reduction in the feedback ratio (and also the gain around the feedback loop) at audio frequency 5.3kHz. The feedback ratio will drop 6dB per octave after that.

If that doesn't work, you can try further reducing the feedback ratio by adding another 47k ohm resistor is series with the two 47k ohm resistors that are already there.

Stu

Note  I have modified my post, immediately above... to correct some errors

Stu

I re-connected the feedback loop and added the .1 mfd cap. I could
adjust the bias to the proper value but as soon as I started talking
into the mike, the Modulator Plate current just about pinned the meter.

So, I dis-connected the feedback loop and listened to myself on another
receiver. Not the best test, but it didn't sound bad, not the quality of
WQXR, but I am not in the broadcast business !!

I'll try a couple of on air tests and see what others think of the sound.

Thanks for all the help,

73, Dick, W1KSZ

Dick

okay...

Good luck!

Best regards
Stu

Dick

One last thought:

Since the Valiant went from working to not working (due to too much feedback):

Check the actual resistances of the two 47k ohm resistors in the feedback path. Since they are each dissipating power equal to about 50% of their rating on audio peaks (maybe more)... one or both of them may have decreased in value, over time, to the point where the gain around the loop is too high.

If so, try replacing them with a new pair of 2W (or 5W) 47k ohm (or 56k ohm) resistors... or a single 100k ohm 10W resistor. Wire wound is okay.

Also consider replacing the two 0.22uF capacitors. If they are leaking (even a little), that will significantly increase the power being dissipated in the resistors, and also change the biasing of the 6C4.

Stu

An interesting discussion concerning HF oscillations in a guitar amplifier when GNFB applied:

http://www.audiobanter.com/showthread.php?t=138909


quote from discussion, "But what I done loads the 6V6 with 9.4k a-a above about 30kHz, thereby stopping the oscillations"   (note a-a means anode to anode or plate to plate)

Perhaps the Valiant needs a series R-C Zobel network across driver or modulation transformer primary?

Jim
Wd5JKO

Could you put up/post/ draw/ a schematic of these W6BM? mods. It would make things a whole lot clearer as to whats going on with your Valiant.

Al VE3AJM

The W6BM mods were in Electric Radio some time ago. I had to
draw it out by hand to see what weas going on.

As it stands today, it's back in the cabinet and I will do some
audio checks with the locals today. If it sounds OK (or even
passable), I'll leave the feedback loop dis-connected.

If I find a quality 0.1 mfd 1 KV cap, then perhaps I'll re-do it.

There's a feedback mod by 'LHR that uses the vacated Clipper
pot that might be worth doing.

Thanks again to all that responded,

73, Dick, W1KSZ

Dick
All

If I were going to try to reconnect the feedback, here is what I would do:

Step 0: Make sure the Valiant's power is turned off, and all power supplies are discharged

Step 1: Leave the feedback disconnected from the cathode of the 6C4. I.e. the (approximately) 100k ohm resistor (i.e. two 47k ohm resistors in series) that is in series with the (approximately) 0.1uF capacitor (i.e. two 0.2uF capacitors in series) is disconnected from the cathode of the 6C4.  Add a 0.1uF capacitor from the 6C4 cathode to ground (i.e. in parallel with the existing 1.2k ohm resistor). You may already have this capacitor in place as a result of our previous exchanges.

Step 2: With the power still off, and the B+ supplies still discharged: use a 330 ohm resistor in parallel with a 0.1uF capacitor to connect the open end of the 100k ohm (i.e. the two 47k ohm resistors in series) feedback resistor to ground.

Step 3: Connect one probe from a dual trace oscilloscope to the cathode of the 6C4.

Step 4: Connect the other probe from a dual trace oscilloscope to the 330 ohm resistor || 0.1uF capacitor combination added in Step 2.

Step 5: Connect a low-level, 1kHz sine wave to the microphone input of the Valiant

Step 6: Making sure that everything is properly connected, and that nothing is inadvertently shorting to anything else: turn on the Valiant

Step 7. Observe the two waveforms on the oscilloscope: The voltage across the 330 ohm || 0.1uF capacitor should have zero DC value. If it does have a DC value, greater than 100mV... then the .1uF capacitor (actually two 0.2uF capacitors in series) is leaking too much.

Step 8. The two waveforms on the oscilloscope should be close to 180 degrees out of phase. The waveform across the 330 ohm || 0.1uF combination should be larger than the waveform at the cathode of the 6C4... maybe 2 times as large... but no larger than that. If the waveforms are not approximately 180 degrees out of phase at 1000Hz, then there is something introducing excessive phase shift in the 6C4=>12AU7=> driver transformer=> 6146 amplifier chain.

Step 9. Increase the frequency of the sine wave toward 5kHz. The phase shift between the two waveforms on the scope should increase by around 45 degrees (i.e. to around 225 degrees), and the amplitude of the sine wave across the 330 ohm||0.1uF combination should be closer to the amplitude of the sine wave at the cathode of the 6C4 (maybe 1.4 times as large, or less).

Step 10. Increase the frequency of the sine wave beyond 5kHz. The amplitude of the sine wave across the 330 ohm || 0.1uF combination should drop to less than the amplitude of the sine wave at the cathode of the 6C4 at a frequency at which the phase shift remains less than 270 degrees. I.e. As the audio frequency is increased, the gain around the (open) feedback loop must drop to less than unity before there is enough phase shift around the (open) feedback loop to cause positive feedback.

Step 11. Decrease the audio frequency of the sine wave to 200 and also to 50Hz... to make sure that, as the frequency is decreased, the phase shift around the (open) feedback loop remains well below 360 degrees until the gain of the (open) feedback loop drops to below unity.

If the gain around the loop remains above unity, when the phase shift increases toward 360 degrees... then increase the value of the 100k ohm feedback resistor.

Stu

I noticed in another post mention of the REA Modulation Monitor.

Does anyone use this ? How does it compare to say an SB-610 or 614 ?
I realize the REA does not show received signals.

73, Dick, W1KSZ