Distortion in the 813 transmitter

[KB5MD]

I have an 813 home brew transmitter that is modeled after the K1JJ transmitter using (2) 813’s modulated by (2)813’s.  Everything worked great for a while but now the audio is very distorted.  I have tried everything including replacing the mod iron, the driver transformer, the audio input amplifier, the microphone and both 813’s and it is still distorting.  Any ideas and suggestions?

[W1RKW]

got grid current? if so, what is it?

[w9jsw]

Are you still getting full carrier output with this distortion? Does screen current look correct? Can you look at the self modulation on the screen with a scope? Assume you are using the choke circuit rather than a screen dropping R?

[KB5MD]

 

I am a retired sign maker and graphic artist, not an engineer, I just have to stumble along in electronics, a hobby since junior high.  I value greatly any suggestions and tips
given on this board by folks much more knowledgeable than myself.

The transmitter diagram is attached

[W1RKW]

sorry, to early in the morning on my last post. I meant cathode current on the modulators. Check the relay contacts on the 50Kohm resistor and the wiring to the relay. Make sure you have good continuity when the relay is closed. check the wiring to the relay contacts.

[w9jsw]

My comments are referenced to the RF deck, not the modulator. Could a failure in the self-modulation of the screen be causing the distortion? Do all of the RF currents look ok when properly tuned up?

John

[K8DI]

All the grids in the modulator are at ground potential in the K1JJ transmitter.  I don't measure anything on the 500ma meter shown in the filament circuit center tap.

This says either there’s no HV on the plates or that the center tap isn’t grounded.  If the mod transformer is open or a connection to it failed, you would see no HV on plates.  If the filament center tap is open, or the meter is open, or the relay is bad, or those connections are open, there’d be no current.  Most of this is what W1RKW was saying, too.  Measure the resistance from cathode pin to ground with HV off, should be a few ohms at most in transmit, and 50k when in standby.

Ed

[KB5MD]

HV is on the plates,  cathode current is present, plenty of modulation but audio is distorted.  The speech amp checks okay on scope, no distortion there.  Good carrier, about 200 watts on bird 43.  I am going to check all caps in rf section tomorrow.  Those are about the only things  that haven’t been changed out.

This thing worked fine for several months, so something has failed.


This thing is driving me nuts!…..lol

[KD6VXI]

" I don't measure anything on the 500ma meter shown in the filament circuit center tap."

Then

"HV is on the plates,  cathode current is present"

This is impossible.  If the meter isn't indicating then the meter is bad or you have no cathode current.

--Shane
WP2ASS / ex KD6VXI

[KB5MD]

I appreciate everyone’s efforts and advice but, since this is not my primary AM transmitter and the fact that I have spent too much time already trying all recommendations;  I am abandoning this project.  I’ll chalk this up to experience!  Thanks again, see ya on the bands. 73

[W1RKW]

Roy,
don't give up just yet.  we need one or two more clues.  since you have an oscilloscope handy and you have good audio from the speech amp to the audio transformer, try to get a look at the secondary or output of the audio transformer to the grids of the 813s.  You'll need to do both taps or each grid connection on each modulator tube which ever is easiest to get at.  Might be worth it at this point in time to add some external test points here for future use so you don't have to disassemble next time there's a need to troubleshoot. Can you take a picture of the envelope with modulation and post it?  Keep at it, we'll finger it out.

[w9jsw]

What are you using for the mod transformer? Have you disconnected it and verified the windings are intact and are not shorting to the case? Do you have spark gaps on the mod iron? What was your implementation for sequencing the rig?

[K1JJ]

Hi Roy,

I've been where you are many times.... :-)  I have given up in frustration and later found the problem by chance. It is usually a simple solution. The master simulation gremlins want you to give up, but don't do it!

What you uncover may help other guys in the future too. The good thing is that the rig was working great before.  Keep faith cuz it's a FB rig - my own 813s X 813s  rig is the best sounding plate modulated rig I own... It will do 900 watts out carrier fully modulated with 3KV HV if pushed...

Let's try a few more things...  Could the mod xfmr spark gaps be pushed together and shorted by accident?  My own 813 rig had that mistake and drove me nuts a few years back.

Is the mod xfmr blown? (shorted or open) You can also disconnect the RF output and put few K load of power resistors on the mod xfmr output and look at the waveform.  A 1 KC sine wave into the audio driver is all you need for all scope tests.

Maybe you have another mod xfmr you can clip lead in there to test. Even a small ART-813 mod xfmr will substitute OK.

Is the mod reactor choke OK?  Measure its reactance and DC resistance for reasonable values.

Now, the real test:  Using the scope with a HV rated high voltage divider or HV probe!! for safety, look at the input sinewave audio to the grids of the modulators and then again at the mod xfmr input and output leads. Where does it start looking funky? Also see what the audio waveform looks like at the input to the RF choke and 500 pF bypass cap.  Is the bypass cap OK and measure 500 pF and of infinite DC ohms?  Subtitute the bypass cap AND the RF coupling cap, both verified at 500 pF.

* I assume the 813 tubes are all OK and were substituted already.*

There's some more stuff I can suggest if this doesn't do it.  You will be so proud of yourself when you find it. We won't let you fail, OM!

Tom, K1JJ

** Could you post a series of pictures of the complete rig?  

[KB5MD]

Well, out of frustration, I disassembled the modulator circuit and rebuilt it using 811a’s.  The darn thing works beautifully.  I never found what the problem was with the 813 modulator.  All is well that ends well, I guess.  Thanks again, everyone!

[K1JJ]

"Well, out of frustration, I disassembled the modulator circuit and rebuilt it using 811a’s.  The darn thing works beautifully.  I never found what the problem was with the 813 modulator.  All is well that ends well, I guess.  Thanks again, everyone!"

FB, Roy.  Whatever it takes to get the rig running....

My guess is it was probably a bad connection, like maybe the modulator tube pins (or soldered pin joints) in the socket or connections thruout the modulator circuit.

But anyway, you did good -  the 811A or 572B triodes are a better choice for modulators - though the 813s are usually less expensive and easier to find.  The 813s are very close in performance when in triode-connected service.

Please post some pics of your final rig config.

Later -

T

[KD1SH]

  Congrats on getting it working, Roy. As someone who has a couple of 813 rigs upcoming, I was following this thread with interest.
  One of my 813 rigs is only in the chassis layout stage at this point, but it will be using Gates BC-1T iron and probably 810's in the modulator. The smaller one—RF deck picked up at Nearfest partially completed—will use a pair of 811's or 572B's in the modulator, with either a UTC VM-4 or Thordarson T-21M65 mod transformer.
  Tom mentioned a high-voltage scope probe; if you don't have one, I strongly suggest you pick one up. They're affordable, and an indispensable part of your workbench setup if you're playing with transmitters and modulators.
  Good luck.

[KD6VXI]

Will a HV probe work as a scope probe?

I never thought to try it.  But I have one of the fluke 40kv sticks. Basically a few gigohm hv resistor in series with a smaller one making a voltage divider.

Should work as a scope probe, no?

--Shane
WP2ASS / ex KD6VXI

[AJ1G]

Don’t want to sound like Captain Obvious, but did you check to see if you have lost one of the push pull modulator 813s?   One of the 811 mod tubes in the ART-13 went bad here a while back, still was modulating but sounded awful and mod envelope looked horrible.  Swapped in a pair of 572Bs (didn’t have any spare 811s at the time) and all was well.

[AJ1G]

Don’t want to sound like Captain Obvious, but did you check to see if you have lost one of the push pull modulator 813s?   One of the 811 mod tubes in the ART-13 went bad here a while back, still was modulating but sounded awful and mod envelope looked horrible.  Swapped in a pair of 572Bs (didn’t have any spare 811s at the time) and all was well.

.

Well that was weird, for some reason I was not seeing the last few days updated posts…glad to hear you got things working again.  Did you ever check the original 813s to see if you actually had one go bad on you?

[KD1SH]

  I don't see any reason why not. You'd need to calculate the divider's ratio exactly, so you could do the volts/division math, since I doubt many scopes have a setting for anything beyond X100. Many of my older scopes don't even have a X100 setting at all.
  There would be no compensation adjustment, either, which might be a problem if you intend to observe high-frequency signals imposed on high-voltage DC.

Will a HV probe work as a scope probe?

I never thought to try it.  But I have one of the fluke 40kv sticks. Basically a few gigohm hv resistor in series with a smaller one making a voltage divider.

Should work as a scope probe, no?

--Shane
WP2ASS / ex KD6VXI

[K8DI]

Will a HV probe work as a scope probe?

I never thought to try it.  But I have one of the fluke 40kv sticks. Basically a few gigohm hv resistor in series with a smaller one making a voltage divider.

Should work as a scope probe, no?

--Shane
WP2ASS / ex KD6VXI

I have one of those, too. I’ve used it as a probe on the RCA’s modulators. The calibration is way off but it’s fine if you do the math (no modulation shows x volts, so 100% is 2x volts, even if x volts is nothing close to the actual HV).  The reason it’s off is it’s designed around a typical Fluke DMM with a 10M input, and the scope is 1M.

I’d be hesitant to connect it to the RF finals and expect decent results, the capacitance would likely significantly affect things…but it works fine at audio frequencies.

Ed

[KB5MD]

The 813 rig modulates great but had developed a chainsaw sound in the receiver in use.  After numerous attempts including relocating some of the wiring, grounding the cabinet, relocating electrolytics and what seemed like a hundred other attempts, I removed one of the rf 813’s and behold,the noise was gone.  One of the 813’s was shorting.  It just took two days to find it……whew!

[DMOD]

I have an 813 home brew transmitter that is modeled after the K1JJ transmitter using (2) 813’s modulated by (2)813’s.  Everything worked great for a while but now the audio is very distorted.  I have tried everything including replacing the mod iron, the driver transformer, the audio input amplifier, the microphone and both 813’s and it is still distorting.  Any ideas and suggestions?

Have you checked the PTT relay to make sure the 50k "cutoff" resistors were being by-passed (shorted out) on transmit?

[Opcom]

A HV probe will work well as a scope probe for the audio range.

For a flatter response to higher frequencies, a compensated probe would be needed, and that is another matter because commercial HV compensated probes are expensive, and home brewing them needs extra attention to safety, capacitor voltage ratings, etc. A side note there is that many olde-tyme HV vacuum rectifiers like the 1X2, 1B3, and 3A3 make nice HV capacitors and are almost free.

Using random HV "TV set" probes we see from the 50's-80's: Usually the probe can be opened, and the value of the HV resistor discovered. This, in combination with the known input impedance of the VTVM or scope, will give you the reading to expect for any given high voltage being measured.

Depending on the current drawn by the probe/meter combination, an amount of power will be dissipated in the probe's high-value resistor. At the 10-40KV range, it's important to know what this will be, so that the resistor does not overheat when making long continuous measurements. Most TV set HV probes are rated for intermittent use because the resistor may be rated only 2-5 Watts at most. At the voltages discussed in this topic that should not be an issue.

It's simple to make your own probe for the voltage ranges in the topic.
Link to a presentation including such HV measurement probes:
https://bunkerofdoom.com/lit/CRT-scopes/resource/hvprobes.pdf

link to a book on probes including high voltage for scope use, see chapter 6, particularly page 114:
https://bunkerofdoom.com/lit/CRT-scopes/resource/zucconi_probes.pdf