12SA7 AM Transmitter Kit Being Noisy [SOLVED]

[Cyberdragon]

I have an Antique Electronics Supply "Pine Board" transmitter that decided to start broadcasting a ton of static all of a sudden. There is slightly more hum than I'd like (might need new caps eventually) but it's not that bad and certainly not compared to the annoying fuzz. It's the type of crackling static you get from something like a bad connection or bad pot, but there are no pots or anything. I reseated the tube several times and it didn't have too much of a difference and I also whacked all components with a non-conductive object with only a tiny effect (which might just be from rattling the tube guts).

Things I've noticed are the noise does die down a bit (not as harsh but still blatent) when everything warms up after a while. The tube also does not sit totally flush with the socket, but does not seem loose. There are two suspect old carbon resistors in the circuit, one in the B+ dropper and the other the cathode bias.

I might have a poke around with a scope later to see if I can find it, whoever built the kit packed everything so tight it's gonna be a PITA to work on.

[Cyberdragon]

 

The scope says the cathode resistor is fine, nice clean sine wave.

However, the output is oddly being phase modulated, making it a bit "fat". Checking the control grid reveals a small signal of the same frequency and an odd waveform (the top is square with a hump and the bottom is a sine peak). This is of course getting mixed with the original sine wave, making the output a bit werid. It is a mixer tube being self exited, there will likely be some interference no matter what, but could that be the hiss I'm hearing? An FM signal being pumped into AM radios? If I touch the output changing the loading, the signal "rotates" modulating in both amplitude and frequency and any radio has real trouble picking it up at that point.

Note that even the scope is affecting this circuit, it's unstable as hell, so that interference could be from the scope.

EDIT while posting, probing the other side of the antenna decoupling capacitor reveals a clean signal! Their are two ceramics and a trimmer in parallel on the antenna output, is this really a case a crappy caps or am I falling into some bad probing trap?

[Cyberdragon]

Sort of a probing trap, I underestimated how weak this transmitter is, it's basically down in the noise floor of all the noisy CFL and LED lightbulbs, switching PSUs, digital noise, exc. So any crap that gets onto the antenna gets a nice little carrier wave mixed with it to go right into your radios. That's probably one of the reasons they don't sell it anymore.

So the output is way too high impedance, now what? I can't just bypass the caps, there's 141V on the other side. Bigger cap?

[Cyberdragon]

Tacking an additional 2nF capacitor into the antenna decoupling seems to have reduced the impedance enough to eliminate most of the noise. 

[Cyberdragon]

 

It was behaving, but now the sputtering and crackling is back (the background hiss/noise is still reduced, so the additional coupling cap did help). It comes and goes, it wasn't doing that during tests so I didn't catch it. But I'm sure the cathode resistor is bad/noisy at this point. Poking it with my fingers makes it pop, but I don't see any obvious bad joints.

I now hate carbon resistors...this might turn me into a carbon resistor shotgunner.* Screw you intermittent faults!

*JK not really, too lazy for that unless they are bad.

[Cyberdragon]

 

 


I said the soldering was fine, I'm blind. Poking the screen resistor (mounted next to the cathode resistor) made a worse noise but it's a modern resistor, so I thought maybe the joint was a bit cold and cracked.

It's a cold joint alright...it's so cold it's a no solder joint. And of course it never showed up on the scope cause I never probed it. TBF, it's a bit of a hard to see join under their, but in such a simple circuit, I should have checked everything.

I assumed...and made an ass of myself on the internet.

EDIT: It actually does have a microscopic amount of solder on it so it doesn't actually wiggle, which is how I was checking joints. But it's still the most likely culprit, no way that joint is sound.