So the design has changed again, I need an amplifier to make up for the losses in the filter and balanced modulator (and to compensate for the fact that the carrier oscillator is running kinda "light" - this gives better stability, which is super important in SSB and CW operation). So there will be both a pre and post filter amplifier, I've actually duplicated the same circuit.
I've had issues with my RF level measurements as my home made probe's diode burns up due to how strong the signal is after my pre filter amp! (Im proud of that stage, as I designed it myself). This caused me to think I roasted that tube - I did have it biased wrong and the screen dissipation was very close to exceeding the tube's specs - but its not damaged, and I added two resistors (the 47k and 22K) to stabilize the screen and lower the screen current. It does not appear it cost me any gain, either.
I ALSO found out that my meter might not be reading correctly as it appears it does not understand RF even when rectified (I will talk about that again below). I've been in component level electronics, even worked as a repair tech for 25 years and I've learned SO MUCH these past weeks. This is so much fun.
Again, I thought I burnt up the pre filter amp tube, but here is a copy of a post I made on another forum, I will simply use this text to round out tonight's update:
The Pre-filter amp is FINE. I blew the diode in the probe again. The amp is putting out so much voltage, it nukes the diode but not instantly and it does not always destroy the diode. The diode conducts heavily in an "avalance" or "zener" mode which means it "goes short" at high RF voltages but when those voltages are removed, if done quickly, the diode either is still OK (so it measures the carrier generator level fine, giving me the idea the probe is fine) or it goes 100% short, which it did today which is how I caught what was really going on.
I found that if I go from a 1N34A to a 1N914, it tolerates high levels better and if I go to a 1n4007, it suffers no damage. Either the diodes are avalanching, or the meter I am using does not understand high RF voltages even when rectified. But the 34A is the only one that is accurate. Accurate, but not tolerant of more than about 3 volts of RF. And as I said, my meter may not like RF and I think THAT is also part of my problem.
In any case I am going to borrow a Simpson 260 which I KNOW can take 4mHz RF. Next weekend I willl get to the bottom of it.
In any case, the pre filter tube is not destroyed by overvoltage - I did not see how it could be. IT is fine, but I did change that stage to 250v and added a stabilizing resistor network (the 20k and 47k resistors) to the screen circuit to throttle the tube back within specs, it was at 95% of screen dissipation before.
I built the post filter amp today as well, but was too crimped for time to test it.
I also added the permanent filament transformer and moved all transmitter tubes over to it, reserving the bias/filament transformers' filament winding for the outboard VFO's tubes and pilot lights.
I learned a bit about how small signal diodes can fail - I had no idea they could act is such unpredictable ways, and I did component level repair for 25 years.
Anyway, today was encouraging and next weekend I MIGHT be able to plug in the external VFO and generate an actual 40 meter (7 mhz) signal and hear myself on my receiver! I do not expect good audio or very linear performance, as the ALC is not built yet (gotta wait to get the driver amp and final amp built for that), but it would give me a feeling of accomplishment.
Photos so far, note how (IMO) the post filter amp construction is prettier than the pre filter amp as I knew when I built the stage that it was verified to work as designed (cause it's a copy of the prefilter amp), and I could look at the pre filter amp and plan it better. As I've said before, this is a transmitter NO ONE has built before.
My weekends sure go by fast these days.
