[TV MAN]

Ok, so I tried an experiment tonight. I put a switch on to kill the feedback or keep it in the circuit. Result? FAIL! That's a no-go. I thought it was great for a moment as the gain was tremendous and the highs and mids came alive, but so did the distortion, hum and all kinds of noise. Probably OK for a guitar amp, but on this HiFi amp, noooop. I also added a really cool Johnson Cinch jack to fill in an extra hole over the bias pot that I can use to check the bias voltage right on top of the chassis. So that experiment is a success. Once the current is set, you can check the voltage too. I have the voltage at -26.4 and current is about 30ma idle each tube. It's weird how the tone changes when you adjust this. This seems to be the sweet spot. It runs cool and had plenty of punch. It's really almost to the "leave it alone" point, and start on the next one. It's working really good right now. I almost hate to mess with it any more.

EDIT: The switch I added to turn on or off the NFB might actually not be a total failure. It could be used to help weed out marginal or noisy driver tubes. I note this in that some of the common 6GH8a's are noisy.

   

Sooo - at this point, time to start building it's twin sister. Now that I have the build down, I will try some more experiments as I go together with this one. Gonna try a better step bit and see if I can enlarge a couple tube holes for the octal sockets. I might attempt using a chassis punch, but not holding my breath that it'll work. If not, then I'll grab the die grinder and do a little more "redneck engineering"  
   
   

[TV MAN]

I finally had a chance to do a test power up on my twin amp build, which is a twin, but not identical twin. With this being my second crack at this, I revamped some things in the layout to simplify construction, clean up & be easier to service. I separated the FB network and put it on it's own terminal strip (upper RH corner) which cleaned up the terminal strip near the driver tube. I also rerouted some of the wiring around the driver to simplify. I pretty much left the output area as it was, except I am trying different screen resistors to see the effect on the G2. I used a 100r resistor on the original, tried a 4.7k and it only dropped it 10v. Just an experiment. I'll probably go back to 100 on the final, although I see a lot of those connected directly to B+. I personally feel there should be something there. I also moved the bias filter cap to the control. Plenty of room, and makes more sense to filter it there right before it's sent to the tubes, and neats up the bias terminal strip. I used different coupling caps, since I didn't order enough of the bigger ones. I'll see if it really makes a difference. They weren't cheap ones either, so I really don't expect anything super different. Same value.

You'll also note that I revamped the power supply. My initial was just that - an initial. I found that my choke fit the existing holes already in the rear of the chassis, and it fit perfectly in that area, which neats up the top of the chassis. I moved the filter caps between the PT and A/C plug, mounted to a couple stickys and glued in place, Stays put, yet not hard to snap off and replace when that day comes. I also used a 5w power resistor instead of that huge 10w. I let it play for 20 minutes and could lay my finger on it with out withdrawing forcefully. I'm using a plain ole 6GH8A driver, Japan, and a couple Russian made 6V6s.

It powered up beautifully without any setbacks other than the fact I must have put the line control in backwards. Simple fix. I did reverse the plate leads on the output tranny and had no feedback issues, so this DOES make a difference! It makes sense as this would reverse the phasing of the transformer. I found the blue lead (at least with this tranny) goes to the low side of the inverter, and the other (brown?) to the high side. I've made this mistake more than once on other amp builds, so learned something. I also had the SAME hum issues from the PT, so moved the trans nuts around and used washers and poly washers as before and made it totally quiet. Not quite as attractive as tying the transformer directly to the chassis, but I'll sacrifice some cosmos for performance. I'm still planning to play around with some other tube types and try a couple different output trannys but I need to set a standard for comparison. Next up is going to be a power, freq response and distortion test. So far my main tester (those 2 things on each side of my head) say is sounds pretty good!

Edit: Mike, I know you mentioned that large yellow cap by the driver tube. It has some pretty thick leads and feels solid. I ran the lead through the terminal strip to the socket. It's not going anywhere.


   

[morzh]

Looks good.

Judging by the value of the caps, your rectifier is solid state?
If so, a small NTC to keep the inrush down could come handy.

[TV MAN]

Mike, I have a CL90 I could put on. Honestly, I question if it really needs those big caps. Probably 47s would be fine. It's what came with the kit and they seem to work pretty good. I did reverse my level control. Ugh! Awful! You just TRY an desolder a 6 pin something without destroying it. I ended up cutting one of the traces and soldering direct to the control pins to make this thing work right. I have it set up so you can connect a stereo something or other and have it sound good. You can run it either way. Mono or stereo (mono).

Make SURE you solder that control on the side of that little PC board that has the square, otherwise - welcome to my world.
   

[TV MAN]

OK, so now down to the nitty gritty. Numbers. While I don't have a a lot of experience doing this, I gave it a shot. I got my scope, signal generator and 8R dummy load out and this is what I have as is where is: Take a look. The numbers speak for themselves. I found the sweet spot for the bias is about -26 something. If I crank it up over -30, I can watch the gain drop and it starts clipping, top I think, if I drop below about -20, it starts clipping again, bottom I think. I also found its pretty responsive across the audible band. I can see a drop around about 25-30 hz. The waveform starts getting pretty jagged, but smooths out around 28-30hz. If my calculations are anywhere near non-scientific correct, I'm pulling right around 18w out of this baby! Now, honesty, my scope is not a high dollar scope by any means, so there could be some distortion I can't see.

Not bad for a budget build using parts of a kit, off the shelf parts and some redneck engineering. But I will say those testers on each side of my head say it sounds pretty good so far. More to come.