Today was the anual Tree Street Yard Sale event

[TheGimp]

Wired the second Edcor transformer up and had to tweak one of the input gain stages to match the outputs at 6Vrms. Then I cranked it up and measured the max output and one channel gives 8.3Vrms and the other 7.2Vrms. Looks like I'm driving the grid positive early on one channel.

I tried swapping tubes between channels with no effect. I finally dug out the tube tester and some tubes are weak. I matched up four out of the eight tubes I have and the results was pretty much the same.

I wired the original transformers back in and got 7.2Vrms on one channel and 3.2Vrms on the other. Looks like a bad transformer so I can either install the Edcors, or scrap it.

I now know why the original design had a balance control, 10% resistors result in really bad channel gain matching.

[TheGimp]

I swapped the original transformer primaries with the feedback disconnected and the problem switched channels. I pulled the transformers and the one that had low output was warm compared to the normal one.

I pulled both and will install the Edcor transformers.

I also bought metal film 2% replacements for every resistor in the amp. I might as well rebuild it completely.

[TheGimp]

I stripped it down and started rebuilding it. TRWs are Polyester and need to be replaced but I'll have to order something to replace them with.

First photo when I first got it running. Second is progress to date. New Filter caps, additional RC stage in B+ filter, twisted heater wiring with NTC resistors and start of adding coupling caps.

I thought about replacing the sockets, but the holes are too large for the replacement ones I have. Then again if I replaced the sockets I'd probably switch output tubes as well.

At least the tubes are reasonably cheap.

[TheGimp]

Finished wiring it tonight.

275V k-a at 32.6mA=8.98W total dissipation per tube, so the plates are running way below max. I may try cranking it up a tad by changing some ps resistors after I run it a while and take some readings.

I've got GNFB but no UL running right now. I may experiment with some Schade if I can figure it out for a PP amp without ITs.

Right now I've got Stevie Ray Vaughan cranking with 'Couldn't Stand the Weather'.

Sounds really nice, better than I expected with the budget Edcore's.

[TheGimp]

I found a major screw up in the original schematic. They have a 470pf cap across the 180K anode resistor. I was wondering why it was there so I looked at the frequency response. Roll off starts below 10KHz and is really pathetic so I took them out.

At 10KHz the 470pF cap only represents 33.86K ohm. I wonder if they were there for a deficiency in the original transformers?

Speaking of the original transformers, I tacked them back in one at a time to once again test them. One is bad for sure. So the Edcor transformers stay in.

With 6dB of gnfb, the lower -3db point is down at 20Hz, and the upper -3dB point is way beyond 22KHz where it begins to sweep up.

I have to go back and measure it with my sine generator and scope as I haven't been able to get beyond 22KHz with AudioTester.

[TheGimp]

I'm still flogging a dead horse.

I ran the power up to 8W today and got a lower -3dB point of 20Hz (although obviously distorted on the scope). Upper -3dB point was beyond the 90KHz limit of the IG-18 I'm using.

I thought these were cheap transformers!

As an experiment I replaced the single cathode resistor (135 Ohm with 20uF in parallel) in the output with two 270 ohm resistors (each with their own 220uF bypass cap). I expected to see distortion go up, and instead it went down.

Then I replaced the two 270 ohm resistors with four 560 ohm resistors and again the distortion went down.

Finally I used clip leads to short the cathode resistors in pairs and finally in quad. Distortion went back up in each case.

The variable amount of capacitance may have impacted the test, however it seems to run counter to what I had read.

All four cathodes tied together with a common resistor and bypass cap should average the voltage since it is summing push pull currents from two channels. This should effectively approached a voltage source on the cathodes which should have produced a more stable operating point and less distortion.

But it didn't.

Why not?

It sounds better in the individual cathode resistor/capacitor configuration as well.

I matched four tubes out of a set of four I got at the ham fest with the four in the amp when I got it. Not a great match, but better than the original set by itself. I have more tubes on the way and will try to get a better match out of all of them.

[Rockmore]

Originally Posted by TheGimp

The "Tree Streets" so named after trees are part of the old town and a historical district. As such they are limited to yard sales once a year, and most houses have a sale on that day. I don't get the paper so I wasn't aware it was today until I drove by around 2pm and saw a sign. I called the wife and said "lets go".

I missed a Sears Console Stereo for $40, but found this for $30.

Webcor 801068-16. Probably 8WPC. 12BA5s in pp with rather small opts. Active tone controls on a separate PCB.

Nice that the all important schematic is included. Interesting of note is the separate feedback winding in the OPTs.

Too bad it does not have an FM stereo tuner in it.

It will need new caps before powering it up, but it looks like it has been kept high and dry although quite dusty inside.

Just wanted to say Gimp- even putting in designer caps on a re-cap job on that old console will do wonders as you know. Did one last winter with all new Nichicon and Elna audios and I could believe my ears. Even using simple Dayton's for the polyesters was well worth it. Pretty fun.

[TheGimp]

I went with all new metal film resistors, Polyester caps for the input gain stage to phase splitter and KY40s from the phase splitter to the output tube grids.

I've got 4 new tubes from the ham fest, 9 from Ebay, and 5 more on their way from ebay. I hope to get at least four complete matched sets of two so that there will be enough for two complete sets of tubes.

It is currently biased at over 9W plate dissipation per tube so I may either increase the cathode resistors or the resistor in series with the screen tap to reduce dissipation below 9W per tube.

Final tweaking will be matching the tubes in sets, as I think everything is as good as it is going to get sound wise.

I just finished measuring frequency response is really impressive for the Edcor transformers specs of 40Hz-18KHz.

Time to take it up stairs and hook it up to the Heresy speakers.

I've got 6dB of gnfb and get this frequency response: