Truetone console pull

[wlgrant6]

And a happy Friday to everyone. I came across an old Western Auto Truetone tube console, that wasn't in the best condition, cosmetically, so decided to see what I could use out of it. The tuner, preamp, and power amp were all on one chassis, so all l could salvage was the transformers and tubes. The amp portion of it was a stereo single ended 7868) (never seen a single ended 7868 before) and the output transformers looked pretty good. 12AX7's and 7868's tested good. So I put together the amp in the pics. I copied the amp and power supply from the original schematic, adjusting the power supply levels to compensate for the tuner and preamp which wasn't being used. The only problem that I have, is that there is too much bass. I can compensate for it by turning the bass way down on the preamp, but would like to correct it. Everything is as shown in the schematic except have a .01 cap, and 470K grid resistor in front of the 12AX7. Any one got any suggestions how I can reduce the base, and possibly boost the high end a little? Bill

 

[gadget73]

SE 7868 is indeed an oddball.

Thats an awful big capacitor across the feedback resistor. It will cause the highs to be cut considerably relative to the lows. Maybe remove it or at least make it a fair bit smaller and see if that doesn't help things out.

tuning it on a scope using square waves would be even better if you have that capability, but failing that try maybe 0.001uf and just see what you get.

 

[wlgrant6]

Thanks, gadget. I'll give that a try first chance I get. What is the procedure for using a scope and a square wave generator?

 

[gadget73]

Set the generator for a square wave at 10khz and set the level so the amplifier outputs about 2 volts peak to peak into a dummy load. Output level isn't super critical, it just needs to be well below the power output limits of the amp.

mess with that cap until you get as close to a square wave as possible. A heavy rounding on the rising edge indicates HF roll-off. Overshoot is too much HF. Wobbles on the top of the square indicate supersonic ringing. I suspect you'll have a heavy round-over on the leading edge from a too-big capacitor. Its probably never going to be perfect but if you get a reasonably sharp corner without a huge amount of overshoot or ringing on top of the wave it'll probably sound reasonable and not be unstable.

That big of a cap I'm guessing was an intentional choice to boost bass and cut highs, maybe to work with the original tone controls or something.

 

[wlgrant6]

I'll try to get set up tomorrow, and give that a try. I'll report back later. Thanks.

 

[wlgrant6]

Ok, here are the results. I have a Sencore Substituter, so it is a simple matter to dial it in. Values in order .05, .02, .001, .0005, and .0001. To my untrained eye, the 4th pic, .0005 looks the best. What do you think?

 

[TriodeLuvr]

None of that is acceptable for a 10kHz square wave. The last three traces are better, but they indicate a lot of phase shift that shouldn't be present at this frequency. How much feedback is being applied (compare the output amplitude difference with and without the feedback loop connected)? Are you using a resistive load?

Jack

 

[Lavane]

It's hard to tell the difference between #3 and #4 from the picture really.

 

[TriodeLuvr]

It's hard to tell the difference between #3 and #4 from the picture really.

That doesn't look like 10kHz to me.

Jack

 

[gadget73]

Is that 10khz or 10hz? The sloped top usually indicates limited low frequency response. That sort of thing shows up when testing below 1khz.

and yeah that first image is really bad, but kind of what I expected to see in terms of rounding.

 

[wlgrant6]

My bad. I went back and rechecked the oscillator, that was 1KC. I had some trouble with the oscillator, and it ended up on 1KC, instead of 10KC. I'll redo the test later

 

[wlgrant6]

OK, same sequence of caps.

 

[wlgrant6]

25 watt, 8 ohm resistors on speaker terminals.

 

[TriodeLuvr]

I don't understand why the first and last traces look so similar, despite being at the two extremes of capacitance values. And the first trace should show the most rounded leading edge, because it uses the most capacitance. Again, you should try to measure the amount of feedback being employed.

 

[wlgrant6]

I noticed that too. I measured the actual values of the capacitors in the substitutor box, and in spite of being about 50 years old, they were pretty close.

 

[wlgrant6]

I'm tempted to order some .0001 caps, and put them in and see if my ears can tell the difference.

 

[gadget73]

picture 9 looks the most reasonable to me. Picture 8 is what I'd expect from the cap being too big. I would expect a pretty audible difference between those two responses.

 

[wlgrant6]

oK, .0005 it is. Never got into feedback much. Good learning experience for me. Thanks to everyone for their help.

 

[wlgrant6]

Got the .0005's in, and put them in, made quite a difference. Calmed the bass down, and brought the highs up where they should be. Sounds nice now. Didn't know what to expect out of a single ended 7868 amp, but sounds nice. I really appreciate everyone's help on this.

Out of curiosity, how much effect do the coupling capacitors have on the overall freq response?

 

[gadget73]

Not vast amounts inside a feedback loop, up to the point that the cap is so small that there isn't enough feedback to overcome the loss in the cap.

This is somewhat odd values though. A 47k grid resistor is a lot smaller than I'd expect wirh a 12ax7 driver. 0.01 into 47k is a -3db point of 338 hz which is fairly high. If you feel like experimenting maybe make the resistor larger instead of the cap though. That will move the rolloff down but also makes life easier on the 12ax7. They aren't a high current tube and it might struggle a bit to deliver clean signal into that load.