SCA 35 Can you replace preamp section with vta preamp section?

i just got a description. i'm looking at schematic now. Basically lifting wire # 7 of PECO from wiper of volume control and lifting wire from connection of wipers of bass and treble. So direct connection is wiper of volume to to non grounded end of balance. Completlely isolated from bass treble pots and peco. Look that over against schematic and tell if it makes sense to you.

That would be correct if you are leaving in the balance. Only the signal wire gets bypassed all grounds stay the same. I'll try to do the switch again. The middle row [always connected] of the switch will have the wires from the volume and balance. The left row would the wires that are attached to the PICO and the wire exiting the bass treble originally attached to the balance control. The right row has a 1 jumper wire for each channel so when on it connects the left volume to the left balance. And the same for the right. So the left row reconnects the tone controls the center row. The right row is the jumper wires that connect the volume to the balance. I bypassed one side [no switch]. If your volume is presently at one o'clock it will be moved back to the 9 o'clock position for the same output. The biggest difference in sound quality was in the upper mids and highs.

Over the weekend I tested my modified channel [see earlier post for resistor and circuit changes]plus tone and balance were bypassed. Then I tested the original channel. I modified one channel and kept one original so I could AB test them. The modified channel tracked perfectly from 20 to 20,000 HZ. I measured every 1/3 octive with a 2 volt output. The original channel was a different story. It was boosted at the 20 hz and trailed slowly off until above 12,500 HZ then it increased back up to 20,000 HZ. Choosing 100 HZ as 100% Going to 20HZ swung up by almost 15% and a low point around 2,500 where it was down about 16%. Think of it as a wave high at 20HZ slowly trailing off to 2500 HZ where it slow started to rise again to 20,000 HZ. In terms of db I guess that would only be about plus or minus 1 db. of swing. But when testing it still seems like a lot. I am going to bypass the tone and balance on the original channel to see if the swing is the controls or the amp itself. As far as sound goes the modified side sounds a lot better. More testing and modifications to come.

This is great stuff, Keep up the good work and the posts. If your still singing the praises at the end. Maybe i'll follow you. Keep posting . Thanks

Ok, I bypassed the tone and balance on the all original channel [see post above]. In the testing it made a big difference, it was basically flat. To my surprise, the difference in volume between the 2 channels is about 3db.
So the 2 channels test alike so they sound alike, right? Nope, not even close. With the tone and balance controls removed both sound better as well as a lot less input needed to drive them. But, what I said about the two channels when I described my modification [removal of the positive feedback] is still true. Before the mod the sound is like being in the 25 row, soft, lack of detail, and direction. With the mod. you are in the first row. You hear the detail, the sound is quick, and you know where the sound is coming from. It sounds very close to my best amp and that is saying a lot.
Why the difference. Easy, the positive feedback was added to boost the output at the cost of sound quality. Second, although the output at the 2 sides of the phase splitter are equal. Part of one side is used to power the positive feedback so the phase splitter is actually out of balance going to the output tubes. By doing away with the positive feedback and replacing the resistors on each side of the phase splitter with a matched set of resistors. The sound is cleaner, tighter and more balanced. The only other thing I did was power the pentode screen from the B+, it was originally powered by the DC from the positive feedback resistor. The 1uf cap needs to stay but could be reduced in size. Again see my post a page or 2 back for more details. More changes to come.

Hey Dave , would it be possible for you to post a before and after schematic? Just a hand sketch, of the portion of the circuit you changed? If its a real problem, please email me. I was anybody else interested could see it. But i'd really like to see what you're doing. Thanks

There will be more changes so unless you have the parts on hand or a local source you might want to wait for all of the changes.
When done, I will try to take pictures and redraw the circuit. Assuming you have a copy of the manual, the changes so far are very simple. Look at the Schematic Diagram. First replace R32 and R34 with a matched pair of 22K, 1 Watt resistors. Remove R33 the 120K resistor do not replace with anything. On the schematic find eyelet 5, this is where the B+ supply comes in. Because there is no extra holes you must attach the new resistor a 1.5 Meg. to either the R30 or R32 resistor on the B+ or eyelet 5 side of either resistor then the other end to the now open hole where R33 use to be on the side that attaches to capacitor C18. Stand your amp on end so you can see both side of the board and you will figure it out.
If I drew a changed schematic the only changes would be to remove R33 and add a new resistor from top line coming from eyelet 5 to the now open end of capacitor 18. That's it.
Future changes may include, increasing the neg. feedback, adding a cap to R37, adding a resistor to one of the output tubes on each channel to help with the unbalanced impedance of the splitter, changing the diodes, increasing the capacity of the power supply and adding the EFB circuit. Maybe more.

I have an updated diode cap board already and the EFB done point to point. Dave Gillespie is now marketing the AR boards (which were I think, caps diodes and EFB in one that mount where caps were). I'm sorry it wasn't available when i did mine. I was ordering stuff from element 14 and mouser for another project. I'll add this stuff to it. If you can't find info on those boards get back to me and i'll get you info.

You were talking eyelet 5 on amp boards yes?

ok i got it i looked at schematic. That's the only change? What exactly are we accomplishing by this? Are there any stability issues?

With the change the amp is a lot more stable and should be able to run the 7199 substitutes without problems.
Before I explain what I did let me tell you what is happening in the driver stage. Follow along on the schematic. This driver has 2 stages The first stage, the pentode part of the tube produces all of the gain of the driver stage. It is then DC coupled to the second stage, the triode which is used as a phase splitter. It actually losses about 9% of the gain so the output at the plate and cathode each are about 45% of the pentodes output and are out of phase with each other. The out of phase is how you get a push pull amp. For the signals at each end of the splitter to be equal the resistors must be equal. If they are not the signal will be weaker in either the push or pull side of the amp. What I have just described is textbook and is how I changed the amp to be. The 1.5 meg. resistor I will explain at the bottom.
What Dynaco did was change the cathode resistor on the phase splitter from one 22k resistor to a 27K and added an the 120k which equals the 22k as they are in parallel with each other. The 120k resistor has both ac and dc flowing through it. The DC is used to power the screen on the pentode pin 3 [very important] and the ac is used as positive feedback to the pentode cathode to increase the overall gain of the pentode. The Cap c 18 blocks the dc from reaching the cathode. The positive feedback was added to give the driver stage a little more gain to overcome the tone control and not to improve sound quality unfortunately it also makes the amp on the verge of unstable.
That takes care of everything but the resistor I added. It had to be added to replace the dc to pin 3 that was originally provided through the 120k resistor. The 1.5 meg. resistor replaces the needed dc by bringing it direct from the B+ a much high voltage source so a much higher resistance value. This screen also needs c18 to couple it to the cathode. If you remove c18 you would turn the pentode into a triode and loose a lot of gain. However the cap no longer needs to be 1uf it could be reduced to.1uf because it is no longer handling ac from the feedback resistor.
I did not describe everything that is happening but it's most of it. You will probably need to read and follow along a few times before it becomes clear.

Hey Dave, I was just curious, I had been discussing your mods with i guy i consider billiant when it comes to this stuff. He seem to feel your mod is to overcome the differences in the russian 7199 as opposed to the american verion. He thought a more interesring test would the modified channel with the russian tube versus the unmodified channel with a good quality american tube. He didn't say what you were doing was wrong, just that your not comparing apples to apples.

Does he question that removing the positive feedback makes the amp more stable? It's true that the other tubes will work a lot better with my circuit change. Mainly because it lowers the gain a little also the triode splitter has a more balanced output to the power tubes, less distortion. If you leave in the tone controls you will probably need the extra gain.
Rays circuit would be the best but the little amp simply doesn't have the room for it.
Dynaco knew how to design much better driver stages than they did but to improve them would have increased the cost of there kits. After all in the early 60's a Stereo 70 sold for $100 a Mac MC275 sold for $444.

Oh, one thing I forgot a $1 in the 60's is worth $10 today. So in today money Dynaco ST70 would cost $1000 the MC275 would cost $4440. I think it's present price is slightly more than that. But one thing is for sure this comparison does show how reasonable Bob's kits are. Better parts and driver stage for less than what Dynaco would sell the kit for base on dollar value today.

Well, got back to playing with the SCA35. If anyone other than I have made the changes I suggested in previous posts or if you have at least bypass the tone controls you should increase the neg. feedback. This will increase the tightness and authority of the bass plus increase stability. The present resistor is a 82K to make it the same amount of neg. feedback as is used in the Stereo 70 you would change it to 12K resistor. I tried it at 16K and it was a lot tighter bass. Also with the 16K resistor you will still have about 3db more gain than the boards using the 12DW7. Don't want to remove the resistor until you try it out? Tack a 20K resistor to the 82K resistor in one channel and switch from one channel to the other and adjust the volume to make the output equal to compare. 82k in parallel with a 20K is 16K.
To the novices the lower the feedback resistor's value the greater the feedback.
Dynaco used such a low amount of feedback in the SCA35 to preserve as much gain as they could at the cost of sound quality and stability. Without the tone controls this is easily corrected.

going all the way back to post #6 on this thread, I have essentially copied the ST35 boards onto a board to fit the SCA35 (the mounting holes are different)
and these are now available from me.  IMHO, better fidelity, but yes, you lose the tone controls.   I hope to put up a new web page of SCA35 goodies this weekend,
including a new I-O board (6 pair of inputs-outputs with wider spacing) and a couple of new phono boards (SCA35 sized versions of my PH6 and PH10).

SCA35 mods

WOW Roy, you got a lot of new SCA products up in a short period of time. Other than your own case and your new phono board you have all you need to sell a complete Amp. Congratulations!

Update on neg. feedback that I mention a couple of posts back. I have very little test equipment so I can only do trial and error and no way to optimize each part so. I found this post and I see no reason not to use it with the standard board [7199 or 6gh8A] if the tone controls have been bypassed. He uses a 30.1K resistor in place of the 82K to increase the neg. feedback and adds a 150PF in parallel for reasons which are stated in the paper. His findings are the same as mine. Tighter bass and more stable circuit. Dynaco knew this but to increase the neg. feedback would have lowered the gain below there target level.
http://www.classicvalve.ca/docs/SCA-35_PUD-0.pdf

The other day I was trying to find some info on the 7199 when I stumbled onto this article written by RCA. There are a lot of things in the article I want to comment on but I will need to wait until tomorrow. There is one thing in the article you need to click on it's under heading [Applications] marked [Fig. 1] which shows the 7199 being used without positive feedback. Basically a similar design to the one I proposed. They claim it to be very linear and stable. Ignore the output stage as it uses a different tube.
I fine the Fig. 4 chart very interesting as Dynaco uses a 270K pentode plate resistor RCA recommends a 220K read why. More to come.
http://www.triodeel.com/7199.htm