Improving my SCA 35

My SCA 35 is a good sounding amp from the factory. Great Bass, Good midrange and a little thin on the highs. Detail is its main weakness. I will list every change in order I made to my SCA 35 and the results. I only changed one channel so a fair comparison could be made. After most changes were tested the unit was changed back except where noted.
By passed the selector switch by plugging into the tape out plug. Result, a little cleaner sound minor change.
Changed out the 4 main caps. Improved clarity but not detail. Left the new caps in.
By passed everything up to the board added a ladder style volume only one resistor at any time in the signal path.
Results, This was the big shock, I was expecting a huge change by passing all of those pots, wiring and resistor capacitor networks, not so about the same amount of improvement as the caps. Still no improvement in the detail department.
I am using Sovtec 7199's which squeal at about 600HZ at start up in this amp. Found this article on Audiokarma posted by Flyquail56.
http://www.audionote.co.uk/articles/..._peak_01.shtml
It described the squeal and said routing the positive feedback [the 1mfd cap]to ground would lower output but solve this problem you also needed to replace a resistor. I routed the cap to ground and then replace 2 resistors per instruction matching them to .1%. Results, There is my missing detail about a 6 db. drop in output but well worth it. Also the squeal at startup is gone. Big improvement This change stays.
My next change will be the rest of the resistors and caps.
More to come as I find time to work on it.

Need to make one correction. It was Flyquail56 who stated that the resistor values needed to be changed. He was very right. I know Flyquail56 also answers questions on this board so if you read this Thank you for great advise.
http://www.audiokarma.org/forums/showthread.php?t=118553

The single best improvement I did to my SCA-35 was to do Dave Gillespie's EFB (bias) mod. A few dollars for parts and a bit of work made a very worthwhile difference in the sound of my SCA-35.

If I ever find the time to build my ST-35, I'll add the EFB to that too, but with the ability to bias each tube individually. I picked up a dozen or so 6BQ5 pulls years ago at hamfests, and this would eliminate the matched quad requirement. Additinally, with EFB, the tubes are biased at 27 mA rather than Dynaco's 35 mA, which means they'll last a lot longer.

The EFB™ mod is highly recommended!
http://www.tronola.com/html/a_new_look.html

Below are Gillespie's "Before and After" EFB measured results. My results were judged by my ears.

Hi 1973shovel, I would highly suggest even if you are using the original 7199's that you try the mod anyway. It only requires $1 worth of resistors to match a pair. Physically you need to unsolder 2 resistors and resolder the new ones. Then unsolder the correct end of the 1mfd cap and add small piece of wire to it so it can be solder to the ground hole 6. Do one channel then compare. Unless my unit is one of a kind for some reason, you should hear far more detail than you have ever heard out of your little amp. On my amp the difference was huge.
Thank you for sharing the above mod. I was aware of it but put it on the back burner for 2 reasons. My speakers are 99db/1watt so I seldom use more than 3 or 4 watts of power. Also I didn,t want to do the larger mod if I couldn't improve the detail first. My SCA is a very clean factory wired unit and if I couldn't improve the detail I was going to Audiogon it. And most buyers hate to buy modded equipment. One question however, did I miss something? I only saw where each channel not each tube could be bios?

sailor wrote:Hi 1973shovel, I would highly suggest even if you are using the original 7199's that you try the mod anyway. It only requires $1 worth of resistors to match a pair. Physically you need to unsolder 2 resistors and resolder the new ones. Then unsolder the correct end of the 1mfd cap and add small piece of wire to it so it can be solder to the ground hole 6. Do one channel then compare. Unless my unit is one of a kind for some reason, you should hear far more detail than you have ever heard out of your little amp. On my amp the difference was huge.
Thank you for sharing the above mod. I was aware of it but put it on the back burner for 2 reasons. My speakers are 99db/1watt so I seldom use more than 3 or 4 watts of power. Also I didn't want to do the larger mod if I couldn't improve the detail first. My SCA is a very clean factory wired unit and if I couldn't improve the detail I was going to Audiogon it. And most buyers hate to buy modded equipment. One question however, did I miss something? I only saw where each channel not each tube could be bios?

Hi Sailor (ohh, that doesn't sound good. It sounds like something a streetwalker would say!),

Thanks for the second link to Flyquail56's post. The first link in your first post leads to a dead-end, at least for me.

I will save this and give the mod a try at some point. It's new to me, so thanks.

Two things about the EFB. The first is that I don't really need the extra power the EFB brings either, since my Horn Shoppe horns are 95 dB, with no crossover. The specs I posted would mean very little to me if there wasn't a subjective improvement in the musical quality of the SCA-35. But there is, and I haven't read one posing about someone trying it and not hearing the improvement in sound. Being able to run Gold Lion EL84 's was my goal with trying the EFB. The improvement in sound quality was simply a bonus.

I agree with you regarding mods and resale. Most people would prefer to buy stock units. The EFB requires no drilling of the chassis, cutting of traces, etc. and is easily removed. Your other observation is correct. The EFB takes your SCA-35 (or ST-35) from having no user bias control (cathode bias) to a single pot for the four output tubes, so a matched quad is still required. Gillespie's goal was to "keep it simple" ala Dynaco, is my understanding. Individual bias to each tube can be done, but it increases the complexity.

One thing I found very interesting while doing my SCA-35 homework was finding posts made many years ago regarding converting each tube to individual fixed bias by adding resistors and pots to the cathode of each output tube. One well known Dynaco guy (I "think" it was Gary Kaufman) said when he did this mod that his amp sounded worse, but he wasn't sure why. Gillespie addresses this in his article, and it has to do with the Z565 output transformers. The details are in the article if anyone's interested.

Good luck with your amp, and thanks again for the link (and thanks to Flyquail56 too!).

http://www.audionote.co.uk/articles/tech_amp_peak_01.shtml

Thanks for catching the mistake The one above should be correct. But I will test it after posting. This is where they mention taking the cap to ground. They however forgot to change the 2 resistors to a matched pair. Which Flyquail56 addresses in his post.

http://www.audiokarma.org/forums/showthread.php?t=118553

I agree that the EFB mod is a good idea but first I needed to address the 3 second squeal on start up and what I consider a real lack of detail. Both solved with the above change. Actually I printed a copy of the EFB mod so I will probably buy the parts to do it when I buy the new resistors.
Sorry I don't understand your first comment about a streetwalker? I guess I said something you didn't like. Not sure what it was. Oh well, hope were cool.

sailor wrote:
Sorry I don't understand your first comment about a streetwalker? I guess I said something you didn't like. Not sure what it was. Oh well, hope were cool.

I was just trying to be humorous. "Hey sailor" is a common joke phrase of how a street prostitute would get the attention of a sailor on shore leave.

We're cool, and no offence intended.

Thanks for the added information.

Note that the Sovtek 7199 is not a true 7199 equivalent, and while it will work fairly well as such in many applications, this is not often the case in an SCA-35. Due to the unusual feedback arrangement, the Sovtek 7199 will often exhibit instability, including outright oscillation (within or outside the audible range), oscillation on power up, and/or spurious oscillation that may occur along with the program material, which will muddy the sound. Just because the oscillation on power up seems to go away, does not necessarily mean there is still not spurious oscillation happening along with the program material! Taking the 1uF feedback capacitor to ground will usually cure the oscillation problem, but, as expected, will also lower the gain. A true 7199, which the circuit was designed around, will typically operate in a stable manner and sound good. If you must use Sovtek tubes, then grounding the feedback capacitor, as described, is the easiest to implement fix, although somewhat of a bandaid solution.

Also note that taking the 1uF capacitor to ground will not appreciably affect the "balance" of the phase splitter. The effective resistance on the bottom side of the splitter is comprised of 27K in parallel with the series combination of 120K and the 560 ohm cathode resistor of the pentode section. This results in 22.06K to match the 22K resistor on the plate of the splitter. Removing the 560 ohm resistor from the combination results in 22.04K, an insignificant change.

IMO, Dave Gillespie's EFB modification is, perhaps, the most significant upgrade you can make to your SCA-35. More output power, lower distortion, as well as reduced bias current for longer tube life and cooler running. My own lab tests easily confirmed Dave's published results.

First thanks to everyone who has commented. Please keep the comments coming as this is an ongoing project and I will try to check out what everyone posts. I did know that the Sovtec was not a true 7199 and that was a problem for the SCA. However the big problem as you are aware is the high gain of the SCA which lets you know how noisy almost all original 7199's are and the lack of any new stock of 7199's. That is why you designed your 6GH8 replacement board. But at one time I did have a good quiet set of old 7199's and the sound out of my SCA was about the same. A little more musical but not as much punch in the bass. Still even with the originals there was a lack of detail.
I found your site months ago and read about the EFB then I also studied your new board. I didn't mention it but I added the extra resistor/capacitor network you added on your board to one of my boards to see if it would stop the squeal or improve detail. It did not. I probably should have added the 10k resistor to the el84 also. But I could see no way adding the resistor would improve detail. As I have said, detail was my major goal the squeal at start up was secondary.
I just measured the resistance of the resistor from the tube side to ground point 6. It read the same as the resistor not the parallel value. I agree with you they seem to be in parallel but they don't measure that way. Probably because of the cap. I will alter one channels. I will ad a resistor to the resistor I changed to bring the value back to the original 27k and do a listening test to see what sounds better. But it could be a few days before I can do it.
Please understand I am not knocking your boards. I haven't heard your finished product. But I would like to. And if I decided to keep this amp I will probably at least add your EFB power supply board. But for now I am simply trying to hear better detail from the amp before I make large changes that are hard to undo.

Dave Gillespie's Enhanced Fixed Bias circuit is very clever and the best thing
you can do to an SCA-35.

I mentioned it on one of the "amateur" audio boards, saying it was far and away more
effective than any amount of tube or capacitor swapping. That got some unfriendly
comments from those who think rolling "magic" capacitors is a change in basic circuitry!

Again everyone I accept that the EFB mode is a good change however it requires 10 resistors to be added to the boards, a resistor removed from the power supply and even more parts added. I think if you are for sure you are going to keep your SCA by all means make the EFB mode. I'm going to, but only after I conclude that my SCA sounds good enough to keep. I am at present trying to find a simple way to get better detail out of my SCA. Having said that the below change has made the greatest difference in my SCA and requires so very little change to make happen. Now back to the Audionote mod.
I've been unsoldering and soldering and listening all afternoon. To check and recheck Audioregensis comments. I tried every possible combination. Here are my findings.
The original factory hookup does not sound as good as my original change which was the Cap to ground and matched resistors. So I changed both channels back to my original modified wiring to make sure they still sounded the same. They did, then I changed one channel back to the original 27K resistor. After listening I concluded that Audioregensis is correct, the original resistor values do sound better and I will now assume are correct.
So, after all the testing the best sounding by far was Cap. to ground with original resistor values.
All that is required for the mod is to unsolder the end of the 1MFD capacitor hooked up to the 560 ohm resistor add a 1.5 inch piece of wire, cover the bare wire with shrink tubing or tape to protect from shorting and solder the other end of the wire to the wires coming out of hole 6. I would suggest doing only one channel so you can AB the the change with the original. I think this very simple change makes sense. Why would you add positive feedback except to increase gain [I think in this case about +6 db]. I can't think of one way positive feedback would improve sound quality.

sailor wrote: by all means make the EFB mode. I'm going to, but only after I conclude that my SCA sounds good enough to keep.

Your logic in this matter seems to elude me. GP49 and scores of others on other forums (including me) have posted that the EFB is the most significant sonic upgrade you can easily make to your SCA-35.

Yet you aren't going to add the EFB until you decide whether or not your SCA-35 sounds good enough to keep? How will you know if it sounds "good enough" if you don't do the mod that is generally considered to add the most musical improvement to the amp?

Let’s say I were trying to tweak the sound of an old, stock ST-70. And everybody here and on other forums told me that adding one of Roy's VTA input boards was the best thing I could do for the sound of the old amp. Would it makes sense for me to say that I want to improve the sound of the stock ST-70 first, prior to deciding if adding the VTA input stage is worth it?

I'm not trying to be critical, but simply trying to understand where you're coming from. The EFB is $10 worth of parts, and a few hours of soldering. Unless you have a museum quality SCA-35, I would believe, given its reputation, having an EFB mod installed would enhance resale value, not hurt it.

My 2¢

Note that the Audioregenesis PC-10A board was not created due to issues with the original design, rather the difficulty and expense involved with obtaining good NOS or used 7199's. The goal of the PC-10A was to create a board that did not degrade the performance, while being able to use much less expensive, and plentiful, 6GH8 or 6U8 tubes. While more stable than the Sovtek 7199 (in the SCA-35 application), there were some examples among the many tubes tested during development that did display some stability problems. An additional R/C phase shift network was added, along with the 10K resistor you mentioned, to ensure the circuit would be stable with all 6GH8 and 6U8 type tubes. I tested many, many different tubes. It is important to note that the these additional modifications are applicable specifically to the 6GH8 and 6U8 tubes, and I would not be surprised that they have little or no affect in your case. In fact, it is even possible they could degrade the performance with the Sovtek tubes.

As to the 27K resistor in the phase splitter, effective "impedance" would be the more correct term to use when describing the result of the parallel combination of components. You would not be able to measure that with an ohmmeter. As far as the actual audio signal is concerned, the 1uF capacitor can be considered a short circuit, and thus the 22K result.

Now, as to the positive feedback, in the case of the SCA-35, Dynaco had to squeeze all the gain they could from the PC-10 to make up for the loss of the passive tone controls which directly precede it, since there is no preamp associated with the high level inputs, only the low level inputs (PC-11). The gain is required to achieve typical expected input sensitivity for the high level inputs. It was a careful balancing act to achieve the required gain while maintaining stability as well as low distortion, all achieved with circuitry specifically designed around the (real) 7199. Amazing they pulled it off so well.

I should also mention that the 7199 was intended for high quality audio applications and originally exhibited a suitably low noise floor. It is my understanding that as production of these tubes was nearing it's end there were some occasional issues with quality, higher noise being one of them.

With the Sovtek tubes, and the 1uF cap. to ground, you will likely not have stability issues, and it could well be that you personally prefer the sound of the Sovtek's. And really, that's the most important thing. It has to sound good to YOU.

Sorry if I have ruffled some feathers but it did get a lot of good and usable information out to everyone. Thank you.
First 1973shovel You hit the nail on the head when you said museum quality. If I find a picture I will try to post it. My SCA is a factory wired, probably 1964, very clean unit with all of the paperwork and original box. Other than the usual finger prints on the back plate and a few period size dots of black in the chassis the rest is super clean. I won't call it museum quality but it's the closest I have ever seen. The seller who sold it to me did call it museum quality. Being factory wired, mounting the regulator is the major problem [rivets not bolts]I would need to drill out a rivet to mount it as described. Not reversible. If it was an average unit I would have had it in pieces and parts would have been a flying. I would have added a choke power supply and a regulated dc filament supply for the driver stage, replaced the selector switch and volume control and removed all of the balance and tone controls. But this unit is so clean I feel obligated to control my urges to redesign it. As far as soldering and modifying I modified my first Dynaco PAS2 in 1974, a PAS3 in 75, Mark3's in 1978 and about 40 other units over the years. I have been soldering for 45 years so I am a very experience at soldering and can do it without damaging the old Dynaco boards.
Audioregensis you are right on everything you have stated. Of coarse most of it I already knew such as no preamp and huge gain almost unstable driver stage.
As far as measuring the 27K resistor I figured that out after I wrote my last posting. I was tired and in a hurry to get the post up and made several mistakes. One was about positive feedback. What I should have said was in the case of the SCA, Dynaco added the positive feedback to get an extra 6db of gain which was a big compromise. Because of the positive feedback and the small amount of negative feedback it is barely stable even with the 7199. That is a known fact. If you don't need that 6db of gain, which I don't, your SCA will sound a lot better and be more stable without it.

[img][/img]
[img][/img]

Well I have lived with the change and I am happy to say I have a lot more detail and better highs than before. Wife agrees, and her opinions can be cold and are very unbiased.
I have done some research to see if there was any other in inadvertent changes caused by my moving the 1 MFD cap to ground. If I am not mistaken originally the cap is also used to short the screen to the cathode along with the positive feedback. This shorting is required to make a pentode work like a pentode. With the cap going to ground I think I have turned the 7199 pentode into a triode. But it still sounds great and my overall loss in gain seems to be about 6 db. I've always heard pentodes make great triodes.
The next step will be the EFB but for now I will need to stop on this project until I find some extra time.

Does anyone know how to fix the rotary selector switch on the SCA-35.
Mine does not line up correctly and I have to move it one way or the other to hit the right spot so both channels work. Especially in the phono selection. Also, in other modes, there is always a hum.
When i got the unit it was greasy and very dirty. I have re-wired the entire unit and changed resistors and caps. I got most of my help and followed his wiring from Bill Thomas at Audio Circle. I cleaned the rotory sw. as best as I could.

Any ideas, help would be appreciated

It sounds as if you want to keep it original. It's probably wore out and needs replacing. There is a used one on ebay for way to much 29.95 plus shipping. If you don't care if it stays original, by doing away with obsolete connections and moving the resistor/capacitor network to the board, plus a few other changes, you can use a modern selector switch. It will also improve the level of noise.