Improved SCA-35/ST-35 Performance

for the DIY ST35, the Dynakit and every other PP EL84

Postby Carlp » Mon Feb 13, 2012 6:53 pm

Hello folks,

This discussion has been really fun to follow along with and try to understand. I had a SCA-35 back around 1977, for a short time. Since I also had the PAS-3/ST-70 combo, I gave the 35 to a friend, who was poor and needed some tunes. Next time I saw him and asked how the amp was working, he said it had developed a problem, so he threw it out. Threw it out! With what seems like a revival in interest in EL-84 amps, I’ve been looking for a Dynaco, but haven’t stumbled on another one in all these years.

But, since all this talk has been so exciting, I stopped waiting and ordered a diy kit from Triode. Figure on building it with the EFB mod ground-up. To keep it simple for my simple mind, the one-pot bias adjustment will do fine. So there will be a lot of leftover kit parts. Hope to find the mod parts in town this week before the kit arrives.

One thing I want to double-check is, in this one adjustment scenario, what should I do with the existing 470k grid resistors (R13-16). Should they be installed as directed, pairs joined to ground? Like Geezer said, I’m a paint by numbers technician.

Thanks in advance.

Carl
Carlp
 
Posts: 15
Joined: Mon Feb 13, 2012 5:19 pm
Location: Rochester, NY

Postby dcgillespie » Tue Feb 14, 2012 7:14 am

Hi Carl -- Thanks for your interest in the EFB modification.

With the single adjustment installation of EFB, the original 470K output tube grid return resistors are not disturbed from their original connections. They should be matched pairs in each channel, and ultimately connected to ground, as per the original circuit. It is only when the individual tube/channel adjustment options for the EFB modification are installed that these resistors employ different connections.

Don't forget to install the grid stopper/screen stability resistors when you upgrade to the EFB modification. These are required components when the modification is made to ensure proper operation.

I hope that clarifies the connections for you. Good luck with your project!

Dave
dcgillespie
KT88
 
Posts: 399
Joined: Sun Mar 08, 2009 5:34 am
Location: Ball Ground, GA

Postby SpyderMike » Sun Feb 19, 2012 8:05 pm

Today I finished my EFB mod of a ST-35 kit build. I had finished the kit build a little over aweek ago and enjoyed the sound. I played it for about 40 hours before deciding to get the EFB mod done.

I have one question regarding the results of my EFB mod. I turned the amp on and let it warm up for about 10 minutes and then adjusted the bias while cross checking voltages from the regulator output pin to tube pins 3 on each side. The bias setting on one side was adjusted to .285 and that brought the other side only to .258 VDC. I was wondering if that was an acceptable variance from side to side.

Thanks,

Mike
SpyderMike
 
Posts: 7
Joined: Fri Jan 20, 2012 3:10 pm

Postby dcgillespie » Mon Feb 20, 2012 8:07 am

Hi Mike -- congrats on installing the EFB modification! I'll be interested to know your impression of its impact on your amp's performance.

Your current draw numbers are certainly acceptable, with the channel balance being just within a 10% variance. I've always advised folks that at long as you are within 1 ma of the 27 ma target (either way), you will be hard pressed to hear any sonic differences within this range. You are slightly outside of this definition, but so close that for all practical purposes, I hardly think it's worth buying new tubes over, unless you just want to. However, the closer the tubes all match, the better each channel will perform, and the better the channels will match each other in performance.

For what its worth, a tight match would have all of the tubes draw within 5% of each other, so you can use that as a guide for future decisions when matching your tubes.

Good luck with your amp!

Dave
dcgillespie
KT88
 
Posts: 399
Joined: Sun Mar 08, 2009 5:34 am
Location: Ball Ground, GA

Postby SpyderMike » Mon Feb 20, 2012 11:15 am

Thank you for your reply Dave. These are brand new matched tubes. Maybe as they burn in things will get closer? As I mentioned, they only have about 40 hours on them at most. I will check the bias after some more time...might even set up the independent EFB biasing scheme for the two sides.


Mike
SpyderMike
 
Posts: 7
Joined: Fri Jan 20, 2012 3:10 pm

Postby Carlp » Sat Feb 25, 2012 12:04 pm

It took a bit more time than I'd hoped to get all the parts together, but once I did, this was a fairly easy thing to assemble on the Triode/DIYtube kit platform. Much simpler with only one bias adjustment.

A Radio Shack 47x72mm component board (276-149) happened to mount diagonally on the output transformer screws, so with an extra nut as a spacer, most of the EFB parts are contained there. I mounted the LM 337 to the chassis at an open hole, where a choke could be mounted, right next to the project board. That spot on the chassis gets very hot, so I hope it is enough heat sink where it's necessary. I installed a 22mf 50v cap in place of the non-existing 100mf cathode cap.

For the amp board mods, I avoided cutting G2 traces by mounting only one end of C1-4 and the 470k resistors, leaving the ends which tie to the added 1k parts free. If you mount one end of the 1k Rs in the unused/unconnected holes of R18, 21, etc, you can tie those loose leads at that point. Then, the other ends of the 1ks will reach over to the solder pads for pin 2.

For the cathode leads, I did the same thing as Marshman (thread page 4), who bridged the + pads of C5,6 and C7,8, and flew the 5Rs and a wire over the board. This leaves a lot of stuff in mid-air, but gravity is pulling in the right direction.

I didn't use the terminal blocks, preferring to solder wires. On the G1 grids, I soldered the 100Rs to the terminal pads, and the UL tap wires to the Rs, again similar to Marshman's approach.

Not all of the readings in the Initial Impedance chart were confirmed, but don't panic unless you've got shorts to ground where they shouldn't be. Bias fell well within range, and matched pretty closely between channels. That was with brand new JJs. I can put my ear to the speakers and hardly hear any noises or hum.

The amp started out sounding very 'treble-y' but after four hours or so it's sounding warmer. I'm pleasantly surprised by the power; even on reggae and organ music the bass is pretty firm and makes my small speakers go pretty deep. Once I'm sure it won't hurt anything, I'll try it in my better system.

Do people find that it takes a while for the tubes to sound right, or the whole amp to mature? When I replaced the power transformers in my MK IIIs recently, they sounded kinda weird for a little while.

It wasn't clear to me last night at 9PM how to implement the choke. I guess it would just replace R34, 35? Should I do that?

The PT is very hot - like those in my Mk IIIs. Is that normal?

Speaking of Mk IIIs, I got the Poseidon boards today. Parts should be here Tuesday. For how long can I resist jumping into that project?

Sincere thanks to everybody for the information/inspiration, and to Shannon and Dave for the great engineering and support.
Carlp
 
Posts: 15
Joined: Mon Feb 13, 2012 5:19 pm
Location: Rochester, NY

Postby dcgillespie » Sun Mar 04, 2012 8:12 am

Carl -- Congrats on successfully installing EFB in your diy ST-35! And, thanks for adding your sonic impressions as well.

I think sometimes when we make significant changes to our equipment, we can initially become overly sensitive in our hearing to the point of questioning everything we hear, and then as the sound settles into a new normal, we assume that the equipment changes with burn-in time. For me, I have found that it was not so much the equipment at all, but my acoustic memory system becoming reacclimated again. However, I am very aware that we all hear differently, and that may not be the case with you. In any event, I'm glad your installation was successful, and I think over time you will realize the benefits that the EFB modification produces.

Within your comments -- if I may -- realize that "G1" or Grid #1 always refers to the control grid; that is, the grid that is traditionally driven from the previous stage with the audio signal to be amplified. G2, or grid #2 then refers to the screen grid, which is connected to the output transformer UL taps through 100 ohm resistors when the EFB modification is installed. The grids are so numbered in this fashion due to the development of the vacuum tube itself. The original grid installed was simply called "the grid". However, with the development of the screen grid (which came later), new terminology was need to distinguish between the two grids. Therefore, the original grid became gird #1, or the Control grid (since it was the first grid, and primarily controls current flow through the tube), while the new grid became the Screen grid, or grid #2 (because it came later and "screens" the control grid from the plate).

I mention this only because there are so many new folks coming into this hobby, and it is important for them to gain a clear understanding of just what the terminology means. In your text, it appears that you are referring to grid G1 (for example), but describing it with grid G2 connections, and visa-versa, so that was the reason for my comments.

Your connections are all described perfectly however, so I have no doubt you did a great job with your work. Thank-you for taking the time to detail your installation, as it will no doubt be a tremendous help others as they work to install the modification in their own amps.

Thanks again for your comments, and congrats on a job well done!

Dave
dcgillespie
KT88
 
Posts: 399
Joined: Sun Mar 08, 2009 5:34 am
Location: Ball Ground, GA

Postby Carlp » Sun Mar 04, 2012 8:23 pm

Thanks Dave. Judgement one week later: worth every penny.
Carlp
 
Posts: 15
Joined: Mon Feb 13, 2012 5:19 pm
Location: Rochester, NY

Postby rmyauck » Sun Mar 18, 2012 9:24 pm

H i Dave,

Could it be implemented on a Pilot SA-232 EL84 paraphase amp? . I see on the Scott 299 with paraphase it is part of the fixed bias though it has AC adjustment.

No Rush!

I suggested it for someones amp to help with hard run output tubes.

Thanks very much in advance!

Regard!

Randy
rmyauck
 
Posts: 89
Joined: Thu Dec 09, 2010 5:13 pm
Location: Canada

Postby dcgillespie » Mon Mar 19, 2012 12:43 pm

Randy -- EFB can certainly be implemented in the SA-232, and would be the easiest way to add fixed bias, since no separate winding is available on the power transformer used.

I plan to investigate what bias setting and operating parameters would offer the best performance when typical pentode designs (such as the SA-232 represents) are converted to fixed bias operation. If an optimum condition can be found that allows for improved performance and reduced quiescent current, then the modification really becomes a win-win result, much like that with the Dynaco EFB installation.

Dave
dcgillespie
KT88
 
Posts: 399
Joined: Sun Mar 08, 2009 5:34 am
Location: Ball Ground, GA

Postby rmyauck » Mon Mar 19, 2012 2:59 pm

Hi Dave,

Thanks! I found a really good copy of the manual on the Pilot yahoo site and in the parts list the PS trans has a 23 VCT tap @ .62A. This isn't shown on the schem. He doesn't have much room so maybe that would help as he confirmed that his trans has that tap! The parts list shows the Output Trans are 6K primary, so ideal for fixed bias!

Thanks very much again Dave!

Regards,

Randy
rmyauck
 
Posts: 89
Joined: Thu Dec 09, 2010 5:13 pm
Location: Canada

Mileage report

Postby Carlp » Fri Sep 07, 2012 5:17 pm

Six months on, love it, love it, love it! There's nothing fancy about the build, just the parts Triode supplied, plus the EFB mod. JJ tube-set. I'm knocked out by the general quality of the sound, but really surprised by how powerful it manages to be.

At first, I figured it was a fortunate accident, that my smallish room and 87dB tower speakers were a good match. Changing over to less efficient monitor speakers hasn't effected that sense that there's more than 17 watts on hand. A little compression sets in when it has gotten too loud anyway, but within its envelope the solidity and depth of the bass is beyond expectation.

To anyone who is on the fence about trying this amp, I don't think you should hesitate. It has tons of magic and musicality, plus drive, transparency, precision, and low distortion. For a tube amp, it's an easy keeper. It's been hot in Rochester this summer, too hot to run my Poseidon Mk. IIIs. But, with the ST-35 on hand, I really haven't missed them.
Carlp
 
Posts: 15
Joined: Mon Feb 13, 2012 5:19 pm
Location: Rochester, NY

Postby tomlang » Tue Oct 16, 2012 8:44 am

Earlier in this thread, Dave mentioned a version of this mod for the Dynaco ST70 and other amps that already have a fixed bias supply. Has there been any progress on this?

Is it too much simplification to state that regulating the negative bias supply voltage is all that is required?

Also, is it fair to say if you created a negative bias supply for the stereo 35 (and regulated it), you would increase the headroom of the amp further by not "using up" positive supply voltage to bias the output tubes?
User avatar
tomlang
 
Posts: 192
Joined: Sat Sep 26, 2009 8:14 pm
Location: Augusta, GA

Postby dcgillespie » Thu Oct 18, 2012 5:05 pm

Tomlang -- Thanks for your interest and quires. They are in fact quite timely.

First -- as to your questions:

1. If you are looking to install EFB in an amplifier that already operates with conventional fixed bias, then in fact you would use an EFB supply regulator, versus an EFB cathode regulator. I mentioned this only briefly in my original article introducing EFB.

Since an amplifier with fixed bias already operates (obviously) with fixed bias, then only the unique "Enhanced" element needs to be added to achieve EFB operation. It is this element that reduces distortion so significantly in spite of significant changes in B+ level, be it either due to changing AC line conditions, or changing dynamic (power output) conditions. This too is an equally simple circuit to install, but in releasing it, I got somewhat distracted along the way, which I will explain further in a moment.

To finish the possible existing scenarios then, the cathode regulator version already released not only provides true fixed bias operation, but also of course produces the Enhance element as well. It is appropriate for traditionally cathode biased amplifiers where conversion to fixed bias is desired, or for amplifiers designed with fixed bias, but where chewing up some of today's higher B+ voltages might be of benefit.

In getting ready to release the supply regulator version of EFB however, I thought it best to back up and also address the numerous requests I've had for installing EFB in pentode output stages as well. To be done properly, this of course has added yet another version of EFB to the EFB stable, so that you now have:

1. The original EFB employing a cathode regulator, which is appropriate for cathode biased UL designs to be converted to (or requiring) fixed bias,

2. The original EFB employing a supply regulator, which is appropriate for fixed bias UL designs,

3. EFB II employing a cathode regulator, which is appropriate for cathode biased pentode designs to be converted to (or requiring) fixed bias, and

4. EFB II employing a supply regulator, which is appropriate for fixed biased pentode designs.

To document the validity of EFB II, I first built a clone of the Fisher SA-100 amplifier which I documented here:

http://www.audiokarma.org/forums/showth ... p?t=466229

I chose this amplifier because it is one of the highest regarded EL84 fixed bias pentode amplifiers, and frankly, because I also wanted one. Its initial intention was to be a development mule for EFB II, but ultimately became much more than that. For the record, my clone uses the version of EFB II described in #3 above, because paleenty of B+ was available, and the cathode regulator version chews some of that up as previously mentioned.

With that project complete now, a kind AK member has let me borrow a stock SA-100 not only for performance comparisons against my clone with EFB II, but also to actually install EFB II in after I'm finished documenting its stock performance. Since the SA-100 is already a fixed bias amplifier, the version described in #4 above will be used, using the supply regulator element you are asking about.

Work on installing EFB II into the SA-100 is underway now, and should (if all the planets align) wind up in another article by the end of the year, not only releasing the simple supply regulator version described in #2 above (which you are asking about), but also both versions relating to EFB II.

I mention all of this not only to provide a long needed update on EFB, but also because most particularly, the version described in #4 above (and being installed in the SA-100) should be of great interest to the gazillions of Fisher, Scott, and Eico integrated amplifier owners (to name a few) using 7591 output tubes. For those units, EFB II allows for lower dissipation levels, increased power output with both channels operating, and of course lower distortion as well. With how expensive the NOS examples of 7591 tubes are, it should not only improve performance, but increase life as well, which is a valuable commodity for those tubes.

And finally,

2. Yes. Because any form of cathode bias chews up available B+ voltage, then converting to the use of a negative voltage supply for bias turns that "lost" B+ into available power output and increased headroom.

Thanks so much for the interest. I hope this answers your questions!

Dave
dcgillespie
KT88
 
Posts: 399
Joined: Sun Mar 08, 2009 5:34 am
Location: Ball Ground, GA

Postby battradio » Thu Oct 25, 2012 2:09 pm

Mark
Image
User avatar
battradio
KT88
 
Posts: 927
Joined: Tue Mar 10, 2009 12:58 am
Location: near ST.Louis MO.

Previous

Return to stereo 35

Who is online

Users browsing this forum: No registered users and 3 guests

cron