Improved SCA-35/ST-35 Performance

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

Postby paart » Sun Dec 19, 2010 8:54 pm

Geek wrote:
paart wrote:The ST-35 that I'm currently building will incorporate power supply caps in the range of 1100-1200 mfd. However, I'm also using an external "ramp-up" device, 3 amp diodes, bleeder resistors, and I'll do lots of careful testing.


Which will save your diodes on charge, but not your transformer in the long run.

Dude, seriously, you do not need that much cap.

Cheers!


I've already run this setup in another amp, and the power transformer is fine. Again, I use an external circuit that slowly increases the AC line voltage over about 30 seconds, which would be similar to bringing it up on a variac. The total running load on the transformer is about the same as it would be without the extra capacitance. The reason that I kept the capacitance in the 1100-1200mfd range is that is about the physical limit that will fit on the original chassis. I've been tinkering with another ST-35, that being built on a Soundtastic chassis which will have quite a bit more!

Seriously, I do need that much. I've compared and heard the differance many times.

Regards,
---ART---
paart
 
Posts: 95
Joined: Mon Jul 17, 2006 8:36 pm
Location: Central PA

Postby dcgillespie » Sun Dec 19, 2010 9:16 pm

Thanks for the input gentlemen! I guess the old school in me doesn't really use over 200 mfd in capacitance at any one B+ point out of shear practicality, although nowadays they can cram so much capacitance into a small package that using more becomes very practical. However, that being said, I have never shied away from using active regulation in numerous pieces of my equipment as the ultimate in capacitance value. That is also very practical to implement now with modern devices. Both approaches have their benefits, and their drawbacks, but there is no doubt that anytime the B+ becomes more stable, it improves the sound without question.

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

Postby paart » Sun Dec 19, 2010 9:54 pm

Active regulation is one of the reasons why the "Soundtastic" chassis job is on the back burner right now! I did something really stupid and blew up two complete regulator circuits!---and raised my level of consciousness in the process. The main point of my post, simply was that one needs to be very careful when dealing with a lot of stored energy!
I'll try to get this back on track tomorrow by showing a possible layout of the EFB bias regulator on the Dynaco ST-35 chassis. I tried earlier and couldn't get the pictures to post.

---ART---
paart
 
Posts: 95
Joined: Mon Jul 17, 2006 8:36 pm
Location: Central PA

Postby dcgillespie » Sun Dec 19, 2010 10:49 pm

Art -- It would be very interesting to hear your take on the EFB modification both with a "typical" (i.e. standard) amount of B+ capacitance, and with increasing amounts of B+ capacitance added, as you plan to do.

Besides the basic conversion to fixed bias operation, part of the success of the EFB modification in the original application is the "Enhanced" element of the modification, wherein the output tube bias is modified by the prevailing B+ conditions within the amplifier -- be it because of changing AC line conditions, a slump in B+ due to increased power output, or both. While it obviously cannot make up for power output lost due to lost available B+, it does do a very effective job of maintaining very low distortion for the maximum power output that is available in both channels from the B+ that is available, which is always notably greater than with any kind of resistive cathode bias when using Z565 type transformers.

The point of all this, is that as the B+ itself becomes more and more stable, the Enhanced element of the design drops out of the equation, such that with perfectly regulated B+, it would have no effect at all, and conditions would simply become that of fixed bias operation, powered from regulated B+ AND regulated bias supplies, which of course is ideal, and all in one easy step. Therefore, within the available power output that an ultimately non-regulated supply can provide, it would be interesting to hear your take on how increasing levels of B+ capacitance affects the audible presentation with the EFB modification in place, given that it is already acting to minimize the measurable and audible impact that a less than perfect power supply has on performance. While EFB action is not exactly the same as adding more power supply capacitance, both are working towards the same end, so your take on it all would be most appreciated!

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

Postby GreatTone » Mon Dec 20, 2010 11:59 am

Thanks! And considering the fact that UF4007s are only 18 cents apiece, I'll put a couple on my next Mouser order. Since I am going to try the EFB mod, I think I'll skip the prebuilt board and just replace the caps myself, and install the mods on the chassis as per the article.

I hope to get some work done on this over the holidays. I'll post some pix if I do.

dcgillespie wrote:That's what these forums are for -- to ask questions, learn, and give feedback. It helps all of us!

The 1N4007 diodes are excellent replacements for the original diodes from a specification standpoint, and will work just fine -- particularly nice since you already have some on hand! My particular SCA-35 in fact still has the original diodes in place.

As you gain experience with your amp and wish to explore other things to enhance it's performance, many feel the use of ultra fast switching diodes (i.e. UF4007 in this case) provides improvement by minimizing the switching noise created in conventional Si diodes. An alternative is to bypass each traditional diode with a .01 mfd cap to minimize the noise, but be aware that these caps need to be rated at 1000 vdc themselves to handle the maximum PIV that appears across these diodes.

Dave
GreatTone
 
Posts: 9
Joined: Fri Dec 17, 2010 11:50 am
Location: Northern Virginia

Postby GreatTone » Wed Dec 22, 2010 9:22 am

I've ordered my parts, so I hope to get something done on this project over the holidays.

Question: is the value of the added 1K grid resistors critical? I have some 1K1 resistors I'd like to use. Thx
GreatTone
 
Posts: 9
Joined: Fri Dec 17, 2010 11:50 am
Location: Northern Virginia

Postby dcgillespie » Wed Dec 22, 2010 9:42 am

They will work just fine.

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

Postby GreatTone » Thu Dec 23, 2010 4:44 pm

I'm considering the replacement boards that are available for the power amp section of the SCA-35. They use 6U8A or 6GH8A instead of the 7199s. Will these different tubes have any effect on the mod? Thx
GreatTone
 
Posts: 9
Joined: Fri Dec 17, 2010 11:50 am
Location: Northern Virginia

Postby dcgillespie » Thu Dec 23, 2010 11:31 pm

As long as the new boards will still allow you add the new required control and screen grid resistors for the output tubes as shown on the schematic, they will work just fine. The use of different AF amp/phase inverter tubes has no effect on the modification.

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

5 Ohm resistors

Postby SilverT » Sat Jan 08, 2011 5:28 pm

Hello everyone,

New member here. I've been following this post with great interest. The parts to complete the EFB™ bias upgrade to my Dynaco SCA-35 have finally arrived.

I ordered a couple of Mills 5 ohm resistors to use for the cathode resistors / bias test points. However, these check anywhere between 5.2 to 5.4 ohms each, depending on which of my meters I believe. The Mills are supposed to be 1% tolerance, which should be a high of 5.05 ohms.

I could send them back, but would rather not if it's not necessary. Are these close enough, and would I have to compensate for them by adjusting to something other than the 0.27 VDC mentioned in the article?

Thanks.
SilverT
 
Posts: 5
Joined: Mon Dec 27, 2010 12:25 pm

Postby TomMcNally » Sat Jan 08, 2011 6:45 pm

When reading low values of resistance, the resistance of the
meter leads and connectors will come into play. You might
try bending the leads of the resistor and plugging it right
into the meter. In any case, a few tenths of an ohm is not
critical.

I got parts together to modify my SCA-35, and did what
Dave suggested, and paralleled two 10 ohm resistors
and they did come out right at 5.0 ... two sets actually.
I believe that's why he said to "hand select" the 10 ohm
resistors so you could come up with a pair that hit the 5 ohms.
User avatar
TomMcNally
Darling du Jour
 
Posts: 2729
Joined: Sat Nov 19, 2005 2:19 pm
Location: Northfield, NJ

Hi Tom

Postby SilverT » Sat Jan 08, 2011 7:39 pm

Thanks for the fast reply.

Prior to testing the resistors, I zeroed out both meters by touching the leads together. Of course they didn't zero out, but the leads added two tenths of an ohm, so I subtracted that to get the results in my post.

I didn't have any 10Ω resistors in my parts stash, or I'd have followed your and Dave's lead and paralleled a few. Even though I knew the Mills might be overkill, they were on sale, and the 5Ω value was in stock. Since I had to order some other parts anyway, it was simply easier.

Note: As I was typing this reply to you, I was thinking about your suggestion to bend the resistor leads, since the value of the resistors is so low. Then the light bulb went on in my head, and I remembered I had another way to test them. I have an ESR meter for checking caps. I hooked the resistors up to that, and they're dead-nuts (whatever that means) at 5Ω each.

Thanks again for your reply, and triggering my V-8 moment of remembering the ESR meter!
SilverT
 
Posts: 5
Joined: Mon Dec 27, 2010 12:25 pm

Postby dcgillespie » Sun Jan 09, 2011 11:39 pm

T -- I would bet the Mills are more accurate than you can actually measure. Even with your meter zeroed out and leads compensated for, unless you are using HP quality gear with documented accuracy, the quality of the Mills resistors is such that I would be entirely comfortable using the components you received, and still use the .27 vdc bias setting.

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

Postby SilverT » Mon Jan 10, 2011 12:42 pm

"I would bet the Mills are more accurate than you can actually measure."


Hi Dave,

I've had good luck with Mills resistors in the past, from both a performance and quality standpoint. On the other hand, I do have a manufacturing background and know that occasionally things can and do go wrong. That's why I've always made it a habit to check component values prior to installation.

Can you clarify the location you used to connect the output of the 337 to the 100 µF leg of the quad cap for me? Does the cap need to be tied directly (as close as possible) to the output to maintain stability, or can I wire the cap to the junction of the 5 ohm cathode resistors, simply for ease of construction?

Thanks.
SilverT
 
Posts: 5
Joined: Mon Dec 27, 2010 12:25 pm

Postby dcgillespie » Mon Jan 10, 2011 2:41 pm

In the SCA35 I worked up for the article, I retained the original can caps at that point, as they still measured full value. Therefore, the 100 mfd can lug became the connection point for (1) the output of the 337 regulator, (2) the lead for the heater bias to the hum balance controls (as before), and (3) the connection to the two 5 ohm cathode resistors.

Also, the ground of the EFB circuit was returned to the ground lug of this cap that the cathode resistor was originally grounded to. Using this configuration then, I found no tendency towards any regulator instability.

Thanks!

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

PreviousNext

Return to stereo 35

Who is online

Users browsing this forum: No registered users and 11 guests