2000s, cap increases, worth it or not

[pbman1953]

On a Sunn  2000s , The owner before me did upgrade the filters. I was wondering if there were any advantages to increase the following values:


Existing are:
Section B- 20uf/500v

Section C from 22uf/ 450v and 40uf/500v


 I have these values to work with:

100uf/450v (Xicon)
80uf/500v (JJ)
50uf + 50uf @ 500v  ( F&T)



The bias cap in section E is now a 2- 100/160v. Should that be considered or left?

Happy 4th!

[shooter]

I was wondering

Is it working?  Is it humming?

[pbman1953]

working with no hum

[shooter]

Play it til you can't stand it and just HAVE to tweak something

[pbman1953]

Ha, I sound bored don't I, 


I look to improve if it's worth it.


Thanks

[sluckey]

That amp is very well designed. I'd leave it alone. If you do decide you must replace the caps I'd advise you to measure DC voltages at each node before changing. B+ voltages run higher than the schematic shows. You would need 600V caps for the reservoir cap and node A cap.

[pbman1953]

The A section is:


2-  220uf/ 350v in series  (110/700v)


2- 100uf/ 450v in series-  50/900V

[sluckey]

The A section is:


2-  220uf/ 350v in series  (110/700v)


2- 100uf/ 450v in series-  50/900V

Wow! I would definitely change those. Those big caps are likely to make those GZ34s light up like July 4th sparklers.

[pbman1953]

I use diodes for SS, does that matter?




What would you suggest to change to?

[PRR]

DO NOT change the value of that first cap off the rectifiers!

This C-L-C filter is carefully designed around rectifier limits. The 20u cap shows huge ripple, but also low peak currents, so the GZ34s working at high voltage won't arc-over, die, and take the PT with them.

This also gets the average DC at less than the peak of the AC, giving the "right" DC voltage.

> I use diodes for SS, does that matter?

Well, yes; but the extreme voltages and currents here will also stress Silicon diodes. And the DC voltage output is still designed around the smaller uFd. Going "generous" just threatens the whole amp.

The choke and second cap get this big ripple down. As you say, it does not hum.

The second cap must also carry the audio current around the finals. 20uFd is adequate.

> I'd leave it alone.

Agree 110%.

> The A section is: 2-  220uf/ 350v in series  (110/700v)

So already tampered. Why? It doesn't hum. It already ran close to the edge. If you truly need MORE POWER, get a sand-state booster. Put the first cap back near 20uFd. Today, probably two 47uFd 400V. I'm less concerned about the second cap.

[pbman1953]

"So already tampered. Why? It doesn't hum. It already ran close to the edge. If you truly need MORE POWER, get a sand-state booster. Put the first cap back near 20uFd. Today, probably two 47uFd 400V. I'm less concerned about the second cap. "



Tampered with in the sense that the filter caps were installed because it did hum, in the past. It's not a matter of power because the power is good. I want to make it right since Sluckey felt the A section was wrong. 

[pbman1953]

Are you saying

this is wrong -  2-  220uf/ 350v in series  (110/700v)This is ok,  leave alone?-   2- 100uf/ 450v in series-  50/900V




Would changing the 2- 220UF/ 350v to 2- 80/500v be better?

[sluckey]

Have you actually measured the B+ voltages on every filter cap?

If you must stick with ss diodes be sure to double up, ie, two in series on each leg. And for sure, change those two 220µFs to two 100µFs.

If you want to make it right then use two GZ34s rather than SS. The GZ34s take about 10 to 15 seconds to warm up enough to put out any B+. This slow start up is a good thing for caps and output tubes. And also change the two 220µFs to two 60µFs.

Early on, Sunn had problems with arcing GZ34s that would fail and also take out the PT. And the first two filter caps were failing because that 525V can needed to have a higher voltage rating. By 1970 they had modified the power supply to use 20µF @ 600V caps for the first and second filters.

Here's a pic of the caps they used in my Sceptre...

     http://sluckeyamps.com/sunn/sceptre1.jpg

[pbman1953]

Diodes are 2 in series for each- NTE- 5809-  2 diodes sets in series per socket.


Will 2- 80/500V be ok too?




Also, now, all "A" caps have a 220 resistor across them. Does this stay? go? or change the value?




Measurements-


A (stand by)-  548 v


A-(output) 542V


B-  446V


C- 326V

[sluckey]

Will 2- 80/500V be ok too?

yes

Just an idea... If I were doing this I would get four 100µF@350V Illinois caps. They are cheap and small (.75" dia, 1.5" long). You could make a neat, compact turret board to hold all four caps and all four 220K resistors. Or just use terminal strips.

Also, now, all "A" caps have a 220 resistor across them. Does this stay? go? or change the value?

You probably mean 220K? They should stay. Anytime you stack equal value caps in series you should put equal value resistors (typically 220K) across each cap to force the voltage to divide equally across each cap. These resistors also serve as a bleeder.

[pbman1953]

You are correct I forgot the "k"- for 220K

After these changes, rebias?

[SoundmasterG]

The B and C sections are mostly RC coupled (for B, one is, one isn't) and if you change the cap values there, then the tone of those stages supplied by B and C will change. I would leave those alone.

I would also leave the bias cap where it is now. Going too large increases the time constant at that stage and it can make the bias too slow to react to changes which you do not want. 100uF is about as large as you would want there.

For section A, it was carefully designed back in the day by Dynaco for hi-fi and essentially copied by Sunn for MI use. Caps were also expensive back then so they used the smallest cap that would get the performance they were looking for. Caps are much cheaper today. That said, as PRR noted the C L C filter in there with the two caps and choke in between is designed to work a certain way and changing any of the values changes its performance. The higher cap values are only appropriate to use if you are going to use solid state rectifiers. If you plan to use tube rectifiers, then the values of the cap at the section before the choke must be reduced to below 60uF, and with today's poor quality GZ34's going lower than that is a good idea too. The early Sunn 2000S's used the dual GZ34's and Dynaco transformers. Then they changed to Schumacher transformers with a slightly higher B+, and dual GZ34's. Then later they got rid of the GZ34's and used solid state diodes from the factory. They never changed the filter values, but caps were still expensive and the performance was fine for the day so that likely explains why they never redesigned the cap values.

Using larger caps at node A before and after the choke will give a bit tighter and more present bass. I know of quite a few Sunn 2000S's that have this mod and they are perfectly reliable with it. The owners seem to like the increased and tighter bass response. I wouldn't do the mod if I was using guitar with the amp but it works fine and sounds good for bass. 


Why do you feel you need to make changes to it if it is and has been working fine with the modded cap values? You can leave it as is or go back to stock as the others are suggesting but I wouldn't go larger in cap values at the A node than what it is now. If it was me I would leave it where it is as I like the feel of the tighter/quicker bass response, and it is perfectly reliable with those values, but its your amp and your time so do what you want to it. 

Greg

[pbman1953]

Greg ,


Interesting write up. You probably remember these were your cap suggestions when I bought this 2000s. The other one I have, and I have to pop it open to check, has the caps closer to stock. My other one was completely gone over before I  bought it. My other one uses copper caps and this one has diodes.


The reason for all this was that I was replacing all the tube sockets and  I was thinking of the cap vales on the B & should be upped. Keep in mind that I used to be from a a car audio standpoint and bigger caps were always better.

[pbman1953]

 


 What should the bias cap value be for Copper caps?  Change to the stock 47uf?


Thanks

[Ed_Chambley]

Rich, do not confuse your Traynor mods with a SUNN. It is like comparing a Lincoln town car with a 69 Ss Chevelle with a 396. The later is the Traynor.


The power section of the SUNN is extremely proven and even though both may be Ul output, they function differently, meaning the screen taps.


Secondly, the Traynor design is ss rectified. I love the SUNN powers section and by design will pass a very balanced signal and preamp responsive as well.


Put it stock keeping in mind many Dynaco amps have been playing hifi wonderfully for years and still people steal from the design and pay very good money for those transformers.


Think of your SUNN as your finesse amp. I really like the idea of upping the voltages in the caps if you must replace them. I series 100uf dual f&t cans for 50uf and 900 volts with 270 balance resistors at the maximum and I do use hexfred diodes for instrument amps and it will make the attack a little more aggressive, but not a lot. The only reason I do this is it is cheaper and easier to build.


My 2 cents

[SoundmasterG]

Greg ,


Interesting write up. You probably remember these were your cap suggestions when I bought this 2000s. The other one I have, and I have to pop it open to check, has the caps closer to stock. My other one was completely gone over before I  bought it. My other one uses copper caps and this one has diodes.


The reason for all this was that I was replacing all the tube sockets and  I was thinking of the cap vales on the B & should be upped. Keep in mind that I used to be from a a car audio standpoint and bigger caps were always better.

Rich,


I do recall that those were the cap suggestions that I posted on the Sunn forum. I clarified those suggestions as being only for bass and only with the use of solid state rectifiers. A Copper Cap is solid state also, but it has other parts in it to emulate the sag of a tube rectifier. The amp will work fine with stock values or with the modded values....it really comes down to what do you want the response to be like. Stock values have slower and less present bass and a little more hum. The modded values have quicker and more present bass and less hum. Part of that is also the addition of the solid state rectifiers and removal of tube rectifiers. Since Sunn eventually went to the use of solid state rectifiers from the factory on the 2000S and the quality of tube GZ34's these days is lousy, I think using solid state rectifiers in a 2000S is a good idea, especially if you plan to use it for bass as the bottom end tightens up considerably with the use of the solid state diodes.


In either case you should leave the other B and C locations stock as since they are RC coupled in those locations for the most part, changing a cap value there will change the tone directly, which you may not like. The bias caps shouldn't be any larger than the 100uF that they are now as going larger slows down their ability to track what happens to the B+. You can leave them where they are or go back to stock. If it was me, I would just play the amp and not worry about changing this or that to make it perfect. It functions well as it is now and sounds good so use it. You save yourself money and time by leaving it as is.


In car audio, bigger caps are always better but in tube amps, that isn't always the case, or at least it might be the case, but within reason. A lot of solid state amps will use 2000uF in the first stage for example. Solid state can control the speaker better than a tube amp, and using a massive cap like that helps too, but I would never go that high with a tube amp. Solid state amps often need high values of caps to behave properly but on a tube amp as you go higher you start to choke the life out of it. A first stage of 110uF is a good compromise for many tube amps as long as they are using solid state rectifiers. The screen tap can also benefit from having a higher than vintage cap value and since they aren't RC coupled like the preamp then it can make sense to do that change. In the Sunn it is ultralinear so the screen is getting it's voltage direct from the output transformer and the A tap. The 2000S also has individual screen grid resistors too which the 200S and the smaller amps don't.


You said you have another one that is using stock values. How would you compare the sound of each and the bass response of each?

Greg

[pbman1953]

Greg,


Actually I was opened the other one and it's the same cap set up. I guess I forgot I changed them because it's been at a friends house for a while and I finally brought it back.


Anyway, both are back in their cases and it's time to rock because they both sound great. I'd say the amp with the copper caps has a bit more transients when you get on it. That's good for when I play along a guitar solo.


Thanks Again!

[SoundmasterG]

Greg,


Actually I was opened the other one and it's the same cap set up. I guess I forgot I changed them because it's been at a friends house for a while and I finally brought it back.


Anyway, both are back in their cases and it's time to rock because they both sound great. I'd say the amp with the copper caps has a bit more transients when you get on it. That's good for when I play along a guitar solo.



Thanks Again!

Rich,


You can modify the one amp with solid state diodes to use the Copper Caps also if you desire. It isn't that hard. You may like having the different options though by leaving one with the solid state diodes rather than the Copper Cap. My early 2000S had solid state diodes in it when I bought it, but it would originally have come with tube rectifiers. I'm not sure yet if I will use the solid state diodes or Copper Caps when I overhaul it, but its moot right now as the amp still works and sounds fine. One of these days I'll overhaul it though.

Greg

[pbman1953]

That's what I plan, one with CC's and one with diodes.