The sunn Forum
Sunn Musical Equipment => Q & A => Topic started by: Isaac on June 27, 2007, 12:57:06 pm
-
SoundmasterG sent this to me a while back:
Basically what I did was use isolated input and output jacks, with a ground buss that only touches the chassis through a lift switch as Kevin O' Connor recommends in his TUT3 book. I also use series connect caps to get a higher voltage capability, but that involves drilling holes and moving the choke among other things. I added a bias pot so now I have one for each power tube. I also added bias tip pin jacks to the outside of the chassis so I can bias it without removing it from the head. I also added a positive bias of about 50v DC to the heater center tap to reduce hum and noise. I replaced every component in the amp with new ones. I used metal film 2W resistors and polypropylene and polyester orange drop and mallory 150 caps. I used a couple Solen caps too. Its all made the amp much quieter and better sounding. For electrolytics I used Xicons, and Sprague. Basically the only mods I did to the circuit per se was to increase the first two cap stages in the power supply to 110uf and 100uf respectively, and also upped the first stage cathode cap to 300uf. It all made it much better for bass with a tighter and deeper bass response. The main thing with doing all of this that was hard was doing a new layout, drilling the holes, and fitting all the parts in properly.
By "the first two cap stages" I assume you mean the 320V 'C' section and the 425V 'B' section, correct?
I don't see the advantage of increasing the first stage cathode cap to 300μF. The stock 250μF in parallel with the 1.5K resistor gives a time constant of 375msec, which bypasses the cathode resistor down to 2.7Hz. Correct? Going to 300μF raises the time constant to 450msec, lowering the frequency to 2.2Hz. Unless I'm missing something, forgetting something from my electronics classes. It has been quite a few years...
-
Actually what I meant by the "first two cap stages" were the 30uf e-cap to ground before the choke, and the 20uf section A after the choke. I would consider sections B and C to be the 3rd and 4th stages respectively, maybe I'm wrong but I would think that would be the way they would be noted since each one filters successively. Section B goes to the phase splitter, and section C goes to preamp and gain stage in front of the splitter, and the preamp stages are RC coupled to that cap so any change in value will affect the frequency response there. Changing the value of the E-caps for the plate and screen nodes will affect how the power supply responds to high current demands, but it doesn't affect the frequency response of those sections of the amp, so going to larger caps like I did makes for a tighter and quicker bass response, and a higher voltage rating.
I didn't get into the math when I increased that cathode cap on the first stage to 300uf, I just listened to it as compared to the 250uf and I liked it's sound better. I know its not scientific, but since we are interested in the sound of it since it is an audio amp, it seemed appropriate. When I got the amp there was no cathode cap at all, so I had to add a 250uf to get back to stock or I could change it up. I played around with other values too but I liked the 300uf the best. Its not a huge difference, as you confirmed with your numbers, but while it is a subtle change, it was one I liked. Go with whatever sounds good to you though. Since everyone hears differently, your opinion of what sounds good is as valid as mine is.
Also, just one correction to make, in that email I sent to you, the second stage is two 100uf caps in series for a 50uf value rather than the 100uf that I said it was. So instead of the stock 30uf/20uf/20uf/20uf @ 525v, my setup is 110uf @ 700v/50uf @ 700v/22uf @ 500v/22uf @ 450v. That third node does have to be 500v too as the voltage is up above 450v still at that point. Sorry for the confusion.
Greg
-
No problem. I was just looking for clarification, as I currently have my 200S open on the living room floor!
-
Okay, so I put a pair of 1N4006 diodes in place of the GZ34. First stage DC voltage is 549V. No problem, as my 200S has a pair of 600V Sprague Atoms there. However, after the coil, the voltage is 547V. The spec for the canned multicap is 525V, but the one I have is only 475V. Unless I can find some other caps, I'm going to have to take out the diodes and put back the GZ34. Even then, the DC voltage will exceed the 475V. What to do, what to do?
-
I pulled the diodes out. Voltage dropped to 528VDC with the GZ34, still well over the rated voltage of the multicap. It's been working, so it probably won't fail right away, but I'll have to get some caps.
-
Do you ever get the feeling you might be looking for magic where none exists?
-
I don't really like the idea of using PCB's in a vintage point to point amp, but for 50 bucks, you can go back to the diodes if you want and it's reversible.
http://www.triodeelectronics.com/sdslabmk3cap1.html
-
Do you ever get the feeling you might be looking for magic where none exists?
Magic? Nah. I'm just fiddling about. The closest thing to magic that I believe in comes inside a bottle of wine.
I don't really like the idea of using PCB's in a vintage point to point amp, but for 50 bucks, you can go back to the diodes if you want and it's reversible.
http://www.triodeelectronics.com/sdslabmk3cap1.html
I'll give that some thought. I'll have to compare the price to discreet caps, but the PCB thing doesn't bother me a bit.
-
I put one of those in my 200s under the carriage so the can is still there and visible. sounds great, gives me the 600 or 800 volts (I can't remember) that this thing spikes to and looks vintage. some modification had to be done to the circuit board to work properly. I didn't do that, I had a great tech in my area go through the amp and modify the new cap board.
-
You could try a Weber Copper Cap WZ34 or WZ68 and the voltage will be close to where you are with the GZ34, but with no filament load on the power transformer, your PT will run cooler than stock, which is always a good thing. Mine is using a Weber CC WZ68, so if you can remember how mine sounds, then there you go.
You'll still need some higher voltage caps on those beginning stages. Your Spragues are ok there, but for the rest, the discrete caps are cheaper and if you run them in series like I did, two 350v caps will give you a 700v capability which is plenty, but to install them neatly in the amp requires drilling a couple holes to mount some terminal strips, or use silicone and/or tie wraps to hold them in place. If you decide to do the series connected caps thing, don't forget the 220k 1w resistors that need to be in parallel with each cap.
The triode electronics PCB setup is good in that those radial caps used on it are even cheaper than the typical axials, and will be around longer in better quality probably given that they're used for modern electronics more.
You may also be interested in a Sprague Atoms thread on the Hoffman bbs that goes into some of the problems with the current small caps inside the large can approach that Spague has been using for some time. Here's the link.
http://www.el34world.com/Forum/yabb2/nph-YaBB.pl?num=1182110842
Greg
-
I don't see the advantage of increasing the first stage cathode cap to 300μF. The stock 250μF in parallel with the 1.5K resistor gives a time constant of 375msec, which bypasses the cathode resistor down to 2.7Hz. Correct? Going to 300μF raises the time constant to 450msec, lowering the frequency to 2.2Hz. Unless I'm missing something, forgetting something from my electronics classes. It has been quite a few years...
I think something is wrong with your math. Since your time constant to frequency conversion is correct, maybe the time constants you are coming up with is incorrect or doesn't consider what you are really trying to determine is the cutoff frequency which is normally defined as when the response is 3 db down.
To determine the -3db frequency - when a cap will have a capacitive reactance equal to a resistor, I simply use:
freq = 1 / (2*Pi*farads*ohms)
For a 250μF and 1.5Kohms, I get a frequency of 0.424Hz. I get 0.354Hz for a 300μF cap.
Using the cathode resistor by itself is incorrect. You need to consider that the cathode impedance is acting in parallel with the cathode resistor. The cathode impedance for this stage would be roughly 1.6K, and that in parallel with the 1.5K cathode resistor would be about 774 ohms. So, using the 1.5K resistor alone will result in a number about an octave lower (i.e. 1/2 the frequency) than actual.
Will you hear the difference? Absolutely not. It will be subtle to the point of non-existance. It'll all be imagination. Here's why.
What would be the gain difference of this stage operating without a cathode resistor bypass cap versus operating completely, totally, 100percent with a fully bypassed cathode resistor?
6db. Actually I rounded up it's a couple/few tenths less.
At the -3db cutoff frequency, 3db out of the 6db maximum available has already been eaten up - the stage is already operating with 3db more gain with the bypass cap than without. How long do you think it'll take the stage to max out that remaining 3db of maximum gain potential?
About an octave, or twice the frequency. For all frequencies above this, the stage is already operating at it's absolute maximum potential. That's all there is and there ain't no more.
With either a 250 or 300μF cap, this stage has already reached it's maximum potential well below 5hz. It will not increase or change anything once you feed in a frequency you can actually hear. Even using the more typical 22μF value for a bypass cap (-3db @ 9.3Hz), the tube will be operating at maximum gain for any usable frequency.
In Kevin's books he consistently makes two statements, which I will paraphrase. The first is one we know from our own experience - a "tight" bass is achieved by reducing bass response. The second is common sense once you think about it - there is no point in increasing frequency response at frequencies that won't make it through the amp.
Look at the output side of this stage. Two values for coupling capacitors can be determined based on the position of the bass boost switch. The value of the capacitor and the volume control resistance in series with the output impedance of this stage will determine how much bass actually makes it out of the stage and into the rest of the amp. This forms a first order high-pass filter.
The output impedance is roughly 38K, but I will bump it up to 50K so I can get better numbers than actual. The volume control is 500K and I'm going to assume the full value, again for better numbers than actual.
With the .0022μF cap alone, the high pass filter has a -3db cuttoff at 131Hz. Frequencies below this will be reduced at a 6db rate. By the time we get down to our lowest frequency, that 6db gain boost we got from the cathode bypass cap is history.
With the bass boost switch on, the two coupling caps are in parallel giving us a 0.009 and change. Heck, let's just round it up to .01μF, again for better numbers than actual. This is much better, our -3db cutoff frequency is now a hair below 30Hz. Our cathode bypass 6db boost will be eliminated at 15Hz and it's all downhill from there. But the good news is that our cathode bypass efforts will be fully recognized at around 60Hz, a fret or so above the open bass "A" string.
Before I rest my case, I'll leave you with a quote from the Radiotron Designer's Handbook, my third edition being printed November 1941, concerning bypass cap values (Ck):
A cathode bias resistor is usually by-passed by a condenser (Ck) in order
(1) to avoid degeneration and loss of gain, and
(2) to avoid hum.
If Ck were omitted, the amplifier would operate with Negative Current Feedback. The capacitance Ck when used to avoid degeneration should normally have a reactance which is low compared with Rk at the lowest frequency required to be amplified. If accurate calculations of this capacitance for specified frequency response are required the formula at the end of this section may be used. For most practical purposes Ck may be a 25μF electrolytic condenser, and although such a high capacitance is often unnecessary for frequency response it is valuable in by-passing hum voltage originating between the heater and cathode.
Yep, there is much more magic in a bottle of wine.
-
Another point to ponder...
I had another thought, but since I was in another room I dragged out my paper schematics to check something on the 200s. Nah, I was mistaken...
However, the schematic I pulled out had a 1969 date. Guess what, no cathode bypass cap - none at all.
I checked a few others. One other 200s schematic didn't show a bypass cap, but it also didn't have a date. I assume Sunn schematics without dates are earlier... as I think they only started dating them in 1969 and later.
So... if you have a late 1968, or most definitely a 1969 or later 200s (or other amp that used the same circuit) the "stock" configuration is no cathode bypass cap. Put one in, and you are really "modding" it, not restoring it to a stock configuration.
Pretty strange huh?
-
I wondered about the absence of the 250uf cap on the first stage, as the schematics I had showed one there, and I remember my '67 had one. I'll have to check the schematics here to see if some of them don't show it, but if you have some schematics that aren't in the archives here, maybe you could get a copy to Tboy of the ones that aren't here so he could add them into the archives?
Greg
-
I wondered about the absence of the 250uf cap on the first stage, as the schematics I had showed one there, and I remember my '67 had one. I'll have to check the schematics here to see if some of them don't show it, but if you have some schematics that aren't in the archives here, maybe you could get a copy to Tboy of the ones that aren't here so he could add them into the archives?
Greg
I no longer freely contribute schematics to this site. Not since 2000, when I finally realized my contributions of schematics (still on this site) wouldn't result in the schematics I wanted to see being submitted by others.
It's a long, boring story, but as the Who sang "We Won't Get Fooled Again". I learned my lesson.
Besides, tboy doesn't provide an email or other link for contributions to the site. Therefore, I assume he doesn't want any.
However, for your purpose, if you have a copy of "The Tube Amp Book", version 4, by Aspen Pittman, this same schematic - the 1969 version of the 200s without cathode bypass caps can be found on page 630.
-
I no longer freely contribute schematics to this site. Not since 2000, when I finally realized my contributions of schematics (still on this site) wouldn't result in the schematics I wanted to see being submitted by others.
I think I have originals for all of the tube stuff up to '69, some with handwritten notes on them but very readable. Is there something you are looking for specifically?
-
Thanks, Joe. Very interesting and informative posts.
In case anyone is interested, the method I used to determine the frequency was to find the RC time constant which, IIRC, is equal to the product of resistance and capacitance (t=RC), then find the corresponding frequency (F=1/t). I did not take the cathode resistance into account.
One thing I get from your log post above is that the Low Boost really isn't boosting anything at all, rather it is restoring frequencies rolled off by the coupling cap coming off the second stage. Is this correct?
-
One thing I get from your log post above is that the Low Boost really isn't boosting anything at all, rather it is restoring frequencies rolled off by the coupling cap coming off the second stage. Is this correct?
Well, the coupling cap is actually off the first stage, but yes you are correct. The action of the switch would be more "technically correct" if the normal position said "Cut" and the boost position said "Normal".
How you hear it all depends on the reference point - it's a relative thing.
-
Yes, of course. Brain fart. First stage.
Odd, that. It really does seem to act as a boost. My speakers are only flat down to about 80Hz, yet my basses sound pretty good through them. Kick in the Low Boost, and it gets really heavy on the low end.
-
Odd, that. It really does seem to act as a boost. My speakers are only flat down to about 80Hz, yet my basses sound pretty good through them. Kick in the Low Boost, and it gets really heavy on the low end.
Thats how Fender's boost switches worked too. s Joe said, its relative, and its a feature. Pack the amp with more features, and people think they are getting more value for the money.
Ed, I'd like to have all the schematics so I can refer back and forth between them, but nothing I'm looking for in particular.
Joe, I completely understand your reluctance to spend a ton of time contributing something that no one else seems to want to do. I'm sure Tboy would be interested in adding to what is there, but it seems that maybe he doesn't go out of his way to allow others to upload.
Greg
-
Ed, I'd like to have all the schematics so I can refer back and forth between them, but nothing I'm looking for in particular.
Me too, that's why I stepped up and bought them! Seriously though, I could send something specific maybe, but I really don't have time to scan everything, probably +/- 30 schematics.
Joe, I completely understand your reluctance to spend a ton of time contributing something that no one else seems to want to do. I'm sure Tboy would be interested in adding to what is there, but it seems that maybe he doesn't go out of his way to allow others to upload.
I did once go through the schematics section of this sight and listed the holes I could fill, but I never followed through. Maybe I'll try it again if we can find a way to get them posted.
-
Ed, I'd like to have all the schematics so I can refer back and forth between them, but nothing I'm looking for in particular.
Me too, that's why I stepped up and bought them! Seriously though, I could send something specific maybe, but I really don't have time to scan everything, probably +/- 30 schematics.
Joe, I completely understand your reluctance to spend a ton of time contributing something that no one else seems to want to do. I'm sure Tboy would be interested in adding to what is there, but it seems that maybe he doesn't go out of his way to allow others to upload.
I did once go through the schematics section of this sight and listed the holes I could fill, but I never followed through. Maybe I'll try it again if we can find a way to get them posted.
Thanks for the offer Ed, but unlike most people I am primarily interested in the solid state amps. :-o
But, if you have a coliseum lead or coliseum 880 schematic... I'd like a copy!! :-D
Seriously, after you've seen a couple/few Sunn tube amp insides, a schematic really isn't necessary. Minor differences at best, even between a 200s and 100s.
-
Thanks for the offer Ed, but unlike most people I am primarily interested in the solid state amps. :-o
In the early 70's, at least in my part of the world, tubes became very passé. Oh, it became solid state this, and transistor that - I even traded an SVT rig for an Acoustic 371. Even guitar players were going transistor, it was the way of the future. And, transistors are better, just ask an engineer! Tubes were antiquated, inefficient, better at generating heat than efficiently amplifying sound.
Ah, but...what sound they do produce!
Luckily, I was cured from the madness, and saw the evil transistor for what it really was; the antithesis to all things thick and warm, the province of harsh and sterile, THE PROGENITOR OF TECHNO MUSIC.
See the light Joe! Come out from under the spell of the dark side! Come on, transistors don't even glow and pulsate, don't emanate the smell of cigarettes smoked long, long ago in a bar far, far away when they get cookin', DON'T HAVE THE MOJO!
Well, actually I still normally use SS gear when I gig, much more practical. I only pull out the "heavies" when it's a special occasion, I'm feeling froggy, I want to impress another bass player, or have roadies available.
-
I went ahead and did some mods.
Changed the first stage bypass cap from 250 mfd to 22 mfd. As expected, no audible difference.
Then I added a 22 mfd bypass to the second stage. Seemed a little louder at the same volume setting, but no discernable change in tone. Again, exactly as expected.
Next, I put a .0047 mfd cap in parallel with the first to second stage coupling cap. Now it sounds as it did before with the Low Boost switch on, and the Low Boost switch has no discernable effect. It should restore the response from 40 Hz down to 20 Hz, but I play 4-strings, so I have nothing below 41 Hz in the first place, and the cab I was playing through doesn't go that low, either. I tried the octave setting on the Bass Synth Wah for some bottom octave signal, but still couldn't hear any difference.
Overall, probably a waste of time, except that it was fun!
-
I don't really like the idea of using PCB's in a vintage point to point amp, but for 50 bucks, you can go back to the diodes if you want and it's reversible.
http://www.triodeelectronics.com/sdslabmk3cap1.html
Ordered one today.
I put one of those in my 200s under the carriage so the can is still there and visible. sounds great, gives me the 600 or 800 volts (I can't remember) that this thing spikes to and looks vintage. some modification had to be done to the circuit board to work properly. I didn't do that, I had a great tech in my area go through the amp and modify the new cap board.
Any idea what the mods were? If not, can you find out?
-
I went ahead and did some mods.
Changed the first stage bypass cap from 250 mfd to 22 mfd. As expected, no audible difference.
Then I added a 22 mfd bypass to the second stage. Seemed a little louder at the same volume setting, but no discernable change in tone. Again, exactly as expected.
Next, I put a .0047 mfd cap in parallel with the first to second stage coupling cap. Now it sounds as it did before with the Low Boost switch on, and the Low Boost switch has no discernable effect. It should restore the response from 40 Hz down to 20 Hz, but I play 4-strings, so I have nothing below 41 Hz in the first place, and the cab I was playing through doesn't go that low, either. I tried the octave setting on the Bass Synth Wah for some bottom octave signal, but still couldn't hear any difference.
Overall, probably a waste of time, except that it was fun!
(Sigh)
You're much too honest to make a tube tone guru. :roll:
:-D :-D :-D
It is most definitely fun!!
-
[/quote]
Any idea what the mods were? If not, can you find out?
[/quote]
I'll see if I can track that down and let you know. proabably a week or so
-
Further mods to my 200S.
I bought the cap board from Tripod Electronics. It’s a kit, just a board and the parts, but I’m an experienced hand with a soldering iron, so that’s okay. I had to drill mounting holes in the chassis. I thought the transformer mounts would be the right distance apart, but that turned out to not be the case. Again, no big deal.
I used the supplied 1N4007 diodes. When wiring it up, I discovered something: when using the diodes, there is no provision for the standby switch. The diode output is directly connected to the first stage of capacitors.
I figured there were several paths to take. Forget about the standby function; cut the trace and wire it in from there; place the switch after the inductor, between it and the second capacitor stage.
I wanted the mod to be as transparent as possible, so going without a standby switch didn’t appeal to me. Neither did cutting the trace, especially since there was no convenient place to connect back into the circuit. So, I went with my third option.
Hooking it up was easy. I just ran the wire from the board to the coil, from the coil to the switch, then back to the board. Then I got to thinking: what about inductive kick? The coil is passing a fair amount of current. When I place it in standby, it will try to continue passing that current as its magnetic field collapses. It has to go somewhere. One solution is a flyback diode. That’s a diode placed in parallel with the inductor which shorts it out when the inductor becomes a current source. I wanted to do that, but couldn’t see a convenient place to put one in. Instead, I opted for a bleeder resistor. Voltage runs well over 500V, more like 550V. Being conservative, I used 600V as an estimate. I figured I had plenty of ¼ and ½ watt resistors on hand, so all I had to do was make sure I didn’t exceed that. ¼ watt at 600V comes to just over 1M. I found a handy 3M6 resistor (3.6 megohms) and ran that from the coil side of the switch to ground. That comes to one tenth of a watt. It should be fine.
I also had to extend the wire from A (the output transformer center tap, which supplies DC to the power tubes). Again, no big deal. Now, what to do with the canned multicap? I could take it out, and leave a gaping hole in the chassis, or leave it in. As before, I wanted this to appear as stock as practical, so I opted to leave it in. I also went ahead and hooked it up. I paralleled two of the four with stages B and C, not because I thought it would make any significant difference, but because, well, it was there. Besides, I might need it someday, and my understanding is that it’s better for electrolytics to be in operation than to sit idle.
I hooked it up to my Model 215, and cautiously powered it up. Cautiously, not because I thought I made any mistakes, but because, if I did, say, get the diodes or capacitors in backwards, exploding caps could result. No problems on that score, though. No problems at all, in fact.
At rehearsal last night, it sounded great. It seemed more solid, with just a little more ‘oomph’. Of course, that’s probably just my imagination. Nothing I did really ought to be audible.
-
Does the cap board use different cap values than stock? If it used higher values, then your impression of the amp being more solid isn't off base as it will be able to supply the lower frequencies quicker thus giving an impression of more bass. The solid state rectifiers will have made the B+ on the tubes a little higher and that will have a subtle effect too. I'm surprised the board didn't have anywhere for the standby to hook up though.
Greg
-
Yes, somewhat higher. Each section comprises two 100mfd caps in series, for 50 mfd, plus a small cap in parallel. Of course, it also has the resistors forming the voltage divider, in parallel with each cap. Stock values are 30 mfd in the first section and 20 mfd in the second, so there's twice as much capacitance feeding the power tubes now.
Anyway, regardless of whether or not it's in my head, it sounds sweet!
-
With those increased cap values, the change in response isn't in your head. Rather than a tone change, it is just that the amp can now respond to high current demands quicker, so it has a tighter bass response and probably a slightly higher damping factor too I would guess. Glad you like the change! I did too when I modded mine and think going with increased cap values in these is a great idea for a bass player.
-
Thanks to EdBass for suggesting the Triode Electronics capacitor board.
-
I've got one of those too, but its for my Dynaco Stereo 70. I'm still surprised that they didn't have a connection point on the board for the standby.
-
I was, too.