Stromberg Carlson ASR-333 filter cap values?

[Ziradog]

Getting ready to redo an ASR-333. The original can is a skinny one with values of 20-20-10-10 mfd, all 450 volt or less, plus two 100 mfd 250 volt singles. There is a CE can with those values but it is a larger diameter and there is plenty of space underneath so thinking I will go that route. My question is, should I stick close to those values or is there any advantage in upping some of them? This has solid state rectifiers BTW. Schematic of power supply attached.

Thanks.

 

[century tek]

I think that I would stick fairly close to the original values but you can always go up in voltage. Just don't go overboard. I believe the engineers allowed something like 10 - 20 % tolerances or some such thing.

 

[RWood]

Those last two 10uFs look like they are in parallel, so 20uF (or 22uF) can be used there too.

Watch the polarity on the 100uF voltage doubler caps

 

[nerdorama]

The two 100uF caps are in a voltage doubler so they have to be separate caps. Can't use two in a common can. At least that is how I interpreted the intention of your post.
Joh

 

[Patrick620]

Good question. I have a 433 on my bench right now getting ready for a recap. I have a new 20/20/20/20 can that I was thinking about using. I discovered that the previous "tech" PTP wired in 4 new caps of the original values but never unhooked the original can. I hate working on an amp that someone else has monkeyed with. The current owner told me that told me that he ran it for about two hours but the PT seemed hot.

 

[Ziradog]

Joh, that was not my intent, if I did a can it would be to replace the 20-20-10-10 only. My question is really about the values.

 

[Patrick620]

Joh, that was not my intent, if I did a can it would be to replace the 20-20-10-10 only. My question is really about the values.

I knew what you meant. On another note, I have read on some post that a good mod might be to increase the values of the doubler caps. and change out the Zener diodes. I have some FR307 fast recovery diodes on hand. Perfect timing for this thread on my bench. Thanks for bring up this thread, Ziradog. I hope it is ok if I join in with related questions about the PS.

 

[nerdorama]

Sorry for misinterpreting. I did a 433 a couple of years ago. I think I increased the values, probably to 33 or 47uF with singles underneath. I also installed larger caps for the doublers.
John

 

[Retrovert]

Increasing the capacitance in a capacitor-input filter greatly stresses the rectifier because the conduction angle reduces. Remember, the capacitor may only charge when the applied voltage is greater than the stored voltage. The shorter the time for charging the greater the instantaneous current draw.

The designers allowed for a certain peak current (not the average current specified by the maximum current) without blowing apart the cathode, and that's what the capacitance limit is really specifying.

So best stick to the limits for the rectifier tube.

 

[Ziradog]

Thanks Retro, this has solid state rectifier.

 

[gadget73]

yeah, voltage doubler. Can cap at 20/20/20/20 will be fine. If you want to bump the 100uf caps you should be able to make them 220uf without any real problems but I wouldn't want to get crazy with it.

 

[primosounds]

Thanks Retro, this has solid state rectifier.

Zira, the state of the art SSR, rectifier for tube amps is a high voltage Schottky type. 1200v . This rectifier has no switching noise which can impart a negative sonic contribution. FREDs, fast recovery epitaxial diode, are also an upgrade but not as noise free.
If your location has high wall voltages over 117 vac, you might consider adding LC , inductor + capacitor, combination ala, HK Citation 2 and McIntosh, to reduce B+ voltage and get better regulation and lower noise floor.

 

[Patrick620]

Thank you, all. Especially ZDog for letting me jump in on this. The timing was perfect.

 

[Retrovert]

Thanks Retro, this has solid state rectifier.

Ok, there's another part to the conduction angle issue which I didn't have time to post. Primosounds has addressed some of it.

The short conduction angle essentially modules the rectifier and transformer as a square wave. That creates an AM carrier wave, a few hundred kHz, which can broadcast mains noise through the amplifier.

The other issue is that the square wave hammers the transformer which causes it to ring, outputting noise. Make Johnson over at diyAudio has designed two gadgets, for which inexpensive (under $10) PC boards exist, which allow snubbers to be created without doing any calculations. Look for Quasimodo and Cheapomodo.

Yet another issue is diode Qrr which causes noise bursts. That's the issue upon which Primo brieftly touches. Using ultra-low capacitance diodes, like Schottky SiC diodes, is a good starting point, then snubbing whatever remains.

RC snubbers are easy to add and completely remove this noise.

 

[Ziradog]

Hm, hadn't really given much thoughts to the diodes. Most of what I have found on the web talks about replacing them with pretty standard stuff (1N4007 or such).

 

[Patrick620]

Right. I have also heard to leave the Zeners in place. The amp in question is for a good customer. I just want to do it the best way. I take much pride in my work.

 

[Ziradog]

Is the FR307 the same as the FR307B?

 

[FlaCharlie]

Hm, hadn't really given much thoughts to the diodes. Most of what I have found on the web talks about replacing them with pretty standard stuff (1N4007 or such).

I'd suggest replacing the diodes with, at the minimum, UF4007s or better yet the Schottkey parts mentioned earlier. The two 100uf caps are in series which results in a 50uf value. You can go up on these since it doesn't have a tube rectifier. I'd suggest at least a pair of 220uf parts but the limiting factor will be space. The chassis on these SC amps is very shallow and you really have to be careful to make sure nothing shorts out against it when you put the bottom cover back on. I wouldn't hesitate to increase the other caps too. Try using 47uf in place of the 20uf and 22uf in place of the 10uf. Vintage electrolytics had a very wide tolerance range specs so going higher isn't as drastic as it may seem.

Right. I have also heard to leave the Zeners in place. The amp in question is for a good customer. I just want to do it the best way. I take much pride in my work.

Those are not zeners, they are regular diodes. Modern diodes will have less switching noise and be quieter.

 

[gadget73]

yep, standard silicon diodes in a doubler configuration. Pretty common stuff for the era. The uncommon thing about these SC amps is the ground arrangement. Not much consumer gear used anything but the chassis for the ground. It does give them the potential to be very quiet.

The not very good phono stage that ran with AC heaters kind of killed that though. Whenever I put mine back together (yeah yeah, its been like 2 years) the phono stage will be from a Harman-Kardon A500 and will run on DC heat in an effort to make that part of it at least work well. The 333 at least doesn't have that problem. No phono stage, so you can simply run a nice quality external.