Filter cap value question (in a Dynaco SCA-35 hi-f amp)i

[Wrongdog]

Okay there are other forums on the net to ask this but I figured some one hear could answer this question from a guitar amp design perspective:

What is the best values for filter caps in regards to tone and dynamics?? Why are guitar tube amps and old Hi-fi tube amps generally based on 40/20/20/20 type values and stages? I know Twins and big amps use 80uf etc...

A buddy just got this Dynaco SCA-35 intergrated hi-fi tube amp for free and wanted me to fix hum and put a grounded cord on it. The stock can caps were typical 60/40/20uf and a 50/50uf all at 450v. I was intrigued by Triode Electronics offering of unloaded cap board upgrade so I bought it. When it arrived with all the parts the 5 caps were ALL 200uf/500v. Is that a lot of capacitance? Don't you loose dynamics like pick attack if you over filter a guitar amp?? Why do bass amps tend to use larger values than a guitar amp generally?

Here is the link to the Cap board:

Dynaco SCA-35 Integrated Amplifier

I guess these are highly regarded entry level audiophile amps. I see there are board kits available to build them from scratch. Anyone have one or built one??

[Chuck H]

IME larger filters increase dynamics. Especially on lower frequencies. This isn't automatically good. WRT guitar amps, big dynamic bass capability can lead to wolf notes and woofy tone. The limitations of most guitar preamps, the speakers and even the instrument itself, and especially in combination, are prone to several conflicting resonant points in the bass and low mids. Smaller filters help squelch this effect. And though "dynamic" is often used as a positive description of guitar tone, truely dynamic amp response for guitar actually sounds cold, flat and uninteresting. Some kind of envalope of attack is needed to make the amp an instrument and not just the bland reproduction of what guitar and pickup are putting out. Contemporary overdriven guitar tone has A LOT more bottom end in the final, post overdrive EQ. The current trend seems to me more filtering in the power amp and more careful EQ sculpting in the audio circuit to get this result. Whereas once upon a time, when larger filters simply weren't available, guitar tone was molded by the limitations of the amp as a whole. Call it a happy accident. So , smaller filters being used in guitar amps probably continued (and continues) to be typical because it works as part of the tone/dynamic shaping formula.

JM2C

[Enzo]

The primary function of the filter cap is to smooth the rectified AC, so you have relatively smooth DC for the circuits.

The larger value the cap, the more it costs. So when Leo Fender made an amp, he was not going to spend extra money for a 100uf cap when a 20uf cap would do good enough.

Caps don't work in a vacuum, they are always in some context. Two 50uf caps in parallel is 100uf, two 200uf caps in series is 100uf. Now you could say look, one has four times the capacitance of the other, but in the context of your use, the two are the same - in both cases the amp sees 100uf.

In the old days, a design engineer would apologize for sag in an amplifier if you pointed it out, nowdays people are trying to add some in. The people designing the old stuff were not all about nuance, they were just doing whatever got the job done.

So what is the "best value" cap to use for tone and dynamics? Please define best tone and best dynamics. One guy likes beefsteak, another likes pork ribs. There is no right and wrong here. Then we start to ask what sort of current and voltage will be needed in the circuit. How sensitive to ripple is the circuit. Etc.

[Wrongdog]

Thanks Chuck. I kind of get it. So In this Hi-Fi situation, what do the big caps offer? I am Sure Ted at Triode and Sheldon at SDS know what they are doing, but I am curious about why and will be curious to hear the results when I get it finished.

Enzo - thanks. You posted before I finished this. So 200uf stages in this hi-fi amp is good?

[Chuck H]

As Enzo was good point out. The primary function of the filter capacitors is to smooth the DC that powers the amp. Bigger filters do this better than smaller capacitors. So, aside from any dynamic and tonal limiting advantage lower uf filters have in guitar amps, bigger is better. You DO want dynamics for a hi fi amp. As Enzo also pointed out, the value of the caps does not always tell you the value of the circuit. I didn't see a schematic for the replacement board so I can't say if you actually have 220uf capacitance or not. If you do, IME it's fine.

[Wrongdog]

okay - cool. I just wondered if it was over kill. The Triode listing did not give any details on the values they would send.

Dynaco SCA-35 Capacitor Board w/Parts

I noticed that the SDS site that the parts are listed now so maybe he updated it. These guys sure are into the Dynaco's!

[Wrongdog]

Tofu! yuk. In this case the best tone is to reproduce the album or recorded music accurately I guess. The circuit is extremely simple.

[Enzo]

For hifi, reproduction is the whole job, but not in a guitar amp - fundamentally different beasts.


And never lose sight of mojo selling - it is everywhere. You convince someone he needs to add immense caps to some amp and then offer to sell them, voila - make your own market for something. I can think of claims for "better" low end response by changing some cathode bypass cap. Never mind that the existing cap has a turnover frequency of 10Hz, but somehow moving it down to 5Hz will be "better." As if a guitar ever made 5Hz.

Where did the 200uf figure come from? As I read the product descrip[tion I don;t see mention of values. As I read it I would expect the parts supplied would have the same value as the original circuit multicap it replaces. They mention that one COULD increase the values - there is room for it.

[Chuck H]

If you go to the link and click on the description of the new board, then scroll down, there is a parts list. The value of the caps is 220uf. Nothing to say how it is wired. But as you say, I'm sure it's appropriate for the application.

[Wrongdog]

Chuck H is right - they are 220uf caps and they must be what SDS recommended so Ted at Triode supplied. Yes when I ordered it they were not specified and only after going SDS site did I see it.

I will stick it together and see! I guess lots of folks use these little amps and modify them. Some recommend bypassing the tone/EQ circuit?? From SDS site: "It's a nice sweet sounding unit that is very musical and enjoyable, particularly if the tone controls are taken out of the circuit."

Here is a how-to: Antique Radio Forums • View topic - Dynaco SCA-35 switchable tone control bypass

[Steve Conner]

In my tube hi-fi amp I have 100uF before the filter choke, and 300 after. When restoring it I got a good deal on some 100+100 dual caps that fitted the old can holes.

It has a diode rectifier. A tube one probably wouldn't be happy with that much capacitance.

In a hi-fi amp you want minimal hum and a nice reserve of energy, so more capacitance is better.

[kleuck]

What is the best values for filter caps in regards to tone and dynamics?? Why are guitar tube amps and old Hi-fi tube amps generally based on 40/20/20/20 type values and stages? I know Twins and big amps use 80uf etc...

Mostly because when amps here fitted with tube rectifiers, and each model of rectifier has a limitation when it comes to the first cap (50µf for an EZ81 for example) and obviously, the cost.
As you're talking bout a Hi-Fi amp, if your rectifier is a silicium diodes one, you can use bigger caps, and improve the bass.
A simple enough rule of thumbs is to use at least 1F for 1A, so if your power tubes are drawing 200mA at max power (see the datasheet) 200µf would be fine.

[Wrongdog]

Yes it is SS rectified. Board kit came with new 1N5408GICT diodes.

[Rhodesplyr]

I have a Dynaco SCA-35 and installed this cap board years ago. MAJOR improvement over the stock power supply. The larger capacitors create a stiffer supply and allow this little amp to produce dynamic transients that make it sound like a more powerful amp than it is.

When replacing the capacitors, I also strongly suggest new, ultrafast soft-recovery diodes, which have become common. The diodes Dynaco used in the 1960s tend to be very noisy by modern standards. Clean the selector switches with DeOxit D-100 and/or silver polish.

The SCA-35 can be tweaked to be a great-sounding amp with a few mods. Its output transformers are among Dynaco's best. It's tone-control implementation, however, is among Dynaco's worst, and the phono section can be noisy.

[Wrongdog]

Cool glad to hear this is worthwhile! Is there a favorite tone circuit bypass method you would advise?

I put the final install of this on hold due to a bunch of guitar repairs I need to do first but I look forward to results. The owner of the amp is our area go-to brass and reed horn repair guy and he is really appreciative of the hi-fi tube warmth on a budget. I may build one of these for myself if I like it too!! Would the Edcor trannies be a good modern route to go I wonder??

[km6xz]

The power supply caps are in the current path return to common for the amplifier so a stiff power supply has good a low impedance to ground. So how much is enough? Depends on your desired low frequency response and transit response. That is not as simple a question as it sounds, tastes in music change with the state of the society. When I was growing up, there was no such thing a deep bass in recorded music, we did not expect it and did not hear it but we always knew exactly where the kick and bass were driving. We heard the second and higher order harmonics from our 6X9 oval car speakers and portable 45 record players so our brains filled in the sensation of the missing fundamental. Having a limited spectrum to deal with with records, AM radio, hi-fi systems, did not diminish the impact of those classic songs that still sound great today even after being remixed and bottom enhanced artificially. So consider the time in which the Dynakits were available. Amplifying anything that was not in the source material was a bad thing, because it was only noise, so limiting low frequency response, while not a goal usually, no one was concerned that the amps where not flat from DC to Daylight.
For modern recordings which have a lot of signal energy that is outside the range of fundamental notes of the instruments playing, extended response in amps and speakers, at orders of magnitude more power needed to reproduce subsonics, are all part of what is in vogue now. You will have to beef up the supply, for that, and many other changes to handle the deep subsonics convincingly. Lowering the power supply impedance at 20-30 hz will mean a lot more capacitance. It also means pretty high inrush currents to tame to prevent diodes, transformers and primary fuses from self-destructing.

Concerts that want 136db levels in 1960 did it with 100 watts, but nothing was expected through the speakers below about 70 hz. The cost of such a system was a couple thousands dollars when gear was more expensive per level of performance than today when power and low distortion are dirt cheap. Even with the tremendous drop in price of gear, 136db flat down to where people house mixers what to emphasis , 20-30hz is 100-500 times more investment to get that 1.5 octave lower. Even guitar players are seeking more bottom which which presents problems with speakers(some are even open baffle cabinets..and expect LF) and amplifiers to be pretty in distortion from the fundamental notes and all their harmonics and up, when mixing them with subsonics. There are a lot of reasons it is hard to get good tone now, when it was actually easier when a guitar rig only needed to be effective from 100 hz and up.
Did anyone NOT get the beat down pat on first hearing of all those great RB records, or Motown, or Stax, Muscle Shoals recordings? The instruments did not share the same sonic space or spectrum and each stood out. Now, when going to a gig in a club, it seems as if each instrument is trying to suck all the spectrum out of the room. When bass rigs, lead guitar and keyboards are running enough bottom emphasis to overtake the kick drum, something is really screwed up the tone sculpting and arrangements. It also makes vocals unintelligible. Excessive volume is the crutch used by less confident players, those who fear or know they have nothing to say.
So the answer to your question is, as usual, Depends. If you want to reproduce modern digital recordings, you are going to need really stiff low z power supplies at the lowest frequency of concern. But it is a waste if your speakers are not really effective at those frequencies. That presents the worst situation, feeding a signal with high levels of subsonics, into an amp that is needing 90% of its power to reproduce signal frequencies that the speakers can reproduce anyway....the result is both the amp and speaker are being bombarded with energy that is modulating all the reproducible frequencies and making them suck. Full range systems don't make any sense. Use your little SCA-35 for a range of frequencies it is good at, with speakers that are good over the same range and remove all the subsonics from the signal path and route that to the 1000 watt amp and gigantic speakers needed to reproduce the subsonic portion at the same level of perceived level. Or, just listen to old recordings. Or wait around another 20years and a new fade in sound will be all the rage.

[Steve Conner]

Well back in the good old days (before I was born ) the bass lines were deliberately distorted to add more harmonics, so you could make them out on those portable radios and so on. I think Phil Spector even had two bassists playing the same line in fifths as part of his "wall of sound" concept.

And even then, the electric bass guitar doesn't have much fundamental energy, compared to a bass synthesizer. The output is mostly harmonics, and classic bass amps played along with that. The speakers didn't have full output down at low E (42Hz) and the tube amp distorted a bit and emphasized the harmonics even more.

That is the recipe for the classic Motown and rock bass sound, and reasonably sized affordable bass rigs still go by it out of necessity. (I was rather fond of my giant 2x15" cab though )

Nowadays, what is somewhat annoying is digital recordings with subsonics that make your woofers flap uselessly (inaudibly) in and out. They're usually electronic dance type music. The producer obviously used a low-shelf EQ to boost the bass, forgetting that it boosts everything down to DC. It is a sheer waste of power, especially if you're listening on a tube amp.

On the bright side, a classic Dynaco style amp could make your Ipod sound almost bearable

And if you're listening to digital recordings, you can just shove them through a digital EQ like the excellent Behringer one and set it to filter out the subsonic crap.

[km6xz]

They were not deliberately distorted, the results were due to every step of the way from board through tape to disc cutter, the medium, playback, thin paper cone speakers driven by 2 watt amps, or a single ended 6AQ5, 6V6 or loctal tube in a car radio with 12 volts on the plates, or AM broadcast FCC limitations etc. All the old classic sounded better than the Spector wall of crud sound. No one liked the sound even than but they liked some of the songs he produced.
Even after that time when independent studios sprang up with more interest in sound quality, we used to spend many hours sitting in cars in the parking lot listening to basics, roughs or mixes on radios or cassette decks. It was pretty standard to have a small AM and a FM transmitter in the studio to hear how kids where going to hear it out in the lot. FM required a bit straighter mixes but even the album rock stations had a lot of compression in the chain that messed us..
Yeah, subsonics in pa systems and recording chains is pure amateur, it ruins the repro for most people, and costs more for tickets in concerts due to having to have such monster systems even for mid level bands. It also has closed off most venues to live concerts due to structural damage. When Paul McCartney played here last in Palace Square, a million in damage was done to the Hermitage from sound. The building survived 900 days of shelling in WWII by the Axis but not a rock/pop concert several hundred meters away. The Stones were kept at much lower levels and everyone there who had seen other Stones concerts said it was the best sounding one they experienced. Mick actually sounded pretty good, which was the biggest shock. If the Stones can rock a crowd at only 134 db why a pop act needs 146 at <30hz?

[Wrongdog]

Steve - how hot does your PT get? I know these run hot but I am talking like 145-150 degrees. New quad of JJ's are not red plating but maybe I should put another 10-20 ohms on the bias resistors to get this running a bit cooler?? They are now 190 ohm. I definitely will tell my buddy to keep it elevated on a hard surface and nothing above it.

Did adding the choke remove much hum? I have a weak 7199 (damn!) and may have to convert it or buy converters 6gh8s. Those 7199s are like gold $$.

[Chuck H]

They were not deliberately distorted

So, just a happy accident since it worked out for the sound.

[Steve Conner]

The PTs in my amp run too hot to touch after a few hours. That is the downside of big filter caps, they push up the RMS current in the transformer windings for a given DC output.

I had 200uF before the filter chokes and 200 after, but I moved one of the capacitor sections to see if it might cool things down any.

[km6xz]

Actually it was not an accident, it was just the reality of the defective technology of the day and it trained a whole generation of what sounds good because that sound was associated with songs they liked. Sound preferences are learned and they become the definition of what later sounds are measured against. Sort of like how any lover is judged by the reference of what you learned with. Objectively the "classics" of rock and roll were not good, drove parents crazy but for kids it became the definition of proper sound. The kids grew up seeking the same goal.
The distortion that some active elements induce happens to fit natural perception of sounds in the real world where even harmonics are dominate, but it still distortion.

[Bill Moore]

I really learned a lot from this thread! My old PA (built 30 years ago), has 18's capable of 35hz, but we also kept the 40hz rumble filter on the mixer activated. I know I'm an old fart, and I like the sound of my EV's, but wasn't sure why I was so annoyed by the DJ's systems that the kids use today. Didn't realise there was so much low freq emphasis. (No wonder I can't tell what the guy is saying!)

[Rhodesplyr]

Cool glad to hear this is worthwhile! Is there a favorite tone circuit bypass method you would advise?

I should have responded sooner, but with the SCA-35, the easiest approach is simply to eliminate the tone controls completely, going straight from the volume control to the 7199s. You can also drop the 7199 input impedance to a more reasonable level by reducing the value of the grid-leak resistor. In stock form, I think it's something like 4.7Meg as a way of preventing additional signal loss after the lossy tone controls. (The SCA-35 has no real line stage, which is where tone controls generally operate.) You can drop the 7199 grid resistor to 470k or lower.

If you're going to build one of these from scratch, I'd build an integrated amp version of the Dynaco ST-35. The ST-35 uses the same transformers, but has an all-triode driver stage instead of the 7199. Most people think that the ST-35 circuit sounds better, and it frees you from having to find a decent 7199. We are scraping the bottom of the barrel these days in terms of NOS 7199s.

[nedhogan]

Actually it was not an accident, it was just the reality of the defective technology of the day and it trained a whole generation of what sounds good because that sound was associated with songs they liked. Sound preferences are learned and they become the definition of what later sounds are measured against. Sort of like how any lover is judged by the reference of what you learned with. Objectively the "classics" of rock and roll were not good, drove parents crazy but for kids it became the definition of proper sound. The kids grew up seeking the same goal.
The distortion that some active elements induce happens to fit natural perception of sounds in the real world where even harmonics are dominate, but it still distortion.

Strange because I know a lot of people who aren't old farts that like distortion, including myself, and also I've never liked sociologists! Seems like a dull view to have of the world, that great sounding music is just great because people are conditioned to like it. Don't dis the best decades of history. You have said yourself that its hard to make something sound good without it. It might have been the limitation of technology in the day that made those great sounds but it was also the height of technology... you are right that distortion mimics real life because anything sounds dead without it. It feels loud and real when it isn't, I think is the main psychological aspect to this whole distortion thesis.

Im very sorry for my counter rant. Im just very "for" the idea that people in the 60's found an opportunity to make good sounds and taken advantage of that. If they wanted to play clean, they could have played more quietly rather than inventing fuzz pedals to put on their 100w guitar amps! I think your being very subjective in your objectivity.

[km6xz]

But were they good sounds? That is pretty subjective and based on learning during formative years. You might have a different opinion if born in the era of the big jazz bands when a whole society was immersed in a different sonic environment. A lot of kids later than the 50s and 60s where surrounded in rock of their parents day so acquired the taste for it, that the grandparents probably did not. Art of all sorts are a product of their era since it was made in the context of the current society. Some people were exposed to early tastes in art and keep an interest in it but few people who did not have early exposure care about past styles at all. I deal with thousands of visitors every summer and most want to visit the Hermitage/Winter Palace and its fabulous art collection. Most say they want to see the Rembrandt rooms and French Impressionists so those rooms are very crowded yet that leaves millions of equally significant works from other periods and schools unseen. Why? Why do visitors gravitate to two schools of art that are not dramatically at a different level than everything else? They were not raised in an art rich environment so only know vaguely those two names and assume the "famous ones" are better and should be treated with more respect. Societies where art is more integrated into daily life would have a broader base from which compare, contrast or simply experience.
I do not see any difference in music after working with it for decades, I find that there is less exposure to different genres than 30 years ago because of the delivery medium has become so personal. A person is only exposed to what they already like because they can do their own programming of their play lists and never have to be exposed to something outside of their zone of familiarity. That was the case before radio opened things up a lot in the 50s and 60s with stations playing a wide variety of music to the same audiences. That was the heyday of Top 40 which as a term was later considered degrading. The reality is that Top 40 was remarkably wide open. The Beatles where competing directly on the same stations with rock, jazz, country, folk, novelty and classical and had each of those types beat them out at various times. Like Al Hurt, Mrs Miller, Serendipity Singers, Jan and Dean, all the Motown artists who made R&B comfortable for white suburban kids, Dean Martin, Mitch Miller, Leslie Gore, Dick Dale, The 4 Seasons, Dave Brubeck, Louie Armstrong, Ramsey Lewis, Herb Alpert, Roger Miller, The Dixie Cups, The Ray Charles Singers and many more who all slugged it out for the same wide market and all had success. What ever was popular was played. Later when FM became dominate and stations began to narrow playlists listeners became much more isolated into genres. Before there were only a couple genres and most met in the middle.
Bill Graham of the SF dance hall fame tried, pretty successfully to expand exposure in his concert halls in the late 60s by booking obscure artists that the kids had never heard as filler between the 4-10 other bands playing on any random night. So between sets by Blue Cheer and the Dead, might be a 80 year old Delta Blues player or Ravi Shankar playing classical Indian music before the Beatles "discovered it" or African ethnic rhythmic groups or Highlife bands. Being exposed to it opened a lot of young people to a wider world of music. That trend did not last long and now the music spectrum is highly fragmented with isolated pockets of interest. After living through that I find it hard to understand why you think tastes are not taught by early exposure.
If you think that is dull, you are living a very sheltered life....there are whole realms invisible to you that you would have preferred if you had experienced it at a younger age.
Personally, I like all sorts of music but do not follow any particular style, prefer live bad bands than great recordings....I spent too much time behind the glass to be in awe of it. Most of the music I deal with now, as a listener, from the vantage point of the dance floor is dance music in clubs plus original jazz, and classical, opera etc. I am lucky to live in such a arts oriented city where all sorts of music, ballet, rock, dance, art and literature are integrated into daily life of most people from a young age. As a result the kids have to flip a coin to narrow their choice between seeing a new ballet staging or go see a rock concert on the same night.
Not liking sociologists? How many do you know?

[philbo]

Okay there are other forums on the net to ask this but I figured some one hear could answer this question from a guitar amp design perspective:

What is the best values for filter caps in regards to tone and dynamics?? Why are guitar tube amps and old Hi-fi tube amps generally based on 40/20/20/20 type values and stages? I know Twins and big amps use 80uf etc...

A buddy just got this Dynaco SCA-35 intergrated hi-fi tube amp for free and wanted me to fix hum and put a grounded cord on it. The stock can caps were typical 60/40/20uf and a 50/50uf all at 450v. I was intrigued by Triode Electronics offering of unloaded cap board upgrade so I bought it. When it arrived with all the parts the 5 caps were ALL 200uf/500v. Is that a lot of capacitance? Don't you loose dynamics like pick attack if you over filter a guitar amp?? Why do bass amps tend to use larger values than a guitar amp generally?

Here is the link to the Cap board:

Dynaco SCA-35 Integrated Amplifier

I guess these are highly regarded entry level audiophile amps. I see there are board kits available to build them from scratch. Anyone have one or built one??

General rules of thumb for power supply filter caps:
Larger capacitance than the original is ok. Larger voltage than the original is also ok.

The devil is in the details... Two other capacitor specs can come into play here.

1) ESR (equivalent series resistance) describes the capacitors internal resistance, which affects the charge and discharge time constants. The lower the ESR the better for high fidelity amps. For instrument amps, it could affect the way the power sags after a large transient. If sag is desirable, you should try to match both the ESR and capacitance of the original part.

2) Maximum RMS current (also called Ripple Current) - - this value is determined from the ampacity (current carrying ability) of the internal connections within the cap. ESR also plays a part in this, because the internal heat generated by the part is determined by the amount of RMS AC current being shunted to ground and the ESR of the part (power in heat = current squared x resistance). In general, a larger value for Max RMS current is better.

Hope you find this helpful.

[Hanger-18]

General rules of thumb for power supply filter caps:
Larger capacitance than the original is ok. Larger voltage than the original is also ok.

The devil is in the details... Two other capacitor specs can come into play here.

1) ESR (equivalent series resistance) describes the capacitors internal resistance, which affects the charge and discharge time constants. The lower the ESR the better for high fidelity amps. For instrument amps, it could affect the way the power sags after a large transient. If sag is desirable, you should try to match both the ESR and capacitance of the original part.

2) Maximum RMS current (also called Ripple Current) - - this value is determined from the ampacity (current carrying ability) of the internal connections within the cap. ESR also plays a part in this, because the internal heat generated by the part is determined by the amount of RMS AC current being shunted to ground and the ESR of the part (power in heat = current squared x resistance). In general, a larger value for Max RMS current is better.

Hope you find this helpful.

Very Useful information... I've been rebuilding a Fender Twin Reverb '75 that's just a chassis and replaced the Master Volume (broke back) with Old Stock and repared the Vibrato with a new "roach" and 50k ra pot. Decided to clean up sound starting with filter caps and was suprized to find Higher values 200 uf instead of the 100 uf's and 100 uf's in place of the three 20uf's. I also found one to the phase inverter broken contact. Soldered it back but these are bp brand electrolytics made in Japan. The 200uf's are in parallel=100uf, but what and why will happen with this and who would do this?????

[philbo]

Very Useful information... I've been rebuilding a Fender Twin Reverb '75 that's just a chassis and replaced the Master Volume (broke back) with Old Stock and repared the Vibrato with a new "roach" and 50k ra pot. Decided to clean up sound starting with filter caps and was suprized to find Higher values 200 uf instead of the 100 uf's and 100 uf's in place of the three 20uf's. I also found one to the phase inverter broken contact. Soldered it back but these are bp brand electrolytics made in Japan. The 200uf's are in parallel=100uf, but what and why will happen with this and who would do this?????

Two 200 uF caps in parallel give a total of 400 uF (the values add directly in parallel, different than resistors & coils). It's probably just a sloppy rebuild job....

[Wrongdog]

What is the voltages of caps in parallel rule? Two 200uf 350v caps parallel is now 400uf at 175v, right?? In series 100uf and 700v???

[philbo]

What is the voltages of caps in parallel rule? Two 200uf 350v caps parallel is now 400uf at 175v, right?? In series 100uf and 700v???

The voltage rating of the parallel combination will be equal to the smallest voltage rating of all the paralleled caps. In this case, if they're both 350V, then the rating would be 350V.

In series (NOT recommended, and usually provides no benefit) the voltage divides up in inverse proportion to the values of the caps. So, if you put a 10 uF and a 100 uF in series and apply 550VDC, the 10 uF cap will drop 500V and the 100 uF cap will drop 50V. That is assuming the individual voltage ratings of the caps will allow this without frying them.
The total Capacitance of that series pair can be found by using the product over sum: (100 uF x 10 uF)/(100 uF + 10 uF) = 9.09 uF

[Wrongdog]

Okay - thanks!!

[Chuck H]

In series (NOT recommended, and usually provides no benefit)

I beg to differ. In an ideal world we would just use a single, appropriately rated cap. But WRT tube amps it can be difficult and or expensive to find and use a single cap. Suppose you have a B+ of 500V with an off load voltage (that will be presented to the caps) of 575V. Ok, I need to use 600V caps... Oh darn. No one makes them!?!

Maybe you could find some. But they would cost too much, may be the wrong size or orientation and you would be limited by whatever qualities (or lack of) they possessed. A totem (series) arrangement solves for this.

Even many 500V rated caps have slipping reputations for quality. The best caps stop at 450V. So should all tube amps be made with less than 400Vp (considering off load voltage that the main filters should see and the inrush spike)??? That doesn't seem practical. In fact, a great many mass produced amps with a B+ over 400 that use a single main filter cap have had on and off trouble with premature cap failure. Component manufacturers just aren't giving much respect to electro's rated over 450V. A series arrangement solves for this too.

Just sayin'

[Enzo]

Series caps are extremely common. One generally uses two of the same value anyway. But even if they are odd value, you put parallel equalizing resistors in the circuit to even out the voltage. A couple of 350v caps in series each with a 220k in parallel very neatly shares the vcoltage equally and acts like a 700v cap.

[philbo]

Series caps are extremely common. One generally uses two of the same value anyway. But even if they are odd value, you put parallel equalizing resistors in the circuit to even out the voltage. A couple of 350v caps in series each with a 220k in parallel very neatly shares the vcoltage equally and acts like a 700v cap.

You are right, it can be done. And, again, you are right (generally speaking) about the equalizing resistors. No arguments there.
But in my experience (working professionally in electronics design since 1974) there is only rarely a need to do so. I have yet to see a schematic, aside from a 40KW radio transmitter power amp stage, that called for 2 caps in series.

BTW (for Chuck): Here are the 60000+ capacitors that are 600V and higher from just one vendor. If you trim the selection to just film capacitors >600V, you still get around 35000 to pick from.

I'm really not trying to be obnoxious; I'm just posting my take on the subject. So, feel free to take it with a grain of salt, or whatever. To be fair, if you are fixing something and HAVE to get it done today, and HAVE to use what you have in stock, then I could see an argument for hacking a couple series caps together. But as far as not being available.... well, the answer's in the link above.