Ah missed that part.
I stay by my opinion that more testing would need to be done, how many amp builder tried different capacitor configurations for the powersupply, including big foil caps. Not many
Design of power supplies (along with PS bypassing) is pretty well-documented. Lots of good resources online.
The ideal power supply would maintain a constant voltage at all times during amplifier operation, regardless of input signal and output load. No more, no less. Running your amp off two Prius batteries in series would get pretty close to this, but there are a lot of reasons I don't recommend this.
The characteristics of all these parts is reasonably well documented, so the circuit analysis to determine the impact of using one type of cap versus another becomes relatively straightforward. If you don't want to do it by hand (or write the python code), you can model all of these nonidealities quite accurately with LTSpice. Those parameters that are inconveniently left out of the datasheet aren't too tricky to measure.
A sanity check monitoring the power supply rail with a scope (do be mindful of probe voltage rating here...) typically reveals any "bad behavior".
In what sense? The high side of the OPT is (ideally) sitting at a constant DC potential. In an ideal world no signal voltage would appear at either node of that cap. Current? Sure, but that's the case with any other bypass cap. That cap can be completely eliminated and replaced with a 400 V battery and the amp will still work.
As for whether I'd use an electrolytic for a coupling cap? If necessary, sure. It's done constantly in solid-state electronics, and if done right doesn't have any real impact on the sound of the amplifier. Care needs to be taken to ensure that the nonlinear characteristics and leakage currents aren't an issue, but in many cases it works just fine.
As for whether I'd use an electrolytic for a coupling cap? If necessary, sure. It's done constantly in solid-state electronics, and if done right doesn't have any real impact on the sound of the amplifier. Care needs to be taken to ensure that the nonlinear characteristics and leakage currents aren't an issue, but in many cases it works just fine.
So I wrote an Email to Steve from Decware and to Allen Wright's Vacuum State, asking for their opinions regarding electrolytic caps vs foil.
Vacuum State are actually in my country only 3 hours by car away from me. They replied to me in german, I will do my best to translate it but will attach a screenshot of the original email.
I've asked Decware this question ""Are there any circumstances where electrolytic capacitors are preferable to film capacitors in valve amplifier circuits?"
Steve from Decware replied: "Many will say the foil caps perform better, but despite the size and cost -- either one would be a detriment, the sound is also different. Sometimes things sound better with electrolytics."
I asked Vacuum State a more detailled questiion "Does the capacitor in an rc-filter have a great impact on the sound?
Does the price of a foil capacitor in this function influence the sound quality or is it irrelevant?
What capacitors do you use for the power supply of your 300B PP amp?"
Thomas Brüntrup, long term partner of Allen Wright and technician replied:
"Sound: yes, in a low-pass filter, maybe less than in a high-pass filter (signal path).
Materials: Naturally, there are justified differences in price. Glass wrapped in Teflon or silver foil in a vacuumed glass case, is naturally (extremely) expensive but definitely sonically better than PVC in a castplastic case.
But without overdoing it, polypropylene foil (PP) or even better, polyphenylene sulphide (PPS), is good enough for starters.
Then there are the so called HiFi High-End Caps. Here, the price/performance is often a matter of luck; Mundorf can surely be trusted. Otherwise, dig through the relevant internet forums to find what could be good. However, there is a certain common consensus.
300B Cs: PP capacitors also exist for higher currents (motor capacitors - MP). Relatively new on the market are PP for solar applications. Rated for up to 800VDC, they are relatively small and square-shaped. They exist from all the big manufacturers (Panasonic, Epcos, etc.). "
Vacuum State are actually in my country only 3 hours by car away from me. They replied to me in german, I will do my best to translate it but will attach a screenshot of the original email.
I've asked Decware this question ""Are there any circumstances where electrolytic capacitors are preferable to film capacitors in valve amplifier circuits?"
Steve from Decware replied: "Many will say the foil caps perform better, but despite the size and cost -- either one would be a detriment, the sound is also different. Sometimes things sound better with electrolytics."
I asked Vacuum State a more detailled questiion "Does the capacitor in an rc-filter have a great impact on the sound?
Does the price of a foil capacitor in this function influence the sound quality or is it irrelevant?
What capacitors do you use for the power supply of your 300B PP amp?"
Thomas Brüntrup, long term partner of Allen Wright and technician replied:
"Sound: yes, in a low-pass filter, maybe less than in a high-pass filter (signal path).
Materials: Naturally, there are justified differences in price. Glass wrapped in Teflon or silver foil in a vacuumed glass case, is naturally (extremely) expensive but definitely sonically better than PVC in a castplastic case.
But without overdoing it, polypropylene foil (PP) or even better, polyphenylene sulphide (PPS), is good enough for starters.
Then there are the so called HiFi High-End Caps. Here, the price/performance is often a matter of luck; Mundorf can surely be trusted. Otherwise, dig through the relevant internet forums to find what could be good. However, there is a certain common consensus.
300B Cs: PP capacitors also exist for higher currents (motor capacitors - MP). Relatively new on the market are PP for solar applications. Rated for up to 800VDC, they are relatively small and square-shaped. They exist from all the big manufacturers (Panasonic, Epcos, etc.). "
Attachments
The Kemet C4AQ line of metallized polypropylene are a fraction of the cost, a much smaller package and sound great in a 6LU8 pentode SE on the bench. Replaced ASC oils.
Incorrect. 130uF 500V Digikey price top, 120uF 400V Solen price bottom.
The Kemet's tighter lead spacing and smaller form factor help keep circuit layouts tighter, beneficial attributes for high impedance tube circuits.
Who were the 'we' in your listening panel?
I've experimented with metallized PP film power supply filter capacitors, and one thing I'd point out is that there's a difference between designing and building an amp with them in mind and trying to retrofit them into an existing amp. I've tried the latter, even building a deeper base for the amp to accommodate them. You can run into problems with the size of film caps, which makes it necessary to mount them a greater distance from the circuit they are filtering than is ideal because you have to connect them via long wire harnesses. I've actually seen people make amps more noisy by hanging huge capacitors all over the place.
It is also the case that electrolytic capacitors have been getting better, partly in response to the demands of modern power supplies. I have an impedance bridge, and I'm finding some electrolytic caps I can now test on the same DF scale as film caps. These are, literally, not your grandfather's FP multisection cans. Take a look at the specs of some electrolytics by United Chemi-con and TDK/Epcos.
It is also the case that electrolytic capacitors have been getting better, partly in response to the demands of modern power supplies. I have an impedance bridge, and I'm finding some electrolytic caps I can now test on the same DF scale as film caps. These are, literally, not your grandfather's FP multisection cans. Take a look at the specs of some electrolytics by United Chemi-con and TDK/Epcos.
How does that happen? It is not possible to have a change in capacitor current without a change in the voltage across the cap.
But are those small ones really suited?
There is a noticeable difference in size and it's not that the space is unused, the Solen has probably more material inside providing better dielectric and better insulation. Again, here we are going into debate "BUT DO THEY AFFECT SOUND?????" but considering you said that they sound good, I am just gonna assume that you indeed believe that different foil capacitors can sound different.
True - Ic = cdv/dt. The key here is that dv/dt is multiplied by c, so when c is large and Ic is modest, dv/dt becomes small. Technically for what I said to be true, the output cap would have to have infinite capacitance. In practice, we size those caps such that the change in voltage is small enough to not worry about. There are quite a few different ways to approach the calculations, but ultimately the physics is all the same.
With a reasonably sized cap the change in power supply voltage becomes pretty tiny, but that might still be of concern due to the poor PSRR of a typical SE amplifier.
If I were to build a single-ended power amplifier (or really any other circuit with poor PSRR) I would seriously consider the use of a regulated power supply. The alternative option is to just massively increase energy storage through the use of huge capacitors and chokes, but that's big, heavy and expensive.
If anything, the poor PSRR of a single-ended amplifier would be an argument against the use of film caps for bulk filtering since their energy density is far lower than that of electrolytic caps.
One cap type that I have come to quite like for various projects are motor run caps. These are oil-filled, typically between 10 and 60 uF, and are excellent quality. The ones I've tested had pretty reasonable parasitic inductance, and they seem to take abuse quite well.
This is what I use - DC Link C4AQ Kemet caps. Vishay DC Links are also good. I tested them in a direct A-B listening test against Kemet 550v electrolytics, common motor run polypropylene caps and Clarity Caps. The blue plastic motor runs were the worst, the Clarity Caps OK but not special, the electrolytics were good but the DC Link caps were smoother and more natural. So I ended up using Kemet electrolytics for the main caps and a C4AQ for the last cap in the chain. I've re-tested various alternatives in a few amps, and this still sounds the best arrangement to me. I tried both the Kemet electrolytics and the C4AQ as cathode bypasses for my 2a3 outputs with the same results - the DC Links were smoother and more natural. DC Link caps are really excellent.
You'd need to do a subjective double-dollar listening test to confirm that - similar to an objective double-blind listening test, but easier to measure.
For some reason poly caps don't work for me in PS. Not even in preamps. And despite huge chokes and regulation. The resulting sound is not something i can live with long term. Or to put it another way, i haven't found a replacement in power supplies for the old black Cerafines.
Some low bass weight is lost, and also some depth, no matter how much film capacitance i add. No doubt the sound is faster and cleaner and with less overhang but for this to be an improvement, the entire system needs to be revoiced: others passives, cables, etc. If i did not have the Cerafines, film might be a justifiable choice as most electrolytics are really horrible. A friend with a different preference for electrolytics ( a Siemens fan) also could not switch over to films. Must be system specific as CJ have been manufacturing preamps with film only PS for a very long time and customers seem happy.