You need both dampings: the transformer and bridge diodes.
And using an R + C snubber is essential for the diodes is important, as the resistor will damp the diode rectification noise pinpointed by the capacitor.
But I am not sure if you can do that with Quasimodo.
BTW: you need an oscilloscope for using the Quasimodo. I do not have one.
Well I recapped my Topping Tp20 mk2 (Tripath amp) this week. I tried a wide range of Panasonic, Nichicon, and Elna. Ended up combining Cerafine and Silmic II.
To my ears, Silmic has the treble shimmer and phat bass, and Cerafine has the forward midrange. On their own, I felt the colorations were wrong for this amp. Together, there is a balancing act with some serious mojo going on
To my ears, Silmic has the treble shimmer and phat bass, and Cerafine has the forward midrange. On their own, I felt the colorations were wrong for this amp. Together, there is a balancing act with some serious mojo going on
Attachments
Great that it work out for you Stell. There’s just too many combis, too many varieties of caps to try & the” most excruciating part is that what works on one circuit does not mean it will do just as well on another. I’ve got some Red Cerafines but never like it’s sound slow & too forward sounding, I do wonder if the black ones sounds different. There was this following back when from tube se guys who worshipping the Cerafines, perhaps it gel well with dht tube stuff.
Hi all,
Just briefly a couple of comments ...
@jameshillj:
Cyril Bateman measured distortion as a function of bias voltage and the Black Gate FK basically were the ones that coped best with higher bias voltages (To my memory 12 V bias was ok with a 50 V rated capacitor).
Bipolars typically have lower distortion levels likely due to their design which - as I understand him - basically is different from a polarized capacitor.
Other capacitors typically could have a (generally speaking here) very high rise in distortion when going e.g from zero volt bias to 18 VDC bias (50 volt types). Not unusual with a 20 dB rise. Second order distortion would be lowest (again generally speaking - with some difference between capacitor types) around 1/12 the specified voltage rating. Again distortion could be much lower at these low bias voltages. There's also been an assumption that if one biased the capacitors at half their rated voltage this would be ok distortion-wise. However, at these bias voltages all the measured capacitors had high distortion levels.
@Monte McGuire: Hi Monte. Thanks for looking into the Cyril Bateman/your measurement practices. I actually didn't notice this while reading the Bateman article but it sounds plausible what you are suggesting - that there's a significant difference in drive levels and thus in distortion levels as well.
Hmmm... Yes & somehow it is interesting that even at these low currents - and hence likely similarly low distortion levels for the capacitors - it is still (to my ears) very, very clear which capacitors are used in a circuitry and which sound "better" in different ways ... Sometimes I think that somehow the measurement methods do not show what is audibly relevant but then that is a different topic ...
I personally have not experimented much with CCS circuitries and other PSRR circuitries so I have only limited experience here ... Have you some listening experience with the degree that PSU component characteristics "couple through" into the sound of the circuitry in spite of the PSRR being high?
Cheers & thanks again for your feedback - good to have it clarified
Jesper
Just briefly a couple of comments ...
@jameshillj:
Cyril Bateman measured distortion as a function of bias voltage and the Black Gate FK basically were the ones that coped best with higher bias voltages (To my memory 12 V bias was ok with a 50 V rated capacitor).
Bipolars typically have lower distortion levels likely due to their design which - as I understand him - basically is different from a polarized capacitor.
Other capacitors typically could have a (generally speaking here) very high rise in distortion when going e.g from zero volt bias to 18 VDC bias (50 volt types). Not unusual with a 20 dB rise. Second order distortion would be lowest (again generally speaking - with some difference between capacitor types) around 1/12 the specified voltage rating. Again distortion could be much lower at these low bias voltages. There's also been an assumption that if one biased the capacitors at half their rated voltage this would be ok distortion-wise. However, at these bias voltages all the measured capacitors had high distortion levels.
@Monte McGuire: Hi Monte. Thanks for looking into the Cyril Bateman/your measurement practices. I actually didn't notice this while reading the Bateman article but it sounds plausible what you are suggesting - that there's a significant difference in drive levels and thus in distortion levels as well.
Hmmm... Yes & somehow it is interesting that even at these low currents - and hence likely similarly low distortion levels for the capacitors - it is still (to my ears) very, very clear which capacitors are used in a circuitry and which sound "better" in different ways ... Sometimes I think that somehow the measurement methods do not show what is audibly relevant but then that is a different topic ...
I personally have not experimented much with CCS circuitries and other PSRR circuitries so I have only limited experience here ... Have you some listening experience with the degree that PSU component characteristics "couple through" into the sound of the circuitry in spite of the PSRR being high?
Cheers & thanks again for your feedback - good to have it clarified
Jesper
Higher voltage caps have lower ESR, but bipolar caps are usually not designed for the lowest ESR.
Rubycon ZL, from an experiment linked by earlier poster, shows the worst subjective performance. But its ESR is very low, so when lowest ESR is what we want (such as stabilizing op-amp and in certain position in power supply) then it can be used.
The treble and bass subjective performance of Silmic II mentioned in this thread is I think in accordance whit ELNA claims regarding the silk behavior. The first time I heard a red cerafine, I didn't think that the sound was the way I wanted, but I thought that the sound was supposed to be that way and that to 'fix' the problem I need to look at somewhere else. May be it is the same case with Silmic II? Accept how it is and adjust somewhere else?
When ELNA produced the silk product line, I believe they copied from Black Gate. Because the great BG-N, having silk material to separate the foils, is a NONPOLAR, then may be it is a good idea to use the Silmic, back to back, for signal coupling. 2x 10uF (and small voltage rating) will gives 5uF suitable value for input amplifier coupling. Compare pos-neg-neg-pos with neg-pos-pos-neg.
@sumotan: Hey ... ;-)
Black Gate Capacitors | Hifi Collective
Cheers,
Jesper
... Have you by any chance compared these with the BG Std & PK types:
Black Gate Capacitors | Hifi Collective
Cheers,
Jesper
Hi Jesper,
You ask the right person. Lol I have been a long time BG guy since I started diy. Those were the days when I was was listening to a Jadis Defy 7 amp so we were caught up with that warmish sound that really didn’t have much highs.
BG caps Std caps are like that but far more visceral & layered. Yes if you want the sound to sound bigger use PK, it’s esr is very high something like 4-5 ohms which means it’s slow you can hear it once installed. I ocassionally play with some 0.47 uf PK as by pass but the slowness put me off but the highs are very sweet & nice sounding. Maybe if we parallel a couple of them together it will sound pretty good. I do use my collection of BG for my build but only 1 or 2 on any circuit just to bring out it’s special quality.
Yes you should purchase some from Hi Fi collective before they’re gone.Do keep in mind that they aren’t that transparent sounding hence only a sprinkle to be used.
You ask the right person. Lol I have been a long time BG guy since I started diy. Those were the days when I was was listening to a Jadis Defy 7 amp so we were caught up with that warmish sound that really didn’t have much highs.
BG caps Std caps are like that but far more visceral & layered. Yes if you want the sound to sound bigger use PK, it’s esr is very high something like 4-5 ohms which means it’s slow you can hear it once installed. I ocassionally play with some 0.47 uf PK as by pass but the slowness put me off but the highs are very sweet & nice sounding. Maybe if we parallel a couple of them together it will sound pretty good. I do use my collection of BG for my build but only 1 or 2 on any circuit just to bring out it’s special quality.
Yes you should purchase some from Hi Fi collective before they’re gone.Do keep in mind that they aren’t that transparent sounding hence only a sprinkle to be used.
Of course the ESR of two series resistance is doubled (conceptually bipolar caps are built this way) as paralleling will halve them. But such capacitor is fine for signal coupling (I think it is intended for this purpose anyway).
But, how did you come to know that?
Because Im a compulsive tweaker Johnego.
Only BG N works well in Super E config for signal coupling, the rest of the caps don’t do well including bipolars. There’s a bipolar cap that does fanatstic as signal coupling & dare I say even better than foil caps like Mundorf Supreme but its my little secret
Only BG N works well in Super E config for signal coupling, the rest of the caps don’t do well including bipolars. There’s a bipolar cap that does fanatstic as signal coupling & dare I say even better than foil caps like Mundorf Supreme but its my little secret
Hi Guys,
It isn't the ESR that's important at all. You won't get anywhere with an ESR meter. What you need to be looking at is the dissipation of the capacitor. There are good LCR meters that will read that, and of course you can make a test jig that uses an oscilloscope and generator as well. There is another test I'm trying that is a little different and uses a square wave as a signal.
So far, my HP 4263A LCR meter has done a great job of sorting out capacitors. It will also indicate ESR (impedance) at various frequencies. The impedance reading isn't that helpful to be honest, but the dissipation readings correlate with what I'm hearing a lot better than anything else so far.
Some of the old impedance bridges will measure "Q", which is the inverse of dissipation. That means that some of you don't have a money barrier preventing you from using a proper instrument for this. Some will also allow you to supply the excitation with an external signal generator. That means you can test the parts at audio frequencies. So, why does this measurement matter? Because, it is a measure of how much energy is lost in the dielectric. That energy comes straight out of the signal, meaning distortion.
ESR meters would only be of value for main filter capacitors at best, but then just looking at the ripple with an oscilloscope will tell you a lot more, and you look at the waveform in normal operation. That tells you a lot more than an ESR meter does. You should all have a working oscilloscope. You absolutely do require an oscilloscope when working even on the smallest power amplifier circuit. Older analog 'scopes are better for what you're doing than the cheap digital 'scopes. If you can score a 100 MHz, dual trace 'scope you will have almost everything you'll ever need in an oscilloscope.
Analogue or digital? I use both on my bench. Some things an analogue does better at (eye patterns for one), and others where a decent DSO would be the better choice. If you can only have one, make it an analog scope.
-Chris
It isn't the ESR that's important at all. You won't get anywhere with an ESR meter. What you need to be looking at is the dissipation of the capacitor. There are good LCR meters that will read that, and of course you can make a test jig that uses an oscilloscope and generator as well. There is another test I'm trying that is a little different and uses a square wave as a signal.
So far, my HP 4263A LCR meter has done a great job of sorting out capacitors. It will also indicate ESR (impedance) at various frequencies. The impedance reading isn't that helpful to be honest, but the dissipation readings correlate with what I'm hearing a lot better than anything else so far.
Some of the old impedance bridges will measure "Q", which is the inverse of dissipation. That means that some of you don't have a money barrier preventing you from using a proper instrument for this. Some will also allow you to supply the excitation with an external signal generator. That means you can test the parts at audio frequencies. So, why does this measurement matter? Because, it is a measure of how much energy is lost in the dielectric. That energy comes straight out of the signal, meaning distortion.
ESR meters would only be of value for main filter capacitors at best, but then just looking at the ripple with an oscilloscope will tell you a lot more, and you look at the waveform in normal operation. That tells you a lot more than an ESR meter does. You should all have a working oscilloscope. You absolutely do require an oscilloscope when working even on the smallest power amplifier circuit. Older analog 'scopes are better for what you're doing than the cheap digital 'scopes. If you can score a 100 MHz, dual trace 'scope you will have almost everything you'll ever need in an oscilloscope.
Analogue or digital? I use both on my bench. Some things an analogue does better at (eye patterns for one), and others where a decent DSO would be the better choice. If you can only have one, make it an analog scope.
-Chris
It's Nichicon, isn't it? Yes I have many secrets too. The problem when you share the finding (if it is a NOS parts) is the price will go up
Very expensive in Brazil. And I do not build that many things to justify spending too much money.
I might have to get someone to test things for me.
Are there any affordable 'scopes that can be used with a computer?
I'm just smart enough to know someone has already thought about any of my 'so-called' secrets ... you see, I'm getting smarter as I get older (and I'd really appreciate it if nobody corrected my delusions ...!)
Thanks guys for the info about the voltage rating/performance of caps - I remember going thru this some years ago with the big Rifa caps in the stud welding machines and the much higher rating caps did function better under heavy use - they still used to boil a bit but didn't burn out as much! and my first experience with the huge Semicron diode bridges - interesting.
Hi sumotan,
I've had these Nichicon KA caps in the supply to my headamp for a few more days now and they're quite surprisingly more transparent and yet smoother than the familiar KZs (and either the Fine Golds or the SuperThrough) - the treble sounds a bit like a softer polyester treble sound - they don't quite have that 'snap' of the KZs but still seem to present that sense of rhythm - interesting things
They do seem to lack the familiar solid bass so I put a BP (reversed the leads with +ve to 'ground' - thanks for that) on the input to the K-Multiplier and that balanced up the final sound to my liking - it's not exactly what you'd call a flat response but I do like a bit of a kick in the low bass without it blurring up the low mids like the KZs seem to do
I've been tempted a bit to get some of those new Kisei (?) caps from HiFi Collective to see how they sound - anybody tried them?
Thanks guys for the info about the voltage rating/performance of caps - I remember going thru this some years ago with the big Rifa caps in the stud welding machines and the much higher rating caps did function better under heavy use - they still used to boil a bit but didn't burn out as much! and my first experience with the huge Semicron diode bridges - interesting.
Hi sumotan,
I've had these Nichicon KA caps in the supply to my headamp for a few more days now and they're quite surprisingly more transparent and yet smoother than the familiar KZs (and either the Fine Golds or the SuperThrough) - the treble sounds a bit like a softer polyester treble sound - they don't quite have that 'snap' of the KZs but still seem to present that sense of rhythm - interesting things
They do seem to lack the familiar solid bass so I put a BP (reversed the leads with +ve to 'ground' - thanks for that) on the input to the K-Multiplier and that balanced up the final sound to my liking - it's not exactly what you'd call a flat response but I do like a bit of a kick in the low bass without it blurring up the low mids like the KZs seem to do
I've been tempted a bit to get some of those new Kisei (?) caps from HiFi Collective to see how they sound - anybody tried them?
Yes, as have been mentioned I simply choose capacitors for power supply based on datasheet Dissipation Factor (Tangent Alpha). Muse KZ and Silmic has around 0.08 for 50V rating at 120hz. The ESR (preferably impedance) cannot be too high though (if for PS). If Tangent Alpha is equal then look at the impedance. For example, Panasonic FC and FM both have tan alpha of 0.10 but FM has much lower impedance so is more preferred from my pov.
I don't think LCR meter can read the DF as it should be with rated voltage (say 50V) which a meter don't have. But with that cheap meter we can measure VLOSS which is useful too.
Of course there are better tools with better accuracy. I can get the sound I want if I have the measurement tools (it is more a two dimensional parameter than a single number). Or I can simply listen to the cap. But I don't think those are urgent or necessary as there are always more critical weak points in the system that can be improved economically and efficiently.
Hi James so the KA treble isn’t that snappy ??? What did you hear after flipping the BP cap ? I was playing with my amp the whole day yestersday using some BP’s crazy thing was cannot connect all with + to ground cause you get too much mids to highs & bass went missing, so ended up using some with + connecting as per norm but very interesting was were to use norm connection & where to use + to ground, that made a huge difference to get to my liking. Now amp sounds so so open & yet have great bass & some warmth which I like.