Hi all - wrt Sikorels hfo in Poland carry some of these and some other NOS capacitors, e.g. Roederstein:
http://stores.ebay.de/hfoelektronik/?_dmd=2&_nkw=sikorel
About their sound - and please remember that this is in my setup - I noted from my listening tests: "A dynamic sound. Insisting and musical, yet somewhat "massive" in its tonality. Upper midrange has a focus and is slightly unnuanced. Upfront perspective. Altogether musically sounding, yet somewhat tonally unnuanced."
Cheers,
Jesper
http://stores.ebay.de/hfoelektronik/?_dmd=2&_nkw=sikorel
About their sound - and please remember that this is in my setup - I noted from my listening tests: "A dynamic sound. Insisting and musical, yet somewhat "massive" in its tonality. Upper midrange has a focus and is slightly unnuanced. Upfront perspective. Altogether musically sounding, yet somewhat tonally unnuanced."
Cheers,
Jesper
Last edited by a moderator:
About those caps in Poland - the axial ones - these are the T125 higher temp caps made for automotive use and aren't so good for sound, even after a good couple hundred hours soaking - the "Rolls Royce" of the Siemens electros were the B51550/51570 screw terminal caps - big expensive things - they/ve been superceded for a few years and not sure if they make any long life 105*C in 'snapin' or radials - the closest thing would possibly be the B41231, 41252 or the B41505 range - the numbers describe the caps and unfortunately, they make a huge range of industrial products that make it rather confusing
When Siemens became Epcos, I thought we might have access to lower sized electros, even under the Matsushita (?) label, but it wasn't to be
I think the SlitFoil BHCs (made by Supertec these days, I understand - more confusion!) are possibly one of 'the best' electro caps for bringing out the midrange capabilities of a power supply but are expensive, have a long breakin and fussy about where they are used in a circuit - they've also changed labels and are getting new numbers
One of the surprising things I found with the SuperThru Nichicons is how well they functioned in the ripple cap position (directly after the diode bridge) - I've used the 6,800uF/80v ones with excellent results particularly if followed by the Elna for Audio 12,000/63v (Thailand) for the F5 amp (25W classA) and the 10,000uF/50V BHC SlitFoil for the F6 (used from =/>150Hz) - I think the old Siemens B41550 are better all round but are just so darn big they're hard to fit
I understand that the Rubycon ZL have a rather good 'sound performance' if you have the patience to wait for them to 'come good' - I recently ordered some (max 1000uF) to try them out and some of the above mentioned Nichicon KZNs - interesting.
I find the BC electros in the C-R-C configuration of a Class A amp's & preamp supply to exhibit a 'loose' bass a bit like the old Roe caps - it doesn't help that there are some of these appearing on eBay that are relabelled rubbish - the Nippon Chem KMH 105 10,000uF/63 (61M04L) is also a bit of a disappointing cap, IMO.
When Siemens became Epcos, I thought we might have access to lower sized electros, even under the Matsushita (?) label, but it wasn't to be
I think the SlitFoil BHCs (made by Supertec these days, I understand - more confusion!) are possibly one of 'the best' electro caps for bringing out the midrange capabilities of a power supply but are expensive, have a long breakin and fussy about where they are used in a circuit - they've also changed labels and are getting new numbers
One of the surprising things I found with the SuperThru Nichicons is how well they functioned in the ripple cap position (directly after the diode bridge) - I've used the 6,800uF/80v ones with excellent results particularly if followed by the Elna for Audio 12,000/63v (Thailand) for the F5 amp (25W classA) and the 10,000uF/50V BHC SlitFoil for the F6 (used from =/>150Hz) - I think the old Siemens B41550 are better all round but are just so darn big they're hard to fit
I understand that the Rubycon ZL have a rather good 'sound performance' if you have the patience to wait for them to 'come good' - I recently ordered some (max 1000uF) to try them out and some of the above mentioned Nichicon KZNs - interesting.
I find the BC electros in the C-R-C configuration of a Class A amp's & preamp supply to exhibit a 'loose' bass a bit like the old Roe caps - it doesn't help that there are some of these appearing on eBay that are relabelled rubbish - the Nippon Chem KMH 105 10,000uF/63 (61M04L) is also a bit of a disappointing cap, IMO.
Feedback & systematic approach to subjective capacitor evaluations?
@carlmart: Some days ago I think you asked me for a schematic of the setup I use for conditioning capacitors. I have now drawn it up ... the sine generator really is an obsolete micro system I bought for this purpose.
Regarding capacitor distortion getting lower as a result of time, when there is a bias across the capacitor, Douglas Self has actually confirmed this in measurement (Linear Audio article). The distortion drops somewhat over time.
@sumotan: ... very, very interesting that there may now be a sort of "common reference" to speak from. As it is I also have some Super Throughs, and IMHO they are actually somewhat different than the Sikorels (3300uF/16VDC, Sikorel 470uF/35VDC - in my setup the actual capacitance shouldn't matter that much). I would say that the Super Throughs are slightly "frail" and a little thin tonally, with a slightly pronounced lower to upper midrange, slightly lightweight sounding, and, yes, not as macro dynamically impressive as the Sikorels. Yet, the Sikorels to my ears are less apt at macro dynamic to micro dynamic shadings. Also the Super Throughs have a spatial rendering which to my memory gives a good focus on the front side of the individual instruments, but overall the spatial rendering IMHO is less integrated and holographic. My preference actually is the B grade KG - less "remarkable", yet IMHO more tonally neutral.
@Monte McGuire: Some days ago I remember asking you if you could outline the setup you use for measuring capacitor distortion (a schematic, if possible, would be helpful). Not to be insisting about this
... but I have used a setup that apparently looks similar to yours, however, my distortion levels have been much higher for electrolytics - similar to the levels Cyril Bateman measured. So it would be clarifying to find out if I have made a systematic error somewhere ...
@johnego: ... You mention Vloss as a parameter. Can I ask you what this is? I don't remember having seen this before - maybe it is the leakage current?
Systematic approach to subjective capacitor evaluations
What I mean with this headline is that I'd like to describe the systematic approach I (would like to) have in testing capacitors. As it is I have been looking into this for some time now and my current thoughts about a systematic approach for assessing the sound characteristics of capacitors in a circuitry is this:
1. Measure the impedance and distortion profile of individual capacitors.
2. Listen to the individual capacitors to get an impression of how they sound.
3. Simulate the capacitor combination to be used, if possible, in a simulator (LTSpice, Kemet K-sim, etc.). Observe non-linear impedance/ESR curves and if possible correct.
4. Measure the impedance curves for the actual capacitor setup - preferably in-circuit (especially with digital circuits) - with something like Analog Discovery:
Analog Discovery by Digilent
5. Measure the actual noise spectrum of the capacitors in the working circuitry with an oscilloscope and/or a spectrum analyzer. Use several frequencies - maybe even music - to see how well-controlled the noise spectrum is in practice.
6. ... Underways ...: Listen to the circuitry and replace capacitors to optimize circuitry and sound.
7. ... Try to gain experience with how measurements may predict sound (everything else being reasonably equal) ...
I realize this is comprehensive but somehow my intuition is that in the end such a systematic approach may actually save time. However, in my personal setup I am missing the Agilent Discovery - and I guess the ability to measure the actual impedance curves high up in frequency may tell something about how it will actually sound. Anyone knows of a less costly yet equally high resolution and bandwidth impedance analyzer as an alternative to the Agilent Discovery?
Cheers,
Jesper
@carlmart: Some days ago I think you asked me for a schematic of the setup I use for conditioning capacitors. I have now drawn it up ... the sine generator really is an obsolete micro system I bought for this purpose.
Regarding capacitor distortion getting lower as a result of time, when there is a bias across the capacitor, Douglas Self has actually confirmed this in measurement (Linear Audio article). The distortion drops somewhat over time.
@sumotan: ... very, very interesting that there may now be a sort of "common reference" to speak from. As it is I also have some Super Throughs, and IMHO they are actually somewhat different than the Sikorels (3300uF/16VDC, Sikorel 470uF/35VDC - in my setup the actual capacitance shouldn't matter that much). I would say that the Super Throughs are slightly "frail" and a little thin tonally, with a slightly pronounced lower to upper midrange, slightly lightweight sounding, and, yes, not as macro dynamically impressive as the Sikorels. Yet, the Sikorels to my ears are less apt at macro dynamic to micro dynamic shadings. Also the Super Throughs have a spatial rendering which to my memory gives a good focus on the front side of the individual instruments, but overall the spatial rendering IMHO is less integrated and holographic. My preference actually is the B grade KG - less "remarkable", yet IMHO more tonally neutral.
@Monte McGuire: Some days ago I remember asking you if you could outline the setup you use for measuring capacitor distortion (a schematic, if possible, would be helpful). Not to be insisting about this
... but I have used a setup that apparently looks similar to yours, however, my distortion levels have been much higher for electrolytics - similar to the levels Cyril Bateman measured. So it would be clarifying to find out if I have made a systematic error somewhere ... @johnego: ... You mention Vloss as a parameter. Can I ask you what this is? I don't remember having seen this before - maybe it is the leakage current?
Systematic approach to subjective capacitor evaluations
What I mean with this headline is that I'd like to describe the systematic approach I (would like to) have in testing capacitors. As it is I have been looking into this for some time now and my current thoughts about a systematic approach for assessing the sound characteristics of capacitors in a circuitry is this:
1. Measure the impedance and distortion profile of individual capacitors.
2. Listen to the individual capacitors to get an impression of how they sound.
3. Simulate the capacitor combination to be used, if possible, in a simulator (LTSpice, Kemet K-sim, etc.). Observe non-linear impedance/ESR curves and if possible correct.
4. Measure the impedance curves for the actual capacitor setup - preferably in-circuit (especially with digital circuits) - with something like Analog Discovery:
Analog Discovery by Digilent
5. Measure the actual noise spectrum of the capacitors in the working circuitry with an oscilloscope and/or a spectrum analyzer. Use several frequencies - maybe even music - to see how well-controlled the noise spectrum is in practice.
6. ... Underways ...: Listen to the circuitry and replace capacitors to optimize circuitry and sound.
7. ... Try to gain experience with how measurements may predict sound (everything else being reasonably equal) ...
I realize this is comprehensive but somehow my intuition is that in the end such a systematic approach may actually save time. However, in my personal setup I am missing the Agilent Discovery - and I guess the ability to measure the actual impedance curves high up in frequency may tell something about how it will actually sound. Anyone knows of a less costly yet equally high resolution and bandwidth impedance analyzer as an alternative to the Agilent Discovery?
Cheers,
Jesper
Attachments
Last edited by a moderator:
Sure. The specs say lousy 'ESR' so perhaps they are essentially useless as HF decoupling but IMHO that is not the whole story. Look deeper into the materials used is the key....I have not yet tried them myself but with prior knowledge in hand yes I expect that that their mere presence across an otherwise reasonably 'perfect' power supply will cause audible changes......for the ultimate betterment or not I am unable to say because I have not yet tried the experiment personally. OTOH I have experimented extensively with BQP, and a 'similar' device of my own devising and the result is that system sound can be changed at will and predictably.
The takeaway from my hints is that every element and every compound has it's own 'sound', and each of these sounds interact...some play 'nice' and some don't as is the subjective experiences related here on this thread. I have experimented extensively with different battery types, and my conclusion is that lead is to be strenuously avoided....inside the battery and also solder joints...lead free solders of which there are a whole bunch of variants and IME they all sound (subtly) different and beneficially different.
Dan.
Last edited by a moderator:
Hi Jesper
Hmm aside from the slight forward focus my impression are opposite to yours,they never sound thin or frail & they have great layering of instruments. Like Max mentioned perhaps its circuit dependent. On power amps these are my findings. I tried them on analog ps on my dac & that didn’t work out too well. Sounded slightly blotted & slow but when I move the caps over to ps of digital side, it sound really good,dare I say analog sounding & the forwardness isn’t there anymore, still it sounded a little slower but I tweak other parts of the dac to compensate .
Hmm aside from the slight forward focus my impression are opposite to yours,they never sound thin or frail & they have great layering of instruments. Like Max mentioned perhaps its circuit dependent. On power amps these are my findings. I tried them on analog ps on my dac & that didn’t work out too well. Sounded slightly blotted & slow but when I move the caps over to ps of digital side, it sound really good,dare I say analog sounding & the forwardness isn’t there anymore, still it sounded a little slower but I tweak other parts of the dac to compensate .
Last edited by a moderator:
@Max Headroom:
@jameshillj: About Rubycon ZL ... I use them extensively and I do find that they have some unusual qualities. One might say that the sound is "there" with the Rubycons (size 2200uF/25 VDC). Kind of "dark" in their tonal quality and slightly overall massive but good spatial rendering and possibly low "self-noise". I wouldn't use them on their own, though - and I am phasing them out in the actual PSUs close to sound circuitries.
Cheers,
Jesper
My goodness ... 3-4 years ago I tried different lead-free solders and since many of them caused the soldering tips on my Weller soldering iron to degrade (some of them) virtually in hours I decided to let go of lead free solder. But I would VERY MUCH appreciate a tip on a good lead-free solder that flows well and doesn't ruin the soldering tips
@jameshillj: About Rubycon ZL ... I use them extensively and I do find that they have some unusual qualities. One might say that the sound is "there" with the Rubycons (size 2200uF/25 VDC). Kind of "dark" in their tonal quality and slightly overall massive but good spatial rendering and possibly low "self-noise". I wouldn't use them on their own, though - and I am phasing them out in the actual PSUs close to sound circuitries.
Cheers,
Jesper
Last edited by a moderator:
Hmmmffff......all of this experimentation yields so many variants......I am happy to regard these results/opinions as metadata, valuable metadata so keep it coming. From experience I find that fine changes at the end of the reproduction chain yield the greatest overall subjective changes, capacitors upstream set a system behaviour that is heard downstream and final downstream (crossover) capacitors are a final determinate.
Dan.
Last edited by a moderator:
Thanks Karl,
Not too sure if I ordered the right ones now! - they just arrived - they're a light blue/grey colour and have Audio KA(M) on them and part number UKA1H332MHD (3,300uF/50V) and UKA1H222MHD (2,200uF/50)
Hi 'sumotan' - love your posts...
I got some SlitFoils awhile back and wanted some more but HiFi Collective doesn't keep them - I was going to ask Denis Moorecroft direct but I figured he's got better things to do than send a couple of caps off in the mail - I do like them for their midrange potential and it allows for a lot of other optimising without sacrificing that 'intelligibility' that's so hard to get without struggling with 'edge' like we intentionally do in PA systems Ah Jesper, I've gone and got some now, damnit! 1000uF, 470uF @ 35V and 220uF @ 50v - it looks like I have some more homework to do here!
This thread is REALLY interesting - thanks guys.
Not too sure if I ordered the right ones now! - they just arrived - they're a light blue/grey colour and have Audio KA(M) on them and part number UKA1H332MHD (3,300uF/50V) and UKA1H222MHD (2,200uF/50)
Hi 'sumotan' - love your posts...
I got some SlitFoils awhile back and wanted some more but HiFi Collective doesn't keep them - I was going to ask Denis Moorecroft direct but I figured he's got better things to do than send a couple of caps off in the mail - I do like them for their midrange potential and it allows for a lot of other optimising without sacrificing that 'intelligibility' that's so hard to get without struggling with 'edge' like we intentionally do in PA systems Ah Jesper, I've gone and got some now, damnit! 1000uF, 470uF @ 35V and 220uF @ 50v - it looks like I have some more homework to do here!
This thread is REALLY interesting - thanks guys.
Last edited by a moderator:
At the risk of seeming like a 'real nutter', I'd thought I might mention something I'm playing with that isn't directly involved with "caps sound' but does produce some startling changes in the power supplies I've used for headphone amps and my SOtM server+HDD.A few years ago, a guy called John Brown (?) of EC Designs was deep in proving up some investigations of using different power supplies in the old 1541A dac chip and those things are all about 'cap sound' Anyway, (chatterbox here!) he wrote about a strange idea and called it "Charge Transfer' power supply - I've not figured out a way to add photos here but the circuit is post #2579 of the "Building the Ultimate NOS dac using the 1541A" thread
It actually makes quite a difference to the sound but I've added this to regulators, C multipliers, etc so not sure just how much can be attributed to the Charge Trandfer technique itself (I'm attempting to scale it up to use on the FirstWatt F6 amp - but this does belong in a power supply thread so ...) Maybe someone might like to try this out ...
It actually makes quite a difference to the sound but I've added this to regulators, C multipliers, etc so not sure just how much can be attributed to the Charge Trandfer technique itself (I'm attempting to scale it up to use on the FirstWatt F6 amp - but this does belong in a power supply thread so ...) Maybe someone might like to try this out ...
Last edited by a moderator:
John Brown is a talented guy. Btw James try contacting this shop THL Audio.
It’s in Taiwan, I bought all my goodies Black Gate etc as well as the BHC Slit Foils from them way back. Perhaps they might still some in stock. I don’t recall Dennis Morecroft had anything to do with the Slit Foil design but only on his T Network caps. Oh well.
It’s in Taiwan, I bought all my goodies Black Gate etc as well as the BHC Slit Foils from them way back. Perhaps they might still some in stock. I don’t recall Dennis Morecroft had anything to do with the Slit Foil design but only on his T Network caps. Oh well.