I have several hundred Solen caps aquired over the years and in a recent application meaured a number of them to create a matched larger value pair for a woofer x -over with an MIT bypass . A number of the Solens (mid 80's production I believe ) had significantly reduced in value, though all the ones from about 88 onwards were fine. The suspect caps I disposed of so I don't know the dates codes but I seem to remember the end epoxy fill was black on the bad caps , not the early red epoxy or the later grey epoxy. You have to sand the leads to get them to solder decently . I use Rel /MIT RTX and PPFXS in my better crossovers - they are far more resolving and dynamic .
Hello All,
I just received this information from Denis Oullet at Solen regarding the bad batch of capacitors that John Curl had:
> I suspect that these old capacitors were made before late 1987. The French
> factory had changed the aluminum metallized film to a zinc metallized film
> without advising Solen, this was corrected immediately.
>
> Aluminum and zinc metallisation are very thin, like .05 micron and all
> metallized film capacitor are using a zinc metal end sprayed contact to the
> end windings. The thickness of the zinc end spray is ususally about .015"
>
> Zinc metallized film capacitor can only be manufacture in metal can or heavy
> wall plastic can, they cannot be tape wrap because zinc will oxidise in time
> and the capacitor will disappear. It is because oxygen diffuse through all
> thin plastics. This problem will never happen with aluminum metallized film
> as aluminum does not oxidise further in time.
>
> Zinc metallized film capacitor are mainly used in high ac current
> application because they can sustain much higher continous ac current
> application without suffering from electroerosion and the end contact is a
> better zinc to zinc contact.
>
> Aluminum metallized film has better conductivity than zinc, it will make a
> better capacitor with less overall loss in lower ac current application but
> will suffer from electroerosion at high continous ac current and the end
> contact is less good aluminum to zinc contact.
>
> I hope that these explanation are of some use and please do not hesitate to
> contact me if you need further technical information on capacitor.
I trust that this clears up this issue. If you have some very old Solen capacitors it may be worth checking them. (They also used to make some models with an aluminum case, and these should be OK even if they were made with the zinc metallization.) But anything purchased in the last 20 years should be fine.
I just received this information from Denis Oullet at Solen regarding the bad batch of capacitors that John Curl had:
> I suspect that these old capacitors were made before late 1987. The French
> factory had changed the aluminum metallized film to a zinc metallized film
> without advising Solen, this was corrected immediately.
>
> Aluminum and zinc metallisation are very thin, like .05 micron and all
> metallized film capacitor are using a zinc metal end sprayed contact to the
> end windings. The thickness of the zinc end spray is ususally about .015"
>
> Zinc metallized film capacitor can only be manufacture in metal can or heavy
> wall plastic can, they cannot be tape wrap because zinc will oxidise in time
> and the capacitor will disappear. It is because oxygen diffuse through all
> thin plastics. This problem will never happen with aluminum metallized film
> as aluminum does not oxidise further in time.
>
> Zinc metallized film capacitor are mainly used in high ac current
> application because they can sustain much higher continous ac current
> application without suffering from electroerosion and the end contact is a
> better zinc to zinc contact.
>
> Aluminum metallized film has better conductivity than zinc, it will make a
> better capacitor with less overall loss in lower ac current application but
> will suffer from electroerosion at high continous ac current and the end
> contact is less good aluminum to zinc contact.
>
> I hope that these explanation are of some use and please do not hesitate to
> contact me if you need further technical information on capacitor.
I trust that this clears up this issue. If you have some very old Solen capacitors it may be worth checking them. (They also used to make some models with an aluminum case, and these should be OK even if they were made with the zinc metallization.) But anything purchased in the last 20 years should be fine.
Ah ha! Inspector Clouseau has nothing on me.
(Glad I got my post in ahead of Charles's)
My Solens all have gray end caps, although I believe a friend of mine has some with other colors. The question is whether that is a reliable indicator of Zn vs. Al.
I'm also curious to know why their leads are so bloody difficult to solder.
When all is said and done, though, I'm glad to know that the caps are presumed to be trustworthy at this time. If I ever get my finances back on track, I might want to try the all film cap thing I mentioned above.
Grey
P.S.: If any news reports of little old ladies being robbed of their bingo money happen to surface in the next month or so, I disavow any knowledge of said actions. Nope. Not me. No way.
(Glad I got my post in ahead of Charles's)
My Solens all have gray end caps, although I believe a friend of mine has some with other colors. The question is whether that is a reliable indicator of Zn vs. Al.
I'm also curious to know why their leads are so bloody difficult to solder.
When all is said and done, though, I'm glad to know that the caps are presumed to be trustworthy at this time. If I ever get my finances back on track, I might want to try the all film cap thing I mentioned above.
Grey
P.S.: If any news reports of little old ladies being robbed of their bingo money happen to surface in the next month or so, I disavow any knowledge of said actions. Nope. Not me. No way.
Grey, I just put out a warning, because who knows where or WHEN you got your caps? I got them as soon as they were made available. We still have to watch out for any E-Bay bargains on older Solen caps. You never know! 😉
Charles Hansen said:
I guess that you don’t sense the irony of your own post.
I made a relevant point, which has been completely ignored and asked a perfectly legitimate question. What I immediately got in response was mostly unsubstantiated baloney. But such baloney complies with the agenda of most here. What else is new?
MikeBettinger said:
I don’t think that anyone would deny the requirement for a properly rated power supply, but I don’t think that you can make generalisations about ‘power supply sound’ of this sort without any mention of the amplifiers design.
A zero nfb phono amp with common source JFET amplifier stages will be much less immune than to rubbish on the supply rails than one typical one designed around a high performance audio opamp. Ditto for some of the class A SE MOSFET power amp designs floating around here, compared to a D.Self blameless, for example.
john curl said:
I agree. I, like many others, tend to recycle hard to get or expensive parts. One result is that they tend to be older than if I bought entirely new parts for every project. Fortunately, the ones I have aren't old enough to be suspect...at least for that problem. Who knows what will surface a year or two from now?
Grey
P.S.: Isn't it amusing that those who claim to hear no differences claim that it's a matter of expectations on the part of the listener...never once facing the fact that they themselves are guilty of exactly that? Double blind? Hey, no problem! They heard no difference. Why? Because they expected to hear no difference!
Just another reason double blind is a failure.
ticknpop said:
Correct.
The incident with the French Chateauroux caps choked the brand enthousiasm it had gained in a very short time frame for crossovers by their very favorable price rate.
The news of the French SCR manufacturing slip-up spread in the audio community overhere in a similar fashion as the electrolytic pop-sicle mishap of the PC industry a few years ago.
It's likely due to the short span of the mishap, the cooperation with Canadian Solen, and the fill color change that the effect of the incident was limited.
Somewhat surprising that the Chateauroux manufacturing error story didn't stick, but nice to read the tid(t)s and bits after 2 decades.
G.Kleinschmidt said:
Hi Glen,
I believe we're commenting on two different effects. In both cases the supply is properly outfitted with a competently chosen amount of capacity, past that aspect I understand that you're commenting on the circuits ability to ignore garbage on the supply rails, I was highlighting the sonic effect of large amounts of additional supply storage capacitance.
The initial comments that I was responding to presented large capacitor banks as a universal positive performance improvement. My experience doesn't agree with this.
Different? Yes. Audible? Yes. Better? Not to my ears. If I'm chosing parts I opt for smaller caps with good high frequency performance and better transformers, etc; keep the supplies fast.
Mike.
MikeBettinger said:
OK, but don’t you think that any potential for an amplifiers ‘sound’ to be influenced by the power supply topology would have a relationship to the degree that variation of the supply voltage modulates the audio signal?
As an aside, personally, I think that it is obvious that anyone who needs ridiculous amounts of filter capacitance to achieve an acceptable sound quality from their design is playing around with a woefully inadequate circuit.
Cheers,
Glen
hi mike,
isn't the ideal for the rectified psu to approach pure dc as in a battery?
so how do you define a power supply that is "fast"? any guidelines by way of calculations as to how to proceed to build such psu, or any rule of thumb that you can share...
thanks,
tony
isn't the ideal for the rectified psu to approach pure dc as in a battery?
so how do you define a power supply that is "fast"? any guidelines by way of calculations as to how to proceed to build such psu, or any rule of thumb that you can share...
thanks,
tony
MikeBettinger said:
And how, pray tell, is a smaller amount of capacitance 'faster' than a larger amount of the same quality cap?
G.Kleinschmidt said:
That's because you've never tried it.
Assuming that you want the power supply to act as a voltage source rather than a current source, then the ideal would be to have 0 Ohms impedance, right? So what's the function of the power supply capacitance? In part, to serve as a shunt to ground for any AC on the rails...whether it's ripple from the rectifier or signal imposed by the circuitry. Over and over you see people choose the amount of capacitance by setting some arbitrary level of ripple they will tolerate from the power supply--at which point they call it quits without ever once considering the signal put into the rails by the circuit.
(This is part of why MikeBettinger's post makes no sense. By definition, smaller amounts of capacitance allow the signal to modulate the rails more. That's not fast, that's an incremental addition to your distortion. Yet another thing that doesn't show up in standard testing.)
Obviously, we're talking in terms of low frequencies. With that in mind, increased capacitance will be of less benefit in a dedicated high frequency amp in a bi-amped system. That goes along with the observation that increased capacitance improves the bass.
For the last couple of years, I've tended to focus more and more on balanced/bridged designs so as to get the load seen by the power supply as close to DC as possible. I don't know as I'd go so far as to call all single-ended circuits 'woefully inadequate,' but it does place more demands on the power supply in terms of AC performance.
Or you can look at it the way most people do--more capacitance supplies more current on demand for peaks. Since most of the heavy current peaks are in the low end, that's where you want the current to be. Yes, a sudden cymbal clash will spike quite nicely on the scope, but the amount of current required to reproduce that is much smaller than the amount required to reproduce a resounding whack on a bass drum.
Grey
Tony said:
No, because batteries aren't 'fast.' They tend to have higher impedances than a decent AC power supply. They do have a nifty benefit in that there's no ripple or high frequency line noise. That works pretty well for something like a phono stage or line stage where the current draw is low, but I'm not all that taken with the idea of using batteries to drive an amplifier.
And that's not even mentioning all the heavy metals required.
Grey
Tony said:
Hi Tony,
Both AC supplies and batteries have sonic drawbacks. Pure DC is the Holy Grail for amplifier applications.
My comment and use of the term fast comes from an observation 30 years ago when using a light bulb as a current limiter during initial powerup of amplifiers after working on them. When playing music amplifiers I considered to have a better sense of articulation in the low frequencies showed a faster, more responsive current draw as represented by the response of the bulb. Others showed a slowerbuild up and decay in their demand on the line. Hence the faster terminology. Experimenting over the years has shown the difference to be in the effectiveness of the supply to supply power to the circuits demands AND sink the back EMF effectively. Part of this equation is the construction of the electrolytics. I have rules that I follow but I doubt (from broaching this subject in the past) the response will be worth the effort. My suggestions would be physical not anything you could calculate.
GRollins said:
I said a proper amount of capacitance, not smaller. Only smaller relative to amounts that don't make sense. Throw a scope on a supply that has very low ripple and you'll notice the DC varying due to line fluctuations that no amount of filtering will smooth out, look at the AC noise components from the line.
Then as Glen pointed out, start drawing power to play music and watch how the supply gets modulated, makes the ripple look kinda unimportant.
This is why I like symetrical designs. Try running a well built amp on 1000uf and see if you can hear any hum. The outputs will run hotter but it won't be readily apparent in the sound.
I've been through this discussion before and told I'm wrong. OK. 🙂 Save your breath. Like I said before, if people want to experiment with the comments in mind. Then maybe there's something to discuss.
I'm out of time now, life is calling, Mike.
And let's not forget that the coupling capacitance to the mains is zero with batteries -- which I find their most andvantageous feature when most of the current through interconnects is psu leakage currents.
- Klaus
Same quality is the key point... with a single cap, its self-resonant frequency is 1/(2pi*sqrt(LC)), and L won't probably ever be any less than 30nH or so (which is already optimistic)... which gives an intrinisic upper limit of a cap before it goes inductive ("slow"), regardless of technology used. And the bandwith needed is that one of the amplifier/circuit...
- Klaus
i asked because i believe there may be an optimum ratio between transformer impedance versus capacitor impedance, wonder if anybody ever investigated this.......
if we can nail down a quantifiable ratio for so-called "fast" psu, then maybe we can come to common understanding of this phenomenon instead of just arguing about it...
GRollins said:
Presumption after presumption. Sorry, but I have indeed tried it out of necessity - back when I was a teenager building crappy circuits from acquired electronics magazines published in the 70's and 80's.
But then agian, why am I kidding myself here, thinking that basic engineering principles have anything to do with true 'HiFi'.
Also, the adequacy of the power supply is one of the easiest things to objectively test (low frequency tone burst testing into a reactive load, HF slewing / squarewave testing into a load, etc, etc.) You’re out of your depth on that one too.
- Status
- Not open for further replies.