Solen Fast Capacitors, from the Solen.ca website:
GENERAL INFORMATION
Type : Metallized Polypropylene Film Capacitor.
Dielectric : High Purity Polypropylene Film.
Construction : Round Tubular "Square Aspect Ratio" Type, Axial Leads.
Coating : Black Plastic Tape Wrapped, Grey Epoxy Resin Sealed.
Electrodes : Aluminium Metallized, Vacuum Deposited.
Winding : Bifilar Extended Metallized Film.
Contact : Non-Inductive, Zinc Thermal Arc Sprayed on Extended Film.
Leads : Tinned Plated Oxygen Free Pure Copper.
Looks as though the plates are aluminum, though the "contacts" are zinc. Aluminum is fairly stable, chemically, under normal conditions, but zinc is pretty reactive. However, if the zinc is only in the area where the leads are attached to the electrodes (which is what I'm assuming they mean by contact), then if the zinc reacted with something (oxygen, chlorine), I would think that it would increase the ESR, not lower the capacitance. Hmmm. I'll think on it some more.
Now, 10uF...that starts sounding reasonable to do in polystyrene, which would be a win/win situation if that small a capacitance would do the trick. For a preamp, 10uF is very reasonable. For an amp...that I'll have to try. Once I get another bank of outputs on each amp, I'll be in a position to test this. Unfortunately, that will be a while. I've got the output devices, but hit the wall financially and won't be able to lay hands on sufficient heatsinking for quite some time. Bummer.
I've not had any failures of Solens and that includes nearly 1000uF per channel in my tube amps, meaning they run at something like 575V. I have had some that were quite hard to solder. For the record all of my Solens have come directly from Solen in Canada, so I doubt there's a counterfeit problem in my case.
Grey
GENERAL INFORMATION
Type : Metallized Polypropylene Film Capacitor.
Dielectric : High Purity Polypropylene Film.
Construction : Round Tubular "Square Aspect Ratio" Type, Axial Leads.
Coating : Black Plastic Tape Wrapped, Grey Epoxy Resin Sealed.
Electrodes : Aluminium Metallized, Vacuum Deposited.
Winding : Bifilar Extended Metallized Film.
Contact : Non-Inductive, Zinc Thermal Arc Sprayed on Extended Film.
Leads : Tinned Plated Oxygen Free Pure Copper.
Looks as though the plates are aluminum, though the "contacts" are zinc. Aluminum is fairly stable, chemically, under normal conditions, but zinc is pretty reactive. However, if the zinc is only in the area where the leads are attached to the electrodes (which is what I'm assuming they mean by contact), then if the zinc reacted with something (oxygen, chlorine), I would think that it would increase the ESR, not lower the capacitance. Hmmm. I'll think on it some more.
Now, 10uF...that starts sounding reasonable to do in polystyrene, which would be a win/win situation if that small a capacitance would do the trick. For a preamp, 10uF is very reasonable. For an amp...that I'll have to try. Once I get another bank of outputs on each amp, I'll be in a position to test this. Unfortunately, that will be a while. I've got the output devices, but hit the wall financially and won't be able to lay hands on sufficient heatsinking for quite some time. Bummer.
I've not had any failures of Solens and that includes nearly 1000uF per channel in my tube amps, meaning they run at something like 575V. I have had some that were quite hard to solder. For the record all of my Solens have come directly from Solen in Canada, so I doubt there's a counterfeit problem in my case.
Grey
They are not fake. It took 10 years in storage at normal temperature and humidity to make the change, but they DID change.
Unless I am very mistaken Mundorf once had a range of MKP caps whith a zinc alloy as the metal film. They were even stating why they were using this alloy but I don't remember the details.
I just happen to have 4 82uF 400V Solen MKPs awaiting use in a tube pre supply in the very near future. I beleive they are 10-15 years old. They served several years in a X-over earlier in life. They may have seen storage temps at least 125 degrees (My Az Garage). So naturally after reading this I had to give them a look...
I have a simple (handheld) Wavetek Cap meter. They are all 79-81uF 😀 Well within their 5% value rating...
I have a simple (handheld) Wavetek Cap meter. They are all 79-81uF 😀 Well within their 5% value rating...
The Solen caps are designed by Denis Oullet, who operates Solen in Canada. Solen caters to small manufacturers and DIY speaker builders, selling drivers and crossover parts. He saw a gap in the market for a reasonably priced polypropylene capacitor that could be used for loudspeaker applications and introduced his own brand of capacitors. They were originally made for him in France, but I believe that he later purchased the machinery to make them in Canada (not 100% sure of this). He also makes excellent air-core coils, both solid and Litz wire configurations.
I don't think there is anything unusual about the construction of these capacitors. As far as I know, they use aluminum for the metallization layer. They, like all plastic film capacitors that I am aware of, are terminated with a "flame-spray" process whereby a composite of zinc, bismuth, and other metals joins all of the windings so that a lead wire can be soldered to it. I have visited the Rel-Cap factory and seen this process in action. If you look at the Wima catalog, they call this "schoopage", which is apparently untranslatable from the German!
I know Denis personally, and he makes a quality product. John, if you have had a problem with a batch of capacitors (and based on the other postings, this seems to be a limited experience), I would suggest contacting Denis directly. I am sure that he would like to perform a proper failure analysis to prevent any future problems and that he would be glad to replace any defective capacitors.
I don't think there is anything unusual about the construction of these capacitors. As far as I know, they use aluminum for the metallization layer. They, like all plastic film capacitors that I am aware of, are terminated with a "flame-spray" process whereby a composite of zinc, bismuth, and other metals joins all of the windings so that a lead wire can be soldered to it. I have visited the Rel-Cap factory and seen this process in action. If you look at the Wima catalog, they call this "schoopage", which is apparently untranslatable from the German!
I know Denis personally, and he makes a quality product. John, if you have had a problem with a batch of capacitors (and based on the other postings, this seems to be a limited experience), I would suggest contacting Denis directly. I am sure that he would like to perform a proper failure analysis to prevent any future problems and that he would be glad to replace any defective capacitors.
I have also used the Solens for many years without problems. When I was at Avalon we used them in several different applications. One was to make a composite 1600 uF PP capacitor! All of the parts were measured and sorted, and we never experienced any problems like that.
On the other hand we once had some Rel-Caps that were mismarked. They were labeled as 0.001 uF but turned out to be 0.01 uF. This led to a funny situation at the CES about 10 years ago. Our prototype 96 kHz DAC had a -3 dB point of 9 kHz instead of 90 kHz in the analog reconstruction filter!
Nothing is perfect, and I'm sure it is possible that John received a defective batch of capacitors. I think that the Rel-Cap PS caps outperform the Solens, but at a significantly higher cost. I have found the Solens to provide excellent value for the money.
On the other hand we once had some Rel-Caps that were mismarked. They were labeled as 0.001 uF but turned out to be 0.01 uF. This led to a funny situation at the CES about 10 years ago. Our prototype 96 kHz DAC had a -3 dB point of 9 kHz instead of 90 kHz in the analog reconstruction filter!
Nothing is perfect, and I'm sure it is possible that John received a defective batch of capacitors. I think that the Rel-Cap PS caps outperform the Solens, but at a significantly higher cost. I have found the Solens to provide excellent value for the money.
My particular units have a very good finished appearance (black outer and grey epoxy), and are model/part# PB-MKP-FC. Also printed on them is "Made in France". I have had a handful of other sizes in the past from the same source and have not detected the issue under discussion. I'm happy with them so far 😀
I use Mundorf also when need be... I'll keep this in the memory bank somewhere. Thanks for the heads up though Mr. Curl 😉
I use Mundorf also when need be... I'll keep this in the memory bank somewhere. Thanks for the heads up though Mr. Curl 😉
I do not wish to be unfair to a manufacturer, and the parts that I did get probably were made in France. However, it was a shock to me when they did change value. It may have been a bad batch, and not repeated. I am glad that it is an isolated incident.
Note that I haven't measured mine. I just noted that I haven't had any outright failures to date. I'm still mulling over possible failure modes that might lead to the capacitance reduction that John noted.
Grey
Grey
Early parts did fail, I have had independent confirmation, but that is probably not true today. That is the best that I can say.
Don't get me wrong, I'd love to do a massive (relatively speaking) bank of polystyrene caps, but the economics of the thing tend to weigh against it. Next best is polypropylene with polystyrene bypasses. Solen is the best dollar per uF candidate that I'm aware of. I like the WIMA polypropylene caps, but they are pretty sparsely distributed--or at least if there's someone here in the US who carries a wide assortment, I've yet to find them.
The RTX polystyrenes have done very well for me, but I realize that I'm behind the times. There are scads of other caps out there now that I haven't worked with. John, Charles, do you have any input on either of two categories:
--Reasonably priced caps, probably polypropylene, such as the Solen or WIMA (and where to find them)
--Top of the line caps to serve as bypasses or in places where smaller values are indicated?
Grey
The RTX polystyrenes have done very well for me, but I realize that I'm behind the times. There are scads of other caps out there now that I haven't worked with. John, Charles, do you have any input on either of two categories:
--Reasonably priced caps, probably polypropylene, such as the Solen or WIMA (and where to find them)
--Top of the line caps to serve as bypasses or in places where smaller values are indicated?
Grey
Grey, I have settled on Rel RT .1 uf styrene caps for most bypassing. However, for some amp designs, it might be better to use a few uf of film, just to catch the risetime portion of the current after a fuse or circuit breaker.
Yes, I think John is reporting some problems from early days.
I do have one dead 50uf Solen from those times but from there on I have had no problems with these caps, I own several dozens of Solens.
Later caps, the ones marked as True Caps are quite good. As coupling caps I bypass them with 0.1uf usually Wimas or others, but generally avoid them for this task.
I do have one dead 50uf Solen from those times but from there on I have had no problems with these caps, I own several dozens of Solens.
Later caps, the ones marked as True Caps are quite good. As coupling caps I bypass them with 0.1uf usually Wimas or others, but generally avoid them for this task.
Charles Hansen said:
This is a very curious failure mode for a film cap. I just ran out to the Calibration Lab and put a resistor decade in series with our 10uF MPP capacitor standard (10.00uF D=0.0012, ESR=0.02 ohms). I added series resistance with the decade and measured C and D with our HP4263A, at 120Hz 50mV, 120 Hz 1V and 1kHz 1V. The 1kHz 1V measurements are below (the other 2 ranges are similar):
R = 0 C=10.00 D=0.007
R = 1 C=10.00 D=0.069
R =10 C=9.99 D=0.622
R=100 C=9.86 D=Out of Range
Even with really large values of ESR the capacitance shift is less than 2%.
It's almost as if the construction was so defective that the cap somehow became two series caps, reducing the value to
CT = C1 × C2/( C1 + C2 )
Best Regards, Chuck Hansen
chascode said:
Agreed. Thanks for studying the effects of added resistance between the leads and the metallization.
It would seem that one potential failure spot would be the electrical contact between the "flame-spray" end terminations and the metallization layer. But even if there were a problem here, it would simply turn the non-inductive construction to an inductive one (as seen on the cheap clear small value PS capacitors), which would simply raise the ESL and not change the capacitance.
I cannot think of any mechanism that would reduce the capacitance at all, especially by a factor of 10x. Apparently JC purchased these capacitors over 20 years ago, so in all likelihood the failure mode was analyzed long ago. I would be curious to find out what it was.
Hi Charles, why don't you just call the cap manufacturer and ask him? He must know about this, as I am not the only example. In fact, you might call Jam (unfortunately a banned member from this website) and ask him about his experience. The manufacturer has not called me yet about this problem. (Mr Blunt) 😉
Charles, I just talked with Bas about the problem, (as well as other things). You might give him a call sometime about it. (name changed to Mr. Thorough) 😉
I keep circling back to the idea that--and yes, this is contrary to the information posted on the Solen website regarding the construction of the cap--what if the plates were zinc instead of aluminum?
If the plates are (or were at some point in the past) zinc, then they'd be more chemically reactive than aluminum. If the zinc were to react with oxygen, just to pick something at random, then you'd have your choice of two failure modes:
--Surface corrosion, leading to a "bulking up" (due to most any zinc compound being physically more volumetric) of the conductive layer which would lead to the plates being farther apart. Bingo! Reduced capacitance.
--Total corrosion of the plate, leading to a reduction of effective plate area. Again, reduced capacitance.
Given that the layer is going to be quite thin, I imagine that the second mode would be more likely.
John, if you happen to still have one of those caps, could you cut it open or unwrap it to see if there are any visual clues?
Grey
If the plates are (or were at some point in the past) zinc, then they'd be more chemically reactive than aluminum. If the zinc were to react with oxygen, just to pick something at random, then you'd have your choice of two failure modes:
--Surface corrosion, leading to a "bulking up" (due to most any zinc compound being physically more volumetric) of the conductive layer which would lead to the plates being farther apart. Bingo! Reduced capacitance.
--Total corrosion of the plate, leading to a reduction of effective plate area. Again, reduced capacitance.
Given that the layer is going to be quite thin, I imagine that the second mode would be more likely.
John, if you happen to still have one of those caps, could you cut it open or unwrap it to see if there are any visual clues?
Grey
- Status
- Not open for further replies.