Has anyone used the very large value MKP caps (>= 20 uF) that are usually in the "DC Link" category? Ideally would like to use them in a coupling application.
Just wondering if there is any measurable difference in distortion or linearity vs the <10 uF sized metalized polypropylene caps?
Just wondering if there is any measurable difference in distortion or linearity vs the <10 uF sized metalized polypropylene caps?
Larger that 20uf at higher voltages are harder to find and get expensive for a quality types. Not sure about the voltage or your specific application. If you are looking for coupling applications in lower voltages and high value caps, you may consider using back to back quality electrolytics. Example, connect the neg posts of two 100uf silmic caps together, then use the two pos posts for the lead as if one cap. The net value is a 50uf cap that has some of the negative features of a single cap nullified. This technique is used by many manufactures including Denon and Marantz on some of the better equipment. You may also bypass this back to back set with a small film cap to improve/change the sound of the passed through signal. Good Modding!
"DC Link" capacitors are designed for use with permenant high Vdc across them, have low esl, esr for high current energy delivery in sw mode power supplies, self clearing electrode metallization
http://www.mouser.com/pdfDocs/WIMA_DC_Link_Capacitors.pdf
for small Vdc DC blocking, small audio signals the special features may not matter
for audio signal coupling, low V DC blocking, into hi Z circuitry just about any film cap will be fine - distortions of low order, very low level - never expected to be audible in Blind, controlled listening tests
all electros measure worse on audio distortion - but still at inaudible levels
non/bi-polar Al electors with full thickness oxide on both foils measures lower in distortion than the old "back-to-back" polar electro recommendation
http://www.mouser.com/pdfDocs/WIMA_DC_Link_Capacitors.pdf
for small Vdc DC blocking, small audio signals the special features may not matter
for audio signal coupling, low V DC blocking, into hi Z circuitry just about any film cap will be fine - distortions of low order, very low level - never expected to be audible in Blind, controlled listening tests
all electros measure worse on audio distortion - but still at inaudible levels
non/bi-polar Al electors with full thickness oxide on both foils measures lower in distortion than the old "back-to-back" polar electro recommendation
Last edited:
Yes I agree with JCX. I assume that the reason for the need for a large value cap is that the load is not a high Z load. If this is not true then of course a MKP cap would be easily and better used.
Thanks, I was just trying to see if the DC link caps had some undesirable properties related to it's high density for an MKP cap. Sounds like there is no problem, other than the cost.
The application is low/medium Z (10k) and low V DC bias.
The application is low/medium Z (10k) and low V DC bias.
large physical size itself can be a problem - makes for larger parasitic C to the world
and added distance between leads adds esl
ideally there will be one electrode "on the outside", marked lead - so you can connect the film cap with the orientation that gives the least operating problems with the outside electrode's parasitic C to the world
if you are only blocking single digit V then you can even get smt 10-20 uF mylar or acrylic parts for least physical size
and added distance between leads adds esl
ideally there will be one electrode "on the outside", marked lead - so you can connect the film cap with the orientation that gives the least operating problems with the outside electrode's parasitic C to the world
if you are only blocking single digit V then you can even get smt 10-20 uF mylar or acrylic parts for least physical size
chris719,
DC Link caps are probably not the best choice for audio, they are intended for high power circuits and are designed to that end. They are extremely large and expensive, and will not perform any better than a good quality, standard polypropylene film cap, and in fact they may perform worse.
And I don't know your intended application here, but 20uF/10k gives a .8 Hz corner frequency, are you sure you need to go that low? 10 uF will give a 1.6 Hz corner at a significant saving in cost and physical space.
Mike
DC Link caps are probably not the best choice for audio, they are intended for high power circuits and are designed to that end. They are extremely large and expensive, and will not perform any better than a good quality, standard polypropylene film cap, and in fact they may perform worse.
And I don't know your intended application here, but 20uF/10k gives a .8 Hz corner frequency, are you sure you need to go that low? 10 uF will give a 1.6 Hz corner at a significant saving in cost and physical space.
Mike
10k is a rough guide, I haven't finished the design.
I have no desire to use them over standard value PP caps but if I needed a value of, say, 47 uF, that leaves me with a choice of several || normal MKP, large polyester, or NP electroytic, other than DC link MKP.
I'm aware there are no audible problems with a big NP electrolytic, but I like overkill.
I have no desire to use them over standard value PP caps but if I needed a value of, say, 47 uF, that leaves me with a choice of several || normal MKP, large polyester, or NP electroytic, other than DC link MKP.
I'm aware there are no audible problems with a big NP electrolytic, but I like overkill.
I’ve used them in HV tube PSUs with excellent results ~30uf.
I like the four legged Vishays, if just for the extra security to the pcb as they get pretty heavy/ big.
A nice big 450V Nichicon might run $8. Two for a bipolar would be $16.
To me it makes sense to spend $13 on a 22uf DC-Link or $17 on a 30uf and never have to worry about replacement.
Size is about the same footprint as two big electrolytics.
The cost on them is low and getting lower, I don’t see why one should use electrolytics at all these days with tubes.
Cost savings just isn’t really there, as much as people may recite otherwise.
For a 50V application, quite wasteful of course.
Can someone give me a credible reason why this would be an inferior choice over two high quality HV electrolytics besides a couple dollars in difference?
MKP1848630454P4 Vishay / Roederstein | Mouser
I like the four legged Vishays, if just for the extra security to the pcb as they get pretty heavy/ big.
A nice big 450V Nichicon might run $8. Two for a bipolar would be $16.
To me it makes sense to spend $13 on a 22uf DC-Link or $17 on a 30uf and never have to worry about replacement.
Size is about the same footprint as two big electrolytics.
The cost on them is low and getting lower, I don’t see why one should use electrolytics at all these days with tubes.
Cost savings just isn’t really there, as much as people may recite otherwise.
For a 50V application, quite wasteful of course.
Can someone give me a credible reason why this would be an inferior choice over two high quality HV electrolytics besides a couple dollars in difference?
MKP1848630454P4 Vishay / Roederstein | Mouser
Last edited:
Motor start or motor run caps in Metal Poly, especially if higher voltages present.
Even the power factor correction caps used in fluro lamp fittings are usually Metal Poly now , but you may need to remove the discharge resistor inside the plastic case.
Even the power factor correction caps used in fluro lamp fittings are usually Metal Poly now , but you may need to remove the discharge resistor inside the plastic case.
- Status
- Not open for further replies.