Capacitors for HP filters in tube amps recommendations?

I’ll soon be installing a capacitor on each of the input jacks of two tube power amps, to create passive 1st-order high-pass filters. Cap values are 19.9uF (20uF will do) and 30.33uF (let’s say 30uF). I’ll need two of the former and four of the latter (balanced/XLR input jacks), and I don’t want to spend more on the caps than the amps cost me ;-) .

I’m all ears for nominees. I don’t need or want any flavoring, "just" neutral transparency. The amps are powering fairly transparent loudspeakers (ESL’s, and Magnetic-Planars with Ribbon tweeters), which will pretty well reveal the character of the caps. In spite of that fact, "most-bang-for-the-buck" nominees are of particular interest, not cost-no-object ones. Thanks y’all.

Um, I think your cap values are off by a couple of orders of magnitude.

What is the input impedance of the amps??

Did I use the wrong abbreviation? The 19.9uF (microfarads) is for 80Hz into 100k ohms, the 30.33uF for 175Hz into 30k. I used the calculator on the site to get those cap values.
Those frequencies are mHz, not Hz (0.080 and 0.175) . Why you want to cut off below 80 to 175 hz?

The filters are not for tweeters, they’re for high-passing main speakers, the lower frequencies going to subs with their own filters and amps. I guess I assumed the mention of mounting the caps on the amp’s input jacks would imply that. I’ve done it before, but not personally---I had Tom Carione at Brooks Berdan, Ltd. do the installation (on a Bedini 25/25).

So, I came up with the incorrect cap values, ay? I sent the amp designer/maker an email asking for them, but if anyone wants to educate me, great!

The specs:

Amp 1: 100k Ohms input impedance; 80Hz x/o.

Amp 2: 30k Ohms input impedance; 175Hz x/o.

Cap suggestions, anyone? I used a Solen Metalized Polypropylene 35uF (sourced from Madisound) in a very low-frequency low-pass filter, for which it was fine. I always look for the "sweet spot", where diminishing returns kick in big. A completely subjective determination, of course. I don’t discourage opinions, I welcome them! Good value caps at $20/ea? $50/ea?

Assuming we're talking line level here, there's no need for monster capacitors.

Amp 1: 100k Ohms input impedance; 80Hz x/o.

20n (0.02uF) created from two 10nF Polystyrene 1% Caps in parallel

Amp 2: 30k Ohms input impedance; 175Hz x/o.

30n (0.03uF) created from three 10nF caps

All for less than $10, you can spend more on exotics but I don't think they'd sound any better.

Yep @pragmasi, line level. So I was by off by x1000?! I gotta figure out what I misunderstood on the cap value calculator. Thanks for the info.
No problem... even if you are tempted to try expensive capacitors, buy ten of these and compare. The only dialectric that may work better than polystyrene for this application is C0G/NP0 ceramic (definitely not any other type of ceramic) which is fine if you can solder surface mount but the through hole versions are ridiculously expensive.
@bdp24 , @pragmasi is giving you good advice. I've had a number of customers do things like this and we've used polystyrenes for the job as there's very little out there that is small enough in any other dielectric.

Thanks fellas. I figured out what I did wrong on the calculator, and with the help of Danny Richie at GR Research got the correct figures. 100k and 80Hz requires .02uF, 30k and 175Hz .03uF. But caps aren't made in those exact values, .022 and .033 being as close as one can get.

The resulting x/o frequencies are 72Hz and 161Hz respectively, close enough considering the shallow slope of the 1st-order filter. Small value caps are priced relatively modestly, even the better ones being not THAT expensive.

But you're right Ralph, some cap designs are huge, even in small values. To fit them onto the amp input jacks will take some doing!

100k and 80Hz requires .02uF, 30k and 175Hz .03uF.
 You parallel up .01uF caps (for caps in parallel you just sum the values), so five for each channel means $4.20 in capacitors.

On the inexpensive, I like Cornel Dubelier a great deal.

If you want to try pricey ... both Audyn (True Copper Max) and the 600V versions of Jansen copper foil caps come in 0.01 values.


Ah yes @pragmasi, you explained that above and I missed it. And the caps you speak of are small enough to fit side-by-side on the input jacks. Thanks, mate. @Erik, I’ll have to look inside the amps to see exactly how much room I have for the caps. Those Jansen’s are huge!
BDP24 - You know, you can easily make yourself an interconnect cable with a cap in line, no reason to mod your existing pre. :)

But good for you for catching the size problem with those! :) I forgot to mention it.
Actually Erik, the caps are going into the two power amps rather than the output jacks of the pre, as each amp is used with a specific loudspeaker. But I take your point---the caps could be incorporated into the interconnect going to each amp. Not soldered onto the outside of each cable’s RCA or XLR plugs, of course, and they won’t fit inside them. Modjeski lays out his amps (RM-10 Mk.2 and RM-200 Mk.2) pretty sparsely, so there should be plenty of room inside for caps.
I'd do it this way just for the sake of ease of changing things. It's far easier to swap out interconnects than to get into an amp/pre and solder a cap, regardless of the size of the cap. 

However, it's your money and effort, so do it however it feels right to you. :)

So Erik, how do you incorporate the caps into an interconnect? Now that I think about it, since each amp is fed via un-switchable cables (one amp is balanced/XLR, the other se/RCA)---my pre has multiple balanced and se output jacks, caps on the cables sounds like a good idea. I wouldn’t know how to put caps into a balanced connection (a cap on each leg) other than inside the amp.

It finally occurred to me (duh ;-) to do an Audiogon Forum search on the subject of capacitors, and holy cow---what a goldmine of information!

I also followed the link to the Humble Homemade Hi-Fi Cap Test, and double holy cow!! That was a lotta work!

Some caps stood out in both the above, and I now have it narrowed down to what appear to be the caps that provide the highest sq-to-price ratio. I’ll share them here, both for anyone else researching the subject, and any reactions.

- Jupiter Copper Foil. Though considered at the very top-o-the-heap, Sonicraft (a favorite vendor of mine) sells the .022uF for $28.50, relatively cheap. Parts Connection sells the same for $38.00. ?

- Clarity CMR (thanks Erik). Sonicraft sells the .022uF and .033uF for even less than the Jupiter.

- Miflex KFPM-01. Recommended to me by Danny Richie at GR Research, who sells .022uF and .033uF matched to 1% for about 20 bucks each. Sonicraft also sells them..

-Audyn True Copper. Parts Connection sells the .01uF, don’t remember the exact price, but as Marge Gunderson says in Fargo, it’s reasonable.

Any of the above are good enough for my needs, so I can’t loose. The only question remains will any of them fit in my amps?!

All of the caps listed will be out performed by polystyrene in this application. Polypropylene are ideal when higher capacitance is required and the only other option would be electrolytic. I think there is a perception that more expensive must be better but in this case that is entirely wrong. If you want the best possible then look at the type of cap used in a top end phono amp in the RIAA section - there is no room for error in this application and any distortion will be massively amplified. I’ll put my money on them being polystyrene or NP0/C0G.

@pragmasi, I was hoping I would see a response from you to the above! Polystyrene, ay? I take it you don’t share the perception that caps made of plastic materials impart a "plastic" coloration? Can you expand on your contention that polystyrene caps will outperform polypropylene or copper foil/film/paper/oil/wax, etc. caps IN THIS APPLICATION (that interests me greatly---see below)? Thanks.

I haven’t received a response to my email to the designer of my two amps, but I already know he "believes" in using the right part for a given application---which doesn’t necessarily require an audiophile-approved boutique part, and that higher cost (or audiophile cache’ ;-) does not necessarily buy one higher sound quality.

Like Frank Van Alstine (another "sensible", old school ee), Roger Modjeski has seen his amps butchered by well-intentioned audiophiles, who replaced the stock parts (chosen and used for very specific reasons) with boutique parts which actually degrade the sound of their designs, not to mention their reliability. Some boutique parts are designed with a specific parameter prioritized, but also possess characteristics such that when (mis)used in an inappropriate application can result in not just degraded sound quality, but also unforeseen and unfortunate consequences.

Thanks again for the tip on polystyrene in general, and to those you above provided the link to. One advantage they afford is small diameter, making possible using 2 or 3 stacked .01uF caps side-by-side, to get exactly .02 and .03 combined values, rather than the .022 and .033 compromises. I’ll take another look at them, and google this polystyrene of which you speak. ;-)

I guess the important factors in this application are that we're working at line level and we are setting time constants in the audible frequency range.

Line level means we don't need to worry about dealing with high currents, voltages or significant power dissipation (large packages are better at dissipating heat than small ones). But it also means that any distortion will be amplified by the following gain stages.

Setting the time constant means that we're attenuating a selection of frequencies, in this case ones we can hear so we're looking for precision and linearity (high linearity = low distortion).
If the actual capacitance deviates from the nominal capacitance then we'll find that the -3dB point is at a different frequency to that specified. In this case, assuming you can adjust the roll-off of your subs then the precision of the pair of filters (left and right) is not overly critical. However the precision between the filters (the difference between left and right) is critical and may well be audible if one channel rolls off at a different frequency to the other.
So we want to know that the capacitors on both channels are close in value but not so worried about how accurate that value is (in an RIAA filter both values are critical).

Linearity is a lack of distortion so the output of a linear component will closely resemble the input (in this case we are intentionally distorting the input by filtering out LF, but we want the audible frequencies to be unaffected). Linearity is measurable in Total Harmonic Distortion and there have been studies into the distortion characteristics of dielectrics,  if you want to learn more (and I completely understand if you don't) Doug Self has researched the topic in depth. C0G/NP0 and polystyrene caps can be considered free of distortion, polypropylene is very low distortion which can probably be said to be negligible, electrolytic is another matter.

So bearing this in mind...

Jupiter Copper Foil: I don't know the linearity of wax and paper but it can't be better than unmeasurable, so let's be generous and assume it's unmeasurable. Precision... the ones I found were 5%, okay but not great.

Clarity CMR: These are polypropylene so we know what we're dealing with. Precision is 3% which is better.

Miflex KFPM: Polypropylene and 2% precision.

Audyn True Copper: Couldn't even find out what the dielectric is, so let's assume it's unmeasurable again. 2% precision.

I couldn't find (admittedly I didn't try too hard) any detailed datasheets for the above. One thing they all have in common is that they are large, which means they can dissipate heat (which we don't need) but the size also means that the parasitic properties of the leads, foils etc are greater.

Polystyrene is available in 1% in a small package and has no distortion to speak of. So it's not streets ahead of the caps above but it will be better for this application (and that's without taking into account cost). Some of the exotics will probably be available in matched pairs which is an improvement... but still not better unless the matching is less than 1%.

I wrote more there than I meant to... I hope it's useful.

Excellent @pragmasi, exactly the kind of information I was hoping for! And lots of it!! Pure ee considerations and concerns, very much aligned with the kind of thinking that went into the design of the amps. No intent or desire to disparage audiophile language and concerns, but you gotta get the ee stuff right first ;-) .

Yes, having the caps matched to 1% is a very high priority (perhaps THE priority). I’ve seen the term "time constants" many times over the years, but never took the time to educate myself on what it means. Perhaps I should do so now! I’m also familiar with the term "parasitic", and DO understand it. It’s bad, right? ;-) . I’ll look up Doug Self (I’m familiar with the name) and his research.

I’m going to try and reach RM (the designer and maker of the amps), He may have the caps I need in his parts inventory. He encourages the installation of caps on the input jacks of his amps as a way to create a 1st-order filter for partnering the amp/speaker with a sub. I’m sure he views the question of cap "sound" just as do you---a matter of their electrical characteristics in any given application.

Thanks again for all your time and help @pragmasi. It is very much appreciated, and of great value and use to me---Eric.

@ramlabs, Roger, baby, where are ya?! When you have time, could you look for my email from last week? I'll call you Monday morning.

Modern film caps are usually spot on in terms of measurements, but you can always order matched pairs.

I've not bought a film cap in ages that wasn't actually within 1%.

I just realised the capacitors in the link above are 5% tolerance, shouldn't be too hard to find 1% ones though... in the UK they are available from both Farnell and RS Components.

I’m also familiar with the term "parasitic", and DO understand it. It’s bad, right? ;-)

Parasitic properties are the characteristics of the 'real world' component that deviate from the ideal. An example in this case would be the capacitor's inductance, that is directly related to lead length so a large axial resistor will present greater inductance than a small radial cap with short leads. Whether anyone can hear the difference is another matter but if you can avoid it then why not.

In reality you could get a good result using 1% polystyrene or 1% polypropylene, the former are smaller and less expensive... it's less clear whether the better spec in terms of parasitics would make an audible improvement.

Hope the project works out well whichever you choose.

I noticed that too. There are some cap vendors who match them in pairs, and Parts Connexion offers hand picked caps that are 1/2 or 1/4 of the factory tolerance figure, but they sell only "audiophile" brands. I think Michael Percy offers the polystyrene caps he sells (Rel and Multicap, each about $5/ea) matched to 1%, but to each other, not tolerance. The caps you provided the link to are sold on ebay, but I’m not going there!

Something suddenly occurred to me last night: as I mentioned, one of the amps has balanced/XLR inputs, requiring a cap for each leg of each XLR. What I forgot was that while the input is 30k ohms, each leg is of course half that---15k. So on that amp I need not .03uF (three .01 caps) per input jack, but rather .06uF (six .01uF caps!) for each leg of the XLR connection! Oy!

@pragmasi, I took a look at the product description of the B4 cross-over (one of which I own) on the First Watt website, and in it Nelson Pass mentions he used Wima Polypropylene film caps in all the filters. If they’re good enough for Nelson.....

I then went on the Wima site, and I’ll be darned: they offer a .01uF poly cap (model no. FKP 2) with a range of tolerances, 1.5% being the lowest "regularly" available, but 1% available if requested.

Mouser sells the 1% version for cheap ($2.11/ea, $17.10/for 10), so I have my caps. For the 100k/80Hz amp, anyway. I’m thinking I’ll just use the B4 on the 30k/175Hz amp, as soldering .12uF onto each XLR may be just too much!

Thanks again for your advice, it was of great help.

Wima is a good choice of PP capacitor, at line level I tend to use NP0/C0G for 10pF to 10nF, Wima MKP for 10nF to 10uF and try to avoid higher values in the signal path. Be aware the leads are short on the Wima caps so you may need to mount them on strip board as point to point soldering might be a bit fiddly although do-able.
I'll be putting just a pair of .01uF (10nF) caps (.02uF apparently isn't available) on each RCA jack, so those short leads will be long enough. I wasn't able to find any 1% polystyrene .01uf or .02uF caps. I'm going to research NPo/CoG (with which I am unfamiliar) right now.
I think you've found a solution that will work for you at a sensible price point... I didn't realise Wima made PP caps in 1% so I've learned something new too. I usually go for 5% as I use them for high pass filtering at sub-sonic levels, knowing they are available in tighter tolerances will come in handy I'm sure.
Thanks for coming to my aid! I’d still be floundering around if you hadn’t, ’cause I haven’t heard back from Modjeski. I figured he might have the appropriate caps in his parts bins, but I think he uses polyester caps in his amps.

You need to be cautious when recommending CGO/NPO ceramics for audio (particularly coupling not bypassing) applications. I know from experience that some NPO/CGO caps (particularly those from TDK ceramics) have a voltage coefficient of capacitance that is quite significant (sometimes as high as 0.5 x C) at DC voltages near rated max. This is more evident in the larger values (and .01 uF is large for an NPO).  Remember, the CGO/NPO spec relates to temperature coefficient only, and while meeting the low TC, are still quite nonlinear when stressed at various voltages. Folks have slammed the use of tantalum caps for years because of this characteristic, and some NPOs are just as bad or worse. Also, you may be hard pressed to find a ceramic cap manufacturer of NPOs providing dielectric absorption specs. Due to the large variety of ceramic compounds, it is risky to extrapolate the product of one manufacturer with an NPO compound of another.

Also, there is data that shows polypropylene has pretty much the same dielectric absorption as polystyrene. With the same metallisation, ESRs and dissipation factors are identical. Polystyrene has a major weakness in that these caps can be easily damaged while soldering, due to the low temperature ratings of polystyrene. I am not a fan of the cheap Xicon polystyrene caps, but those made by RelCap are quite good.   
You need to be cautious when recommending CGO/NPO ceramics for audio (particularly coupling not bypassing) applications.

I have question for you.

Can you use a PIO russian .01 or a pair of .015
they come in 100-1000 v.  There were a few different model numbers., I I think Y77????, not painted, glass tube..
It's been 3-4 years since I ordered 50 or so @ 600v. I matched 16 as close as I could.  less than .5 % var. I used .01 as bypass in addition too
1.0 uf coupling. They were about .25" x 1" 1.5 inch tails. I took out EROs at 400v and 10%, they were all over the map. The PIO russian surplus worked wonderfully.
Took a bit of time to break in but really softened (I think that's the right word) or took the edge off the blistering HF issue, that just kills my ears.
I'm one of those people that UHF literally makes me clench my fist and walk in circles. Cotton in the ears, or a hoodie a LOT of the time in public.

So will these caps work as an inline XO (of sorts) on line level interconnects if that is what someone wants to do?? Your thoughts

I'd like to take care of this problem in a cable format if possible. On the line in  where the values are much smaller and WAY less expensive..

Your thoughts


Dear @bdp24  : I know that now maybe is to late but I will post anyway because I just made a huge caps discovery.

"  If they’re good enough for Nelson..."""

My discovery was exactly WIMA caps that not only for N.Pass but for electronics in all the world WIMA is the true standard in the industry and for very good reasons I never experienced before till I tested and learned why is the industry standard.

I was using at the input of my Levinson monoblocks the V-caps teflon Cu for the same application that you are doing.
Well I just changed that truly expensive V-cap ( top of the line. ) for the humble Wima FKP 1 and Wima outperformed very easy.

Before that electronic cap change I changed all the boutique expensive and heavy colored caps ( Duelund, Mundorf, V-caps, Jantzen, Sonicraft, that I own it. ) in my 3-way crossover speakers. I use here the Wima MKP 10 model and happened the same: Wima outperformed all those " high-end "  caps.

Nothing compares against Wima, no matter what.

I came to this specific forum looking for an advice for a harmless/signatureless caps for my speakers and posted a thread about.

I posted the thread with out any agenda but looking for help and through that thread and with out any Wima recomendatiopn I pull the triger for it.
You can read that thread and know how and why I decided for Wima.

Btw, the FKP 1 and according the Wima site is the top cap for our applications and you can find out therough Mouser that is where I bougth all my Wima caps and I don't care to much of the Wima tolerance because it measured really near the capacitance values even if the tolerance is 10% or 20%.

Regards and enjoy the MUSIC NOT DISTORTIONS,
I pulled WIMA MKG-4 .47@ 250v in a VTL 300 Deluxe, replaced them with a really good VCap. LOL, two months later thinking it was a breakin thing, pulled them and put the USED WIMA caps back in.. Sounded wonderful again. Crazy I'm with you they are a great cap, my first Watt B5 is packed with WIMA, My new Mcintosh PACKED with WIMA. High quality, LOW prices in comparison... 5.00 USD, Vcaps 100.00 USD at the time..

Dear @oldhvymec : Wima comews not only in your electronics mentioned or in my Levinson monobloks but comes too in very expensive/cost no object high end electronics:

or in this 250K+ amplifier:

Passive high-end/boutique parts as caps are only the snake oil we " high end " audiophiles are looking for. All of them are full of " nice " colorations when Wima humble ones are true signatureless and harmless caps to truly honor MUSIC.

Differences between high-end audiophiles and real MUSIC's lovers are like the differences between Duelund and Wima caps where obviously Wima is for the music's lovers.

Over 40+ year in audio I just learned what I said in that last statement and I did it through first hand experiences.

Well, everyone's "first hand experience" will be different. Wima makes excellent industrial caps. I used them in my Spectral DMC 10 rebuild as the final output coupling caps and they sounded good. But the Mundorf silver gold oils sounded much better in the DMC 10 by comparison. If there really was no difference (or the WIMA was superior) I doubt if Mundorf would continue to get the prices they charge over the Wima. As Nelson Pass states, audiophiles vote with their wallet.

I can't see ANY poly propylene cap out performing a teflon cap like the VH. This is quite commonly accepted and reinforced in reviews like the Humble Homemade  Cap tests. What a manufacturer uses in his product may be the result of a number of factors, sound quality being only one. Costs, size and availability may be others. For example, Bryston uses an electrolytic cap in the output of their BDA products, much to my chagrin. Speaking with their engineers, they "reassured" me that this cap was chosen from listening tests. My own tests with that cap removed proved they were sadly mistaken. The BDA sounded much better with that cap removed. BTW, Bryston only uses this cap to remove the very small DC offset, no other reason.
The Dahlquist DQ-LP1 crossover used a passive HP section.      It’s manual contains a chart (and formula) that you may find of interest.                When I was using my modded version, Michael Percy was my go-to guy, for Polypropylenes, Polystyrenes, etc, that he’d match to 1%, plus: a wide variety of other nice, parts/pieces/materials.      His service and expertise were consistently excellent.      Can’t swear that he’s still in business, since it’s been years, but- his site’s still up:         Personally; I never paralleled capacitors, unless they were identical, to avoid any possible time-constant disparities.
Personally; I never paralleled capacitors, unless they were identical, to avoid any possible time-constant disparities.
That's not a concern... capacitors in parallel will just sum in value (like resistors in series)... the source and load won't see any difference. In fact there are good reasons to parallel capacitors. It's easier to match the values for left and right channels (more combinations) plus statistically you'll get a tighter tolerance than specification (although you'll need to parallel quite a few to make much difference). You can achieve higher capacitance than is commonly available for a particular dielectric (I have used 5 x 10uF wima for DC blocking). Also multiple smaller packages will have a larger surface area than a single large package so heat dissipation is better.
If you want to go down the route of matching your own for around $25 you can get a cap meter that will measure 1pF to 20,000uF although you might want to factor in a few precision caps to use for calibration.
I said, "personally".        You’re welcome to your opinion, if you have one regarding time-constants (which is to what I referred, but- you never addressed) and anyone else’s, that you value.
I’ve read concerns from a couple of amplifier designers about time constant ramifications resulting from the use of bypass caps. They state that contrary to the opinion of some audiophiles, bypass caps are to be avoided whenever possible. The obvious question is: "What designer(s)?" I don’t at the moment recall, but it was a couple of biggies, on the level of Nelson Pass, if not in fact he himself.
I just prefer not to give myself a reason to think about possible timing/phase anomalies, anywhere in my music's path.      That's been my stance, since the early Eighties.       Like everything else; opinions vary.       
By the way, I too have bought parts from Michael Percy, the last time earlier this year. I hope he's doing okay, he's a great source.
Absolutely and good to hear, he's still in the game!    I always enjoyed our conversations, appreciated his ear and wisdom.
He's been in the biz a long time; I've been patronizing him since he was in Northern California. His online catalog reflects a wealth of knowledge and experience.
I said, "personally".       You’re welcome to your opinion, if you have one regarding time-constants (which is to what I referred, but- you never addressed) and anyone else’s, that you value.
I was trying to be helpful, not make anyone wrong... I'm sorry if it didn't come across like that.

The time constant of an RC filter is the cutoff frequency, this is set by the product of the total resistance and total capacitance in the network. So the accuracy of those two values will determine how close you are likely to get to what you're aiming for. Low tolerance components (e.g 5%+) mean there's a larger margin of error in relation to the target cutoff point - sometimes that's important (e.g. RIAA filter), sometimes not so much (DC blocking filter).
Components can also change their value due to temperature changes, ageing and other factors. So when setting the time constant it's most important to select components with the required tolerance and stability which is mostly down to the dielectric (PP, polystyrene, polyester etc.) and the quality of manufacture.  Using caps in parallel doesn't affect any of these qualities providing they've got a good electrical connection.