Coupling Caps - what sound the best?

[alnight]

This is very interesting, thank you for sharing!

I was under the impression that polystyrene caps were generally not used in guitar amps because they are less tolerant of heat, leading to a higher likelihood for durability issues to arise. Having no first hand experience with this, I am certainly not claiming to "know" one way or the other.

You have certainly gotten me curious.

[GrayDigger]

Alnight,

Interestingly, the max temp rating for the ETR 630V Polystyrene caps I’m using (+85C) is the same as for 715/716P and Mylar caps. The leads are much thinner and the PS caps are a squatter cylinder; the leads, though axial, are eccentric. They do “feel” less robust than your typical axial cap, and using a heat sink clip while soldering like with any film cap is probably a good idea…

I’ve measured the temp inside my chassis’; the warmest bit always lies under the rectifier socket, followed by power tubes. But neither area gets much over about 50-60C max. Cathode bias resistors or zeners can certainly exceed 85C, so I keep any caps or other temperature sensitive components well away from those and make sure there is ventilation…

Don’t know if they’ll last 50 years, we’ll have to see!

[martin manning]

~~~~~~~~~~~~~~~~~~
Martin, in the end, I found nothing earth shattering. The effects of different types of coupling caps on sound is just as the hi-fi guys characterized for PP, PE, PIO, electrolytic, tantalum, and silver mica.

Essentially, “Polystyrene rules, Polyester drools”. After all my goofing around, I ended up replacing every signal path (PP and PE) and tone (PE and Silver Mica) cap in my DUT with polystyrene, and have been practicing and gigging with such since. The difference in the amp has been… Astounding. Not subtle. Not even close.

This is a bit sensational, don't you think? If PS caps were all that, there would be a lot more interest in them. Instead, they are disappearing.

[Roe]

Roe,

In plexis, the mustards sound too dark for the normal channel, so there I prefer something else (e.g., jensen PIO)

I get what you’re saying, sorry for mis-interpreting. Based on my testing, to see what effect those mustards are having might require replacing all the caps in that channel’s signal chain with PS (or PP for the larger values) to get a “gold standard” baseline, as even one PE cap along the signal path will be audible. The PIOs I tested in my amp were quite vibey, but at the cost of being woofy and loosing some clarity as well. Unfortunately, each of the PIOs I tried were also different, and each tested quite differently across different frequencies on my LCR.

I just use NOS Jensen PiO because I have a few hundreds lying around. They tend to read a bit high and to sound very bright and clear. The mustards are darker. Darkest are the early/first Sozo caps. It is just about getting the right balance.

[nuke]

Polystyrene caps are very good, have a lot of good properties. They are all film-and-foil, since styrene won't take metallization well. The downside is you really must heat-sink the leads and use care when soldering. They're also physically large for their capacity.

The downside is I think they are all but out of production. I don't know of any PS capacitors in SMD format and I think it would be hard to have PS survive reflow soldering in the oven. That's why they're hard to find anymore.

Film and foil polypropylene is also very good in audio coupling and tone circuits. But foil construction is harder to find anymore, most are metallized film, which isn't necessarily bad.

[GrayDigger]

Martin,

This is a bit sensational, don't you think?

Yup, I guess. But maybe only from a guitar amp standpoint as it goes against conventional dogma. (Yet I don’t think it can be considered sensational when guys way smarter than me concluded the same, based on meticulous measurements and extensive testing, more than 45 years ago, right? I just independently arrived at the same conclusion in a far less rigorous manner). And - being most excellent conversation facilitator you are - I was just innocently responding to:

Can you share what you've observed?

I never in my lifetime would have thought to use polystyrene caps throughout a guitar amp. Curiosity and the cat, I guess.

During phase 1 of my testing, the analogy of my signal chain being like a multi-component camera lens came to me. If all of the component elements were “imperfect” (i.e. PE), if substituting a single element was noticeable (it was), the image obtained would still clouded by all the other “imperfect” elements in the lens. The only way to assess any conclusions from phase 1 would be to bite the bullet and replace ALL the component elements in my lens with “perfect” glass (i.e. PS). This became phase 2.

(In my “observational testing validity” brain, a “gold standard” true positive had been established, permitting the consideration of a valid 2x2 contingency table. Now test positives (“transparent” caps) and negatives (“colored” caps) can be judged against true positives (the amp with all PS installed) and true negatives (my “control” amp with all PE) for both false positive and false negative results. More realistically, since my DUT “lens” now has all perfect elements, I can evaluate whether I like any “colored” filters in my signal chain. But I’m so digging how my amp with PS caps sounds now, I see no reason currently to add dirty glass to it. I’m in the process of re-capping its twin I used for phase 2 A/B listening. A cross-over group if you will.

If PS caps were all that, there would be a lot more interest in them. Instead, they are disappearing.

I’m not a hi-fi guy, but it seems like their market size is about the same as ours (~$500 million/year). But since none of the guitar guys use PS, the audiophile market alone seems pretty darn small. I AI’d your question and this seemed as good an explanation as any: https://cdn.shopify.com/s/files/1/0085 ... 070719.pdf

Thanks for allowing me to share findings.

Nuke,

The downside is I think they are all but out of production.

I filled in a lot of values I didn’t have here:
https://www.justradios.com/

And here:

https://www.elliottelectronicsupply.com ... itors.html

[martin manning]

I was just innocently responding to:

Can you share what you've observed?

I never in my lifetime would have thought to use polystyrene caps throughout a guitar amp. Curiosity and the cat, I guess.

I appreciate the sharing and detailed feedback!
PP came in second in your evaluation. They are known to have better temperature stability and lower loss (on paper a more ideal cap), but at higher cost and larger physical size. MI amp people say they are "too bright" when compared to PET (polyester/Mylar), but IMO that could be addressed with EQ or selective component value changes.

[pdf64]

LCR make robust polystyrene caps in Wales, UK.
https://lcrcapacitors.co.uk/capacitors/fsc-polystyrene/
Distributed by Farnell worldwide.

[GrayDigger]

PP came in second in your evaluation. They are known to have better temperature stability and lower loss (on paper a more ideal cap), but at higher cost and larger physical size. MI amp people say they are "too bright" when compared to PET (polyester/Mylar), but IMO that could be addressed with EQ or selective component value changes.

100%. The effect of Mylar on muffling upper mids and highs was apparent testing one section at a time, and really apparent when all were replaced with PS. The mid and upper range has now been “Free Bird”- ified. (Interestingly, any harshness was removed as well.). The distorted tones are more detailed, creamy, and lush at the same time. This did require adjustment to my “normal” EQ settings, but I figure that’s why God created the tone knob

LCR make robust polystyrene caps in Wales, UK.
https://lcrcapacitors.co.uk/capacitors/fsc-polystyrene/
Distributed by Farnell worldwide.

Right on!

Alnight, I did just order a bunch of Teflon caps from a nice chap in the Ukraine. When (if) they get here, I’ll pop them in the warmer bits of my amp just to assuage any nagging concerns about temperature related non-linearity.

Parenthetically, i always wondered why my measured EQ responses with a ‘scope didn’t quite match what was predicted in LTSpice. It will be interesting to see if there is a closer match in my all-PS chassis.

Take the polystyrene challenge! It’s like putting headers and a tuned exhaust on your old V-8. Brings a big smile every time you fire it up!

[ngw]

I have been using LCR Polystyrene caps for a long time and I like them.
The problem I have had is getting them in high enough values at 630V. The highest value I can get from Farnell/element14 is 7n5.
How do you manage to do all the coupling caps in an amp; do you just parallel smaller values to get the higher values, or do you have a supplier with higher values?
Where do you get yours from?
Cheers, Noel.

[GrayDigger]

Hello Noel,

I’ve only been replacing all my caps since my experiment over the last couple months. The biggest PS 630V caps I have are 10nF. I believe they are made by ETR (Taiwan, maybe China). Unlike the NOS PS caps I have, they do not have the foil end marked, but the foil lead can be seen through the capsule. I’ve only been replacing caps in my own design amps so far so haven’t needed larger than 10nF, but for my Fender/Marshalls I’m planning on using them in parallel.

I get the 10nF from:

https://www.justradios.com/polystyrene.html

[ngw]

Thanks Graydigger,
I will check out Just Radios.

[GrayDigger]

I also recently obtained a variety of 500-630V NOS axial PS caps (ASC, MIAL, Mallory) on eBay. Compared to the (presumably Taiwanese) current production caps I have, these NOS PS caps have heavier leads, appear more robust in construction, and have true axial construction (instead of the wire lead directly connecting to the film/foil layer leading to eccentricity, these NOS have a little “jumper” doing this, so the leads can be centered) with the foil end conveniently marked. On my DE-5000, they test like the current production PS caps. Finding values over 10nF is difficult but not impossible.

FWIW, A month or so ago, I also got a shipment of steel-cased Russian K72Π-6 500V 10-22nF Teflon caps (late 80’s to late 90’s manufacture) from a chap in Ukraine at a very reasonable price. These test “perfectly” (θ= -90 and DF=0 from 100-100Khz, ESR essentially = 0) but are physically large. I just built my first couple of customer amps with them, attaching them to the chassis near the “warm” bits (power tubes + rectifier) and using the PS caps on the “cooler” preamp side of the chassis. In this design, using antique salvaged chassis’, all the bits are pretty close together.

I’ve left one of these amps on non-stop for the last couple of weeks. (In the past, I did this for a week to settle the tubes’ gm, and also to “break in” the Mylar caps so they sound less brittle and warm up prior to delivery). This time I was curious to see if there would be any tonal change after 350+ hours of continuous power and heat. Prior to this experiment, I configured these two most recent builds to sound the same A/B’d. Then I left one on for two weeks, and A/B’d again. They still sound identical, so I’m reasonable confident the PS caps should deal with the ambient heat just fine located on the cooler section of the circuit.

After the two-week continuous power test, I took a peek at the components, not that I expected to, but I couldn’t see any visual changes.

[Faze]

Your findings are interesting. I’m following along and ready to experiment. I’m interested I. Some subjective qualities of your test and also objective info. First what kind of style are the guitar phases used for evaluation. Blues , jazz, rock, metal. Are they chords, solo lines, combo. What is the looper used? They also have a tone. And to continue with objective, the amp shows octal tubes mostly. What is the typology? Seems like volume and tone only for controls. What is the speaker used for testing? What was the mic used? I’m sure you’ve stayed consistent with the variables in your setup. Just wanted some info on the setup and style used. Thanks for all the effort you put into this project.

[Faze]

One other thing. What kind of guitar and what pickups were used in your loop test? Thanks