1. Things you need to know about the new ‘Conversations’ PM system:

    a) DO NOT REPLY TO THE NOTIFICATION EMAIL! I get them, not the intended recipient. I get a lot of them and I do not want them! It is just a notification, log into the site and reply from there.

    b) To delete old conversations use the ‘Leave conversation’ option. This is just delete by another name.
    Dismiss Notice

Replacing 100uF electrolytic signal path caps with film types

Discussion in 'd.i.y.' started by John_73, Jun 10, 2018.

  1. John_73

    John_73 pfm Member

    Hi folks.

    I have a circa-1986 era grey Quad 34 which I use with a Quad 606.

    Over the past few years I’ve updated the op-amps, the cmos switches, and have removed most of the electrolytic 100uF coupling caps, replacing them with wire links. Very pleased with how it sounds. However my OCD side would also like to finally ‘finish’ the project and dispense with the remaining 100uF electrolytic couplers with film caps for longevity (they shouldn’t ever need to be replaced) and possible sound quality benefits.

    I will use the Panasonic ECWF 250V polypro caps as I’ve been very happy with them in other projects in signal coupling positions, they’re very reasonably priced and not stupidly large like many ‘audiofoo’ brands. I’d also respectfully request this thread doesn’t turn into any objectivist/subjectivist battles over cap types ;)

    There’s a pair of 100uF couplers just before the ALPS volume control (to protect from any wear due to DC) at positions C30/C31, and there’s also a pair on the Quad 34s outputs too, C77/C78. Circuit diagram viewable from Keith Snook’s website below:

    http://www.keith-snook.info/amplifier-hifi-schematics/QUAD 34 Schematic Circuit.pdf

    My question is, what film cap value would be suitable for replacing the 100uF electrolytics without audibly cutting the bass response? Would 4.7uF be a good value? It’s a value I see used a lot on various pre-amp outputs. The Panasonic ECWF film caps go up to 6.8uF in value.

    Many thanks.

  2. david ellwood

    david ellwood Kirabosi Kognoscente

    For an improvement in sound quality and reasonable improvement in longevity try using a 100uf and a 1uf tant together in parallel.
    John_73 likes this.
  3. John_73

    John_73 pfm Member

    Thanks David, but I want to get rid of the electrolytes in the signal path altogether, if possible. Cheers.
  4. martin clark

    martin clark pinko bodger

    Between those caps and the output is about 5kohms to 0v; that will be in parallel with the input impedance of the power amp that follows. Summed result, a load impedance of c. 4.2Kohms 9assuming 22K for power amp input - Quad PAs are usu 47-100K)

    4.7uF into 4kohms is f-3db at 8Hz. Just fine. Go for 6.8uF if it fits, and you are neurotic. But 4.7uF is fine.
    John_73 likes this.
  5. John_73

    John_73 pfm Member

    Many thanks for working that out for me Martin - I’m hopelessly at sea when it comes to the theory side of audio DIY, anything heavy on maths and equations and formulae. I envy those with an aptitude for it.

    Would the 6.8uF value also be fine for use in the C30/31 positions too Martin (i.e. just before the volume pot)? Thanks.
  6. martin clark

    martin clark pinko bodger

    Unfortunately the resistance of RV1 doesn't appear to be given on that schematic unless I've missed something (entirely possible...); so can't directly help there...

    Answere of likely yes - but these things stack; if you have two series 'high pass' filters with a turnover of 10Hz, the circuit overall is now -3dB at sqrt(2) x f-3db, roughly: or about 14Hz.

    This is one reason you find values used in coupling applications that set the -3dB point well below 10Hz; have a few in series in a system (easily done) and the turnover point, and its effects on phase, might become a problem.

    The other is that distortion in caps used in coupling applications is effectively proportional to the signal AC voltage appearing across them; so using 100uf or more (as Quad do here) is a cheap and v effective way around this (v common elsewhere, too). If you can't find a polyprop to fit, look at 100-220uF bipolar electrolytics instead...
    John_73 likes this.
  7. fatmarley

    fatmarley Registered User

    I think most of us feel that way, but In some cases electrolytics subjectivley sound better and can be surprisingly good.

    If you find something sounds bright or a bit thin after fitting film caps (I've had that happen a lot), try Nichicon FW. In gear that film caps don't work, they sound much more neutral than Nichicon Muse KZ (bright) and Elna Silmic 2 (dull) to my ears.
    John_73 likes this.
  8. Arkless Electronics

    Arkless Electronics Trade: Arkless

    Nice. +1
    John_73 likes this.
  9. John_73

    John_73 pfm Member

    Thanks for the very helpful comments. The 6.8uF caps arrived today, and I’ll be trying those in the C77/78 output positions tomorrow. By the looks of the DaDa upgrade sheets, and reading Keith Snook’s pages too, it looks like I can just bypass C30/31 which simplifies things. Will report back after I’ve given it all a try. Easy to revert back if need-be.

    P.S. Marked on the back of the ALPs volume pot is 50KX2, which I assume means RV1 is a dual gang 50K unit.
    Last edited: Jun 13, 2018
    fatmarley likes this.
  10. Jim Audiomisc

    Jim Audiomisc pfm Member

    Worth adding that using a smaller cap can lead to excess peaking of some waveforms. FWIW I've modded the 34s I use in various ways. Some of the caps can be replaced by a link or resistor if you know the input source has no dc worth worrying about. But large electrolytic caps have always seemed fine here, and as explained, can be low distortion, etc.
  11. 337alant

    337alant Negatively Biased

    I have been thinking about this in my Studer A810 as some of the only 40 year old caps I havent replaced are the solid aluminium types as I was advised tat they will all be OK
    I have ordered some of the Panasonic OSCON solid aluminium polymer SEPC through hole types as an experiment. Being in the signal path they dont need to be high voltage so they should be fine for me but check the actual voltage requirements in your application ?

  12. Jim Audiomisc

    Jim Audiomisc pfm Member

    FWIW I'm still using the same caps in my 35-plus year old pairs of Armstrong 730/732 amps, I've checked the performance every now and then over the years and they seem to have remained fine. In general, people advise replacing electrolytics after a decade or two. But in practice if they are well made and used well within their limits they can work fine for many years. Poor caps OTOH of *any* type can give problems from the start or degrade early. So I've tended to focus on reliable quality rather than take for granted that any "electrolytics are bad" mantra has to be followed.
  13. Juancho

    Juancho pfm Member

    Completely agree with this, although the main psu caps are worth replacing if 20 years old as they have to withstand the larger ripple of the power supply. This goes for valve amps and even more so for solid state. However, in some bad designs of amplifier and CD player there is a strong heat source eg a regulator very close to an electrolytic cap which can dry it out and necessitate its early replacement
  14. markvdv

    markvdv pfm Member

    Panasonic list the use for coupling under prohibited circuits, does that make the Oscon a popular choice maybe ? Forbidden fruits ?

    Guidelines and precautions (OS-CON)
    1. Circuit design
    1.1 Prohibited circuits
    (a) Leakage current of the OS-CON may increase in the following conditions.
    (1) Soldering
    (2) When voltage is not applied ...

    (b) Avoid the use of the OS-CON in the following type of circuits because leakage current may increase.
    (1) High-impedance circuits
    (2) Coupling circuits
    (3) Time constant circuits
    (4) Other circuits that are significantly affected by leakage current
  15. Arkless Electronics

    Arkless Electronics Trade: Arkless

    Yep I'd avoid them as coupling caps. Best results probably from just leaving the original electrolytics in. Excellent article from Wurth Electronik on these caps in last months "Electronics Europe News" so I'll guess that it's probably available from their web site.

    Personally I don't regard the 34 (or any Quad pre really) as a good enough amp to be worth bothering with though... no matter what you do to it. The very fact that it uses CMOS switching, no matter how well done, compromises it fatally to me... and the all op amp design, although less important than the CMOS switching, means it will never be in the the top flight of pre amps...
  16. fatmarley

    fatmarley Registered User

    I'm pretty sure at least one person tried Oscon Sp for coupling In a Naim preamp years ago and reported a harsh or bright sound. Great for decoupling though...
  17. markvdv

    markvdv pfm Member

    When power comes from eg LiPo batterypack the Oscons for decoupling can seriously reduce playtime, I read a festival boombox thread once LOL
  18. 337alant

    337alant Negatively Biased

  19. fatmarley

    fatmarley Registered User

    I don't think all aluminum polymer caps are made equal (were Oscon SP polymer?). Not sure If It's connected but the old Oscon SP was very low ESR and people didn't seem to like them as coupling or feedback caps.
  20. markvdv

    markvdv pfm Member

    Well poster linked to Panasonic, so I copied and pasted Panasonic warnings. I could have posted exact same warnings by Nichicon, Nippon/United Chemicon, NIC Components (prior to Sanyo there was NEC and NIC producing these kind of aluminium solid electrolytics, so NIC probably has longest experience with these solid types), Rubycon, Elna... do I forget a decent brand, google and you'll see these warnings. Wurth seems to sell chinese capacitors with their name printed on them, maybe the chinese forgot to tell the germans LOL or the germans decided it didn't fit their marketing brochures.

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice