Bypass caps

May 15, 2022
 by Paul McGowan

48 years ago, in 1974, Stan Warren and I got an insider tip we considered at the time to be solid gold.

Bypass caps.

This was back in the days when Bob Carver’s Phase Linear 400 amplifier was what anyone “in the know” was using to power their speakers. That Phase Linear amp was sooooo much better sounding than any other solid-state amp of that day (SAE, Crown, and a myriad of Japanese brands) that if you could get your hands on one that’s what you did (unless you could afford Audio Research which none of us could).

No sooner had we acquired a Phase Linear to replace Stan’s SAE 31B than a friend came over with his own PL 400 for us to compare. Holy Moley. The friend’s identical model amplifier was stunningly more open, airy and musical. And not just by a little bit. Everything was better about his Phase Linear 400 than ours.

He popped open the lid and pointed to two small blue capacitors tied across each of the screw terminals of the amp’s power supply capacitors.

“That’s the difference?” we asked.

It was and to prove his point he unscrewed one leg of the small blue caps and fired back up the amp. Darkness was again upon us. The airy openness and improved transients we had heard before were muddled and lost.

That small 0.1mF film capacitor strapped across the amp’s sluggish power supply electrolytics made a huge improvement. They work because electrolytics aren’t great at higher frequencies. By paralleling a small film capacitor to fill in where the bigger caps aren’t that good makes an amazing improvement.

To this day, every electrolytic capacitor in any PS Audio product has been lovingly bypassed with a smaller hand-picked film capacitor.

Of course, this drives the measurement crowd crazy since it’s not in the signal path and (horrors) only in the power supply. This measurementist’s view ignores the notion that an amplifier is a power supply with an output valve that modulates that power supply.

But then, you’d know that simply by listening.

Because it matters.

Subscribe to Paul's Posts

26 comments on “Bypass caps”

  1. Anything that “drives the measurement crowd crazy” is usually a
    win for audio (aural) sanity… & tends to put a big grin on my face 😉

    **Off Topic**
    Ukraine wins ✨Eurovision 2022✨
    “All we are saying, is give peace a chance” ✌

  2. A good story / experience. A great example of finding something that works applying it and sticking with it.

    I for one like the measurement crowd. Others here not so much. Seems to me measurement is very easy… Take 1 farad of electrolytic capacitance add 1mF of ‘special’ thin film across the electrolytic – end up with 1.000001 farads

    What products of PSA have big electrolytic electrolytic caps other than the BHK amps and regens?

    1. Well, pretty much all our products do. Even the preamp has good sized electrolytic power supply caps. The only ones that don’t are the Class D SMPS models that we buy from ICE power (like our Stellar amps).

    2. Mike,

      The key measurement missing from adding the two capacitance values together is ESL (or equivalent series inductance). This “feature” of a capacitor is much larger (bad) for big electrolytics and much smaller (good) for smaller value film capacitors. ESL gets in the way of the “perfect capacitor” and effectively removes it from the circuit at higher frequencies.

      BTW, I think Paul meant 0.1uF (microfarads) not millifarads. A Farad is a large unit = 1,000,000 uF which is more than most amplifier power supplies provide for DC bus filtering.

      Best,
      Kurt

      1. Thanks Kurt,

        What I was eluding to was the fact that capacitance can be measured, and I would assume ESL also. Your explanation is well received.

        I too got confused by the m versus u.

        You’ll notice that when I did my math I added 1uf – but up on rereading I notice that if milli farads is what Paul was referring to then I should have written 0.0001 farads Oops ✌️ 😀

        Have a great day and thanks for the response.

  3. The power supply is very important to the way an amplifier sounds, if it isn’t up to the task across the whole frequency range the amp will never reach it’s potential.

    If you look at all the great sounding amps and preamps a lot of care and thought goes into the power supply section because it can make or break a product.

  4. How long is it going to take audiophiles to realize that the power supply is the real amplifier; the juice comes directly from the power supply, not from what we thihk of as the circuit. The circuit is a varaiable valve that opens and closes the power supply producing the enlarged signal based on the input signal. A cheap power supply doesn’t affect specs much and can save lots of money, but it does affect reproduction. The better the power supply the better the amplifier sounds since I repeat the power supply is the amplifier.

    A simple circuit with a great power supply will out perform a super circuit and ordinary power supply. That was the basis of the original Naim amp circuits.

    1. Agreed!
      But not just amps.
      30 years ago I swapped out the factory 600mA power supply that
      was supplied with my Audio Alchemy DAC for an after-market
      6A beast & the improvement in sound quality was phenomenal.
      That was all the confirmation that I needed.

      1. The most costly parts of an amp are the chassis(and front plate) and the power supply. It doesn’t take that much of a power supply to measure well on conventional measurements like harmonic and IM distortion so you can save tons of money on a power supply and still advertise great specs. But it takes great power supplies to handle the sudden huge dynamics of great music and that can tear a power supply apart unless it’s well designed(money). Voltage drops on swings and current supply can be insufficient and the the amp is not the amp it is under simple conditions and the reproduction changes significantly.

        1. The power casework and power supply are not just the most costly parts, but also the heaviest parts. On my Hercules II stereo power amp ( dual mono ) most of the casework is 1/8″ aluminum, and there two 3000 are VA toroidal transformers ( left and right ) and the big PS capacitors look like those 10 oz beverage cans that the airlines use to use.

  5. Paul just curious if you are forgetting the Dynaco 400? Extra capacitance made this amp a great alternative to the Phase Linears of the time

  6. I’m pretty sure you can measure these effects. Should be fairly easy if you’re measuring the right thing. The differences are published in the capacitor specs. Walt Jung was an early advocate too, with a series he and Robert Marsh (IIRC) wrote in the 70s for Audio magazine.

    I do believe this is just good audio engineering practice that most manufacturers follow (though I can’t claim I’ve heard significant differences, even back when my hearing was very good). Bypass caps are also generally used in digital and other circuits, for various reasons.

  7. In the late 70s I was selling the Blaze Linears, but us budding audiophiles at the shop saved our pennies up for the Yamaha and Sony V-FET gear or Luxman. That was back when some of the Japanese companies were putting out great gear, or at least that’s how I remember it!

  8. thanks, Paul for elucidating the function and implementation of bypass caps. not even being able to read schematics, and not having a complete overall understanding of electronics, it is a big help towards that.

    i remember seeing your ads in Stereo Review and Audio Magazine classifieds along with the Tigersaurus amp and other items of interest for a budding audiophile. ALL preamps had phono sections, not to mention receivers. i went to separates quickly soon after my Sherwood s8900a receiver started oscillating in one channel 3 months prior to the 3-year warranty ran out. as soon as it was repaired, i sold it and got my friend’s old Quatre preamp and Spectro Acoustics 100wpc amp, hence no need to get a PS Audio phono.

    the friend i got those components from had a Phase 400 and i took my two pairs of Dynaco A25s and we stacked them ala Advent style and they not only played too loud to stay in the room for with very low distortion, the lights in the house would dim on peaks.

    fun times!

    1. Loved my A25’s at the time. My pioneer receiver went into some weird oscillation while playing Live At Leeds at 11 but they didn’t blow. Ended up giving them to my parents and picked up some large Advents.
      Memory lane.

  9. My apologies for a second comment, but Paul’s message of the day has made me quite nostalgic. In my first comment, I mentioned the Dynaco ST400 which taught me about the importance of power supplies and bypass caps early in my listening journey. I had an ST 400 driving Tympani T1C speakers. I first added the external C-100 Capacitor bank and the improvement to bass performance was substantial. A second improvement was bypass caps and it resulted in some of the best sound I have ever had in spite of later more expensive and refined equipment. Granted the room was very good, but I vividly remember goosebumps and being transported to the musical venue….

    1. Hello Willem,
      No need to apologise as nostalgic stories are always appreciated here.
      Well behaved rooms & goosebumps are indeed wonderful bonuses 🙂

  10. The Phase 400 & 700 were the first super powered transistor amplifiers on the market. They proved to be a revelation with the many low sensitivity, acoustic suspension speakers (such as the AR3a, LST, and a legion of imitators) which were the rule of the day.

    They also tended to power (if not ideally) the completely unconventional, astonishingly revelatory Dahlquist DQ10 speakers, but those same speakers, along with the Magnaplanar Tympany 1B, tended to reveal differences between amplifiers rather routinely.

    Subsequent designs, such as the James Bongiorno’s GAS Ampzilla, the Harman Kardon Citation 16, Dynaco 400, and others, were hailed for thier sound superior to Phase Linear, which, in high end circles, was relagated to mid-fi status. Second generation SAE amplifers were also highly regarded.

    1. Wow, does this brief history lesson bring back memories. I lusted over the LST like teenage boys did over Playboy centerfolds. My late brother-in-law had a pair of the DQ10’s. Remarkable speakers.

    2. This post brings back memories of my first decent system: Crown preamp, Phase Linear 400 and DQ10 speakers. Many hours of rock and roll in the 80’s.

  11. I did IT.
    I have made a DYI linear power suply for my small Yamaha streamer. Toroidal transformer, ready to use AC-DC convertor board from amazon. I added extra large electrolytic capacitors. Big difference compared to original power source.
    Today I bought and mounted a small film capacitor – cost about $0.1.
    O.M.G.!
    Thank You Paul. This is almost magic. And also the cheapest tweak so far!

  12. I saw a video quite some time that you did about this same subject. I searched and could never find it again. I sent you a message a few months ago asking this same question. Thanks for answering it. Very interesting and informative. I wish there were more people like you and Bob Carver in today’s world. The Phase Linear and Cube amps were very interesting. I also have one similarly designed (an Adcom GFA-1).

Leave a Reply

Stop by for a tour:
Mon-Fri, 8:30am-5pm MST

4865 Sterling Dr.
Boulder, CO 80301
1-800-PSAUDIO

Join the hi-fi family

Stop by for a tour:
4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram