greetings ladies and gentleman...
i've done a deep research regarding capacitors on the power supply and other stages within the tube amp, to understand the role of larger capacitance electrolytics and the options to swap them out when recap time comes by;
the most reasonable approach when recapping with the same specifications electrolytics is not an option, is the following answer:
-there is no generic rule, it depends on the circuit design;
so... as attached, the diagram/schematic of the device is here for analysis and here goes the question and this topic subject:
when replacing the electrolytics for MK-types rated for run, with voltage tolerances above or equal to the stock electrolytics, how lower can i go for the capacitance, since some electrolytics are rated for 10.000uf and i don't want to parallel many motor run caps to achieve that sum?
(i am aware that lowering the capacitance will introduce some effects such as sag, decreased bass response, but anyway... how lower can i go on the capacitance?)
much thanks
ps: i've already reaped the c1 and c4 electrolytics and installed 1uf MKT blue bricks @250v on each spot... since c4 was originally 22uf, i thought it could change much, but i confess the device is running much better now, but what about c14, c15, c16, c17, c18, c19, c20, c21 and c26?
i've done a deep research regarding capacitors on the power supply and other stages within the tube amp, to understand the role of larger capacitance electrolytics and the options to swap them out when recap time comes by;
the most reasonable approach when recapping with the same specifications electrolytics is not an option, is the following answer:
-there is no generic rule, it depends on the circuit design;
so... as attached, the diagram/schematic of the device is here for analysis and here goes the question and this topic subject:
when replacing the electrolytics for MK-types rated for run, with voltage tolerances above or equal to the stock electrolytics, how lower can i go for the capacitance, since some electrolytics are rated for 10.000uf and i don't want to parallel many motor run caps to achieve that sum?
(i am aware that lowering the capacitance will introduce some effects such as sag, decreased bass response, but anyway... how lower can i go on the capacitance?)
much thanks
ps: i've already reaped the c1 and c4 electrolytics and installed 1uf MKT blue bricks @250v on each spot... since c4 was originally 22uf, i thought it could change much, but i confess the device is running much better now, but what about c14, c15, c16, c17, c18, c19, c20, c21 and c26?
Attachments
Last edited:
Substantially reducing the value of a cathode bypass capacitor will cause the gain to drop
significantly in the bass. The frequency where it starts to drop is 1/(2Pi x R x C).
The amount of gain loss depends on the particular circuit.
thanks, i had read some theory about it before and now that you've said it then i'm more confident that its a fact, if not much to ask, what could you tell me about lowering the power supply caps and the infamous c26, rated for 10.000uf (at least i've seen 10k micro, but i might be wrong, who knows eh?)
the c14 and c15 @100uf are the coupling ones or are they bypass caps? how lower can i go within desireable audible range?
isnt c26 10000uf? im lost now ;_;
Last edited:
Yes, C26 is (10000 uF). However, C26 has nothing to do with the audio signal going through this amp; the only thing C26 does is provide DC for the heaters for V1 & V2. The other heaters in the amp are fed with AC.
This design looks like a guitar amp, and for a guitar amp, you do NOT want full audio bandwidth. First of all a six-string guitar in standard tuning doesn't put out any signals lower than 83 Hz, and secondly, most guitar amps roll off the signal well above 83 Hz, to keep the sound from becoming thick and muddy.
Presumably Blackheart already put some work into finding the right capacitance values to produce the sound they wanted. My suggestion is not to change any capacitor values - not unless you have already played through the amp in stock form, and have found that you are not happy with the stock frequency response.
-Gnobuddy
Electrolytic caps are a blessing. No, they don't live forever, but they are MUCH cheaper than any other type.
While I am not one to argue "tone", I suspect that in a guitar amplifier, the (now small) "faults" of electrolytic caps are not significant.
E-caps only last decades, but are *cheap* to replace in 2028, 2048, etc. Many of us expect maybe one more e-cap replacement before we switch to heavenly harp (or hammers of hell). Younger guys have the dexterity to replace e-caps between sets (though it is easier off-stage).
While I am not one to argue "tone", I suspect that in a guitar amplifier, the (now small) "faults" of electrolytic caps are not significant.
E-caps only last decades, but are *cheap* to replace in 2028, 2048, etc. Many of us expect maybe one more e-cap replacement before we switch to heavenly harp (or hammers of hell). Younger guys have the dexterity to replace e-caps between sets (though it is easier off-stage).
the BHHD-15 came up pretty cool as stock, it's designer (Pyotr Beltov) did a solid job and i have no complaints
it started when i had to swap tubes like 4 years ago... that time, the tubes were already very worn because i used to play about 10hrs a day for years til then...
since i had financial problems and little knowledge on tech/gear, i've purchased the cheapest tubes available:a pair of 6n2p + 6n14p.
i didnt know the 6n2p had to tweak the pins, and the problem started... i ended up burning alot of the pcb (of which is pretty strong, rugged... but nothing outstand a 80w soldering iron rated for anything but electronics)
then i started experimenting some stuff such as swapping the 22n tone caps for PIO, but the PIO i could afford came from a drifed bath and later on a friend measured them and they had 28-30n capacitance... he also gifted me a hikari electronic iron, claiming he felt pity for the amp and said it was a big waste for a crazy dude like me to keep torturing and burning such fine build device with an huge 80w iron
before that discovery of the drifted capacitance, i thought the output transformer was gone because the sound was saturating, heavily constipated or blocked, whatever...
the woman who build my custom output transformer did a great job btw and it improved alot the tone
since the amp was all screwed up, i've kept experimentating and for my surprise, it gets even better than stock at every mod i do (although, Pyotr claimed he designed this amp to be modded)
after getting rid of the 'lytics (high esr, bad treble smoothing, evil and short lifetime) for the sake of the tone and to be sure i'd never have to worry about it again, i'll install a baxandall tone stack after the PI and use the mid pot as a ground lift. i think it will sound killer;
thats my story and i understand what you've said
i agree with you and maybe my life would be simpler if i call it a day and leave it as is
but i could enjoy the results of replacing the lytics for polypropylene with lower capacitance:
-absence of further recapping
-more high harmonics passing through the device, meaning less need to increase treble/presence, resulting in a warmer mid crank without the 'blanked muffled syndrome' or shrill highs when maxing this controls, of which also thin the sound because this TMB stack scoops mids when other controls are increased
-slight sag in the sound, of which can be pleasing
-less muddyness on the sound due to less bass
you shall know my clean settings aint clean at all (ac/dc territory) and my dirt settings is a fuzz higain (i slam the front end already hot with 2x ecc83 gain pedal and believe me... no buzzsaw mushy tone... plenty of headroom lol)
for this reasons i humbly ask your help to understand how low can the capacitance go, so i can slowly recap. i quit physics university after almost 2 years failing differentials math, its hard for me. i prefer to play and songwrite, but i don't wanna be a leech after a magic formula... at least we learn something, and i hope this topic piles up some knowledge to help as many
thanks!
yeah, i've read some stuff saying the lytics improved alot since the early production, and not saying that you are this kind of guys, but when people say 'oh, guitar amplifier... don't mind minor faults... you know... guitars are all buzzy, they will never be hifi' stuff, it doesnt apply for me... maybe im getting koo-koo, my friend PRR, just like that guitarrist... i think his name is eric johnson, that claims he uses rubber rings instead of screws for his pedals because it alters tone
but subtle changes when combined can achieve a pleasant execution that turns me on into playing harder... i dont want to pursuit an endless quest for tone, the microphonics and hum are gone, the headroom and mids increased... i think resolving this lytics and adding a post PI baxandall stack will satisfy what i feel its lacking to call it a day... the baxandall i've learnt how to make and i'm ok with it, so theres this caps thing...
when you say '2028, 2048' you are being somehow optimistic... i don't even know if there will be a tomorrow, but i hope so
i agree its easier to recap offstage and this is also a good reason to get rid of the evil and infamous electrolytics, haha!
speaking of the future... sometimes i wonder how it will be and whats gonna come for us to have fun playing. i've got amused with the samples of this new seymour duncan chorus, the catalina. i had one ehx small clone, the one with 90's chip. pretty good, but lacked headroom (the sound shrunk or got too muddy... too bad i traded it for a DB01 before trying to power it with 18v by two 9v carbon batts in series... maybe it would get me what i needed, since i am a big fan of chorus, but kurt cobain made it really huge and i couldnt do it)
thanks!
thats fine, i thought for a moment that i was mistaken... you know... tiny android device
anyway, wanna speculate which value could be the lowest acceptable?
thanks!
I am not good at lying, so I will be honest: these things will not make an amplifier better. What they will do is suck money out of your wallet.
So why not focus on those things that you love? Keep working and practicing and improving as a player and songwriter. That will take you much further in your development as a musician than any amp-tinkering ever can.
You are not cuckoo, it is normal human nature to fool ourselves when we rely only on our senses. For example, it is well known from decades of research with Hi-Fi loudspeakers, that people will automatically hear better sound from a loudspeaker if it is larger than the other speakers in the comparison. They will also automatically hear better sound if they know the loudspeaker is more expensive.
In your case, you hear better sound because you know the PIO or polypropylene cap is more expensive, and other people have told you it will sound better. So you have a strong confirmation bias that these caps will sound better - and you will then hear exactly what you expect to hear. It is not cuckoo, it's just our error-prone human brain at work. But it is a mistake to trust that fallible brain.
Eric Johnson was, in my opinion, the first shred guitarist who was also musically interesting to listen to. While other shredders sounded as boring as a sequencer driving a synthesizer, Johnson managed to put some expressiveness into his playing by not always playing fast all the time. He was/is an amazing guitarist, and I would love to have as good guitar technique as he does.
But I would not trust Eric Johnson to operate on my heart, develop a new submarine, calculate a trajectory from earth to Saturn, or write a great novel. Nor do I take his supposed beliefs regarding the sound of rubber rings or different brands of 9-volt batteries seriously. If he really believed those things, he was wrong.
They are neither evil nor infamous, though some people have fooled themselves into believing those things.
Think about this for a second: virtually every piece of music recorded in the last century has gone through an enormous number of electrolytic capacitors before it ever reached your ear. There are electrolytic capacitors in microphone preamps, in mixing desks, in all the other audio processors (compressors, FX units, etc), in reel-to-reel records, in A/D converters, in D/A converters, in PC sound cards, in audio preamps, in audio power amps, even in many loudspeaker crossover networks. And let's not forget they are also in every power supply.
So if electrolytic capacitors were in fact doing horrible things to the sound, then every recorded song you have ever heard would have sounded horrible, including Kurt Cobain and other musicians you admire...yet you admire his sound, so isn't that the proof that electrolytic capacitors do not damage the audio signal going through them?
Saying that electrolytic capacitors are evil is like saying that major triads are evil. In fact, major triads are a crucial building-block in Western music, just as electrolytic capacitors are a crucial building-block in electronics circuitry. You may manage to write a song without a single major triad in it, but it will be a weird and compromised song. You may manage to build an audio circuit without ever using an electrolytic capacitor, but it will be a weird and compromised circuit.
Forget the audiophool nonsense...hang on to what matters. Your music, and your muscianship. Put your time, energy, and money there. That is where you will get the best return on your investment.
-Gnobuddy
No it's not. I'm not going to discuss the logical flaws in the statement above.
Back on topic: I will generally agree that "damage" caused by well selected and well biased electros is vastly overrated.
The main reason for replacing electrolitics with motor run caps in guitar amps is because electros dry out and fail, eventually.
Whether that's a good investment is another thing altogether. I'll stick to electros for my.
With a passive, unregulated power supply, lowering the supply capacitor values could cause
low frequency instability (motorboating), increased hum and noise, and phase and amplitude
aberrations in the bass. You might prefer the sound with film capacitors of similar values, though.
Lowering the C26 would cause more ripple voltage on the filaments, and more noise and hum.
I've seen as low as 2000uF used for this function, but use 10000uF myself.
C14 and C15 are bypass capacitors for the EL84 cathode bias resistors. Reducing their values
would again reduce the gain in the bass by the same formula. They must both be the same value.
Last edited:
I am not good at lying, so I will be honest: these things will not make an amplifier better. What they will do is suck money out of your wallet.
So why not focus on those things that you love? Keep working and practicing and improving as a player and songwriter. That will take you much further in your development as a musician than any amp-tinkering ever can.
You are not cuckoo, it is normal human nature to fool ourselves when we rely only on our senses. For example, it is well known from decades of research with Hi-Fi loudspeakers, that people will automatically hear better sound from a loudspeaker if it is larger than the other speakers in the comparison. They will also automatically hear better sound if they know the loudspeaker is more expensive.
In your case, you hear better sound because you know the PIO or polypropylene cap is more expensive, and other people have told you it will sound better. So you have a strong confirmation bias that these caps will sound better - and you will then hear exactly what you expect to hear. It is not cuckoo, it's just our error-prone human brain at work. But it is a mistake to trust that fallible brain.
Eric Johnson was, in my opinion, the first shred guitarist who was also musically interesting to listen to. While other shredders sounded as boring as a sequencer driving a synthesizer, Johnson managed to put some expressiveness into his playing by not always playing fast all the time. He was/is an amazing guitarist, and I would love to have as good guitar technique as he does.
But I would not trust Eric Johnson to operate on my heart, develop a new submarine, calculate a trajectory from earth to Saturn, or write a great novel. Nor do I take his supposed beliefs regarding the sound of rubber rings or different brands of 9-volt batteries seriously. If he really believed those things, he was wrong.
They are neither evil nor infamous, though some people have fooled themselves into believing those things.
Think about this for a second: virtually every piece of music recorded in the last century has gone through an enormous number of electrolytic capacitors before it ever reached your ear. There are electrolytic capacitors in microphone preamps, in mixing desks, in all the other audio processors (compressors, FX units, etc), in reel-to-reel records, in A/D converters, in D/A converters, in PC sound cards, in audio preamps, in audio power amps, even in many loudspeaker crossover networks. And let's not forget they are also in every power supply.
So if electrolytic capacitors were in fact doing horrible things to the sound, then every recorded song you have ever heard would have sounded horrible, including Kurt Cobain and other musicians you admire...yet you admire his sound, so isn't that the proof that electrolytic capacitors do not damage the audio signal going through them?
Saying that electrolytic capacitors are evil is like saying that major triads are evil. In fact, major triads are a crucial building-block in Western music, just as electrolytic capacitors are a crucial building-block in electronics circuitry. You may manage to write a song without a single major triad in it, but it will be a weird and compromised song. You may manage to build an audio circuit without ever using an electrolytic capacitor, but it will be a weird and compromised circuit.
Forget the audiophool nonsense...hang on to what matters. Your music, and your muscianship. Put your time, energy, and money there. That is where you will get the best return on your investment.
-Gnobuddy
i totally understand, in my 'defense' *laughs* today i use pio caps where i need HF attenuation such as dull/hf bleeding, etc... due to its snail pattern, of which causes inductance.
i agree that i shalt avoid getting into gimmicks and i know that 'lytics are the base for all audio gear since a long time, and some lytics problems such as dullness can be fixed with simple things such as placing a low capacitance ceramic/mica/whatever in parallel with lytics, for example
i still would like to get the formula ready and written in my book, regarding the lowest capacitance values for replacing the lytics, because when recap time comes by, then i'll be safely ready to put in motor run caps and also benefit from less bass and a more defined tone since i use my gear with hot standards, this amp is older than 10 years and there was a time that i used to play long hours daily with it;
i must be realistic that someday i will need to recap it and i liked so much your kind advices into focusing the songwriting/playing. never having to recap the amp again would get me an upstart. i also wouldnt enjoy to have the amp failing onstage, it would put me down, talking about efficiency/reability and putting aside the tone things that could enter the gimmicks area;
talking about your tastes and shredding stuff...
check this out
00:00 Hashashin // 11:50 Leoes Chorando // 25:11 Insanidade Nacional by Netuno - Listen to music
this one above was recorded 6 years ago. 26 minutes of live freestyle shredding... i played the guitars, the drums were performed by my wife and she was learning to play back then
the link below is a pre release of our demo album with 9 songs, recorded this year with no PC, no interface... i just mic'ed straight into the android phone's headset pin with adapters... between the mic and phone i've boosted the signal with a blackstar ht-dual pedal on clean settings... the solo is pretty cool and i improvised live
YouTube
i hope you enjoy it, and thanks for your feedback... sorry for being stubborn about installing motor run caps with lower capacitance instead of electrolytics, i totally understand your point and also agree with you;
trust me i dont seek to be a gimmicks fool, i only want to sharp my 'swiss army knife' into a standard that i can rely into when the war call heeds me, to cut alot of throats without losing the edge or getting too tired making a dull knife enters the flesh;
too much bass and it gets muddy on high gain/volume and it also can conflict with drums/bass
too much highs and the sound gets unpleasant, too thin and shrill, and also can muffle the clarity of vocals/lyrics
thats also a reason i'll have to remove the tone slope and install baxandall TS. for the mids without becoming muffled... i've been fighting VS my tone stack for like many, many years. i've had some audible improvements that aint no gimmick and also recordings to compare;
about the speakers, i agree too... for example, i have an el84 SE PTP amp that sounds way better with a pair of 15/20w 8 inches speakers than with the wider range 12 inches... its all about matching whats handy
thanks!
yeah, i agree with you.
focusing on stability and reability and putting aside any tonal characteristics, its valid to install motor run caps, specially when theres no more than 6-8 to be replaced;
money wise, i see outside my country things have different price rates.
for example, the catalina chorus by seymour duncan costs no less than what i've paid for the BHHD-15
the lytics cap can cost almost the same thing as motor run caps.
for ex: motor run caps: 5-8 BRL... nichicon gold electrolytics: 4,5 - 6 BRL
then if consider the recapping someday, without worrying about the failure rate (of which i agree is overrated) or claiming about tonal improvements over electrolytics, its already a good investiment
thanks!
.
I respect musicians' ears. As I said, I won't debate "tone".
From an engineering view, "too small" B+ filter caps will give buzzy DC and buzz in the speaker. They also allow bass to sneak-back from high-signal later stages to hi-gain early stages and allow motorboating.
Calculating these effects requires math, of the sort you admit is not your style. It IS hard dull math and won't be done for free.
There is an easy and "correct" experimental method. Start replacing 47u and 22u caps with 10uFd. That amplifier is WAY "over-filtered" and won't crap-out even with much less uFd at every stage. C21 is the main cap. A smaller value here is a little less raw B+ with significantly more buzz. Even so you could almost get away with 10uFd into push-pull pentodes and an open-back cabinet. However this amp has an added level of filtering R29 C20. Even 10uFd here would take the high harmonics off the buzz. So that much may work. Keep going down the line C19..C16. I suggest only changing one at a time, then playing at ALL tone and volume settings to be sure the thing is not tending to excessive instability, howl, thump, whatever.
That 10,000uFd is not in the audio path directly. It "lays next to" the audio path, running to the heaters in every tube. In hand-wiring it is possible to have low hum with pure 50/60Hz AC heat. On PCB construction, hum happens. The low cost of rectifiers and large e-caps encourages designers to use "DC" heat on low-level stages. However even 10,000uFd here has high 100Hz ripple: maybe 0.5V. With strong buzz overtone to past 500Hz, which is far more audible than 50Hz hum. If you don't mind the background hum, leave this as-is; for church/wedding work I would consider a modern switched DC supply for lowest heater crap.
From an engineering view, "too small" B+ filter caps will give buzzy DC and buzz in the speaker. They also allow bass to sneak-back from high-signal later stages to hi-gain early stages and allow motorboating.
Calculating these effects requires math, of the sort you admit is not your style. It IS hard dull math and won't be done for free.
There is an easy and "correct" experimental method. Start replacing 47u and 22u caps with 10uFd. That amplifier is WAY "over-filtered" and won't crap-out even with much less uFd at every stage. C21 is the main cap. A smaller value here is a little less raw B+ with significantly more buzz. Even so you could almost get away with 10uFd into push-pull pentodes and an open-back cabinet. However this amp has an added level of filtering R29 C20. Even 10uFd here would take the high harmonics off the buzz. So that much may work. Keep going down the line C19..C16. I suggest only changing one at a time, then playing at ALL tone and volume settings to be sure the thing is not tending to excessive instability, howl, thump, whatever.
That 10,000uFd is not in the audio path directly. It "lays next to" the audio path, running to the heaters in every tube. In hand-wiring it is possible to have low hum with pure 50/60Hz AC heat. On PCB construction, hum happens. The low cost of rectifiers and large e-caps encourages designers to use "DC" heat on low-level stages. However even 10,000uFd here has high 100Hz ripple: maybe 0.5V. With strong buzz overtone to past 500Hz, which is far more audible than 50Hz hum. If you don't mind the background hum, leave this as-is; for church/wedding work I would consider a modern switched DC supply for lowest heater crap.
Logical flaw - the capacitors could "damage" the signal in such a way that it sounds glorious. That's a "flaw"?
Keep in mind that if I were trying to decide for myself if magic mermaid-oil capacitors would change my sound, I wouldn't settle for any kind of purely logical reasoning as a proof or disproof: tens of thousands of years of human history has shown beyond any doubt that reasoning is subject to being terribly wrong.
So I would follow the scientific method. I would look for trustworthy experimental results, measured frequency responses using both types of capacitors and the same RC time constant. If the two frequency responses were within, say, 0.3 dB or less over the frequency range of interest, then there is no audible difference between the capacitors. QED.
Audio once used to be an engineering field. Now what's left of it on the Internet has degenerated into mostly superstition / religon / philosophy, where we might as well be reading tea leaves at the bottom of the cup and arguing about "horseness", the magical quality that makes a horse a horse, as medieval philosophers once did.
Exactly. Overrated in the way that deadly attacks by flying shoals of barracuda and rabid unicorns are overrated.
While it's unlikely I'll live long enough to need to replace the electrolytics in anything I've built in recent years, I do like the thought of using capacitors that have lower DC leakage currents and smaller physical size in the same capacitance value. If such capacitors were available without audiophool-ransom pricing, I would happily use them.
Guess what, such capacitors are indeed available. So for the last few years I've been using multi-layer ceramic caps, which are easily available up to 22 uF in voltage ratings up to 100 volts DC or so, and are perfect replacements for cathode bypass caps in valve guitar amps. Here's an example: https://www.mouser.ca/ProductDetail...EpiMZZMukHu%2bjC5l7YVfyOzdN3dBI72fO6f%2bPOhc=
I will add that the 22uF cathode bypass caps that Leo Fender used for the 12AX7 triodes in his amp designs are far too big. A more appropriately sized 2.2 uF, 25V, multi-layer ceramic cap costs 32 cents (Canadian) each if you buy ten: https://www.mouser.ca/ProductDetail...sSlwiRhF8qqG20SskRxtTkF0PVl8VPtOTxS0/ej6CWg==
To be very clear: these ceramic caps sound exactly the same as all other caps of the same capacitance. Capacitors have no "sound" of their own.
Decades ago, ceramic capacitors were sometimes microphonic (the ceramic dielectric inside is a close cousin to the piezo ceramic material used in guitar piezo transducers and undersaddle pickups.) I haven't found any trace of microphony in these new multi-layer ceramic caps though.
-Gnobuddy
For me, this is the only way to go with new builds.
It's a pity 6.3 volts DC doesn't happen to be a common voltage for contemporary switch-mode power supplies, but that is easily worked around. You can either get a power supply that puts out a little more than 6.3 volts and drop the excess voltage in a series resistor, or tweak a 12V power supply to put out about 12.6 volts, and use either 12V valves or groups of two identical 6.3 valve heaters wired in series.
-Gnobuddy
thanks, i've wrote it down... i think the motor run caps here go up to 20uf for sure, maybe a little more... so i think its gonna be within range and i feel relieved to know, thanks for applying your precious knowledge for free
if theres space left inside the chassis, i'll try to filter closest to the original specs... i know its wrong to use long wires and connect the caps outside, far from the pcb... lets say where the tubes and transformers are assembled, so i won't assemble in the wrong places because it could be worse than recapping with specs electrolytics;
about the heater subject, i'll take up more info about it to choose wisely and i appreciated your explanation and i'm very thankful to you;
Logical flaw - the capacitors could "damage" the signal in such a way that it sounds glorious. That's a "flaw"?
Keep in mind that if I were trying to decide for myself if magic mermaid-oil capacitors would change my sound, I wouldn't settle for any kind of purely logical reasoning as a proof or disproof: tens of thousands of years of human history has shown beyond any doubt that reasoning is subject to being terribly wrong.
So I would follow the scientific method. I would look for trustworthy experimental results, measured frequency responses using both types of capacitors and the same RC time constant. If the two frequency responses were within, say, 0.3 dB or less over the frequency range of interest, then there is no audible difference between the capacitors. QED.
Audio once used to be an engineering field. Now what's left of it on the Internet has degenerated into mostly superstition / religon / philosophy, where we might as well be reading tea leaves at the bottom of the cup and arguing about "horseness", the magical quality that makes a horse a horse, as medieval philosophers once did.
Exactly. Overrated in the way that deadly attacks by flying shoals of barracuda and rabid unicorns are overrated.
While it's unlikely I'll live long enough to need to replace the electrolytics in anything I've built in recent years, I do like the thought of using capacitors that have lower DC leakage currents and smaller physical size in the same capacitance value. If such capacitors were available without audiophool-ransom pricing, I would happily use them.
Guess what, such capacitors are indeed available. So for the last few years I've been using multi-layer ceramic caps, which are easily available up to 22 uF in voltage ratings up to 100 volts DC or so, and are perfect replacements for cathode bypass caps in valve guitar amps. Here's an example: KTD250B226M55A0T00 United Chemi-Con | Mouser Canada
I will add that the 22uF cathode bypass caps that Leo Fender used for the 12AX7 triodes in his amp designs are far too big. A more appropriately sized 2.2 uF, 25V, multi-layer ceramic cap costs 32 cents (Canadian) each if you buy ten: FG28X5R1E225KRT00 TDK | Mouser Canada
To be very clear: these ceramic caps sound exactly the same as all other caps of the same capacitance. Capacitors have no "sound" of their own.
Decades ago, ceramic capacitors were sometimes microphonic (the ceramic dielectric inside is a close cousin to the piezo ceramic material used in guitar piezo transducers and undersaddle pickups.) I haven't found any trace of microphony in these new multi-layer ceramic caps though.
-Gnobuddy
i agree with you regarding some people expect miracles from mermaid oil caps filled with star-alloy foil, and since a design is well made, with stable components, the results will surely humiliate the so called 'gimmicks gurus'
you mentioned 'leakage' in a different and humorous point of view, per se, in defense of electrolytics... should i have leakage concerns when installing motor run polypropylene caps?
about capacitors have 'no sound on their own' i have no study to have the balls to disagree or agree with you, and probably you're right, since science/electronics is exact (as my old experimental physics teacher used to say) and current is current. 92736492732 electrons flowing in x rate, etc, etc, will be the same no matter which source makes it happen, yes... its like buying bottled water... no matter if its from coca-cola, unilever, pepsico, its all water, regardless the bottle material, color, shape, water source... it will all quench our thirst, no matter if x water has y% less strontium or z company owns a water font with less sodium, i back up your simplicity and effective way... this path surely leads to a happier life with better, practical results, theres no doubt about it
yet, on my ignorance, i dare to state something in defense of the components variety, and absolutely NOT considering gimmicks, but the wise choices, for example:
pio caps tend to act like inductors, their absortion and emission is more subtle, which could be good in parts where attenuation is required, sparing the use of inductors or resistors, this could lead to less components, less loss, noise, etc...
micas have lower internal resistance, fastest transfer, polystyrene and mkp/mkt are more neutral
ceramics have the unique behaviour of vibrating on high flows or frequencies
i agree with you that the differences on 1 cap type vs another (like one on one duel) are ignorable, yes
but 20 capacitors in one circuit can lead to audible differences, since some low side effects such as 1-3% THD being amplified by the circuig and then pushed by the speakers can be noticed and depending on the application, desireable (lowest % than differences from many caps)
for example, i've swapped ALL non polar caps between 10 and 100 nanofarads for PIO in my amp, it sucked and i regretted it. the device was dull, lifeless... at least it wasn't irritating, spikey, but i couldnt even concentrate on the songs
so, like chess, if we strategically place the right parts... theres a bigger chance to achieve the goal of which is mainly victory (unless you're trying to teach your chess partner, prove a point or just make the match long and boring)
thats why i came up to ask all of you for help by advising me:
-the polypropylene motor run caps as experimented by other 'audio scientists' had success on coupling, bypass and power supplies by their reliability, lowest resistance, neutrality, speed and ability to pass wider frequencies of which can benefit the tube amp
-but the capacitance values for safe and useable operating conditions were a concern, since capacitance availability per single unit is restricted to a narrow list and the useful spots on chassis could be limited if the need to associate multiple capacitors in parallel to sum the optimal capacitance... and since less is more for tube amplification, even if multiple caps in parallel could fit... i think it shouldnt be of expertise to stack more components... i prefer to reduce all i can in the amp
overall, i appreciate very much your quotes since it helps not only me, but contributes heavily to pile up more and more info on this subjects... some people dislike when other fellows come to disagree in the topics they create, or disencourage them saying reasonable truths in order to help them, or even by not answering directly what the topic creator wanted to know, but i salute you and i really enjoyed typing words with you and i encourage you not to be disencourageated when someone gets mad with the truths you said... if someone doesnt want to learn or to be reasonable, keep it up because someone will, such as the way i've enjoyed it;
thanks!
Last edited:
very thanks my friend, the posts delayed and i couldnt see your wise advices until now;
so i could lower some stuff to 20-30uf, but c14, c15 and c26 demand exact values for a decent operation, of which leads to ponder that tube related filters such as bypass and heater filament need a higher, tighter tolerance to designer's specs, but the power supply ones can take a little drift downwards with the negative side effect described as 'sag' if not going lower than 20uf?
of course, with the exception of preamp/ecc83 triodes as our friend Gnobuddy said and i also previously tried and related on first posts, regarding the c8 22uf into 1uf, with great results
much thanks
Last edited:
None of these values have to be exact at all. A larger C26 gives less hum.
For C14 and C15, much smaller values (but still kept equal) will reduce the
lower bass, but with film capacitors instead, you may prefer the sound.
The C8 and C9 are 0.22uF (22nF) coupling capacitors, and can be made smaller
(but still kept equal), with the same reduction of low bass.
it seems to be a very intelligent approach and i'll look forward into it, considering the high capacity demand, it looks like the DC regulated PSU is gonna be rated for 10A... do you know if any of that old national TO-3 metallic transistot with 2 pins that require a large dissipator is in specs to nail the job?
sorry i meant c4 22uf electrolytic that i've already decreased to 1uf
the c8 and c9 are currently blue block MKT of the same specs as original (22nf)
it seems that for c26 its gonna be either the good 'ol 'lytic or a regulated psu...
intresting fact that i didnt know... c26-wise, would it be any benefits in putting another 10kuf in parallel or swapping in a 20kuf if theres any easily available?
for heater, putting in parallel one really small mica, lets say 10nf or some 1/10th of the lytic (1uf or 2uf MKT) would improve anything on the tube performance, or the quantic relationship between high frequencies and the filament could cause microscopic vibration that could lead to undesireable effects such as feedback/noise/microphonics? if so, the high ESR of the electrolytics could be a plus on this application. if the answer is yes, i dare asking if some pio cap in parallel or inductor in series could tame even more this micro event;
or could the high frequency be good for the heater?
... or... the heater filament is invulnerable to mechanical tainting?
At some point, increasing C26 will increase the noise, due to the charging pulses getting
higher and narrower. The diodes could be stressed as well.
Why not try the small cap and see? It shouldn't hurt anything.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.