Hello,
How to determine value of bleeder resistor?
Bleeders paralelled across all PSU caps, or at last cap in filter?
(paralelled across all caps helps to "balance" voltage, but what is the improvement of this?)
How to determine value of NTC thermistor? (amps, or by "open" and "closed" resistance)
Where to locate said NTC in circuit, after mains fuse on PT pri, or on PT secondary HV winding?
How to determine value of "bypass" caps paralelled across electrolytics to improve sound. What type of sonic improvement am I listening for, anyways?
Thanks, as always
Scott
How to determine value of bleeder resistor?
Bleeders paralelled across all PSU caps, or at last cap in filter?
(paralelled across all caps helps to "balance" voltage, but what is the improvement of this?)
How to determine value of NTC thermistor? (amps, or by "open" and "closed" resistance)
Where to locate said NTC in circuit, after mains fuse on PT pri, or on PT secondary HV winding?
How to determine value of "bypass" caps paralelled across electrolytics to improve sound. What type of sonic improvement am I listening for, anyways?
Thanks, as always
Scott
Bleeders aren't always necessary; if your caps are reasonably sized (why I should even have to say that, ehh.. 🙄 ) and the amp is always connected, it'll discharge it within a minute. (As the tubes cool down, emission goes down, but there's usually enough left to bring the capacitors down.)
If you have series connected capacitors, bleeders might be a good idea. Typically, aluminum electrolytics don't technically need this because, if voltage rises above whatever maximum voltage the capacitor is at*, leakage current increases, which will pull more voltage across the other caps, keeping it naturally in balance.
*I say this, and not simply rated voltage, because capacitors 'soften' after a while. If ran under the rated voltage for several years, they will take a set and if you use one at a higher voltage, leakage goes up temporarily - reason being the dielectric decomposes over time; voltage keeps it fresh. Bringing a cap back up to a higher operating voltage is known as reforming.
So... for safe keeping, and for other capacitors such as film types which have low leakage, bleeders are a good idea to keep balance on series strings. This works by creating a voltage divider which swamps the capacitors' natural leakage currents, forcing their DC voltage into line.
As for bleeders in general, as I mentioned, they may not be necessary. If they are (for whatever reason; external PSU which may not always have a load connected which can discharge it, ludicrous caps, etc.), the typical rule of thumb is a bleeder 10% of the output. Oh, and bleeders are also useful for choke-input supplies, to keep a minimum current so voltage doesn't soar. So for a 200mA supply, you'd want 20mA through the resistors.
In general, figure what time constant you want before it's fully discharged and safe to work on - 5 * R * C is a good value.
As for NTC's, these aren't necessary for any reasonable sized amplifier. Once you start drawing a few amperes continuous through the outlet, then you can find a good value. Can't help you on selection off the top of my head.
And finally, lytic bypass? Pffbt. Lytics have 500 times lower ESR than absolutely ANY tube circuit you'll EVER see.
Tim
If you have series connected capacitors, bleeders might be a good idea. Typically, aluminum electrolytics don't technically need this because, if voltage rises above whatever maximum voltage the capacitor is at*, leakage current increases, which will pull more voltage across the other caps, keeping it naturally in balance.
*I say this, and not simply rated voltage, because capacitors 'soften' after a while. If ran under the rated voltage for several years, they will take a set and if you use one at a higher voltage, leakage goes up temporarily - reason being the dielectric decomposes over time; voltage keeps it fresh. Bringing a cap back up to a higher operating voltage is known as reforming.
So... for safe keeping, and for other capacitors such as film types which have low leakage, bleeders are a good idea to keep balance on series strings. This works by creating a voltage divider which swamps the capacitors' natural leakage currents, forcing their DC voltage into line.
As for bleeders in general, as I mentioned, they may not be necessary. If they are (for whatever reason; external PSU which may not always have a load connected which can discharge it, ludicrous caps, etc.), the typical rule of thumb is a bleeder 10% of the output. Oh, and bleeders are also useful for choke-input supplies, to keep a minimum current so voltage doesn't soar. So for a 200mA supply, you'd want 20mA through the resistors.
In general, figure what time constant you want before it's fully discharged and safe to work on - 5 * R * C is a good value.
As for NTC's, these aren't necessary for any reasonable sized amplifier. Once you start drawing a few amperes continuous through the outlet, then you can find a good value. Can't help you on selection off the top of my head.
And finally, lytic bypass? Pffbt. Lytics have 500 times lower ESR than absolutely ANY tube circuit you'll EVER see.
Tim
Hi,
NTC: Negative Temperature Coefficient resistor, a non-linear resisitive element that exhibits a high resistance value at low temperature eventually becoming 0 resistance (or close to) as temperature rises.
In tube circuits these are usually put in front of the rectifiers (silicon) to give the circuit a slow start and thus preventing high inrush currents.
Most NTCs hace an AC rating in the order of 120VAC to 240VAC at a given current.
Maybe there's more variety nowadays, haven't looked around for NTCs in ages...
"Lytic bypass" I assume means putting a smaller value (or several ones) low ESR cap in // with an electrolytic cap in order to reduce ESR at higher frequencies.
With modern electrolytic caps having much improved ESR compared to even just 10 years ago this bypassing is hardly useful.
Besides, if you care for sound, it more often than not sounds worse instead of better.
The argument that ESR of a tube circuit is a factor of 500 higher than the caps feeding it is of course only valid for those listening with earplugs or the auditory impaired.......Or both.
Cheers,😉
NTC: Negative Temperature Coefficient resistor, a non-linear resisitive element that exhibits a high resistance value at low temperature eventually becoming 0 resistance (or close to) as temperature rises.
In tube circuits these are usually put in front of the rectifiers (silicon) to give the circuit a slow start and thus preventing high inrush currents.
Most NTCs hace an AC rating in the order of 120VAC to 240VAC at a given current.
Maybe there's more variety nowadays, haven't looked around for NTCs in ages...
"Lytic bypass" I assume means putting a smaller value (or several ones) low ESR cap in // with an electrolytic cap in order to reduce ESR at higher frequencies.
With modern electrolytic caps having much improved ESR compared to even just 10 years ago this bypassing is hardly useful.
Besides, if you care for sound, it more often than not sounds worse instead of better.
The argument that ESR of a tube circuit is a factor of 500 higher than the caps feeding it is of course only valid for those listening with earplugs or the auditory impaired.......Or both.
Cheers,😉
Thankz man!
When you mention those modern improved ESR caps... you dont mean those darn-make-a-hole-in-my-wallet Black gates ahh.....?
So u usually forgo the bypass caps and just use single good resestor aka gainclone?
When you mention those modern improved ESR caps... you dont mean those darn-make-a-hole-in-my-wallet Black gates ahh.....?
So u usually forgo the bypass caps and just use single good resestor aka gainclone?
To do the calculations, you have to measure the inductance in the PS accurately. Otherwise, just experiment with different small cap values. The bypass takes care of the high freq ringing in the PS which eventually will leak into the sound. There is currently a thread in the gainclone section which discuss this topic.
Hi,
Nah...Would I do that to you?😀
While I've no idea what you can get your hands on locally, generally speaking Philips' more recent efforts are surprisingly good (BC Components is the name of the daugher company, IIRC) but there are others.
When in doubt consult the datasheet of the cap.
If you don't need big C values (the higher the B+ the less Cs you need for an equal amount of stored Joules) the better ones besides the to be fased out BGs are polypropylene in oil (Unlytic for example) or the plainer vanilla PIO or MKPs.
These are relatively exensive compared to "lytics" but substantially cheaper than BGs and in some respects better even.
A decent amount of Iron Henrys (PS choke) is also hard to beat if you can get it cheap enough... 🙂
I fail to see how a resistor can replace a cap but yes I forgo bypass caps as they sound horrible whatever values and/or technologies I chose for them...??
Care to explain that?
You mean "a la" GC I suppose. Do you mean the so called snubber circuit?
Cheers, 😉
Nah...Would I do that to you?😀
While I've no idea what you can get your hands on locally, generally speaking Philips' more recent efforts are surprisingly good (BC Components is the name of the daugher company, IIRC) but there are others.
When in doubt consult the datasheet of the cap.
If you don't need big C values (the higher the B+ the less Cs you need for an equal amount of stored Joules) the better ones besides the to be fased out BGs are polypropylene in oil (Unlytic for example) or the plainer vanilla PIO or MKPs.
These are relatively exensive compared to "lytics" but substantially cheaper than BGs and in some respects better even.
A decent amount of Iron Henrys (PS choke) is also hard to beat if you can get it cheap enough... 🙂
I fail to see how a resistor can replace a cap but yes I forgo bypass caps as they sound horrible whatever values and/or technologies I chose for them...??
Care to explain that?
You mean "a la" GC I suppose. Do you mean the so called snubber circuit?
Cheers, 😉
fdegrove said:
And of course, you have once again forced me to demand strict proof that the average tube circuit encountered has an impedance any less than 500 * 0.1 or so (to be generous). Heaters and filaments do not count.
Tim
Hi,
And of course you don't read too well...as usual. Or is it write too?
As any thoroughly indoctrinated "engineer" would, you once again assume perfect components in a perfect world....
Not surprisingly then, ESR won't bother you as you keep moving the goalposts as it best suits you, n'est-ce pas?
Maybe it's time to either "unlearn" a few things or "learn" some more...Better still forget about audio altogether?
Cheers, 😉
And of course you don't read too well...as usual. Or is it write too?
As any thoroughly indoctrinated "engineer" would, you once again assume perfect components in a perfect world....
Not surprisingly then, ESR won't bother you as you keep moving the goalposts as it best suits you, n'est-ce pas?
Maybe it's time to either "unlearn" a few things or "learn" some more...Better still forget about audio altogether?
Cheers, 😉
sorry error:
*the word resestor should be replaced with capacitor.
Ya, I meant ala gainclone: one single good cap as near as possible to the pins, and leave out all the snubbers.
Im going to try the gainclone philosophy in my 12B4 preamp: Move the PS caps as close as possible to the tubes. Problem now is to source for good caps at a reasonable price.
And yap, Im using choke input.
*the word resestor should be replaced with capacitor.
Ya, I meant ala gainclone: one single good cap as near as possible to the pins, and leave out all the snubbers.
Im going to try the gainclone philosophy in my 12B4 preamp: Move the PS caps as close as possible to the tubes. Problem now is to source for good caps at a reasonable price.
And yap, Im using choke input.
If you are using a choke input filter, a bleeder that draws an extra 10% current is a good place to start, i.e. , if the total load on the HT circuit is say 3500 ohms, use a 35K bleeder after the last cap before the OPT primary. Brute force regulated power supplies sound excellent and are simple to implement. Skroo the attitude of the engineer who barely made his degree and is only able to keep his job by making improvements in "efficiency" at the cost of real performance.
Tough question. Play around with several until you get one that has reasonably high cold resistance (>25 ohms) but is small enough to get hot enough to reach ~0 ohms.
I put them on the primary of the plate transformer.
Modern electrolytics don't need them.
John
Hi,
If you start out with wrong figures yourself how do you expect an answer?
Request? I dare say you were demanding strict proof.
How much proof does an engineer need other than the knowledge that capacitor ESR is not a linear resistance?
What do you think bpass caps are trying to achieve anyway?
Oh, since when are capacitors in an OTL "Special"?
Stop making things up as you go along, will you?
Cheers, 😉
If you start out with wrong figures yourself how do you expect an answer?
Request? I dare say you were demanding strict proof.
How much proof does an engineer need other than the knowledge that capacitor ESR is not a linear resistance?
What do you think bpass caps are trying to achieve anyway?
Oh, since when are capacitors in an OTL "Special"?
Stop making things up as you go along, will you?
Cheers, 😉
fdegrove said:
Meh. Request, demand, makes no difference to your denial. It's sometimes quite fun to tease people in denial, as you will see.
Ah, so now the question is somehow not that the impedance can even affect anything, but that it is nonlinear too!
Waffling, yet saying I am the waffler...classic.
I don't know. I haven't been notified (hint, hint) of any cases where the additional low ESR would be necessary.
I was simply qualifying OTLs as a special case: if you used an "ordinary" capacitor (a small one such that might be used on a 12AX7 cathode), with potentially high ESR, in the low impedance circuit of an OTL, it may potentially meet the requirements (under 50 ohms Z, over 5 ohms ESR). Although this appears to be "moving the goalposts", it is simply scaling. An OTL is several amplifiers paralleled, and I gave absolute specifications, not relative values. Also, the other condition, much more subjective unfortunately, is that the capacitor is a reasonable value for its use. In an OTL, the only low-Z place you might use a capacitor is for filtering, or coupling the speaker, and for both these applications, giving reasonable performance, inherently require large value capacitors with correspondingly low ESR values.
That makes three posts with no numbers, Frank...
Tim
P.S. Edit: looking over posts, it seems I have changed things a bit! I'm looking for any evidence at all that ESR is a value anywhere near Z such that it might affect it (10% is a well accepted value for such measurements), hence 5 ohms ESR and 50 ohms Z. My original statement held a factor of 500, i.e., 50 ohms Z and 0.1 ohm ESR. I will revise my specification to this, even though it holds no practical purpose (50.1 ohms compared to 50 is such a diminishingly small value), but merely to see if you can find any documented cases of this. Again, 50 ohms Z to 0.1 ohm ESR, a factor of 500.
Hi,
You keep on inventing stuff as it best suits you, don't you.
Where on earth did you read that anyway?
Denial of the fact that you absolutely have no interest in audio. Period.
Just a lot of wishful thinking on your part that's all there's to it really.
Oh, not that I care one way or another, but you once again contradict yourself with the OTL example....
Keep on digging,😀
You keep on inventing stuff as it best suits you, don't you.
Where on earth did you read that anyway?
Denial of the fact that you absolutely have no interest in audio. Period.
Just a lot of wishful thinking on your part that's all there's to it really.
Oh, not that I care one way or another, but you once again contradict yourself with the OTL example....
Keep on digging,😀
fdegrove said:
Oh, you showed no objections at all in the beginning of the thread. And elsewhere.
*Listening to, and enjoying, Tool - Lateralus*
Snicker... not a good arguer are 'ya Frank.
Tim (snickering some more)
Hi,
Goes along way to showing that you can't be bothered to read what's being posted.
Prejudice and conjecture, those are the words that come to mind....
Snickers and Mars bars, 🙄
Goes along way to showing that you can't be bothered to read what's being posted.
Prejudice and conjecture, those are the words that come to mind....
Snickers and Mars bars, 🙄
- Status
- Not open for further replies.