| Joseph0 |
attached is a transfomer coupled anode follower circuit from an AudioNote DAC (2.1x BAL).
the original capacitance connected to the cathode pins of E188CC is 470uF.
I was thinking of trying two BG N 470uF caps in super E: 940uF.
The circuit is hardwired however and I'd rather get it right the first time.
What does changing this capacitance do technically?
What does changing this capacitance do sonically?
Thank you to anyone who responds.
 |
|
|
| Robski666 |
Hi Joseph,
looks like a common cathode buffer using parallel triode sections, with self bias mmm. anyway the caps I take it you are referring to are bypass capacitors accross the 180R cathode resistor. the value is chosen to effectivly increase the AC gain ie audio gain and mostly effects the lower frequency response. that is the point is where Cx reactance equals the cathode resistance. so increasing the value effectivly decreases the low frequency response, to a point.
thats the technical.
In reality I'm not sure if this translates to better sound given that it is large electrolytic in the the signal source and some people would say it would be better avoided.
what you have 470uf Cx @ 20hz ~ 17r almost 1/10 of Rk
after you change it to 940 Cx @20hz ~ 8.5r almost 1/20 Rk
Personally I don't think I could tell the difference. |
|
|
| Robski666 |
Hi Joseph
I re read this and it made sense at the time but!
so increasing the value effectively decreases the low frequency response, to a point.
decreases the -3db point not the low frequency response so should read
so increasing the value effectively improves the low frequency response, to a point.
Robert |
|
|
| Joseph0 |
thanks for the help, Robert.
now, say I'm using 470uF BlackGate N, non-polar, 'completely symetrical', capacitors and have accidentally reversed the leads, hooking up the +, long lead, to the ground, and the -, short lead to the signal.
I don't feel like redoing it.. |
|
|
| Robski666 |
Hi Joseph,
Ok I don't know that brand/type, however if it is a non polorized cap there should be no notion of polarity. the long and short leads are simply a function of manufacture i.e. same assembly as regular caps.
incidently is 470uf the standard value? |
|
|
| fdegrove |
Hi,
| quote: | | now, say I'm using 470uF BlackGate N, non-polar, 'completely symetrical', capacitors and have accidentally reversed the leads, hooking up the +, long lead, to the ground, and the -, short lead to the signal. |
Since the cap is a non-polar type there is no actual polarity so from a mere electrical POV it doesn't really matter which end's up....
Mind you, those Black Gates defy all laws of physics so be ware....
However, one of the leads is connected to a shielding foil which is directional. You'd want that foil connected to ground (-) or it won't have much of a shielding effect.
No big deal either way but just so you know.
The Jelmax "Tech sheets" are probably written by people lacking any technical knowledge, stink at English or any random combination of the above.
IOW, things can get confusing................
Now, regarding your question about "Super Eing" the bypass cap(s), assuming that's what it is as my reading of Modigliani paintings sux big time, I wouldn't bother.
There must be better places to waste money on..... Must be.
Besides, if you put two caps in series the total capacitance does not double but halves...
Unless the kids at Jelmax pulled a magic rabbit that is, in which case the value would remain the same. At best.
It's been a while since I last amused myself wading through their technobabble so you'd better measure the actual capacitance before you hook up those caps that way.....
Did I confuse anyone?? Probably just myself......................
Cheers, ;) |
|
|
| serengetiplains |
| quote: | Originally posted by fdegrove
Did I confuse anyone?? Probably just myself...................... |
Articulate, cultured and of subtle good humour, as I'm sure many appreciate.
;) |
|
|
| JojoD818 |
Frank,
As always, very informative and straight to the point.
:cool:
JojoD |
|
|
| Joseph0 |
with two BG N's in series, the capacitance doubles. I've measured it.
BG N's may be a waste of money, but they sure sound good. |
|
|
| Robski666 |
Hi Joseph,
You indicated that the caps would be in parallel as per the drawing, what I am intrigued with is the sub 1 hz -3db point your trying to achieve using increased k bypass capacitance. What was the original value? I'd guess somewhere between 47uf and 100uf??
Have you considered this might cause issues with short term drift, perhaps saturation of the O/P transformer core I don't know the specs on the units fitted.
Sy had a good response on this subject about a month ago he indicated too much LF response might in itself cause problems and was a good read.
fdegrove,
I didn't realize these black gate caps were sooo expensive this means I have to choose between ofc cable or BG caps ;o((
Robert |
|
|
| Joseph0 |
thanks robert.
the original caps were Elna Starget 470uF 16V. |
|
|
| JojoD818 |
| quote: | Originally posted by Joseph0
with two BG N's in series, the capacitance doubles. I've measured it.
BG N's may be a waste of money, but they sure sound good. |
are these special type of caps? what happens in their capacitance when they are in parallel?
thanks!
JojoD |
|
|
| fdegrove |
Hi,
| quote: | | with two BG N's in series, the capacitance doubles. I've measured it. |
Thing is you see, they're not really in series and they're not really non-polar either.....
Magic, right?
What you do with this Super E connection is similar to what some of us did to create a better cap using electrolytics for loudspeaker X-overs long before non-polar electrolytics were available.
I'm talking like more than thirty years ago already..........
Whatever you decide to do, keep in mind that there's little point in doubling the existing cap's value.
Little point in using the Super E connection method either if you already have a BG-N type cap across the cathode resistor.
Be ware of the voltage across the cap as well, provide for a reasonable safety margin like roughly 30% over the static cathode voltage.
I honestly don't know how such a connection behaves voltagewise and I doubt Jelmax have it covered somewhere.
| quote: | | BG N's may be a waste of money, but they sure sound good. |
Did I say they were a waste of money?
Did I say they sounded bad?
No, I didn't.
I just wanted to make you aware of what it was you're doing with a spare cap that may find better use elsewhere relative to what you intend of doing with it here. That's all.;)
I hope I learned something....:clown:
Cheers, :cool: |
|
|
| Joseph0 |
I meant when connected like this:
 |
|
|
| Joseph0 |
[QUOTE fdegrove]There must be better places to waste money on..... Must be.[/QUOTE]
maybe you were saying Modigliani paintings are a waste of money. : ^ P
 |
|
|
| fdegrove |
Hi,
That's what they call Super E?
A connection like that would be electrically equal to 2 caps in //.
What you achieve with that, assuming two identical caps, is twice the capacitance and whatever directionality each cap would have would now be cancelled out.
Voltage rating is the same as that of a single cap.
Whatever internal shield was present is now connected individually, one at the input, the other at the output rendering only the shield that goes to the ground end of things effective.
I'm still confused...How 'bout you?:cannotbe:
Cheers, ;) |
|
|
| JojoD818 |
| quote: | Originally posted by Joseph0
I meant when connected like this:
|
so that means the caps would be in parallel and not in series. no wonder the capacitance increased.
thanks man! |
|
|
| Joseph0 |
I decided to keep the capacitance at 470uF.
Here are some photos (the circuit includes part of an auto de-emphasis circuit along with the circuit on diagram):
the parts of the d/a section, minus the output transformers:
the long leadout wires are for LEDs.
and yes, I used copper foil oil caps in the de-emphasis filter. I made the decision to do this before I understood their purpose (i originally thought they were part of the analogue filter).
now I can listen to certain pink floyd cds and 1 in 150 others in need of de-emphasis..like never before.
they make a nice fit however.
I hope no one is getting emails for my edits. I apologize if so. |
|
|
| fdegrove |
Hi,
| quote: | | The article in the first link uses a 1000uF cap connected to the cathode, Cc in the diagram. |
What?
Where on earth do you read that??
Find me a value for Rc where it would require a 1000uF cathode bypass cap if F0 = 10Hz using the following formula:
C= 1/(63.Rc)
Second note:
There's no cap like no cap.
Small caps usually sound better than big ones (due to increased R/C time constants of the bigger ones)
A little local feedback does less damage than global loops so why not use some where appropriate?
| quote: | | I believe you were thinking the cathode to be the output, but I'm certainly not an expert. |
If the cathode is taken as the output node, putting a cap across the cathode resistor effectively shunts the AC component to ground, ergo no useable output.......
Sorry if I sound like a pedantic pain in the behind. Just trying to help.
Ciao, ;) |
|
|
| Joseph0 |
hi Fdegrove:
"Application of a 1000uF elco for Cc showed a clear advantage in SRPP. The treble becomes more sophisticated, the placement of sound sources becomes more precise both in width and depth. When it is applied in case of the anode follower, it even gets more involvement, and also adds to treble and precision. We tried several brands, and Philips elcos gave very good results."
it is the second of the four points at the bottom of the first link.
your replies are no problem. I haven't been working with electronics for long and make a lot of mistakes. So please correct me when I'm wrong. |
|
|
| fdegrove |
Hi,
| quote: | | "Application of a 1000uF elco for Cc showed a clear advantage in SRPP. The treble becomes more sophisticated, the placement of sound sources becomes more precise both in width and depth. When it is applied in case of the anode follower, it even gets more involvement, and also adds to treble and precision. We tried several brands, and Philips elcos gave very good results." |
That's a mere subjective assessment within the defined context of their system.
My own assessment of similar tests have shown that a bigger cap, bigger than strictly needed, slows down transient response audibly and smears the audio signal.
This is especially obvious on phono preamps having much higher gain.
No cap is perfect so use the best you can and as little as possible.
I still have to find a cap that actually "improves" anything by passing the audio signal through it...........
A bigger cap than calculated for cathode bypass duty may at best improve low frequency response but if that cap starts to "improve" treble response, as they seem to claim, then I suspect something is not quite right or they're comparing apples to oranges.
IOW they used a cap with a better HF response and it had nothing to do with the value of that cap per se.
Cheers, ;) |
|
|
| Johan Potgieter |
Frank,
I am not quite with you here (perhaps it is just the new year - and NO, I did not partake of the liquid of the vine)
You say| quote: | Originally posted by fdegrove
My own assessment of similar tests have shown that a bigger cap, bigger than strictly needed, slows down transient response audibly and smears the audio signal.
But then:
A bigger cap than calculated for cathode bypass duty may at best improve low frequency response but if that cap starts to "improve" treble response, as they seem to claim, then I suspect something is ..[/B] |
I agree that something appears to be amiss when a bypass capacitor as meant here improves the high frequency response - but then what do you mean by "slowing down transients". Is that not also in the high frequency domain?
I have checked for signal over cathode bypass caps in the past because of qualitative statements that electrolytics become inductive and do all sorts of things to high frequencies, and cannot truhfully say that I have found anything visible there to cause me concern, even with square waves. Have I just been fortunate in using "good" capacitors, or does one have a degree of urban legend here? As you said, listening tests are subjective. I have the greatest respect for individual experiences, but difference of opinions as well as the result of blind tests make it difficult for a designer to come to a universal conclusion wrt such tests.
Regards. |
|
|
| fdegrove |
Hi,
Johan,
| quote: | | but then what do you mean by "slowing down transients". Is that not also in the high frequency domain? |
Their claim goes:
Putting a bigger cap in, in casu 1000 microF iso the calculated 50 microF value improves highs. That would include perceived transient response.
My experience:
Rather the opposite is true as the increased value of the cap alone will broaden the LF response giving a subjective feeling of slowed down, bass heavy sound.
This can be improved by the use of better quality caps but usually, all else kept equal, a bigger cap means more of the same bad things.
That's all I wanted to point out.
Hence my surprise to read the opposite of what I'd expect to read.
Cheers, ;) |
|
|
| Johan Potgieter |
Frank,
No fine - thanks for reply. I have largely misread your reference to transient response, in that I never had that experience. But with one very large capacitor, yes - I should have seen that implication. I have never used that large a single cap but know (i.e. have read) that all is not well that way. I have on occasion used several smaller caps in parallel, also because it is simply easier to accomodate on a p.c., which could be the reason for my not having experienced problems. But in an analysis published a few years ago in Electronics World by Cyril Bateman there appeared to be substantial differences between different makes. Perhaps I was just fortunate.
Thanks again,
Johan |
|
|
| Joseph0 |
| deleted post. -I apologize. |
|
|
| kevinkr |
The whole point behind the super E cap configuration claims on the Jelco website is that this back to back parallel connection cancels the ESL of the capacitors.
Obviously that can't be the case, at best it would effectively halve it. (Two inductors in parallel.) In order to actually cancel the inductance the ESL of both capacitors would have to couple 100% to each other in the magnetic domain and given that they are separate components that is impossible.
All of the BG capacitors I have used seemed to be significantly better in key measurable parameters like linearity than many of the cheaper ones they were replacing. I believe they do actually sound better in a lot of instances. It is a shame that such outrageous marketing claims are made for them, some solid engineering went into their design. The marketing approach is reminiscent of the claims made for patent medicines over a century ago.. :D
Frank made an interesting comment about large bypass capacitors and I have observed similar behavior with large capacitors. I was able to observe ringing in some waveforms that was clearly due to resonances in the capacitor itself. (Perhaps both mechanical at high current levels and electrical otherwise.)
As an aside some metal film resistors "umm" generate harmonic distortion that is measurable as a consequence of the nature of their end terminations as well.. Not exactly very linear behavior from a "linear" component. (Old Mepco Electra RN60 MF )
No passive component is totally linear, some are just a lot better than others, which I guess is one of the points I am making. No capacitor in many instances is the best capacitor.. |
|
|
| Johan Potgieter |
| quote: | Originally posted by kevinkr
It is a shame that such outrageous marketing claims are made for them, some solid engineering went into their design. The marketing approach is reminiscent of the claims made for patent medicines over a century ago.. :D |
Kevinkr,
One must resist the temptation to reply to everything, but this is a particular sore point of mine. You are so correct. I have often said that the outrageous claims efforts at technical description by manufacturers of how their own products work or by subjective reviewers, have sometimes been the deciding reason why I did not invest in them. I already mentioned a series of articles on capacitors by Cyril Bateman in Electronics World a few years ago, where he also found that BGs were slightly better than the best of the rest, but that the extra expense was not merited. But that not to dissuade - if folks can afford them..... |
|
|
|
