I like the way you stretched the capacitor wires Mr PP . Been there , done that . Polystyrene caps , that's dedication . When I repaired amplifiers I liked to put all the components neat . I would always use a scope to be sure it wasn't a choice . I note how carefully the diodes are joined . I remember a time when I did that .
For DVV and other would be sea dogs .
Richard Burton reads S.T. Coleridge's 'The Rime of the ancient Mariner' - Upload 1/3 - YouTube
For DVV and other would be sea dogs .
Richard Burton reads S.T. Coleridge's 'The Rime of the ancient Mariner' - Upload 1/3 - YouTube
I followed the original PCB, that's all.
this "I just follow" thing reminded me of the time (circa '90 something again) when I was building this modulator for my ZX-Spectrum clone, only to be puzzled by the two paralleled output caps, one 1uF ad one 10nF or something. I went asking my "guru" how come I need to parallel 1/100 cap to a large tolerance electrolytic and he replied, much to my frustration "just follow the schematic, damn it".
I clearly remember one old book mentioning styrene caps for the high-pass portion of the Baxendall tone control, I've known that since I was a kid. and all implementations I've seen (although based on mediocre active devices) obeyed that so I said to myself who am I to disagree. fast forward to cap listening LOL
anyway, just nostalgia, sorry for thread hijacking.
Not at all and I am sure others would agree Mr PP . I built my Sinclair calculator and would love to still have it , LED's you see . Sinclair should be respected for two things . His designs although a bit optimistic were often genuinely ahead of the typical shop bought designs . Number two if I get this right he had a company worth £150 000 000 back then . OK he got unlucky . It is unbelievable that he got so far on what were mostly kits . I seem to remember he was a writer in the early 1960's electronics magazines . Like myself making a living out of thin air then . I bet like me he enjoyed it . To market a class D amp in 1969 much as I distrust class D I still have to respect that . I am told they were better than they should have been . On paper they couldn't work all .
Actually, in audio and appearently not only in audio, it makes good sense to parallel say a 220 uF electrolytic in the feedback loop with a metallized foil capacitor of say 470 up to 2.2 uF.
Preferably (but difficult to find) polycarbonate - if there is a faster capacitor on this Earth and on the free market, I don't know of it. Those boys do like 3,000+ V/uS, DAMN fast.
If making an AC coupled amp, I also bypass the input metallized foil Wima 10 uF cap with the same but 100 nF.
Preferably (but difficult to find) polycarbonate - if there is a faster capacitor on this Earth and on the free market, I don't know of it. Those boys do like 3,000+ V/uS, DAMN fast.
If making an AC coupled amp, I also bypass the input metallized foil Wima 10 uF cap with the same but 100 nF.
DVV is salty .......
Whazzat supposed to mean, hombre?
And SMILE when you say that, stranger!
If there is, I am not aware of it.
I do mine by ear, but then, I can afford to do so because I usually build mine for a friend, and only now for myself.
But if there is, I'd love to hear more about it. I love leraning new stuff.
An extract of the PDF below
[ One commonly used improvement is to load the input differential stage with a current mirror. This doubles the available drive current and therefore appears to double the slew rate. There are however other effects, it also doubles the gain, and to maintain the original stability margin we need to either double the compensation capacitor, which returns us to the original slew rate and so has no apparent benefit, or we could double the 100R degeneration resistors, then the gain is unchanged, the slew rate limit is doubled, and input stage distortion at lower levels is reduced. The impressive slew rate specification however is in practice unatainable in the example above, it could only be achieved with Vdiff = 2V, so with a 1V peak input signal that could never happen with any signal whether sine, square or whatever. ]
Slew Rate
I would love to see a very simple high slew rate amplifier ( 1000W would be OK ) . My mind says it needs nothing complicated . DC offset adjustment perhaps if running plenty of input stage current ?
The only thing I will say against bypassing is sometimes a high grade capacitor will be even better . People liked to go down in 10's . One friend does it is 1.618 ratio as best he can ( Phi ) .
I hope Andrew T is reading this . I said how curious Pi cubed is ( circa 31 ) . If needing a quick and fairly accurate value for Phi . 6/5 ( Phi . Phi ) = Pi . To be certain 1/Phi = ( Phi - 1 ) . Phi . Phi = 1+ Phi . Thus to a very good approximation 1.2 ( 1+ Phi ) = Pi .To the Egyptians I suspect it looked exact . I would also suggest no engineering problem on this planet would need it to be better ?
[ One commonly used improvement is to load the input differential stage with a current mirror. This doubles the available drive current and therefore appears to double the slew rate. There are however other effects, it also doubles the gain, and to maintain the original stability margin we need to either double the compensation capacitor, which returns us to the original slew rate and so has no apparent benefit, or we could double the 100R degeneration resistors, then the gain is unchanged, the slew rate limit is doubled, and input stage distortion at lower levels is reduced. The impressive slew rate specification however is in practice unatainable in the example above, it could only be achieved with Vdiff = 2V, so with a 1V peak input signal that could never happen with any signal whether sine, square or whatever. ]
Slew Rate
I would love to see a very simple high slew rate amplifier ( 1000W would be OK ) . My mind says it needs nothing complicated . DC offset adjustment perhaps if running plenty of input stage current ?
The only thing I will say against bypassing is sometimes a high grade capacitor will be even better . People liked to go down in 10's . One friend does it is 1.618 ratio as best he can ( Phi ) .
I hope Andrew T is reading this . I said how curious Pi cubed is ( circa 31 ) . If needing a quick and fairly accurate value for Phi . 6/5 ( Phi . Phi ) = Pi . To be certain 1/Phi = ( Phi - 1 ) . Phi . Phi = 1+ Phi . Thus to a very good approximation 1.2 ( 1+ Phi ) = Pi .To the Egyptians I suspect it looked exact . I would also suggest no engineering problem on this planet would need it to be better ?
I read a paper that concluded that the minimum capacitance RATIO when paralleling an electrolytic with a film capacitor was 16:1
I did not understand all the maths but it did seem to make sense.
That was in the days before I saw what others are finding with ringing when doing such with low ESR capacitors.
If there is some truth in that conclusion, then 100:1 may well be a good ratio to use that gets one well away from the "ringing" condition
Not really, I pop in now and again.
There is so much to read and to understand that I found I could not spare the time that this Thread deserved from me. So I gave up reading all of it about 8weeks ago. There are three other Threads that I have had to give up with, simply due to lack of time.
I friend I made at the Frankfurt hi fi show who owns one of the USA's more esoteric speaker brands asked me which filter curve he should use for real music ? I said none , try Phi . He did and said it works correctly . He wrote to me recently and said life is grim in the real world of selling . Out of respect I better not say who .
Andrew . It proves coincidence is normal . The French define it as things that " will " meet if my understanding is correct ? That's the problem with English . We pirate words and never know quite of what we are saying ? Salty sea dogs all ( only the English perhaps ? )
Andrew . It proves coincidence is normal . The French define it as things that " will " meet if my understanding is correct ? That's the problem with English . We pirate words and never know quite of what we are saying ? Salty sea dogs all ( only the English perhaps ? )
Last edited:
The trouble is, Andrew, when there is a minimum, there is also a maximum. The whole story would be to know both.
For example, I usually bypass a quality electrolytic cap of 220 uF with a metallized film Wima capacitor of 2.2 uF, or 100 times lower. So far, it has always worked.
Could you give me a precise link to one of those posts, or some other way of tracking down the relevant material? I am interested in what the circumstances were in total: for example, if there was another film cap in the electrical "neighbourhood" of the first film, with a "bad" ratio match, then I would expect ringing.
Another mistake that is made is to have the leads of the bypass cap long, so that it physically fits in. This is bad in every sense; a rule of thumb is that the smaller a cap, the shorter its leads need to be. Otherwise, the exercise is close to useless ...
Frank
memory no good.
Some are posted by our Member that claims to be a digital PSU and digital Amp designer.
I think Cordell did a circuit/test procedure to measure the effect of fast transients on the PSU caps.
And yes to lead lengths.
The straight through must be the smaller cap value. The high value cap then gets tacked on afterwards.
It's the same philosophy that is applied to the on board decoupling. The small caps must be right next to the power Pins (= short route lengths).
Some are posted by our Member that claims to be a digital PSU and digital Amp designer.
I think Cordell did a circuit/test procedure to measure the effect of fast transients on the PSU caps.
And yes to lead lengths.
The straight through must be the smaller cap value. The high value cap then gets tacked on afterwards.
It's the same philosophy that is applied to the on board decoupling. The small caps must be right next to the power Pins (= short route lengths).
Last edited:
No, I am NOT saying that.
Please note that the ratio is not "much bigger than 100:1", it is EXACTLY 100:1.
but you said
What do we do if there is a maximum (ratio)?
You have put the doubt in our minds.
I am asking you:
Are you saying that there is a maximum ratio?
I am not asking:
What ratio do you use?
Back to "what are you saying".
Are you telling us there is a maximum ratio to go along with the 16:1 minimum ratio? If the 16:1 ratio is right, it might not be.
What are the consequences of not knowing what the maximum ratio is?
- Status
- Not open for further replies.