most of the standard ceramic disc capacitors are NP0/C0G type?
http://bright-components.co.uk/WebR...C111/083C/D120/C0A8/1909/4CE2/ceramicdisc.jpg
http://bright-components.co.uk/WebR...C111/083C/D120/C0A8/1909/4CE2/ceramicdisc.jpg
For Miller I would choose Polyester and for signal input stages. Ceramic is fine for input stages if you can fine one that is not microphonic.
Not all ceramics are NP0/C0G. Check the datasheet. These types are fine for compensation or filters etc. Higher permittivity ceramics will not be, as they are nonlinear and may be microphonic. Polyester is slightly nonlinear but may be OK.
I would use NP0/C0G but you do have to check ceramic for the specific type
multilayer ceramic have very low inductance, you lose some advantage if you use through hole parts though
film types can have different constructions that may give resonances in the MHz depending on foil edge termination style
multilayer ceramic have very low inductance, you lose some advantage if you use through hole parts though
film types can have different constructions that may give resonances in the MHz depending on foil edge termination style
Nothing wrong with using good NP0 dielectric ceramics for filters and compensation. Kemet, AVX and Vishay have a good range.
There is nothing inferior about Wima or Kemet products unless you believe low cost and wide availability are minuses.
If you will feel better by spending more on obsolete dielectrics, try polystyrene film/foil: FSC 160V 100PF 2.5% - LCR COMPONENTS - Film Capacitor, FSC Series, 100 pF, ± 2.5%, PS (Polystyrene), 160 V | Newark element14
or silvered mica: CD15FD101J03: 100pf Dipped Silver Mica Capacitor: Jameco Valuepro
Personally, I would never use low voltage polyester (MKT or greencaps) for compensation caps. Mica and polystyrene types may also not be technically as good as PP film types but some find the small distortion increases useful. Just keep voltage ratings as high as possible within sensible size limitations and performance will be acceptable. In any case, typical VAS compensation caps will need to be rated ≥ the rail-rail voltage supply of a conventional audio amp.
If you will feel better by spending more on obsolete dielectrics, try polystyrene film/foil: FSC 160V 100PF 2.5% - LCR COMPONENTS - Film Capacitor, FSC Series, 100 pF, ± 2.5%, PS (Polystyrene), 160 V | Newark element14
or silvered mica: CD15FD101J03: 100pf Dipped Silver Mica Capacitor: Jameco Valuepro
Personally, I would never use low voltage polyester (MKT or greencaps) for compensation caps. Mica and polystyrene types may also not be technically as good as PP film types but some find the small distortion increases useful. Just keep voltage ratings as high as possible within sensible size limitations and performance will be acceptable. In any case, typical VAS compensation caps will need to be rated ≥ the rail-rail voltage supply of a conventional audio amp.
I have used polystyrene capacitors for comp but the sound became more unforgiving. It was not hyper detail but it was showing the mistakes of the mix engineer. I would prefer the polypropylene may be in that location or need to try MLCC Radial C0G type cap.
its entirely practical to have 30-40 dB excess loop gain @ 20 kHz with dominant pole compensation
any "error" of the compensation cap is reduced just like any other measured by the feedback loop - by that 30-40 dB @20 kHz, and by proportionately more at more musically significant lower frequencies
if concerned then use "2 pole" compensation for another order of magnitude of 20 kHz gain and compounded by the reduction of signal across the cap at the VAS input over audio range
of course these factors presume a genuine high loop gain design
any "error" of the compensation cap is reduced just like any other measured by the feedback loop - by that 30-40 dB @20 kHz, and by proportionately more at more musically significant lower frequencies
if concerned then use "2 pole" compensation for another order of magnitude of 20 kHz gain and compounded by the reduction of signal across the cap at the VAS input over audio range
of course these factors presume a genuine high loop gain design
I don't use polystyrene for Miller compensation. I think it is "distortion", not being "unforgiving".
Problem with polystyrene is I think not every manufacturer follows the coding standard. If "220J" were thought of as 220pF while actually it is 22pF, of course the effect is unexpectedly negative.
Also, if voltage rating is important (as suggested by Ian), not all specify this voltage rating on the body 🙁
Can such thing reach a hundred pF? Another interesting experiment is using variable capacitors found in many RF circuit. This will make it easy to dial the perfect capacitance (using CRO). I think Hugh Dean had mentioned about variable capacitor but I have never seen this RF variable capacitor implemented in DIY amps so I'm reluctant to try.
its not just the value yes I have calculated it does gets about 100pf in that size and mainly im interested to use copper electrodes than just aluminum and Wax paper in foil as dielectric.
That's a physically large cap. You will invite all kinds of EMI, hum and instability effects with it.
It may work fine - then again, it may not.
Jan
I have a question that most of the comp caps are available in 100VDC range so how much ACV can be applied on the cap? 1.414 times multiplied?
Most Vas compensation capacitors see the total supply voltage, so if you have 50V rail, a total of 100V. I would not use a 100Vdc capacitor on 100V.
The advantage of these NPO and COG ceramics is that you can get 250V to 1kV ratings in a 1206 package. This will have much better behaviour at MHz frequencies than many boutique parts and that is what matters for compensation
NPO and COG are indeed very good - stable over time, temperature and humidity and especially so against voltage. Film are good, but at high voltage swings, they do have voltage dependent non- linearity issues - see self for example (poly prop, polystyrene and poly carbonates are discussed in 'Small Signsl Design').
I've used Mica 500V types with good results, but some people don't like them - I found nothing objectionable.
I've used Mica 500V types with good results, but some people don't like them - I found nothing objectionable.
- Status
- Not open for further replies.