Yep, always read the manufacturers stuff (and in the case of silicon, the Errata), but if there was significant voltage nonlinearity in common C0G parts the RF crowd WOULD have noticed!
RF Folks REALLY care about linearity, far more then the wiggly DC audio people generally do, for the sime reason that quite often a receiver is trying to dig signal out 100+dB below another few signals in the front end passband, and that REALLY don't work if there is any significant non linearity present.
If c0g makes you nervous (or you just want to spend some more money), ATC have quite excellent porcelain parts (Intended for RF power use), but that is almost certainly overkill of audio.
RF Folks REALLY care about linearity, far more then the wiggly DC audio people generally do, for the sime reason that quite often a receiver is trying to dig signal out 100+dB below another few signals in the front end passband, and that REALLY don't work if there is any significant non linearity present.
If c0g makes you nervous (or you just want to spend some more money), ATC have quite excellent porcelain parts (Intended for RF power use), but that is almost certainly overkill of audio.
I think I'll follow your advice and start with C0G and later switch to a PS.
Thank you all for your comments!
Thorsten
No, oscillator drift is what they care about with C0G or mica.
Alas you are clearly ignorant of how RF systems work. High linearity power amplification is very hard to achieve at high frequencies, and certainly not needed for many narrowband systems (FM and PM).
Selectivity is achieved by superheterodyning (mixing down to an IF) so that the wanted signal can be passively filtered out.
For a receiver the signal levels are usually very small, so non-linearity is seldom an issue, the exception being when multiple very strong signals are present which actually intermodulate each other. This can only be solved by passive filtering.
One real issue is phase-noise on the local oscillator - this will modulate with any strong signal and tend to deafen the receiver at nearby frequencies.
The front-end mixer is an inherently non-linear device (although its optimized to mainly generate 2nd order products between the LO and RF input, it will intermodulate far worse than any sensible audio circuitry).
Transmitting RF amplifiers are either class C or "linear". -40 to -60dB linearity is typical (intermodulation products, harmonics are not an issue, they get passively filtered). IM3 is a crucial parameter for linear amps, its the notional level at which 3rd order IM distortion reaches 100%. Its usually a dozen or so dB above an amps max power (linear amps usually have to run below full power as a result).
With very broadband RF (such as direct conversion SDR), linearity is much more important and 14bit DACs and ADCs are used, which set a limit to the dynamic range far less than 100dB. Such SDR has limited dynamic range as a result.
Again RF systems select signals by frequency using filters. They mix signals non-linearly. Broadband techniques struggle as linearity can actually matter with broadband.
A good old-fashioned radio system with preselector, high level mixer, IF strip with multiple bandpass filters can have a dynamic range of perhaps 160dB, but no part of it need be very linear!
Last edited:
RF people care about linearity in passive components.
Is this an attempt at humour? You clearly have no idea about high specification RF design.Mark Tillotson said:
TDK FA np0 capacitor RIAA phono.
Hi all. I have been searching for a good economical substitute for Silver Mica caps used in the RIAA section of my valve phono stage. Opinions seem to be very mixed when it comes to RIAA cap choice.
Has anyone used np0/c0g with good results. TDK manufacture through hole leaded versions, (FA series) and they are inexpensive, high temp tolerant, 100 volt rated, and priced at about 10% of Silver Mica caps.
Thoughts please. Schematic attached. Thanks
Hi all. I have been searching for a good economical substitute for Silver Mica caps used in the RIAA section of my valve phono stage. Opinions seem to be very mixed when it comes to RIAA cap choice.
Has anyone used np0/c0g with good results. TDK manufacture through hole leaded versions, (FA series) and they are inexpensive, high temp tolerant, 100 volt rated, and priced at about 10% of Silver Mica caps.
Thoughts please. Schematic attached. Thanks
Attachments
Silver mica capacitors can be microphonic and have a fairly high temperature coefficient - I used to use them and another type with the opposite temp co to stabilise a vco circuit.
Good quality mica caps should not be microphonic, and the temperature coefficient depends on their class (1 to 4): it varies from +/-200ppm/°C for class 1 to +/-20ppm/°C for class 4.
Even +/-200ppm/°C is not what I would call fairly high: it is just a bit more than PS.
Even +/-200ppm/°C is not what I would call fairly high: it is just a bit more than PS.
COG are used in microwave targeting radars, microwave steerable Ethernet antenna's and used in critical oscillator circuitry... Only problem is generally in SMD only available in low values, other than that an excellent cap, even if it is a ceramic!
+1.
You could also look at film caps, other than mica. They should be available in those values and are quite good and stable.
Even available 2% tol less than $ 1 although large: mouser 80-F462AK822G1K0C
Or mouser 80-MMK5822J63J01TA18
For the other for instance mouser 667-ECW-HA3C242H4 + a small cap parallel.
Jan
You could also look at film caps, other than mica. They should be available in those values and are quite good and stable.
Even available 2% tol less than $ 1 although large: mouser 80-F462AK822G1K0C
Or mouser 80-MMK5822J63J01TA18
For the other for instance mouser 667-ECW-HA3C242H4 + a small cap parallel.
Jan
Last edited:
I forgot to mention a very reliable film and foil part that is great, the Wima FPK2.
Had several years of listening to a cog part in a Nakamichi amp, and the wima was a very welcome upgrade, less harsh and sibilant. I chose that over a polystyrene one because I wanted a more punchy sound with this setup.
I guess if a thermally stable part makes you feel good, then the cog would be a good choice.
Had several years of listening to a cog part in a Nakamichi amp, and the wima was a very welcome upgrade, less harsh and sibilant. I chose that over a polystyrene one because I wanted a more punchy sound with this setup.
I guess if a thermally stable part makes you feel good, then the cog would be a good choice.
Member
Joined 2009
Paid Member
that’s a very interesting observation - in that you find a change in this capacitor eliminates a harsh and sibilant sound
assuming of course that they are the same value I’m quite surprised by this observation
assuming of course that they are the same value I’m quite surprised by this observation
Metal dome tweeters in a car, where everything is revealed.
However I should say that the offending part may not have been a COG, and possibly a lesser part, come to think of it.
Production amps have the cheapest parts everywhere sometimes, and sometimes it even makes a difference.
However I should say that the offending part may not have been a COG, and possibly a lesser part, come to think of it.
Production amps have the cheapest parts everywhere sometimes, and sometimes it even makes a difference.
Yes, because if you check the data sheets for COG and the WIMA you specified there is little difference and those would probably only be of a concern if your up in the MHz/GHz
In practice, C0G is hard to beat. Small size and surface mount, specs on par with PP, and most likely a more consistent manufacturing process.
Probably wasn't in a mass-market car application. What was the value?
Probably wasn't in a mass-market car application. What was the value?
Thanks for the replies. Think I will give the TDK FA series a go. They have leads. Nothing to lose at the price and probably sound ok. I will order the Wima FKP2 caps also and compare the 2 types. Much cheaper than silver mica alternative. I am pretty much building a point to point version of the audio note kit m3 phono stage but avoiding all of the boutique parts in the original (caps at £30-40 a piece, resistors £6-10 etc.). A friend has the standard kit which sounds very nice. Should be interesting to compare my 20% budget version to the original.
Thanks for the replies. Think I will give the TDK FA series a go. They have leads. Nothing to lose at the price and may well sound decent. I will order the Wima FKP2 caps also and compare the 2 types. Much cheaper than silver mica alternative. I am pretty much building a point to point version of an audio note kit phono stage but avoiding all of the boutique parts in the original (caps at £30-40 a piece, resistors £6-10 etc.). A friend has the standard kit which sounds very nice. Should be interesting to compare my 20% budget version to the original.
Didn’t see this until now; the application is a 220pf in series with a 1.5kr feedback capacitor in an old Nakamichi car amp.
Trying to get 500 volt COGs with enough pF to be useful as an audio amp Miller cap is like pulling teeth. 500 volt micas are all over the place. 100 volt caps are fine if you’re into building low power amplifiers. My stuff often needs a wee bit more.