If you're referring to the caps to bypass high frequency interference to ground, here are a few thing to watch out for:
* ESL and ESR parasitics are the two most important factors (equivalent series inductance and resistance); the lower, the better
* you will put this capacitor in vain, and even possibly create worse of a problem if the leads are long: they will function as an antenna; the reason why you will see tiny capacitors right next to the v+ supply pin of an IC is not necessarily because they are cheap, but also because they will be the best fit from other points of view. The best capacitor with the best ESL, if left with long leads (even 1cm) will probably have more capacitance in the leads than internally.
* since this is not exactly in the signal path, it will not have such a huge effect on the sound, so a monoblock ceramic can fit the bill amazingly well; I know salas recommends mkp, so do try that, but do not cheat yourself and try to find the size (physical) that will allow you to have almost no leads.
* best is to use three of these in parallel; it will decrease parasitics and also will extend the filter frequency band by a huge margin; something like 100nF || 1nF || 100pF. You already have an amazing shunt regulator and large filtering capacitors before the regulator, so low frequency filtering is not what you're guarding against here; it is high frequency that will be picked up by any wire functioning as antenna.
* try this at every vin and vout of every board and you will have reduced RFI almost completely.
Please feel free to comments or correct me on any of these points. It's just my 2c after digging quite a bit through the RFI reduction literature.
* ESL and ESR parasitics are the two most important factors (equivalent series inductance and resistance); the lower, the better
* you will put this capacitor in vain, and even possibly create worse of a problem if the leads are long: they will function as an antenna; the reason why you will see tiny capacitors right next to the v+ supply pin of an IC is not necessarily because they are cheap, but also because they will be the best fit from other points of view. The best capacitor with the best ESL, if left with long leads (even 1cm) will probably have more capacitance in the leads than internally.
* since this is not exactly in the signal path, it will not have such a huge effect on the sound, so a monoblock ceramic can fit the bill amazingly well; I know salas recommends mkp, so do try that, but do not cheat yourself and try to find the size (physical) that will allow you to have almost no leads.
* best is to use three of these in parallel; it will decrease parasitics and also will extend the filter frequency band by a huge margin; something like 100nF || 1nF || 100pF. You already have an amazing shunt regulator and large filtering capacitors before the regulator, so low frequency filtering is not what you're guarding against here; it is high frequency that will be picked up by any wire functioning as antenna.
* try this at every vin and vout of every board and you will have reduced RFI almost completely.
Please feel free to comments or correct me on any of these points. It's just my 2c after digging quite a bit through the RFI reduction literature.
My comment will be: Don't sweat it that much, because this phono is self bandwidth limited due to high miller JFETS, no overall loop feedback, and non 50kHz pole RIAA. A nice fast compact Wima stacked PP will do most of the job IF there is an especially RFI polluted environment situation, long PSU leads etc. Of course you may use best theoretical solution. Just my 2C.
Ikoflex, thanks for the advise. It was highly illuminating.
However I was talking for the RIAA caps, not the bypass one.
As bypass I already added one 100nF MKP per channel. I will try to get it closer to V+ and add some lower values in parallel. The lead to the ground has to be longer though. I have some spare silver micas, hope they are ok.
regards
George
However I was talking for the RIAA caps, not the bypass one.
As bypass I already added one 100nF MKP per channel. I will try to get it closer to V+ and add some lower values in parallel. The lead to the ground has to be longer though. I have some spare silver micas, hope they are ok.
regards
George
Give me a few days more to comment.
Just changed the out put caps to Mundorf silver oils and today I was doing some work on my speakers, so no music!
Tomorow theya re going to be in place and start listening again.
I heard that Mudorfs need a few hours to brake in.
Initial impressions are very promissing though.
Just changed the out put caps to Mundorf silver oils and today I was doing some work on my speakers, so no music!
Tomorow theya re going to be in place and start listening again.
I heard that Mudorfs need a few hours to brake in.
Initial impressions are very promissing though.
Hi vgeorgevgeorge said:
I believe the cap in the signal line of the riaa is determinant in the type of sound you apreciate.
PIO caps tend to impose a very strong signature ( I find them very high-fi), teflon caps are much more transparent so you can hear the output caps better.
What type of bypass are you using in the Mundorfs ? (PIO, Teflon ?)
Ricardo
I advised only Russian silver mica for the Riaa caps. I use PP bypassed by silver mica if space challenged especially for the 47n. The 16n is the crucial one though. Not only the Tan Delta but also the value precision is not where a PIO excels. Its good to have for taming some harsh cheap couplers. FT3 is king for interstage. For output I lean towards Obliggato and Auricap than Mundorf.
Hi again guys.
Wow, this thread has really moved fast!! I've got a huge amount to catch up on. No doubt afterwards, I'll have a huge amount of tweaking to apply too
I'm still using my original build of this RIAA stage. I find it sounds lovely, really balanced and detailed in my system. I've got some Meridian 105 power amps in now, which I've serviced with new caps and these are very revealing.
I also take delivery of my second-hand Technics 1210 + new Jelco 750 arm next week, so vinyl tweaking will again be on the agenda.
Cheers, Lee.
Wow, this thread has really moved fast!! I've got a huge amount to catch up on. No doubt afterwards, I'll have a huge amount of tweaking to apply too
I'm still using my original build of this RIAA stage. I find it sounds lovely, really balanced and detailed in my system. I've got some Meridian 105 power amps in now, which I've serviced with new caps and these are very revealing.
I also take delivery of my second-hand Technics 1210 + new Jelco 750 arm next week, so vinyl tweaking will again be on the agenda.
Cheers, Lee.
SE circuitry has its merit too. Less components, and it can be phase reversed and mixed. The availability of symmetrical devices can give DC coupled schematics using servo offset control. It was made exactly because of the polar nature of solid state. But someone said that PNP and NPN are so complimentary and mirror alike, as much as a woman and a man. You know, exactly the same people, just mirrored. Yeah, you bet! 