I've found a lot of great information on bypass capacitors in regards to the type that people prefer and the potential benefits and disadvantages of paralleling capacitors. My question stems from what I have on hand in my parts bin...
I need a 220uF cathode bypass cap on a 12AT7. I dont have have a 220uF cap on hand BUT I do have a variety of caps that will get me there. Specifially, a 100uF tantalum capacitor, a 100uF electrolytic and a 100uF film cap - plus some smaller caps to get me to 220uF. What are people's thoughts on paralleling different types of capacitors in this application?
For example, what are the potential advantages and disadvantages of paralleling a 100uF tantalum, a 100uF film and a 20uF film?
Of course I can go buy a cheap electorlytic and call it day, but I'm here to learn.
I need a 220uF cathode bypass cap on a 12AT7. I dont have have a 220uF cap on hand BUT I do have a variety of caps that will get me there. Specifially, a 100uF tantalum capacitor, a 100uF electrolytic and a 100uF film cap - plus some smaller caps to get me to 220uF. What are people's thoughts on paralleling different types of capacitors in this application?
For example, what are the potential advantages and disadvantages of paralleling a 100uF tantalum, a 100uF film and a 20uF film?
Of course I can go buy a cheap electorlytic and call it day, but I'm here to learn.
It is unlike to matter what type you use. The voltage on your 12AT7 cathode is presumably trivial (tants are usually low voltage rated but I guess well above what you need here), and any supposed capacitor distortion will be microscopic compared to the tube distortion. Also, it is quite likely a single 100uF will work fine in your situation. Depends on the circuit details of course, but we're not building life support machines. Certainly you don't need to worry about the 'last 20uF', just a pair of 100uF would suffice.
I did a bypass capacitor shootout on a 6AH4 stage with some well regarded caps like Elna, Nichicon, AN, Kemet etc. The best was Kemet DC Link caps, but I did get one good combination:
The two best results, using 100uF, were:
1. DC Link Kemet capacitors - see RS online. I tend to use 2 in parallel, like 45 + 45 or 40+40 or whatever is in stock.
2. A combination of Elna Silmic and Audio Note Kaisei, each 100uF. This was surprisingly good, well recommended if you can't get DC Link Kemets.
https://www.hificollective.co.uk/catalog/audio-note-kaisei-polar-low-voltage-16v-160vdc.html
https://www.hificollective.co.uk/catalog/-c-61_68_324.html
I tried other combinations, but that was the only one that sounded really good. I don't think there's any science about what to combine - just a case of try and and listen to it.
The two best results, using 100uF, were:
1. DC Link Kemet capacitors - see RS online. I tend to use 2 in parallel, like 45 + 45 or 40+40 or whatever is in stock.
2. A combination of Elna Silmic and Audio Note Kaisei, each 100uF. This was surprisingly good, well recommended if you can't get DC Link Kemets.
https://www.hificollective.co.uk/catalog/audio-note-kaisei-polar-low-voltage-16v-160vdc.html
https://www.hificollective.co.uk/catalog/-c-61_68_324.html
I tried other combinations, but that was the only one that sounded really good. I don't think there's any science about what to combine - just a case of try and and listen to it.
Thanks for this... I'm not quite sure what I was imagining would happen in a worst case scenario. Vaguely remember something about speed... Figured I'd put this out there as they're fundamentally different in their construction and must effect something at some level. If it's audible or worth worrying about is another question.
Regarding the circuit, I'm settling on a 2A3 build and looking like Gordon Rankin's Baby Ongaku. Schematic below for reference in regards to the 12AT7 cathode bypass cap at 220uF. What would be the downside of dropping to 100uF?
100uF will cause a little more roll off at LF, but I doubt you'd notice it compared with 220uF. Try it, you can always add the extra later!🙂
I looked up an online calculator for cathode bypasses and it seemed to me that 100uF was adequate. This matters in the case of the more expensive caps e.g. Kemet DC Link caps - considerable difference in price.
All makes sense. Ill start with the 100uF and go from there. Main concern is layout and leaving enough room if I want to go with a larger cap (physical size or uF) or parallel a couple of caps down the road.
100uF has 100 Ohms of capacitive reactance at 20Hz.
The cathode resistor is 333 Ohms.
The 12AT7 cathode impedance is 1/Gm (1/5500 microMhos = 181 Ohms; plus the additional (1 + RL/u).
The SRPP makes the bottom triode plate load fairly high, so the bottom cathode impedance is going to be much larger than 333 Ohms.
I expect that the -1 dB rolloff will be much lower than 20Hz, and the -3dB rolloff will be subsonic.
When it comes to the sound differences of different reasonable quality bypass capacitors . . .
I suggest conducting a very well designed and implemented Double Blind-Fold Test.
Your mileage, and what you hear, may vary.
The cathode resistor is 333 Ohms.
The 12AT7 cathode impedance is 1/Gm (1/5500 microMhos = 181 Ohms; plus the additional (1 + RL/u).
The SRPP makes the bottom triode plate load fairly high, so the bottom cathode impedance is going to be much larger than 333 Ohms.
I expect that the -1 dB rolloff will be much lower than 20Hz, and the -3dB rolloff will be subsonic.
When it comes to the sound differences of different reasonable quality bypass capacitors . . .
I suggest conducting a very well designed and implemented Double Blind-Fold Test.
Your mileage, and what you hear, may vary.