I was asked why I used cathode bias vs. Fixed bias. Well, started thinking. In cathode biased tube DC voltage across a coupling cap between a driver an a power tube is lower than in fixed bias (grid biased with negative DC voltage related to cathode). I wonder if there are any measurements of caps "behaviour" which give data to compare two topologies?
I think you are over-thinking this. You have not yet even put it in a real context. The grids on a power tube will be further negative with fixed bias, yes, but by how much? Your circuit could be anything from a 6AQ5 single ended to a pair of 8417s in push pull. In one example, the difference might be less that 30 volts, in another it might be over 100 volts.
The two topologies have enough performance differences to study on their own, without wondering too hard about the coupling caps at various voltages. Much more to worry about the cathode bypass cap, at least in cathode bias.
The two topologies have enough performance differences to study on their own, without wondering too hard about the coupling caps at various voltages. Much more to worry about the cathode bypass cap, at least in cathode bias.
If a coupling cap uses a significantly non-linear dielectric or has mechanical stability problems then increasing the DC voltage across it could increase distortion. The solution is to buy a better cap, which could easily be a cheaper cap if the original was a peculiar 'audio' cap. Alternatively, buy a bigger cap so that there is less signal voltage across it (provided this won't upset LF stability).
In any case, this will be a small issue.
In any case, this will be a small issue.
From J. Broskie
The Tube CAD Journal,Design Idea: Polarized Connections -pg 1
[...]
But, on the other hand, certain audiophiles, whose ears I trust, tell me that in fact this device does make a very noticeable improvement. What could be going on here? I have read where a manufacturer of short-wave radios found that by applying at least a small polarizing voltage on all the capacitors used in the unit brought about a very measurable increase in high frequency response (RF).
"[...]
The Tube CAD Journal,Design Idea: Polarized Connections -pg 1
[...]
But, on the other hand, certain audiophiles, whose ears I trust, tell me that in fact this device does make a very noticeable improvement. What could be going on here? I have read where a manufacturer of short-wave radios found that by applying at least a small polarizing voltage on all the capacitors used in the unit brought about a very measurable increase in high frequency response (RF).
"[...]
OK, you choose:
1) John Broskie's unnamed audiophile friends' ears, which claims an improvement,
2) Circuit theory and component properties, which suggests virtually no change.
I know which I would rather believe.
I have never heard of the story about SW radios. Most radios are designed using circuit theory. I suggest that either JB or the person who told him that story has got the wrong end of the stick.
1) John Broskie's unnamed audiophile friends' ears, which claims an improvement,
2) Circuit theory and component properties, which suggests virtually no change.
I know which I would rather believe.
I have never heard of the story about SW radios. Most radios are designed using circuit theory. I suggest that either JB or the person who told him that story has got the wrong end of the stick.
Last edited:
I treat the blog entry as a hypothesis not truth. Moreover, belief or blind choice is not the way I do things. I am not going deep into physics, either, as I finished my technical university some time ago. However, I would rather see numbers and will be looking for proof or denial.
DF96, nothing personal.
DF96, nothing personal.
I guess the only way to find out is to measure the frequency response of an AC signal sweep through the cap repeated by the same measurement a couple of times of an AC signal sweep superimposed upon several different DC levels and compare the measurements.
Anyone here who has the equipment to do this?
Anyone here who has the equipment to do this?
Last edited:
you can use a good amp meater inline wth caps to see how the cap is holing up and the amps being drawn through the circuit
Some of the class 1 dielectrics do however exhibit measurable intermod at surprisingly low imposed voltages, and that is before you even think about something awful like Y5V (But why anyone whould use that for anything that mattered is an interesting question).
I have even measured IMD in thick film resistors used in an RF attenuator (THAT caused much headscratching, especially as said attenuator was being used for IMD tests)!
For me the key thing with capacitors of the coupling sort is to make them large enough that they do not have any meaingful signal voltage across them, even electrolytics are fine for audio with a little standing bias and only a few mV of signal voltage across them, filters of course ask rather more of caps.
Regards, Dan.
I have even measured IMD in thick film resistors used in an RF attenuator (THAT caused much headscratching, especially as said attenuator was being used for IMD tests)!
For me the key thing with capacitors of the coupling sort is to make them large enough that they do not have any meaingful signal voltage across them, even electrolytics are fine for audio with a little standing bias and only a few mV of signal voltage across them, filters of course ask rather more of caps.
Regards, Dan.
RF IMD is measuring levels that are way beyond anything a real speaker or cartridge can achieve.
Only filter circuits have large audio voltages appearing across the capacitor and nobody would use X7R or worse ceramics for that. Sometimes electrolytics end up being used in filters as plastic types are not sensibly available in the value. Amplifier coupling does not fall into this category as the value should be large enough to have negligible ac voltage in the audio range
Only filter circuits have large audio voltages appearing across the capacitor and nobody would use X7R or worse ceramics for that. Sometimes electrolytics end up being used in filters as plastic types are not sensibly available in the value. Amplifier coupling does not fall into this category as the value should be large enough to have negligible ac voltage in the audio range
Yep, the RF guys deal with dynamic ranges and levels of selectivity that make us hyper sensitive to things like IMD, because when designing a high performance rx it **really** matters (IMD3DR is possibly **THE** figure of merit).
For audio linearity is almost always dominated by the transducers in play (condenser microphones, even high end ones, are interesting to measure IMD on, particularly at high frequency, the reputation the things have for a richer sounding mid range suddenly makes a lot of sense!).
I am in agreement that for any coupling cap the trick is simply to make it so large that no meaningful signal voltage that matters appears across it (And if electrolytic to ensure a little bias voltage is present).
There are now C0G ceramic caps in values interesting for audio filter networks operating at sane impedances which is nice as they are available is reasonable SMT packages.
Regards, Dan.
For audio linearity is almost always dominated by the transducers in play (condenser microphones, even high end ones, are interesting to measure IMD on, particularly at high frequency, the reputation the things have for a richer sounding mid range suddenly makes a lot of sense!).
I am in agreement that for any coupling cap the trick is simply to make it so large that no meaningful signal voltage that matters appears across it (And if electrolytic to ensure a little bias voltage is present).
There are now C0G ceramic caps in values interesting for audio filter networks operating at sane impedances which is nice as they are available is reasonable SMT packages.
Regards, Dan.
I'm not sure what was the point of the link in post 14. It is well known that high-k ceramics are very non-linear dielectrics, which is why they are only used where this does not matter. The writer of that piece from the cap manufacturer seemed to be labouring something obvious; perhaps he was a junior engineer who had only just realised it himself? High-k ceramics are never used for coupling in serious audio systems.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.