Please go easy, I'm still learning. 
So far, I understand that a big reason for people preferring tubes over solid-state rectifiers has to do with zero-crossing noise.
As I understand it, the rectifier "drops out" below 0.7V or so, resulting in a sharp edge full of nasty Fourier components. Please correct me if this isn't quite right.
I had an idea on how one might correct this. Is it legitimate?
I decided to connect half of a 6AL5 vacuum diode in parallel with each rectifying diode in the belief that when the SS rectifier "dropped out," the 6AL5 would pick up the job of conducting. At about 0.7V, the current demand on the 6AL5 would be negligible. When the SS rectifier is conducting, the 6AL5 is effectively shorted and relieved from duty (which would kill it anyway).
I would imagine that the 6AL5 would "fill in" the parts that the rectifier is missing.
I can't tell if this makes any difference, but I'm curious if anyone else has thought about or tried this. Searches turned up nothing, and I lack the equipment to verify if this actually does anything.
The motivation behind doing this was that the same power transformer could be used, and the filament transformer only sees another 300mA.
So far, I understand that a big reason for people preferring tubes over solid-state rectifiers has to do with zero-crossing noise.
As I understand it, the rectifier "drops out" below 0.7V or so, resulting in a sharp edge full of nasty Fourier components. Please correct me if this isn't quite right.
I had an idea on how one might correct this. Is it legitimate?
I decided to connect half of a 6AL5 vacuum diode in parallel with each rectifying diode in the belief that when the SS rectifier "dropped out," the 6AL5 would pick up the job of conducting. At about 0.7V, the current demand on the 6AL5 would be negligible. When the SS rectifier is conducting, the 6AL5 is effectively shorted and relieved from duty (which would kill it anyway).
I would imagine that the 6AL5 would "fill in" the parts that the rectifier is missing.
I can't tell if this makes any difference, but I'm curious if anyone else has thought about or tried this. Searches turned up nothing, and I lack the equipment to verify if this actually does anything.
The motivation behind doing this was that the same power transformer could be used, and the filament transformer only sees another 300mA.
Last edited:
Yes, charge storage is the problem. The SS diode will conduct briefly in the reverse direction until all the charge carriers are removed from the junction region, then it abruptly stops. This abrupt stop has high frequency Fourier components which may get into the audio circuitry. Paralleling with a thermionic diode won't make any difference - adding a thermionic diode in series might help because it reduces the current more gradually. The correct solution is either to use a different SS diode designed to reduce this problem, or add a parallel RC snubber to damp the spike.
Easier still, just keep the PSU well away from signal circuitry and reduce the area of any loops carrying charging pulses.
Easier still, just keep the PSU well away from signal circuitry and reduce the area of any loops carrying charging pulses.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.