|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
Hi,Will a hydrid rectifier with 2 silicon diodes and 2 vacuum diodes offer acceptable performance? I have seen what seems to be a general concensus on this forum that silicon bridges have too much hash to offer top performance. I am interested because I am slowly collecting parts and designs for a transmitting tube amp to be built someday and I have a possibly suitable power transformer, except that it is not center tapped. I am not ruling out an all tube bridge, but the additional two filament transformers are a bit of extra hassle, weight, heat, etc. What sort of silicon diodes would be suitable, if any? Or should I keep looking for a centertap transformer? Voltage is undecided, as are topology (SE or PP) and tubes 211,813,845,GM70), but would be in the usual 850-1500V range. Suggestions as to designs also much appreciated.
Follow Ups:
Hi,Is any particular type of diode favored for this application? Or anything rated for voltage and current will work equally as well.
1N4007 or BY255 is what i will be using in my PSU
My own circuit to ease rectifier requirements uses a SS full wave bridge (four diodes) with a rectifier tube in the negative lead. The tube's anodes are paralleled, so it acts as a single diode. I place the main filter choke between the rectifier tube and ground to minimize hash that might otherwise be conducted to the chassis. This also relieves the voltage breadown requirements for the choke (an important consideration when anode voltages approach 1KV).
"I place the main filter choke between the rectifier tube and ground to minimize hash that might otherwise be conducted to the chassis" means that the choke is in the ground leg and not shunted from rectifier output to ground! ;)Brian.
nt
From Lundahl's website. Said to perform equally well as all tube. Tube rectifiers offer isolation against SS recovery noise.
What prevents diode switching noise in this circuit from traveling through the ground reference to active stages? Given that noise impinging on a cathode contributes to the grid signal, the ground reference might even be more sensitive to diode hash than B+.
One big advantage of the hybrid rectifier arrangement is there is very little or no reverse recovery spike to deal with.To generate a reverse recovery spike it takes reverse current. The rectifier tube does not have reverse recovery issues hence no reverse current at switching. The 2 diodes are in series so if no reverse current flows back through the rectifier tube, there is no reverse current in the SS diode either. With no reverse current the switching hash becomes a non issue.
> The 2 diodes are in series so if no reverse current flows back through the rectifier tube, there is no reverse current in the SS diode either. With no reverse current the switching hash becomes a non issue.Even if the tube is removed, each SS diode will be reverse biased to nearly the full voltage of the transformer once each cycle. Won't that create reverse current?
Triode Kingdom asks:"Even if the tube is removed, each SS diode will be reverse biased to nearly the full voltage of the transformer once each cycle. Won't that create reverse current?"
The reverse leakage caused by the reverse bias is not the issue.
The real issue happens at the point where the SS diode is conducting the full load current then is asked to switch off. Their is a finite time it takes to sweep the minority cariers out the PN junction. What this means is when the diode is reverse biased at switch off, current starts flowing backwards through the SS diode. When the diode finally does switch off there is some current flowing backwards into the secondary of the output transformer. Just like in your car's ignition system, when this current is interupted a narrow high voltage spike is generated. This is the issue that all the snubbers and other tricks to get SS diodes to sound good tries to address.
Shottky diodes, Vacuum diodes, and the new sillicon carbide high voltage diodes don't have minority cariers like PN junction diodes do. This is why they don't generate the reverse recovery spike.
Gary
I'll try to make this a little more clear. Here's a prop for the info presented.
The circuit is a loop. When the cap is getting charged all 4 elements: transformer secondary, vacuum diode, filter cap, and SS diode are in series. Current is charging up the filter cap. When the voltage on the secondary reverses both diodes disconnect.
The real excitement happens at the moment that the secondary just starts reversing. With SS diodes there will be a short moment when both diodes are conducting backwards. When the SS diodes finally do disconnect we get the reverse recovery spike as the magnetic field setup by the reverse current field collapses.
With the vacuum diode in the curcuit things change. The vacuum diode does not have reverse recovery current. By the fact that all 4 elements are in series the current through each element is identical, (first approximation here!).
I suspect that there might still be a small amount of noise generated by the SS diode working against the secondary inductance and capacitance, but it should be quite small compared to what the reverse recovery spike can be.
If anyone else has info to add please do!
I understand the concept you're presenting. However, your example is very different from the hybrid circuit under discussion. The hybrid has two SS diodes, not one. As one diode switches on, the other switches off. This means that, for a brief period, they will both conduct in the manner you have described. Why won't this action create a spike?Incidentally, if you're not certain about the hybrid, just say so. I'll cobble up a quick proto and test it. Seems a useful circuit if it works. I'm just skeptical. :)
I only showed half of the full wave bridge to make it eaiser to follow. You can add the second SS diode and the second vacuum diode if you like. It doesn't change how the circiut works. Only one SS diode and vacuum diode at a time are conducting.I fully understand how the circuit is layed out and works. It is how the supply is setup in both my 300B amp and the david monoblocks.
Gary P,
I think you might not be right on this issue.
If I go back to basics, current flows in loops and there has to be a return connection from the first filter capacitor for the power supply to work. My analysis suggests that when we charge up the 1st filter capacitor, you need to have current flowing through the tube diode, charging the filter cap, and returning through one of the solid state diodes.
Perhaps I'm missing something. Your thoughts?
Thanks,
---Gary B
> The real issue happens at the point where the SS diode is conducting the full load current then is asked to switch off. Their is a finite time it takes to sweep the minority cariers out the PN junction. What this means is when the diode is reverse biased at switch off, current starts flowing backwards through the SS diode.Thanks for the explanation. I've always seen the spikes, but assumed they were just noise generated by the rapid on-off switching.
I still don't understand why the hybrid circuit won't create the same effect. The SS diodes in the schematic are performing the same task you've described. With or without the tube, the SS diodes deliver current, and are then suddenly asked to turn off. Why wouldn't the reverse current during that time be supplied by the power transformer? Are you certain the circuit eliminates the current spikes in the lead connected to ground?
Done it works great!
Works good, is a common upgrade for guitar amps. I use fast 1A, 1kV diodes to establish the ground. I put this mod in a Blues Deluxe and it actually got rid of the rectifier buzz that was audible at full volume. I suppose you could use even fancier soft recovery diodes. If you install a 3A 5V filament transformer then you can try various 5V rectifiers. 5AU4 tube is robust and a bargain.
Yes, you can use a tube-diode combo bridge. If you use an indirectly-heated tube rectifier, you'll get a slow ramp up, which may be useful, and you will get more of the "naturalness" of an all-tube rectifier.
This post is made possible by the generous support of people like you and our sponsors: