The rectifiers are switching AC all the time, and no horrible B+ modulation takes place. This shows you that it's completely determined by your LCLC stage, and only very minimally by the type of rectifier, influenced by its series resistance. You can simulate this using LTspice or you can use the standard rules of thumb or nomographs in the old texts.
The loud 'pop' at startup would be from your single power supply voltage rising suddenly into a circuit that has nearly zero power supply rejection. Again though, your LCLC stage is designed to slow B+ voltage transitions down, so it might be sufficient to slow the rise of B+.
How much is a 'loud' thump depends on the circuit and how much gain it has on the B+ line. If the B+ line rises too quickly, you could put a wirewound resistor between the rectifier and the LCLC filter and use a relay to short it out once B+ has fully come up. This will give you a slow start that goes away once B+ is fully charged.
The loud 'pop' at startup would be from your single power supply voltage rising suddenly into a circuit that has nearly zero power supply rejection. Again though, your LCLC stage is designed to slow B+ voltage transitions down, so it might be sufficient to slow the rise of B+.
How much is a 'loud' thump depends on the circuit and how much gain it has on the B+ line. If the B+ line rises too quickly, you could put a wirewound resistor between the rectifier and the LCLC filter and use a relay to short it out once B+ has fully come up. This will give you a slow start that goes away once B+ is fully charged.
Hi Monte,
Thanks for the reply.
Obviously, the advantage of the 5U4G rectifier I am currently utilizing is the graceful turn-on/ramp-up of the high voltage supply and the application of the B+ to the output tube, and the rectifier tube filament and HV AC can of course be energized simultaneously.
I had the thought of soft-starting the 300B plate supply; my thinking was to put the B+ limiting/dropping resistor between the center-tap of the high voltage winding to ground, and either manually or via a time-delay relay, short the resistor out after a second or so to obtain full B+. The soft-start resistor should ideally be in the primary circuit of the plate supply, but the power transformer I am using provides both filament voltage and HV AC to the rectifier tube, so the center-tap concept becomes the most practical.
The voltage dropping resistor value can be calculated so that the initial turn-on voltage to the 300B plate is somewhat greater than the negative grid bias voltage, so the 300B is conducting and drawing some current from the supply. Another possible advantage is it limits the inrush current to the filter capacitors for a few milliseconds as they charge up, but this is not a significant concern in a power supply with a relatively small amount of distributed energy storage.
The issue I have with including the soft-start circuit entails the available real estate within the power supply chassis, which is separate from the amplifier chassis. I have virtually no room to add the time-delay relay and a high-voltage contactor for shorting out the voltage dropping resistor.
I really would like to hear from users of amplifiers with MV rectifiers and high sensitivity speakers who are turning on the HV to the output tube by way of closing the center-tap of the plate transformer HV winding to ground (and no soft-start circuitry), and the magnitude of the transient they hear in their loudspeakers on HV application to the output tube plate..........
Thanks & Regards,
Bruce
The 866A or its "little brother" the 816 should work fine for your application with full wave center tap rectification (two valves) with LC filtering. The filtering should be inductor input to keep down surge currents which could cause arc back and damage the valve. Also, application of HT should be delayed approximately one minute to allow sufficient mercury to vaporize and minimize risk of arc back or damage by cathode bombardment.
The mercury vapor rectifiers have a constant 15 volt drop regardless of load...these are excellent for amplifiers using valves in class A, AB or B audio applications.
My primary line of work is RF.... I worked with transmitting equipment using mercury vapor rectifiers including the 816, 866A, 8008, 575A, and the venerable 857B.
I have never experienced a mercury vapor rectifier explode where the glass envelope is broken. Protective devices on the HT transformer should be used to shut down power if a short or arc back occurs.
The mercury vapor rectifiers have a constant 15 volt drop regardless of load...these are excellent for amplifiers using valves in class A, AB or B audio applications.
My primary line of work is RF.... I worked with transmitting equipment using mercury vapor rectifiers including the 816, 866A, 8008, 575A, and the venerable 857B.
I have never experienced a mercury vapor rectifier explode where the glass envelope is broken. Protective devices on the HT transformer should be used to shut down power if a short or arc back occurs.
RMB, thanks for your input! I am very familiar with MV rectifiers, having used them in high power radio transmitters, etc., such as yourself, as well as in my homebrew phono preamp.
My real question is such that I really would like to hear from users of power amplifiers with MV rectifiers and high sensitivity speakers who are turning on the HV to the output tube by way of closing the center-tap of the plate transformer HV winding to ground (and no soft-start circuitry), and the magnitude of the transient they hear in their loudspeakers on HV application to the output tube plate..........
Thanks & Regards,
Bruce
I have 866's running in PP 6CB5A amps. The plate TX primary is switched by an Amperite relay of 45 seconds delay. These guys run the mids in my tri-amp set up, and there is no 'thump' on start. There is a big FLASH when the relay closes and the caps are empty...🙂 No noise either. Or thump when used full range either. The next amps are alos going to get Hg, in the form of 866's...🙂
cheers,
Douglas
cheers,
Douglas
Briefly used 866As before moving onto bigger and better things. No issues with power-on/off thumps in a PP IT 2A3 driving 92db Wilsons.
I'm a user of DCG4/1000, which are identical to 866 in terms of parameters IMHO.
I have separate heater and B+ power supplies that I switch separately at a manual sequence. The B+ is initiated through a resistor that serves as a passive soft start. Seconds later, I give a full B+ turn-on.
I have some 200 hours on my MV rectifiers and they work flawlessly.
I have separate heater and B+ power supplies that I switch separately at a manual sequence. The B+ is initiated through a resistor that serves as a passive soft start. Seconds later, I give a full B+ turn-on.
I have some 200 hours on my MV rectifiers and they work flawlessly.
I am starting to use the 866A now in my power-amp stage (a 300B PP).
Compared to normal FW rectifiers you get a very high resolution and a much stiffer tonal character, for sure not the tool to make an amp sound warm, but a great way to get a few layers of veil away and look much more into th recording. The warmth need to come from somewhere else. Its a bit like a Hexfred, but sounding at the same time like a tube.
I have read somewhere you better should match the 866A to reasonable pairs...how would you do this actually ? Is there a simple test circuit for this ? or can this be done in the amp ? They always have their 15V loss, no matter howmuch current you take from them...so how do match them ?
Compared to normal FW rectifiers you get a very high resolution and a much stiffer tonal character, for sure not the tool to make an amp sound warm, but a great way to get a few layers of veil away and look much more into th recording. The warmth need to come from somewhere else. Its a bit like a Hexfred, but sounding at the same time like a tube.
I have read somewhere you better should match the 866A to reasonable pairs...how would you do this actually ? Is there a simple test circuit for this ? or can this be done in the amp ? They always have their 15V loss, no matter howmuch current you take from them...so how do match them ?
I'm not sure about matching 866A's. I used them in my 211 amp and did not match them. I also used CLC (20uF/10H/150uF) and the 866A's did not complain despite a lot of people saying that the power supply should be choke input.
I used 816's as well without matching. I currently have 3B25's in there now and I like the way they sound. I swap out the 866A and 3B25 every now and then.
You could put a small value resistor in each plate lead to the tube and measure the AC voltage across the resistor. This would give you an indication of current through each plate lead from the transformer.
I used 816's as well without matching. I currently have 3B25's in there now and I like the way they sound. I swap out the 866A and 3B25 every now and then.
You could put a small value resistor in each plate lead to the tube and measure the AC voltage across the resistor. This would give you an indication of current through each plate lead from the transformer.
Last edited:
Very interesting...do you experience any sound difference within one type lets say Sylvania vs. Rca 866a or American types vs. Europe types or ST vs. Globes ?
I can't say that I have heard much difference between different brands of 866.
The 3B25's seem to have a slightly different "RF hash" signature, but it is hard to tell if there is really a sonic difference between the 866 and 3B25.
The 3B25's seem to have a slightly different "RF hash" signature, but it is hard to tell if there is really a sonic difference between the 866 and 3B25.
Bruce
You can design a relative simple circuit with simple diodes to charge the C and rest of the PS module. While the rectifier is on hating. So the rectifier tube will come to already charged Cs... That will significantly increase life of the rect. tubes. And I am almost positive that will be no "boom" on turning on. And the stand by time will be just the one that the audio tubes deserve?
...
Issues could be that You have to set the close HV to one that tube rect. giving. So some drop at the secondaru for diodes, or so should be added.
Another thing is that relay or manual switch for tube/diodes, have to be from high quality.
.
cheers
You can design a relative simple circuit with simple diodes to charge the C and rest of the PS module. While the rectifier is on hating. So the rectifier tube will come to already charged Cs... That will significantly increase life of the rect. tubes. And I am almost positive that will be no "boom" on turning on. And the stand by time will be just the one that the audio tubes deserve?
...
Issues could be that You have to set the close HV to one that tube rect. giving. So some drop at the secondaru for diodes, or so should be added.
Another thing is that relay or manual switch for tube/diodes, have to be from high quality.
.
cheers
Last edited:
I've been running a pair of 872s in a 2kv supply for about 2 years now, with no issues. I also have an 83 on my 6L6 amplifier. It works nicely, and one bulb 2 rectifiers.
Main thing is a seperate 5v tranny for the filament, so they can be pre heated prior to HV to the anode. And on first run of a new tube, I like to just run the filament for about 30 or more minutes. To insure that all the mercury is fully gone from the filament, and plate.
Sure, there will be a flash as the caps charge, but, that's what they do. With a choke input filter, they work very well.
Main thing is a seperate 5v tranny for the filament, so they can be pre heated prior to HV to the anode. And on first run of a new tube, I like to just run the filament for about 30 or more minutes. To insure that all the mercury is fully gone from the filament, and plate.
Sure, there will be a flash as the caps charge, but, that's what they do. With a choke input filter, they work very well.
I have simply built a 1k5 resistor serial into the ct - earth connection which will be bridged after 7sec of HV is on by a relais (555 timer midule from china for 3$). This way the HV will be slowly applied, no flashes here.
I feel I must advise that employing a TD tube as you are, is extremely hard on a thermal TD relay !!! The easiest 'better' would be to simply install an appropriately sized thermistor in series with the TD contacts and the load. Acts like a resistor until it gets hot and basically takes itself out of the circuit looking like a piece of wire in under a minute. Soft start on the whole HV supply will improve longevity.
I dislike the contact closure bounce that many of these thermal TD relays (SPST switches actually) seem to have. When used gently to close a contactor relay to handle the load, this settling can be easily heard as contactor clatter. It is this bouncing whilst trying to energize a low Z inductive load with high inrush current that generates a wicked back EMF which ionizes the opening contacts after each bounce, vaporizing the switch contacts in the impressive light show you described.
There are numerous solutions to this issue. That said, I do not wish to interfere. 'If you like your light show, you can keep your light show.' 😉
Cheers,
rca
I found this article quiet helpful
https://www.die-wuestens.de/iz/RECT.pdf
..they do recommend to use some small caps on the secondaries to the CT and small HF-choke onto of the Anodes...what is small in this context, has anyone done this and can recommend a type ?
https://www.die-wuestens.de/iz/RECT.pdf
..they do recommend to use some small caps on the secondaries to the CT and small HF-choke onto of the Anodes...what is small in this context, has anyone done this and can recommend a type ?
I've used small ferrite cores on the plate leads. Run the lead through the core and put some heatshrink over it to hold the core in place.
How small is small...like these here:
FR3 FERROCORE - Inductor: ferrite | TME - Electronic components
or more like
B82733V2701B001 EPCOS / TDK | Mouser Deutschland ...?
FR3 FERROCORE - Inductor: ferrite | TME - Electronic components
or more like
B82733V2701B001 EPCOS / TDK | Mouser Deutschland ...?
- Status
- Not open for further replies.