Buy some of these. Slow warm up, higher mA rating, less voltage drop, and quite high reliability. Should be plug and play it your circuit. They used to be around $10 a piece. Oh well.
New Shuguang 5AR4 GZ34 Rectifier Vacuum Tube 2pcs 5U4 | eBay
New Shuguang 5AR4 GZ34 Rectifier Vacuum Tube 2pcs 5U4 | eBay
Wait... this means that in this amplifier, the rectifier tube has absolutely no purpose. It does not rectify and it does not delay B+. A resistor would do the same job in that position.
It's only there for marketing.........
Eko2, you should redesign the PSU to use your tube rectifiers. Eli has already drawn the schematic.
You could download Duncan's PSUD2, measure your transformer voltage (off load) and primary/secondary resistance and then simulate everything to get the voltages and filtering you need.
If I were you, I would keep the transformers and the chassis, trash the rest, and rebuild a better amplifier.
It's only there for marketing.........
Eko2, you should redesign the PSU to use your tube rectifiers. Eli has already drawn the schematic.
You could download Duncan's PSUD2, measure your transformer voltage (off load) and primary/secondary resistance and then simulate everything to get the voltages and filtering you need.
If I were you, I would keep the transformers and the chassis, trash the rest, and rebuild a better amplifier.
Thanks for the link, I'll buy 2. I have 3 Philips GZ34 but not very happy to use them in this kind of power supply.. anyway I tried these modifications:
- 2 NTC in series (cold resistance about 90 ohms) between cathode and the big 100R.
- 100R between GZ34 plates and bridge rectifier (current limiting ?)
- 230VAC secondary instead of 280VAC
At power on, voltage rise gently from 0 to 330V after 20 seconds, then about 260V on EL84 plates (was 295-305V before).
I could use the 250VAC secondary instead of 230V or remove the 100R between GZ34 plates and bridge to rise B+. I am not sure if it's a good idea to add this resistor..
- 2 NTC in series (cold resistance about 90 ohms) between cathode and the big 100R.
- 100R between GZ34 plates and bridge rectifier (current limiting ?)
- 230VAC secondary instead of 280VAC
At power on, voltage rise gently from 0 to 330V after 20 seconds, then about 260V on EL84 plates (was 295-305V before).
I could use the 250VAC secondary instead of 230V or remove the 100R between GZ34 plates and bridge to rise B+. I am not sure if it's a good idea to add this resistor..
Attachments
Hello,
I didn't notice any change after these modifications (lower B+, GZ34 instead of 5Z2P) except the PT is a little bit cooler now.
Now the better power supply following Eli D. proposal.
Did I make a mistake on the schematic or all is right ?
Thanks
I didn't notice any change after these modifications (lower B+, GZ34 instead of 5Z2P) except the PT is a little bit cooler now.
Now the better power supply following Eli D. proposal.
Did I make a mistake on the schematic or all is right ?
Thanks
Attachments
Neither the 50 Ω resistors nor (especially) a variable resistor are needed in the hybrid bridge setup. You control the B+ rail voltage by adjusting the value of C8. Start with 4.7 μF. for C8 and see what you get. The rail voltage will climb, as the value of C8 grows.
Unnecessary heat is the enemy of all things electronic. Removing "useless" resistors, obviously, lowers the amount of heat. The 1st cap. in a CLC filter should be only large enough to keep the rail voltage up. Smallish 1st caps. reduce I2R heating in the power trafo.
UF4007s generate MUCH less switching noise than 1N4007s. Less noise, to begin with, has to be better. Make that parts switch. While the vacuum rectifier keeps SS diode switching noise out of the B+ rail, noise can still sneak into the signal, via the heater supply. Put the film snubbers I showed in. Those snubbers can also reduce ringing in the rectifier winding.
Unnecessary heat is the enemy of all things electronic. Removing "useless" resistors, obviously, lowers the amount of heat. The 1st cap. in a CLC filter should be only large enough to keep the rail voltage up. Smallish 1st caps. reduce I2R heating in the power trafo.
UF4007s generate MUCH less switching noise than 1N4007s. Less noise, to begin with, has to be better. Make that parts switch. While the vacuum rectifier keeps SS diode switching noise out of the B+ rail, noise can still sneak into the signal, via the heater supply. Put the film snubbers I showed in. Those snubbers can also reduce ringing in the rectifier winding.
Did I make a mistake on the schematic or all is right ?
I think you should simulate with a full wave rectifier in PSUD and not a Graetz bridge - the voltage drop is not the same.
Measuring real transformer offload voltage and windings resistances is also necessary for an accurate result.
I don't know.
You could measure with a scope the ripple you already have with the original PSU, or you could simulate it (take into account the vacuum rectifier's plate resistance in the CRC filter).
If it's not enough, you can add another R-C stage after the L-C, if you have enough voltage to spare.
Not sure.
http://www.diyaudio.com/forums/tubes-valves/227184-hybrid-rectifier-possible-psud.html
30 ohms between 0V & 280V secondary. The good news is that the PT has several secondaries: 230-250-265-280V. I didn't do any measures offload yet.
Unfortunately there is no enough space for a big choke
See post #45, I added two NTC in series between GZ34 cathode and 100R, I hope it makes life easier on the rectifier.. but it's only a "quick fix", now I have to make a better power supply
The only doubt I have is about "high ripple" on screen grids finals. Well I'll try and see..
Surely these would have no measurable benefit since the DC resistance of the transformer secondary is likely to be roughly a couple of orders of magnitude greater, and that will act to limit any inrush current. The valve rectifier also has a rating for series resistance which is typically also typically at least 50 ohms (again the transformer winding resistance can be used for all or part of this as needed).
If you want reduce g2 B+ ripple, increase C10 to 47 μF.
The remarks about a larger inductance than 1.5 H. are correct. I'd prefer a 6 H. Triad C-14X, but I don't know if it can be sourced and if sufficient space is available.
Keep in mind the final value of C8 is TBD. Starting with a lower RMS value into the hybrid bridge definitely allows a larger capacitance to be used. Simulations have their place, but actual measurements made on the bench rule. Experiment with several configurations. Good old "cut and try" worked just fine for decades and it will resolve this situation too.
The remarks about a larger inductance than 1.5 H. are correct. I'd prefer a 6 H. Triad C-14X, but I don't know if it can be sourced and if sufficient space is available.
Keep in mind the final value of C8 is TBD. Starting with a lower RMS value into the hybrid bridge definitely allows a larger capacitance to be used. Simulations have their place, but actual measurements made on the bench rule. Experiment with several configurations. Good old "cut and try" worked just fine for decades and it will resolve this situation too.
Well, now you can see what the insides of a tube are like
How tight is that shield-base on the tube socket? Looks like it might have an edge on the inside touching the tube. Thermal expansion of the glass pressed it against the socket and gave you a nice clean break.
Might happen to the next tube, too, if not looked after.
How tight is that shield-base on the tube socket? Looks like it might have an edge on the inside touching the tube. Thermal expansion of the glass pressed it against the socket and gave you a nice clean break.
Might happen to the next tube, too, if not looked after.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.