Pair of 6AU4s can give 380 mA, but need 3.5A heater current. Octal, at least. Good enough for Marantz...
Thanks for the tip. I still do not know if I have to do anything special to make sure they perform OK in parallel?
Cheers
Ian
Assuming I can still only use a 32uF capacitor after a pair of 6AU4s for instance, with a 300mA load I calculate that equates to over 90V pp ripple.
Cheers
Ian
Cheers
Ian
Yeah, I simulated it, and peak repetitive current is over 1A, (pretty much regardless of cap value) - most any tube rectifier will be a problem. 5V3A is good for 1.4A PRC, though.
I would use a Si rectifier in this task, both for reliability nd for the ability to filter well. If voltage allows
a regulator circuit too.
a regulator circuit too.
Folks only need be concerned about rectifier "sound" if they haven't tackled the the real issue, ringing in the transformer's secondary circuit. It's like looking for your lost car keys under the street lamp because the light is better. Fix the problem rather than choose a pretty color for the band-aid.
All good fortune,
Chris
All good fortune,
Chris
Wow, that's heavy. So yes, you can parallel two EZ81s, either tying their anodes together or using them as two full-wave rectifiers. A great alternative would be two PY82s which will happily produce 0.18 A each. These are essentially EZ81s with the anodes paralleled. Choke input is a must at these current levels to reduce peak currents. The PY82s will need 19 V filament voltage (at 0.3 A).
This is how German musical instruments manufaturer did it in their M 40 amplifier:
It can' be easier.
Best regards!
It can' be easier.
Best regards!
Thanks Kay. THat is very straightforward. I notice they use 2 x 50uF as the reservoir which would reduce the ripple to a more manageable 30V pp.
Cheers
Ian
Cheers
Ian
Yes, it is. I forgot to mention the MI manufacturer's name Klemt Echolette. Sorry!
Btw, the M 40 is a rather valuable and sought-after device here in Germany, mostly due to it's decorative »golden cage« case, but also due to it's versatility for it's time.
Best regards!
Btw, the M 40 is a rather valuable and sought-after device here in Germany, mostly due to it's decorative »golden cage« case, but also due to it's versatility for it's time.
Best regards!
Two things.
If you parallel the tube plates, you'll have twice the peak cathode current per tube. Using them as dual FW is preferred IMHO.
Second, I've used 6D22S in parallel by using 47R plate resistors to ensure sharing. The 6D22S has a cathode cap however.
This was a FW choke input PSU using 4 6D22s into an LCRC filter made of 10H, 820uF, 10R, 820uF. They powered a PP 6P36S into 2k2 Ra-a triodeed amplifier with 800mA peaks without issue.
Another idea would be to use say 6 6Z4P in parallel with 22R plate resistors. More complex but the tubes are cheap,
I find they last longer with a choke input, too.
I have run a 6Z4P at 70mA steady state in a headphone amp. They last about 6 months. Paralleling a pair roughly quadruples that to 2 years (24/7).
Of course, I would probably just use HER208 but that's me.
If you parallel the tube plates, you'll have twice the peak cathode current per tube. Using them as dual FW is preferred IMHO.
Second, I've used 6D22S in parallel by using 47R plate resistors to ensure sharing. The 6D22S has a cathode cap however.
This was a FW choke input PSU using 4 6D22s into an LCRC filter made of 10H, 820uF, 10R, 820uF. They powered a PP 6P36S into 2k2 Ra-a triodeed amplifier with 800mA peaks without issue.
Another idea would be to use say 6 6Z4P in parallel with 22R plate resistors. More complex but the tubes are cheap,
I find they last longer with a choke input, too.
I have run a 6Z4P at 70mA steady state in a headphone amp. They last about 6 months. Paralleling a pair roughly quadruples that to 2 years (24/7).
Of course, I would probably just use HER208 but that's me.
Last edited:
I believe the original question boils down to simply can you parallel rectifier tubes for increased current capacity? I have done this in the past without thinking about it at all... and it always worked out fine. Also, there are a few large, hi wattage guitar and bass amps out there that have dual rectifiers for this very reason and they are still around.
Here's how Mesa did it with their dual rectifier... Parallel FW to keep peak cathode current within design centre max but interestingly, no plate resistors...
to Ian
but if the use is a mixer professional why don't consider a proper ss stabilized regulation?
In this way you can get a ripple=0, low Zout and a perfect stabilization.
This is the one I am using for years. Two way, indipendent.
It has the bridge on board but it is possible to use a tube rectifier and the go this board
In this case you can use one EZ81 that drive one way and the other 81 the other way; so you can split the power for. p.e., two banks of mixer.
With 350 mA of Ia the lost in voltage is less than 1 volt, each way.
Walter
but if the use is a mixer professional why don't consider a proper ss stabilized regulation?
In this way you can get a ripple=0, low Zout and a perfect stabilization.
This is the one I am using for years. Two way, indipendent.
It has the bridge on board but it is possible to use a tube rectifier and the go this board
In this case you can use one EZ81 that drive one way and the other 81 the other way; so you can split the power for. p.e., two banks of mixer.
With 350 mA of Ia the lost in voltage is less than 1 volt, each way.
Walter
Attachments
Because these are not needed. A vacuum tube rectifier has 10-20 V voltage drop across it. Agree that fullwave is better regarding cathode peak current. Also, if one tube fails your power supply will still work albeit with reduced performance.
Google this for a start-
https://www.daenotes.com/electronics/basic-electronics/diode-in-parallel
The info applies equally to vacuum diodes
As another point of reference we could consult the data sheets covering 6AS7G/6080.
There we find info on paralleling these tubes as the lossers in regulated PSs.
These are active of course but the mu is only 2.
As another alternative we could consult something called a 'book',
Many texts on engineering cover the diodes in parallel problem from A-Z.
Because right now I am just examining alternative rectifiers. Of course after that will come ripple reduction and stabilisation but they are separate matters. Is the board you pictured available anywhere?
Cheers
Ian
I am not sure this makes sense. Let's work it out. The 5U4G data sheets states that for a capacitor input filter with 300V rms input, a 300mA load current and a 40uF reservoir capacitor, the dc output voltage is 290V. Now 300V rms is 424V peak. At 300mA and 100Hz, a 32uF reservoir will have 75V pp ripple. Add half this to the 290V dc value gives us 327.5V peak dc out. Hence the total peak voltage drop across the tube and the transformer winding resistance must be 424 - 327.5V = 96.5V. At 300mA this represents a resistance of 320 ohms. But the data sheet says the total effective plate resistance for these conditions is just 21 ohms. I just measured the secondary dcr of a random HT transformer and it measured 260 ohms so this ups the total to 280 ohms which is not that far from 320. Given I measured a 250V rms secondary one might expect a 300V rms one to have a proportionally higher dcr of about 320 ohms. SOopresumably I am correct in concluding that the majority of the required series resistance is in fact the secondary dcr of the transformer?
Cheers
Ian