DC Heater voltage question

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

rphancock1

Member
Joined
Aug 14, 2018
Messages
17
I’m working on the second of a pair of Berlant Concertone preamps and have a question about the heater voltages I’m finding.  The first sounds great after I set up new input and output transformers, however the second passed a little quiet and distorted audio before going quiet.  Since I didn’t see any glowing filaments in the preamp tubes I checked the voltages against the functioning one. What I was surprised by was the voltage in the functioning preamp... as spec’d In the heater schematic below, there is about 25 volts going into the rectifier.  Then, the voltages vary widely on different tubes heater pins, from 14 volts on a 4,5 pin combo to 4 volts on a pin 9, even zero volts in one case.

Can anyone explain this to me? Happy to provide a tube by tube voltage if it helps, but I was expecting to find something closer to 12 or 6 volts on each pin. I’ve only built ac heated circuits before.
Thanks
 

Attachments

  • 1A7D030F-A11E-4E84-9B1C-C42BDE52DA23.jpeg
    1A7D030F-A11E-4E84-9B1C-C42BDE52DA23.jpeg
    995.8 KB · Views: 44
You should have 6.3 volts on the filaments. Normally the filaments will be fed 6.3 volts a.c. or d.c. Normally only the low level valves will have  the d.c. You should check the schematic, and make sure all valves have 6.3 volts. Also check that some may have a.c.
 
From the schematic it looks like all the heaters are connected in series. Between what points did you measure the voltages you mentioned?

Cheers

Ian
 
6.3 v was what I was expecting... that’s why I was surprised.

I was checking each pin relative to 0v ground, should I be checking pins 4,5 relative to 9?

Thanks both for response!
 
Ah I see. Checking pin 4,5 to 9 on the functioning unit is 5.37 volts. I thought since it was dc it should be ground referenced. But for an electronics engineer I’m a great furniture maker haha.

On the non functioning unit I’m getting a much lower voltage from the PT going into the rectifier than on the functional one. Any ideas what would cause that? I’ll post a schematic shortly..
 
Specifically I’m getting 24.4 volts into the rectifier on the functioning unit and 14.4 volts into the rectifier on the the non functioning unit, producing 2.7 volts dc on the filaments.

Any ideas where I should look for the reason the voltage is low? I swapped the good tubes from the functioning unit into non functioning and no difference.. being overloaded somewhere else on the circuit could drop all the voltages, no? Schematic below...
thanks for any help.
 

Attachments

  • 6370EA26-2329-4608-8A21-35723EC804AE.jpeg
    6370EA26-2329-4608-8A21-35723EC804AE.jpeg
    282 KB · Views: 27
Just checked and all other voltages are consistent between the units and the schematic (adjusted for wall voltage increase). Just the heater voltage is low from the PT on the non working unit. ...
 
Also removed heater supply from rectifier and the voltage is spot on, so the PT is good. Checking the rectifier next and I google “how to check a selenium rectifier” haha.
 
> I thought since it was dc it should be ground referenced. But for an electronics engineer I’m a great furniture maker haha.

I don't make furniture. But in house-whacking, the foundation is "ground referenced". A precision steel building may be ground referenced. But I'm putting an upper floor on a 1948 self-built house. Up there I am NOT looking at "ground reference". I'm just referencing the top of the lower story.

Your amp is a 5-story house with 6.3V rise per story. Take your tape story to story and check for 6.3V from floor to ceiling.

You say you expect 24V and get 14V? The house carpenter expects shortages, but it is unlikely a long-working transformer will shed 10V. Usually all or nothing. Pull a tube to take the load off; does it change?
 
It’s the selenium rectifier. The PT filament winding reads perfect voltage when removed from the circuit.  Replacing it with silicon diodes now... will do the same in the other unit since the voltage has dropped a bit there too below 6.3.

I found this in “unworking” condition, so all options were on the table for failing, but it’s working fine with new diodes and dropping resistor in series.
Thanks!
 
Take care replacing the selenium rectifier with silicon diodes because you will probably end up with a higher dc voltage due to a lower voltage drop across the silicon rectifiers. You might need to add a small value series resistance.

Cheers

Ian
 
Selenium diodes and bridges can go bad , probably best practise to replace ,eficiency and voltage drop means people often put in a resistor in series with silicon  to drop away the same number of volts the selenium item did , if your replacing the psu filter caps with modern compact types with a higher value  you could expect the voltage to go up a bit too.

Just a small word of advice,mainly for people who are just getting into electronics  ,I always treat the heater supplies with the same respect as other high voltage parts of the circuit ,in other words I simply never ever make physical contact with any part of the circuit while its energised , Sometimes heater supplies are referenced to voltages above ground and so can present  a hazard, it could be all to easy to misread the circuit and presume its safe to touch  something with 6.3 volts dc on it ,you could get a nasty surprise .Its simply safer to always treat the heater supply as dangerous.
 
ruffrecords said:
Take care replacing the selenium rectifier with silicon diodes because you will probably end up with a higher dc voltage due to a lower voltage drop across the silicon rectifiers. You might need to add a small value series resistance.

Cheers

Ian

Maybe in this case the increase in voltage due to the Silicon diode or bridge replacement will work to the OP advantage, since he was reading 5.37 volts before on the filaments, maybe after the change to Silicon Diodes this voltage is closer to 6,3V

What do you think?
 
Whoops said:
Maybe in this case the increase in voltage due to the Silicon diode or bridge replacement will work to the OP advantage, since he was reading 5.37 volts before on the filaments, maybe after the change to Silicon Diodes this voltage is closer to 6,3V

What do you think?

Quite possibly. On the other hand, the caps are only rated at 25V so using silicon diodes may expose them to voltages above their rating which will not be good for them.

Cheers

Ian
 
Tubetec said:
Just a small word of advice,mainly for people who are just getting into electronics  ,I always treat the heater supplies with the same respect as other high voltage parts of the circuit ,in other words I simply never ever make physical contact with any part of the circuit while its energised , Sometimes heater supplies are referenced to voltages above ground and so can present  a hazard, it could be all to easy to misread the circuit and presume its safe to touch  something with 6.3 volts dc on it ,you could get a nasty surprise .Its simply safer to always treat the heater supply as dangerous.

Not to mention meter terminals.  I some valve limiters (Gates SA39 & SA39 spring to mind) the meters can be at a couple of hundred volts .....
 
Remember as well you might have been dealing with ancient smoothing components ,giving way more ripple and less Dc than it should have , Id say for minimum hum aim for X10 original values on  the caps on the LT ,even at that it should be no problem physically to fit modern high value small sized electros into it.

Just looking back on your schem , its a bit of an odd ball series parrelell heater arrangement ,if a heater goes open ,your going to read wrong voltages everywhere, and probably stressing  the heaters with continuity . Fillaments /heaters  almost always emit a dectectable glow , you may require low ambient light conditions to see it though .check each valve's heater for continuity with Ohm Meter.
 
Thanks for the responses..

I made a quick rectifier with some 1n4007 diodes on hand and thought that I’d need a series resistor, but the filaments are reading 5.7 volts now on this unit. That’s what my other unit is also reading, which still has the selenium rectifier. I have new filter caps coming which I suspect will help given that these are 70 years old. 

The filaments wired in series don’t seem to draw enough amps (0.7 on this unit) to exceed the 1A rating on the 1N4007’s but still I have others on order. If I’m misunderstanding something there correct me... in parallel they would draw much more I understand.

Amazing how quiet the unit is even so... sounds great actually, I replaced the original 11pin input transformer socket with an octal and wired it for a Dukane 3A55, and rewired the output socket for an Altec 15095A.

These were designed for ribbon mics, so have tons of gain and no phantom power. Nice for DI bass or of course Ribbons.
 

Attachments

  • 0AF6655E-1C97-4A70-8FFC-2D2A667FEE89.jpeg
    0AF6655E-1C97-4A70-8FFC-2D2A667FEE89.jpeg
    1.3 MB · Views: 17
In many-many cases, 12AX7 can be significantly "under-heated".

It is a 10mA cathode (similar to 12AU7/12AT7) and rarely run much over 1mA.

Dynaco ran 12AX7 down near 5.5V.

I recently saw another Fine Brand working them even cooler.

Diodes: the spiky waveform stresses the rectifiers far beyond simple DC current. 1N4007 will fail sooner rather than later. I would go at least 3A. Considering what diodes cost today, and the "cost" of a failure just when the musical talent got their act together, I would round-up generously.

EDIT: it appears that 1A diodes are OK at 0.3A of DC. Still, I once made a salary in-part by replacing marginal rectifiers. If nobody is paying you per-hour, round-up plenty.
 
Interesting result , 5.7volts maybe around  -20% tollerance ,probably work just fine in this instance ,if your noise is low and its all sounding good there might be no reason  to disturb anything else.
you could  try  soldering new caps in paralell with the existing ones after the bridge, remember removing whats there could leave you with another job to mount the new components. Even though your existing caps are 70 years old and their value may well be out of spec,they might be adequate for another lifetime of use .  If the units withstand longer term testing, hold up ok ,and your happy with the sound maybe just the new diodes are all thats required .
 
Tubetec said:
Interesting result , 5.7volts maybe around  -20% tollerance ,probably work just fine in this instance ,if your noise is low and its all sounding good there might be no reason  to disturb anything else.
10% of 6.3 is 0.63. 6.3 minus 0.63 is 5.67 so they are within the nominal 10% tolerance.

Cheers

Ian
 

Latest posts

Back
Top