What's the best way to reduce voltage on heaters?

[frank57]

This Hiwatt, whether it's due to factory laziness building the transformer,
consistently has heaters too high.
7.2-7.8 in that area.
And yes, they eat up tubes in about 3-4 months.
Only thing I found was here:
http://www.freewebs.com/valvewizard/heater.html
Any ideas?

This amp also has hum issues.
Seems to have some tremolo like sound creeping in.
Bad filtering?

[kagliostro]

May be  you can solve the problem with the solution proposed by Merlin

only put two other diodes (connected the same way of the first two) in series to the other

this way the filament voltage decreases of 1.4v instead of the 0.7v

so assuming 7.8v - 1.4v you have 6.4v

Kagliostro

[The_Gaz]

I've used series resistors to drop the heater voltage before. I usually only need to drop half a volt or so, and use .05 ohm/5 watt resistors in a 50-watt amp. The PTs were overkill for those 50-watts - 8 amps for the heater winding - and I don't think the tubes could load down the transformer enough. Just a hunch.

If you really need to drop 1.5v and the amp is 100-watts, the resistors would need to be .22 ohms/11 watt, but probably an even high wattage for a good safety margin. Not really economical in you case....

Here's a great chart, which shows exactly what your seeing with your tube-life. It seems to me a lot of people will neglect a small discrepancy in heater voltage, including myself, but this chart really shows how crucial it is!

http://tone-lizard.com/images/Tremaine.gif

[phsyconoodler]

Very doubtful it's eating tubes.I have had Hammond transformers with consistently high heaters and they never eat tubes.
But,there must be some reason for your high heater voltages.Are you using the correct tap on the PT primary?
   Center tap or 100 ohm resistors?
for example,if you have 100v,110v,125v,200v or similar primaries and you are using the 100v by mistake,it will have high heater voltages like yours.
  I did that with a Hammond once by mistake and it never ate any tubes after a couple of years use gigging.I corrected it since.
Possible?

[The_Gaz]

Yes, perhaps something else is going on, but I assume the OP hooked up the primaries connected as he didn't report of any othe voltages being out of spec.

The chart shows that his tube life would be cut down to about 200 hours from 10,000 running at 125% the rated voltage. I'm not surprised at all that his amp eats tubes. This is what I'm sayin' - no one takes heaters seriously, not even Psychonoodler!

[bibi]

Do I read that chart correctly that running the heaters at a lower voltage (85%) would decrease the life of the tubes to around 3 hours?  Where is that chart from and what are the circumstances? 
According to the chart, 90% of 6.3VAC will bring the life expectancy down by 1/3?  Sorry that doesn't compute.  The manufacturers rating is within +/-10% last time I checked.

[eleventeen]

The dual diode approach works well. They have to be HIGH CURRENT diodes, obviously they have to handle all the filament current plus a margin.

Twin Reverb: 4 * .9 amps plus 6 * .3 = ~~6 amps, should use 10 amp diodes.

[Fresh_Start]

The dual diode approach works well. They have to be HIGH CURRENT diodes, obviously they have to handle all the filament current plus a margin.

Twin Reverb: 4 * .9 amps plus 6 * .3 = ~~6 amps, should use 10 amp diodes.

You could use four, 5 amp diodes with two in series going each direction correct?

Chip

[The_Gaz]

Do I read that chart correctly that running the heaters at a lower voltage (85%) would decrease the life of the tubes to around 3 hours?  Where is that chart from and what are the circumstances?
According to the chart, 90% of 6.3VAC will bring the life expectancy down by 1/3?  Sorry that doesn't compute.  The manufacturers rating is within +/-10% last time I checked.

The quote is from this book

http://www.amazon.com/Audio-Cyclopedia-Howard-M-Tremaine/dp/0672206757

I first heard of it from Randall Aiken's site on his recommended reading list, and that graphic is from the Tone Lizard site. You know, I never actually considered that side of the graph! It is is very extreme, and I'm curious. At a local book store they have the book, which is over a $100.00 used, but I'll look that up 

[Merlin]

Do I read that chart correctly that running the heaters at a lower voltage (85%) would decrease the life of the tubes to around 3 hours?  Where is that chart from and what are the circumstances?
According to the chart, 90% of 6.3VAC will bring the life expectancy down by 1/3?  Sorry that doesn't compute.  The manufacturers rating is within +/-10% last time I checked.

That graph has been doing the rounds for years. It's basically nonsense, and appears to have been borrowed from really old info on valves under saturation conditions- nothing like the way we use receiving valves. It is well known that a slightly reduced heater voltage increases life, provided you're not operating anywhere near saturation (temperature limited emission).

[eleventeen]

You could use four, 5 amp diodes with two in series going each direction correct?

That doesn't do it, for me. I have never understood the purpose/function of series diodes (EXCEPT as multiple .6 - .7 volt voltage droppers...meaning, you want a 1.3 or a 2 volt current-independent voltage drop) Perhaps it gives a degree of surge protection....BUt all the filament current goes thru both series-connected diodes, thus either one of them has to handle "max amps".

You *could* however, divide the tubes into two completely separate strings, let's say (twin rev) 3.6 amps for 4 qty 6L6 and 1.8 amps for 6 qty 12a_7 and get away with 5A diodes.   
 
 

[phsyconoodler]

I don't believe the heater voltage is that high unless there is a mistake somewhere in the PT wiring or maybe a meter with an optimistic readout.

[Fresh_Start]

You could use four, 5 amp diodes with two in series going each direction correct?

That doesn't do it, for me. I have never understood the purpose/function of series diodes (EXCEPT as multiple .6 - .7 volt voltage droppers...meaning, you want a 1.3 or a 2 volt current-independent voltage drop) Perhaps it gives a degree of surge protection....BUt all the filament current goes thru both series-connected diodes, thus either one of them has to handle "max amps".

You *could* however, divide the tubes into two completely separate strings, let's say (twin rev) 3.6 amps for 4 qty 6L6 and 1.8 amps for 6 qty 12a_7 and get away with 5A diodes.

I thought all the amps you see with 2 or 3 diodes in series for rectification were wired that way to reduce the current rating for the diodes used.  Is that wrong?  YES

Sorry for the sidetrack,

Chip

[sluckey]

I thought all the amps you see with 2 or 3 diodes in series for rectification were wired that way to reduce the current rating for the diodes used.  Is that wrong?

Correct. That is wrong.

They are wired in series in the HV rectifier circuit to increase the PIV rating. For example, the old Fender TR used 3 diodes in each leg. Back then, the diodes were mostly rated for 400PIV, so a single diode (and probably even 2 diodes) would fail due to exceeding the PIV rating.

[eleventeen]

With series diodes, all the current has to go through each and every diode.


                              ---------->|---------->|---------->|-----------

Theoretically, parallel diodes could divide the total current, but that is seriously not recommended because one diode will inevitably conduct first and take ALL the current.

          |---->|------>|------>|------>|
      ---|                                         |----------
          |---->|------>|------>|------>|

You could invoke several manners of resistors or caps across the diodes to try to counter the "one-conducts-first" effect but it ain't the way to go. IMHO. The results are unpredictable, and I think that's exactly what you don't want in a power supply. You just have to pop for the high current diodes or divide the heaters into 2 separate strings.

[Geezer]

I thought all the amps you see with 2 or 3 diodes in series for rectification were wired that way to reduce the current rating for the diodes used.  Is that wrong?

Correct. That is wrong.

They are wired in series in the HV rectifier circuit to increase the PIV rating. For example, the old Fender TR used 3 diodes in each leg. Back then, the diodes were mostly rated for 400PIV, so a single diode (and probably even 2 diodes) would fail due to exceeding the PIV rating.

and, when diodes fail they usually fail "shorted", so you have another diode (or 2) to keep the rectification happn'n & your PT happy/safe, and if they fail "open", then the amp stops working, but your PT is again "safe".

G

[Fresh_Start]

Voiltage // Current... same thing, right?  NO! Wrong (again).

Over on AX84, a guy was complaining about having 8 VAC on his heaters and Paul Fawcett was nice enough to provide links to a couple of other threads relating to dropping heater voltage.  This one seemed useful:

http://music-electronics-forum.com/t1833/

I'm not clear on how putting a reverse zener on the heater center tap would lower the voltage across the heaters though.  I do understand how that works with the HV secondary and have used it successfully in that application.

Cheers,

Chip

[Merlin]

I'm not clear on how putting a reverse zener on the heater center tap would lower the voltage across the heaters though.

It won't. To drop the heater voltage you need to put something in series with the heater feed wires.

Resistors are used when you want to keep the waveform the same, but simply reduce the voltage.

Diodes will also drop the RMS voltage, but they do so by 'coring out' the middle of the waveform, which can result in more hum. (Many people do this without getting any hum problems, though)

[phsyconoodler]

Ok.I'm not leaving this one alone.How is anyone getting those high heater voltages in the first place?
  I've worked on may vintage amps that were designed for 110v and with as much as 123v in my wall socket I still don't get over 6.8v on the heaters of ANY vintage amp yet.
 There must be something else involved here.When using Hammond 'Classic tube" iron I do get 7v or slightly more,but that's mostly becase of the huge current handling on the 6.3v taps.Not enough tubes to draw it down a bit.However at 7v I have not had one filament related tube failure to date after 6 years.In fact I don't recall seeing ANY filament related failures on any preamp tubes in the last few years at all.
  Where is that 7.8v coming from?

Quote from OP:"And yes, they eat up tubes in about 3-4 months."
I have hard time believing that statement.I'm not trying to be negative or too argumentative,just trying to get a handle on this.
  As far as I'm concerned 7.2v on heaters is not too high.  7.8v may very well be.Take a look at tubes in old automotive tube radios.They ranged all the way from 5v to 9v with no seriously ill effects.I'm not buying the filament life voltage thing much at all.at the very least I don't think the life ratings are as drastic as the old chart claims.

[Merlin]

Ok.I'm not leaving this one alone.How is anyone getting those high heater voltages in the first place?
 

It's not totally unbelievable. The heater winding may not be loaded to capacity, and might have been designed for 220V mains, now getting 245V or something.

[phsyconoodler]

I did say that in my previous post,but it still doesn't convince me that it's a huge problem that needs addressing at this point.
  The 220v thing is another story.European stuff...all beats are off.
I still haven't seen any filament failures on any tubes in the past 6 years or so,even with the elevated heater voltages(7.2v in some of my amps)
  The OP stated it was between 7.2v and 7.8v. Then he said it kills the tubes in 4 months.Are you sure it's filament voltages that is killing those tubes?
  I say no,but lets get some more input on this.Again,I'm not trying to come off like a know-it-all here,just digging a bit.

[eleventeen]

From my 1951 Sylvania Tube manual (conveniently pictured at left) under 6L6, 6L6G, 6L6GA...

"Precautions should be taken to prevent the heater voltage from exceeding 7.0 volts during line voltage fluctuations. A minimum potential difference between the heater and cathode should be maintained."

and...

"The maximum grid-circuit resistance when self-bias is employed may be .25 megohm if the heater voltage doesn't exceed 7.0 volts."

I too have never seen (as far as I can recall in all my years an open 6L6 filament, though I have definitely seen absolute dud tubes with ZERO emissions. I actually have one now, a Sovtek that is just simply DEAD. Pretty weird, really, you'd think if it lights up it should emit something, eh? Even on an emissions tester, nada.

Maybe this "excess  heater voltage" thing is something left over from when the tubes were a lot less rugged or tolerant of modest over-voltage conditions. We're not talking about 1200 volts on the plate, after all.

[Merlin]

I think you're looking at the wrong issue. A bit of excess heater voltage isn't likely to burn out the heater. Instead, the increased cathode temperature causes more rapid deterioration of the cathode (by barium evaporation). That's why the old tube manuals say that life will be reduced with excessive heater voltage.

[frank57]

Here's the complete board.
There were comments by at least two techs that the transformer might be wrong for the amp.
I don't know what I can check, but the plate voltage is fairly high about 374-377 range.
I have it biased at about -15.5 or so for now.
Bit under 60 percent.
Other notable voltages were high. Instead of 3.15, 3.8.
I believe I posted that elsewhere but can double check.
There was also too much voltage on the phase splitter.
I think it was 385 on one part 235 the other.
Now they're roughly the same.
That was caused by either a crap connection on the 1m resistor which I replaced.
There are some leads here and there that are just poking through.
There is a hum and buzz problem on these amps which seems to be centered on v2.
Turning the power transformer doesn't do much.
Going straight to the first filter cap from the standby switch didn't do that much either.

The tubes don't last that long I can assure you.


I might try to take the first filter cap and diodes off the main board too.
Would the in4004 be causing problems?
I'm going to replace then with in4007.

[Midtex]

Wanting to learn more about this subject of using diodes to drop the heater voltage. I totally understand that each diode needs to be rated for the full current draw of the heaters - my question is whether the diodes will generate much heat in this application? Since the voltage drop seems to be due to switching losses and not internal resistance, I'm guessing they won't get very hot? I'm also guessing this means that a general purpose diode will drop more than an ultra-fast type? Thanks!

[eleventeen]

The action of the diodes in this usage just takes into account the characteristic ~~.65 volt drop that occurs across a forward biased silicon junction. If we used germanium diodes, we'd expect about a .25 volt drop. Yes, there is the switching action..occurring when the positive half-wave of AC forward biases the diode and it goes from non-conducting to conducting. But we're not really using the diode as a switch, per se.

Properly rated diodes use their leads as their heat sinks. Until diodes have to pass something over 5 amps, their leads are adequate for dissipating heat. They do not generate "zero" heat but they don't generate much heat at all.

I don't think there's an appreciable difference between the characteristic voltage drop between vanilla and high-speed diodes.