Transformer to Reduce Line Voltage for Amp

The filament transformer is not used to provide 5VAC to your heaters; it's used to drop ("buck") voltage from the incoming line voltage. If you swap wires, it will boost line voltage. The actual voltage drop (or boost) is somewhat greater than the rated voltage of the transformer. For example, I'm using a 5VAC CT transformer, with a switch to select 1/2 of the winding, or the full winding. My voltage drops from about 121-122VAC incoming to 117-118VAC on the center tap setting, and 114-115VAC on the full winding setting.

If you check the schematic and read R.G. Keen's explanation in my previous link, you'll get a better understanding.

I'm not a fan of running heaters below rated voltage, based on the cautions I've read much about.

So guys remind again, is this just basically to extend tube life, i.e., running at ~118-120VAC? I mean transformers may run a little hotter, but is that a problem?

I know guys like LeGrace had some issues, but wasn't that mostly because of the higher inrush when powering up with too high VAC, with no CL90 or maybe the yellow sheet diode mod in place, then the rectifier gets weak and either fuses blow then power tubes next, likely because bias then becomes unstable somehow?

I think it is more of an issue with stressing or exceeding the rated values for capacitors etc especially when starting up or experiencing power spikes. There is already like 500 volts going on even at "normal" voltages when you add five or eight volts of input it raises everything proportionately and sometimes something goes kabam. I believe the new replacement multicaps for the ST70 and MkIII are rated higher for that reason and to provide some added headroom.

OK - it is all about tube life. Sure, the caps in an OEM Device may be running at the bleeding edge, but not in a VTA device, and even so, the difference is not critical. However:

In "Power Vacuum Tubes Handbook" by Jerry C. Whittaker, McGraw-Hill, 1999 the assertion is made that a 3% rise in filament voltage will result in a 20 degree C rise in filament temperature, a 20% increase in emission, and a 50% decrease in life expectancy (see page 119) The opposite should hold true for a 3% reduction in filament voltage. Page 499 of the same book details such a program of filament management.

That is a summation of a 47 page treatise on the subject.

I'm a simple man. I don't know the cause or nuances but after two catastrophic red plate meltdowns and ensuing fireworks, I got things fixed, a third new quad set of KT88s and a Variac...problem solved. Heat, voltage, caps, who knows??

wildiowa wrote:I'm a simple man. I don't know the cause or nuances but after two catastrophic red plate meltdowns and ensuing fireworks, I got things fixed, a third new quad set of KT88s and a Variac...problem solved. Heat, voltage, caps, who knows??

Yep!!

Working at the proper voltage simply works. Analytics notwithstanding.

Just like - If I like the way it sounds, that is enough. Analytics notwithstanding.

I was most concerned with the very high PT temperatures my M-125s with KT120s were running (too hot to touch for more than a couple of seconds). From what I've read here from Bob, these transformers were not designed for 121-122VAC, but around 118VAC. Reducing my supply voltage to that target has brought the PT temps down to where I can hold onto them. Unfortunately, I didn't measure before/after temps with either my IR or thermal devices, so can't give better data.

Peter W. wrote:OK - it is all about tube life. Sure, the caps in an OEM Device may be running at the bleeding edge, but not in a VTA  device,  and even so, the difference is not critical. However:

In "Power Vacuum Tubes Handbook" by Jerry C. Whittaker, McGraw-Hill, 1999  the assertion is made that a 3% rise in filament voltage will result in a 20 degree C rise in filament temperature, a 20% increase in emission, and a 50% decrease in life expectancy (see page 119)  The opposite should hold true for a 3% reduction in filament voltage.  Page 499 of the same book details such a program of filament management.

There has been much written about the damaging effects of heater overvoltage. I'm not sure one can assume the opposite holds true. Tolerance for under- or overvoltage of heaters as shown in many of the tube manuals ranges from 5% to 7%, up to 10%, but that figure doesn't seem to be accepted in real world usage, based on reduced tube life.

See:
http://www.emissionlabs.com/Articles/TECH-BULLETIN/TB-05-Heater-Voltage/TB-05-Heater-Voltage.htm
http://www.valvewizard.co.uk/heater.html
https://books.google.com/books?id=VMjkDQAAQBAJ&pg=PA106&lpg=PA106&dq=effects+of+low+heater+voltage+on+vacuum+tubes&source=bl&ots=LgIt7V1DtJ&sig=GhC6MLipb-BFJVyIerwoFlD62Ws&hl=en&sa=X&ved=0ahUKEwj49YSm0d_ZAhVLzFQKHdtTCGQ4ChDoAQhgMAk#v=onepage&q=effects%20of%20low%20heater%20voltage%20on%20vacuum%20tubes&f=false
https://www.w8ji.com/filament_voltage_life.htm

Maybe my imagination, but my ST120 tends to sound a bit pinched at 124vac. Seems to be much more relaxed at the 117vac provided by the bucker.

Granted ... I tend to have a pretty active imagination ...

I'm about to pull the trigger on a ST-120, and am concerned that my 122-123 vac will be too much.

Does anyone know:
1. what the filament voltage will be for that input vac?
2. or the B+ DC voltage on the tubes at that input vac with a solid state rectifier like the Weber WS1 or something even cheaper?

I have a old Tripp Lite LC-1800 line voltage regulator.  If I manually monkey with the transformer windings (bypassing all that auto-adjust circuitry), with 123 vac input the nearest thing I can get is ~114 vac input -- about 9 volts less.

Will that be TOO LITTLE of a line voltage?

And I've searched the forums and read a whole lot about the CL-90 NTC thermistor for the transformer input.  Many people say it's preferable to use in the ST-120 because of it's higher cold resistance.  But's its only rated for 2 amps.  I plan to probably push the amps hard and log occasionally when driving Maggie MMG's.  I'm concerned about melting down that CL-90, cold start resistance be damned.

Opinions here?

And BTW, I plan on also using the TDR board regardles of any of the above.

Thanks in advance for detailed answers to all the questions above.

Tada ... the CL90, on an ST120 custom build.



Granted, the CL80 has a higher amp rating, but I think the size of the CL90 has an advantage for heat dissipation. These things generate a LOT of heat, so make sure you maximize clearance to any other sensitive bits. Mine bridges a barrier strip on the hot side of the wall input. Been working fine for several years now, and I tend to push it hard. If one were to fail, they fail open, basically cutting power to the unit without doing any damage downstream.

All your test voltages are listed in the instructions, with a fairly wide range allowed. Mine are on the low side of the range at 117vac, but still more than acceptable.

PS - not to forget - the new kits are built for a 120vac standard, so you should be able to get away with 122-3vac with no problem. The CL90 will also drop what the amp sees about 1vac as it never opens completely. That said, I still think the tubes are happier at the lower old school 117vac.

sKiZo's solution is elegant. Bravo, sKiZo!
However -- your projected use of SS rectifier, esp. the Weber WS-1, could be a problem.
I'm no expert, but the B+ voltage might well be not only the critical factor for tube longevity, but crucial for satisfying SQ.
With my ST-120, 5AR4 rectification sounded great at first. My line AC is between 116 and 118 - low. My B+ was on the low side - 465 - 475.
When I switched to the Weber WS-1 my B+ rose to an ideal 490 VDC. Sound is robust and clean. No output tube failures in over a year.
Your problem is potentially the opposite. I suggest you follow Bob Latino's suggestion - when the incoming AC is as high as yours, opt instead for a gentle rectifier tube like the 5U4.
And your choice of the TDR board is wise no matter what.

Thanks for the responses.  I'm looking forward to building this thing -- I've done a lot of work and repaired every sort of thing through the years, but this will be my first true 'kit' since I built a Hafler DH-200 in 1981.  And I still have that Hafler (although heavily modded).

I suppose what I'll do is build it, then check B+ at different input voltages.  I have a bench variac (huge) to experiment with.

No tube rectifier here, unless I find a stash of cheap or free big bottle rectifiers (GZ37, etc.).

If my 122-123 vac results in a too high B+, I'll experiment with the Tripp Lite 114-115 vac output, or make a bucker.  Possibly try one of them cheap variacs.

What about filament voltage?  Is it bad to be too high there?  Shorter life?

It continues to amaze me that bucking transformers are not available commercially. That said, I made one based on all the youtube videos for about $25. But I would think there should be quite a market for ready-made units for the tube crowd.

eickmewg wrote:It continues to amaze me that bucking transformers are not available commercially.  That said, I made one based on all the youtube videos for about $25.  But I would think there should be quite a market for ready-made units for the tube crowd.

But they are, and not very expensive, either:

https://dongan.com/portfolio/buck-boost-transformers/

Is one of several suppliers.

Didn't see anything in the catalog with a 6v secondary? 12v would put you into brownout territory unless your wall power is dangerously burn the house down high ...