Lowering Filament voltage

[Cjteleforum]

The AC voltage on the filaments of the 6V6 in my Champ is reading 7.2 V. Please let me know if this the proper way to lower it to 6.3 V.:

7.2V - 6.3V is .9V. 9V/.45 Amps(the current draw of a 6V6)=2 ohm resistor. The wattage is the voltage drop, .9V * .45 amps=.4 watts * 5=2 watts.

So I need to wire two 2 ohm 2 watt resistors from the filament wires to the heater pins, 2 and 7, of the 6V6?
Is that right?
Also, is is safe for the PT? All the other voltages in the amp are reading right.
Any advice much appreciated.

 

[Jon Snell]

Instead, why not place two diodes as shown ... no heat dissipation to worry about and you will get 0.7v drop giving the valve 6.5volts which is well within specification.

 

[King Fan]

The dual-diode method is cool; proper resistors will also work well. Both are covered nicely by Merlin in his short, clear chapter on heaters. In the section on dropping heater voltage, he shows you how to calculate resistor values and has this useful note about the diodes:

"The diodes need to have an average current rating that exceeds the heater current (1N4007s are only rated for 1A!). Sometimes it is easier to buy a beefy bridge rectifier than indvidual diodes, in which case you can wire the bridge to create a pair of anti-parallel diodes. Also, large bridge rectifier packages often have a hole in the middle so they can be conveniently bolted to the chassis."

http://www.valvewizard.co.uk/heater.html

FWIW I liked this big beefy bridge rectifier and it didn't cost much.

 

[Lowerleftcoast]

The wattage is the voltage drop, .9V * .45 amps=.4 watts * 5=2 watts.

The 5F1 has a 12A_7 at 0.35A and the 6V6 at 0.45A, so the current draw will be near 0.8A.
Imo two 0.5R, one in each leg of the AC heater line, to help keep balance, would bring the voltage down to near 6.3V. These resistors get hot so I would use 5W to help disperse the heat.

Although not ideal, I could live with 7.2V on the heaters. OMMV

 

[Pete Farrington]

If the 6.3V heater supply is high it probably means that the HT and 5V winding voltages are also excessive. Can you check what those voltages are?
If so, the mains supply to the primary is higher than intended / the transformer is operating with significantly underloading.
If so, a mains bucker seems the obvious way to resolve the root cause of the issues
http://www.geofex.com/Article_Folders/vintvolt/vintvolt.htm

5E3 Mods

5E3 Deluxe Amplifier Modifications

robrobinette.com

 

[Cjteleforum]

The 5F1 has a 12A_7 at 0.35A and the 6V6 at 0.45A, so the current draw will be near 0.8A.
Imo two 0.5R, one in each leg of the AC heater line, to help keep balance, would bring the voltage down to near 6.3V. These resistors get hot so I would use 5W to help disperse the heat.

Although not ideal, I could live with 7.2V on the heaters. OMMV

Thanks, exactly what I wanted to know. But wow .5 ohm 5 watt resistors are not cheap.

 

[Cjteleforum]

If the 6.3V heater supply is high it probably means that the HT and 5V winding voltages are also excessive. Can you check what those voltages are?
If so, the mains supply to the primary is higher than intended / the transformer is operating with significantly underloading.
If so, a mains bucker seems the obvious way to resolve the root cause of the issues
http://www.geofex.com/Article_Folders/vintvolt/vintvolt.htm

5E3 Mods

5E3 Deluxe Amplifier Modifications

robrobinette.com

I don't know how to check the "HT" voltage.

But the voltage across pins 2 and 8 of the rectifier is 5.6 VAC. So that's about the same percent over as the heaters.

B+ 1 is 348 VDC
B+ 2 is 308 VDC
B+ 3 is 265 VDC
So those are all within 5%.
My outlet reads 124 VAC. The PT is a Merc. Mag. Tone Clone and the spec sheet shows the primary power as 120 vAC.

 

[dan40]

I don't know how to check the "HT" voltage.

B+ 1 is 348 VDC
B+ 2 is 308 VDC
B+ 3 is 265 VDC

You just did...all of those measurements are the HT (high tension or B+) voltages.

 

[Cjteleforum]

The 5F1 has a 12A_7 at 0.35A and the 6V6 at 0.45A, so the current draw will be near 0.8A.
Imo two 0.5R, one in each leg of the AC heater line, to help keep balance, would bring the voltage down to near 6.3V. These resistors get hot so I would use 5W to help disperse the heat.

Although not ideal, I could live with 7.2V on the heaters. OMMV

I've noticed that ever since I learned about biasing and cooled it down a bit, I haven't had to change a power tube. Before it seemed like I had to change them every year(I built this amp a long time ago). Do you think bias has more affect on tube life than heater voltage?

 

[2L man]

Instead, why not place two diodes as shown ... no heat dissipation to worry about and you will get 0.7v drop giving the valve 6.5volts which is well within specification.

When diodes are used to waste same voltage they will dissipate about the same heat than resistors will but there might come ”switching noise” to audio.

Resistor or two is good method to waste excess filament voltage. They also can increase tube life when filament startup current come lower.

There come less than 1W heat so two 0.5 ohm 2W resistors are fine!

If you connect them between wires put heat shrink tubes over the solders and tie wires about where the solders are so that resistors do not vibrate and that there come air room around.

 

[Pete Farrington]

Imo two 0.5R, one in each leg of the AC heater line, to help keep balance

Fender Champs generally don’t have balanced heaters. As this is a build I suppose it may have

I don't know how to check the "HT" voltage

I was thinking of the mains transformer secondary winding voltages, so meant the V AC red - red wires from the transformer, probably about 650V.

the voltage across pins 2 and 8 of the rectifier is 5.6 VAC. So that's about the same percent over as the heaters.

It may be a good idea to address that too then?
The bucker would reduce all secondary voltages.

Do you think bias has more affect on tube life than heater voltage?

My experience is that open circuit heater failure is incredibly rare. Other valve failure modes might involve the heaters, but excessive heater supply voltage seems unlikely to be a contributory factor to them.

 

[Cjteleforum]

Fender Champs generally don’t have balanced heaters. As this is a build I suppose it may have

I see in my original Fender layout the PT has two heater wires, one to ground and one to pin 7 of the 6V6 and on to pin 4 of the 12AX7. My PT has the same wires, but whatever layout I copied when I built it has both of them going to the pilot light with 100k Rs and on to both heaters of the tubes.

I was thinking of the mains transformer secondary winding voltages, so meant the V AC red - red wires from the transformer, probably about 650V.

I tried measuring those across the pins, and my meter started beeping. It only goes to 600V.

It may be a good idea to address that too then?
The bucker would reduce all secondary voltages.

I might try that. Or variacs seem interesting.

My experience is that open circuit heater failure is incredibly rare. Other valve failure modes might involve the heaters, but excessive heater supply voltage seems unlikely to be a contributory factor to them.

I might just leave well enough alone then

 

[King Fan]

I've noticed that ever since I learned about biasing and cooled it down a bit, I haven't had to change a power tube. Before it seemed like I had to change them every year(I built this amp a long time ago). Do you think bias has more affect on tube life than heater voltage?

So many interesting variables, even on the lovely little Champ.

If I may: How did you cool the bias? What bias measurements are you getting now? Do you happen to know the cathode bias resistance on that 6V6 and the plate and cathode voltages on that socket?

It may just be something about the PT, but it seems odd to have such high 5 and 6.3V voltages, and ordinarily I'd go with the bucking transformer suggested by @Pete Farrington .

I don't know if high heater voltage is bad for tubes, but it's widely said to be, and to also increase heater hum.

In theory a *literal* tweed clone PT would expect 115V from the wall; you'd be about 8% high at 124V. But your heater voltages are more like 14% high. OK, If MM built that PT to compensate for modern wall voltage being 120, your 124V is only 3% high, and we might expect the heaters to be at ~6.5VAC (which would be fine) -- but then it's even weirder they're so high. There are other variables, like exactly how a clone is 'reverse engineered', and actual load, plus of course, I don't know if the original PTs hit all their voltage targets.

You're right, it sounds like you have balanced heater wiring -- as noted, the single strand of the originals is often avoided in modern builds.

It can be useful to measure your HT voltage at the rectifier. The trick is to clamp the negative probe of your meter to the ground anchor for the PT's HT center tap (often red/yel, but not always) and measure VAC from just one pin (say pin 4) to ground. This will then be way under 600V. Does that make sense?

 

[Lowerleftcoast]

Do you think bias has more affect on tube life than heater voltage?

Yes. At 7.2v.

 

[King Fan]

@Lowerleftcoast , @Pete Farrington , I always respect your opinions and experience. Do you know of any evidence / expert sources that support letting heater voltages run over 7V? For evidence, I've seen old monographs (from the all-tube era, eg, Electronic Engineering 1954 v.XXVI p.10) showing that keeping heaters (eg, in early computers) at or below 6.3V could prolong tube life. (Any computer nerds out there? This was in the ENIAC.)

Let's admit tube operation in those experiments may not mirror that in guitar amps. So for my expert, I'll cite Merlin (linked above), who suggests voltages ≤ 6.9 or preferably 6.5V.

But even beyond the theoretical question of tube life, I often use heater voltage as a barometer to decide on a proper supply voltage in vintage amps, and that's confusing me a bit here, where B+ is nominally 'not bad' but both heater circuits are very high.

 

[Lowerleftcoast]

Do you know of any evidence / expert sources that support letting heater voltages run over 7V?

The *non-expert* evidence is just the many amplifiers that have high heater voltages. Their tubes seem to have sufficient life spans. Ime, at least 75% of the amps, I have had on the bench, have had heaters >6.5v. OMMV. And no, the high heater voltage is not why they were on the bench.

 

[Cjteleforum]

If I may: How did you cool the bias? What bias measurements are you getting now? Do you happen to know the cathode bias resistance on that 6V6 and the plate and cathode voltages on that socket?

I changed the cathode resistor to a higher value. It's now at 476 ohms. Before it was too low. I also swap out tubes. With these tubes the plate voltage is 314 VDC. The voltage drop from cathode to ground is 17.8 VDC. So, plate dissipation is 11.7 watts, which is cool for this tube, a NOS RCA 6V6GT. With a NOS Tung Sol 6V6GT it's 12.9 watts, which I like the sound of, but that tube is rated for a maximum of 12 watts. I've had that one in there for a couple years at least. That rating between 12 and 14 watts seems kind of waffley anyway.

It can be useful to measure your HT voltage at the rectifier. The trick is to clamp the negative probe of your meter to the ground anchor for the PT's HT center tap (often red/yel, but not always) and measure VAC from just one pin (say pin 4) to ground. This will then be way under 600V. Does that make sense?

Yes. I did that and got 328 VAC.

 

[Lowerleftcoast]

I changed the cathode resistor to a higher value. It's now at 476 ohms. Before it was too low. I also swap out tubes. With these tubes the plate voltage is 314 VDC. The voltage drop from cathode to ground is 17.8 VDC. So, plate dissipation is 11.7 watts, which is cool for this tube, a NOS RCA 6V6GT. With a NOS Tung Sol 6V6GT it's 12.9 watts, which I like the sound of, but that tube is rated for a maximum of 12 watts. I've had that one in there for a couple years at least. That rating between 12 and 14 watts seems kind of waffley anyway.

Did you use the correct numbers when you made the calculations?
On a cathode biased amp, instead of plate voltage we use *plate voltage minus cathode voltage*. Your *plate to cathode voltage* is ~296. (314 - 17.8 = ~296)
When using the cathode to find current, the plate and screen current are included in the result. Have you subtracted the screen current when you made your calculations?

Rather than 11.7W, your tube may be near 10.4W

 

[King Fan]

With a NOS Tung Sol 6V6GT it's 12.9 watts, which I like the sound of, but that tube is rated for a maximum of 12 watts.

Lots of topics coming up here -- bear with me on this little one. The old data sheets for most 6V6s all said 12W; later data sheets sometimes say 14W, but the tubes didn't change at all. The 12W figure was 'design center' and the 14W 'design max.' (Did I say this here already? Sorry, I've been saying it constantly the last few weeks.) I've read intelligent arguments for using one or the other, but I actually use both -- 'safe' bias exists in a range, ideally decided by ear, and in a single-ended amp like the Champ that range extends upward to 115 or even 120%. I bias my NOS tubes where they sound good, and in an SE amp, if I hit 14W, , that's about 115-117% of 12W. If I hit 14-16W, like in my Vibro Champ, that's ~115% of 14W. So far, I have yet to burn through any 6V6s. (These liberal guidelines also work great for JJs, a robust variant, but may be too much for some reissue tubes, specifically the new T/S.)

 

[Cjteleforum]

When using the cathode to find current, the plate and screen current are included in the result. Have you subtracted the screen current when you made your calculations?

Rather than 11.7W, your tube may be near 10.4W

I have not. How do I find the screen current?
I've either used Uncle Doug's tutorial video(which he admits results in a higher value than actual)or Rob Robinette's Bias Calculator. But I would really like to know more detail.