Hi Chris. Thanks for the prompt reply.
My understanding is that each LM317 should be able to handle a minimum of 1.5 amps. (According to National specsheet, the TO220 is guaranteed to handle at least 1.5 amps at voltages below 15v, with average of 2.2 amps and max of 3.4 amps. Since there are only 2 valves, each rated at 1.5 amps heater draw, I should have had current to spare. I used 4 devices instead of 1 or 2 because of heat issues.
My understanding is that each LM317 should be able to handle a minimum of 1.5 amps. (According to National specsheet, the TO220 is guaranteed to handle at least 1.5 amps at voltages below 15v, with average of 2.2 amps and max of 3.4 amps. Since there are only 2 valves, each rated at 1.5 amps heater draw, I should have had current to spare. I used 4 devices instead of 1 or 2 because of heat issues.
Hi hihopes,
The thing is that you are pushing each device above it's ratings. They would need to be on a large heat sink to run cool enough. You will learn that you can't run things at max or over. Life just isn't like that.
That is why I recommended a pair of switchers rated at 5 A each. Your heat loss would be greatly reduced. You may want to consider running each bank of two tubes off one secondary to the center tap. That should give you approx. 6.3 VAC. You can fine tune that with a resistor in each leg (not in the center tap).
-Chris
The thing is that you are pushing each device above it's ratings. They would need to be on a large heat sink to run cool enough. You will learn that you can't run things at max or over. Life just isn't like that.
That is why I recommended a pair of switchers rated at 5 A each. Your heat loss would be greatly reduced. You may want to consider running each bank of two tubes off one secondary to the center tap. That should give you approx. 6.3 VAC. You can fine tune that with a resistor in each leg (not in the center tap).
-Chris
Hihopes, dont worry abotu your heater voltage- it is fine. Some people say that if you run the heater less than 10% below rated voltage, they "suffer". This demonstrates gross misunderstanding of how a valve actually works.
Running OVER voltage obviously shortens the life of the heater, like it would in any lightbulb.
But running BELOW voltage EXTENDS life. However, it REDUCES saturation current. Now, in some circuits, like transmitters, that's a bad thing. But in audio amps the valves are NEVER pushed to saturation anyway, so the reduction in saturation current is irrelevent.
Once you get below 80% rated voltage, gm does start to fall, but only a small amount- and liek I say, it usualyl increases linearity, so what you lose in one thing, you gain in another. But a low heater voltage DOES NOT shorten the life of a valve. I routinely run my heaters between 5V and 6.3V.
Running OVER voltage obviously shortens the life of the heater, like it would in any lightbulb.
But running BELOW voltage EXTENDS life. However, it REDUCES saturation current. Now, in some circuits, like transmitters, that's a bad thing. But in audio amps the valves are NEVER pushed to saturation anyway, so the reduction in saturation current is irrelevent.
Once you get below 80% rated voltage, gm does start to fall, but only a small amount- and liek I say, it usualyl increases linearity, so what you lose in one thing, you gain in another. But a low heater voltage DOES NOT shorten the life of a valve. I routinely run my heaters between 5V and 6.3V.
Hi anatech,
Thanks for the headsup. Maybe I haven't explained myaself clearly enough. There are no "banks of tubes" - only 2 x EL34's (total expevcted current draw circa 3A). Second, there is no centre tap. This is a shop-bought 12v supply with only a +12v output and nothing else. I got it years ago to power a car-audio type active crossover, because it was cheaper than building one.
I still don't understand why the tubes failed.
Thanks for the headsup. Maybe I haven't explained myaself clearly enough. There are no "banks of tubes" - only 2 x EL34's (total expevcted current draw circa 3A). Second, there is no centre tap. This is a shop-bought 12v supply with only a +12v output and nothing else. I got it years ago to power a car-audio type active crossover, because it was cheaper than building one.
I still don't understand why the tubes failed.
Hi hihopes,
Hi Merlinb,
Rest assured that we are in complete disagreement. If what you say is true, most the high end designers would be doing that. As a technician, I don't believe you at all.
-Chris
Well, about 3.2 A at full voltage. Okay, that explains better what things look like.
Okay, you may be able to adjust it internally. Either that or just pic up a straight transformer and use that. Lastly, you could use a switcher type supply. Either way there is a way to correct the situation. Increasing the regulated voltage would cool the regulators a little as well. You should be fine with 4.
The schematic will help here.
Hi Merlinb,
Rest assured that we are in complete disagreement. If what you say is true, most the high end designers would be doing that. As a technician, I don't believe you at all.
-Chris
Manufacturers rate their valves for a combination of long life with decent emission. If the heater is not heated to the correct temperature the emission will not be as designed.
In the late sixties I conducted a series of experiments with output valve heater voltages. The trial amplifier was a stereo Mullard 5-10 using EL84 output pairs.
Operating the output valve heaters at 6.3V AC centre tapped and connected with twisted pair produced clean response with no apparent induced hum.
Earthing one side of the supply and removing the centre tap to produce a heater voltage of 6.3V one sided increased the hum significantly.
Rectifying the supply to provide 6VDC to the output valve heaters did not improve the audio performance. (Running the input EF86 valve on 6VDC significantly improved hum performance)
If the supply to the output heaters was taken from the 6.3V heater winding and rectified and the voltage fell to 5.3VDC even with a 6,800uf reservoir (the rectifiers were large germanium power diodes on heatsinks). At this voltage distortion of peaks when playing orchestral and opera passages was extremely noticeable. As the amplfiers were U/L AB this is not entirely surprising.
HTH
Bergo
In the late sixties I conducted a series of experiments with output valve heater voltages. The trial amplifier was a stereo Mullard 5-10 using EL84 output pairs.
Operating the output valve heaters at 6.3V AC centre tapped and connected with twisted pair produced clean response with no apparent induced hum.
Earthing one side of the supply and removing the centre tap to produce a heater voltage of 6.3V one sided increased the hum significantly.
Rectifying the supply to provide 6VDC to the output valve heaters did not improve the audio performance. (Running the input EF86 valve on 6VDC significantly improved hum performance)
If the supply to the output heaters was taken from the 6.3V heater winding and rectified and the voltage fell to 5.3VDC even with a 6,800uf reservoir (the rectifiers were large germanium power diodes on heatsinks). At this voltage distortion of peaks when playing orchestral and opera passages was extremely noticeable. As the amplfiers were U/L AB this is not entirely surprising.
HTH
Bergo
Thanks for the input. Perhaps if these were being used as output valves, they might be working harder, but in their current application, driving a pair of mosfets each, there are no distorted peaks. Distortion only occurs with very complex waveforms, and is apparent even at quite low levels, so I think the low heater supply is not the cause.
It is definitely on my to do list to improve it though. I will stick to AC I think. (Unless I can figure out why my attempt at DC killed my valves, or some kind soul can come up with a feasible explanation)
I am thinking of replacing the trafo. I can get one wound for as little (or less) than buying one. I was thinking of going for a 50VA. Any idea how much voltage drop I should expect from a 50VA? Should I specify higher than 6.3v? Will a 50VA be OK in the long term?
It is definitely on my to do list to improve it though. I will stick to AC I think. (Unless I can figure out why my attempt at DC killed my valves, or some kind soul can come up with a feasible explanation)
I am thinking of replacing the trafo. I can get one wound for as little (or less) than buying one. I was thinking of going for a 50VA. Any idea how much voltage drop I should expect from a 50VA? Should I specify higher than 6.3v? Will a 50VA be OK in the long term?
I suspect most high end designers simply aren't aware of the effect of underrun heaters, or if they are, they know eveyone ELSE isn't, and will therefore probably shun their designs if they dared deviate from the sacred 6.3V. It is so rarely touched upon in textbooks, after all.anatech said:
As a technician, it's a little gem of information I always like to spout!
Hi Merlinb,
I can say from experience that when my heater supply in my Fisher 400 preamp was failing (selenium rect.) the voltage drop caused very odd behaviour. I had unstable gain that was very severe in the area of voltage you are talking about.
In the interest of learning something here, where did you get this information? My own experience seems to contradict what you are saying. Also, beware your rp is going to go way up by doing this.
-Chris
Edit:
Well, I do a lot of reading in technical publications, and have over 35 years on the bench. Considering signal tubes, a slight reduction in heater voltage can bring about lower noise. Power tubes used as a signal tube are not something I know a lot about. I will then assume their performance is similar to signal tubes on a different scale.
I can say from experience that when my heater supply in my Fisher 400 preamp was failing (selenium rect.) the voltage drop caused very odd behaviour. I had unstable gain that was very severe in the area of voltage you are talking about.
In the interest of learning something here, where did you get this information? My own experience seems to contradict what you are saying. Also, beware your rp is going to go way up by doing this.
-Chris
Edit:
I doubt they could resist the temptation to crow about forever tube life. Many new tube "designers" have not got the foggiest idea how a tube works. Running heaters way down in voltage would seem to be a perfect way to relieve the prospective buyer that he is in for maintenance hell.
You're right, rp will go up- though not always by as much as you might expect, until you start operating below about 80% Vh. Obviously you would need to decide whether any increase in rp is actually a problem. In a preamp it usually won't be, whereas in power triodes it might be, and power pentodes perhaps not, unless it goes so high as to affect damping.anatech said:
I can't remember where I FIRST heard about this effect, but here are some links where it is mentioned. I Know somewhere on the web is a study someone did of the ECC83 at different heater voltages, clearly showing the increase in linearity, but I can't seem to locate it on google at the moment.
http://www.diyforums.org/MAX/MAXheaterResistor.php
http://www.tubecad.com/july2000/page10.html
Hi Merlinb,
What they failed to mention in the first link was that certain tubes were designed to be run in an automotive environment. The heaters were designed to run over a wider range. Locktals are the same. I'll keep my eyes open, but I am pretty sure that normal HiFi tubes are designed to have their heaters run within 10% for the listed values.
If you run the heaters lower, you essentially have a unknown tube type. None of the specs are valid anymore and you should then test and rerate those tubes at your lower voltages. In particular, the maximum cathode current will be reduced for sure. This may or may not be an issue, but you should be sure.
Also, different brands of tubes (different runs actually) may behave totally differently to reduced heater voltage.
-Chris
What they failed to mention in the first link was that certain tubes were designed to be run in an automotive environment. The heaters were designed to run over a wider range. Locktals are the same. I'll keep my eyes open, but I am pretty sure that normal HiFi tubes are designed to have their heaters run within 10% for the listed values.
If you run the heaters lower, you essentially have a unknown tube type. None of the specs are valid anymore and you should then test and rerate those tubes at your lower voltages. In particular, the maximum cathode current will be reduced for sure. This may or may not be an issue, but you should be sure.
Also, different brands of tubes (different runs actually) may behave totally differently to reduced heater voltage.
-Chris
anatech said:
To acheive their rated spec, granted.
But for small decreases the changes aren't that wild. Bare in mind that even at the rated voltage, valve specs (gm, ra and mu) are only accurate to +/- 40%! Now I'm not suggesting we run the heater at half voltage or anything- then we really would get crazy performance. But I'm happy to go as low as 80% in non-super critical circuits.
I think I mentioned that in the first place, but as you say, in audio we shouldn't be running valves anywhere near saturation in any case.
Modern valves all seem to behave differently anyhow!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.