• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

transform simple low voltage OTL tube amp into lower impedance by paralleling tubes

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi, in this page Blog For My Hobbies the PL504 headphone amplifier is presented that uses a pl504 and no output transformer.
Can I somehow parallel more than one PL504 to reduce the output impedance so as to be closer to 8 ohms? For example if I parallel two tubes will it be 300R?
How shall I parallel the tube, does each tube need to have it's own resistors or just parallel them pin by pin?
Is the output power doubled by paralleling tubes?
 
Well, that is a one person's point of view, and many would disagree with him. He obviously has a bit of an "anti OTL" bee in his bonnet.

I have had great success with a variety of different OTL designs. I have never run into problems caused by the negative feedback. I have built OTLs with a very low output impedance, very low distortion, and quite decent output power. (For example, 25W using just a pair of 6C33C tubes.) I've been running some of my OTL amps for years and had no problems with tube failures.
 
Oh please do share that schematic. I'd love to know how you aren't vastly exceeding peak cathode current ratings :)

In reality it's a cost and efficiency thing. An OTL is big, hot, heavy, and expensive, but then again so is an OPT tube amp compared to a modern class D design... To each their own, I guess.
 
There are many schematics available if you search on the internet. Hans Beijner and Tim Mellow, for example, have presented 6C33C OTL designs. I didn't say that the nominal peak cathode current ratings aren't exceeded at times. But the fact remains that in practice one seems to get very decent longevity out of the tubes, as well as excellent performance. Tubes can be pretty tolerant of exceeding their nominal ratings.

I think I recall calculating that if one hammered the amplifier and had it driving at 25W continuously into the 8 ohm load it would imply a mean power dissipation of about 120W per tube, while the rated maximum power dissipation for the 6C33C is 60W. But in practice when listening to music as opposed to driving sine waves at maximum power into a dummy load, the averaged power dissipation will be a lot less. And as I say, and many other people have found too, the lifetime of the tubes is amazingly high.
 
Disabled Account
Joined 2013
Based on the schematic in your link I have simulated it based on your request on parallel tube or amp. You can read more about this on Bridged and paralleled amplifiers - Wikipedia

I added cathode resistor to raise the bias swing and to reduce distortion, grid stopper resistors and adjusted the coupling capacitor values for lowest load. The gain is < 1 and distortion remains quite constant under low load, the downside is not enough power to drive normal 8 ohms speaker, unless you have sensitive 8 ohms headphones.
 

Attachments

  • el95-2.png
    el95-2.png
    77.3 KB · Views: 589
  • el95.asc
    10 KB · Views: 35
An OTL with 2 6C33C in inverted Futterman connection can easily give 25W in 8 ohm with high reliability and long tube life. 25W in 8 ohm means that the tubes need to give a peak current of 2.5A and that the average current through the tubes of 2.5/PI or about 0.8A. This exceed the allowed max average current for the 6C33C which is 0.6A and it will also mean that the max allowed anode dissipation of 60W is exceeded.

However as for any amplifier used for playing music the maximum power is never used contionously, music depending on type have a crest factor, (relation between peak and average power level) of 10 - 20dB. So in reality any amplifier even at playing just before clipping level give out a power that is 10 to 100 times lower than maximum.

For the OTL example above with 25W in 8 ohm and a crest factor of 10dB the amplifier give 2.5W average, (if played up to clipping), the peak current will then be 0.79A and the average current through the tubes 0.79/PI = 0.25A which is well inside the limit. The Anode dissipation with 160V anode voltage will then be 40W which is well below the allowed max.

So thats the story on how an OTL can be built with few tubes, (of the right type) and still have good reliability and long tube life.

For our commercial amplifiers we say as an estimate that tubes normally last for 4 years or more, that is based on an average use of 4 hours per day.

Regarding allowed average and peak current for the 6C33C only average current is specified in the data sheet, it is unusual for tubes to be specified for both average and peak current, one example is the PL509 which is specified for 0.5A average and 1.5A peak, (however 6C33C has much larger cathode area)

What is limiting the safe peak cathode current is the total area and emissivity of the cathode, The 6C33C has very large cathode area and high heater power. In litterature it is mentioned that safe peak current for oxide coated cathodes is 0.5A/sqcm and as the cathode area of the 6C33C is more than 12 sqcm 2.5A it is then well into the safe area.

As an aneqdote I can describe the experiment we did with one of our own 25W OTL amplifiers where we let it run with 25W output power continously for 24 hours. If you entered the room with lights off you could see the anodes glowing very dark red and everything including the chassis was very warm but the amplifier performed without problem and with no reduction in power or any other ill effects.
 
Define "Normal". I expect my amps to be able to pass a sine wave at full power for more than 2 years before the tubes need changing. If you passed a 25W sine wave out of that 6C33C amp, how long do you think it would last? It's well known that, like a light bulb, a tube can work at over it's max design at the expense of reduced service life, right? I can use 100W bulbs for 400W and the light is much brighter and purer, but it'll only work for a few minutes. Even the 6C33C data has ratings for 750h and 3000h :)
 
Last edited:
I tend to design my amplifiers to perform reliably for years at full output 24/7
Why? Do you never sleep?:)
I expect my amps to be able to pass a sine wave at full power for more than 2 years before the tubes need changing.
Wow! This is much better than ordinary tube amplifiers, what tubes are you using?, this would be 17000 hours at full blast with a continous sine wave. Normally power tubes in ordinary power amplifiers often need changing after less than 5000 hours with normal operation, (and that is with average power 10 to 100 times less than max).

Even the 6C33C data has ratings for 750h and 3000h
Yes but have you looked into what the conditions are for that spec? It is while fulfilling all specs of anode dissipation, cathode current, acceleration, shock etc. The 6C33C designed as it is for millitary use in tanks, fighter jets and missiles is specified much more strict than most power tubes. I think it is safe to say that 6C33C have better quality than most ordinary power tubes.

My estimate of 4 hours daily average use, (meaning average, i.e some days maybe 8 or 24 hours, some days 0) for 4 years is what I say to my customers as many have heard that OTLs burn tubes very quickly, In reality the tubes usually last much longer, my first prototype finished in 2000 still use the original power tubes even though I use it almost every day.
 
Last edited:
Why? Do you never sleep?:)

Well, with that full power sinewave playing 24 hours a day, it is probably pretty hard to sleep!

But seriously, I suppose one could, if one wished, double up the output tubes and then the ratings would not be exceeded even at full 25W power. But there is presumably little point in doing this, since the tubes seem perfectly capable of lasting a long time in the present configuration.

In fact, if one did double up the tubes one could presumably in practice push the power output to much higher levels, given a suitable driver. Roughly speaking, if the limitation is the current the tubes can pass, I suppose the maximum output power should grow as the square of the number of tubes, and so something like 100W for four output tubes instead of two.

Did you ever try experimenting with paralleled 6C33C output tubes, I wonder? Presumably the driver stage would need modification to handle the required larger voltage swings?
 
Did you ever try experimenting with paralleled 6C33C output tubes

Yes, in one of my commercial amplifiers I use 4 tubes and this give 80W in 8ohm, I use equalising resistors in the anodes to even out differences between tubes and avoid thermal runaway if one tube goes bad. Driver is similar but I use cathode followers as buffers between the split load inverter and power tubes, in that way I can optimise the splitter for high output voltage and still drive the power tubes to 0V on the grids. See here for a description on my commercial web page Eabout | Sverige | Tillverkare rorforstarkare
 
Last edited:
6P3S and 6P1P both work for years at high power levels in my set up.

I also have a commercial tube receiver that was never turned off and the ECL82 tubes lasted 5 years before I changed them to 6F3P. They were weak but still functional. They were the original tubes from the 70s, so they lasted 25 years before I got the amp in the first place. 30 year tube life :)

What is the class A limit of your 25W design? How much power would it make into 64 ohms?
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.