I saw a thread on here that was close to what I am looking for but not close enough.
Should I expect distortion in the 2% range , out of the first stage of a preamp tube with a u of 17 (12AU7)or so, given a low voltage supply like 30V at most? I know there is a ton of factors in distortion causes but should I expect to see ~ 2% at best for that low of supply?
Second question is how are most of you measuring distortion? And is anyone using LabVIEW for it? I have a Heathkit HD1 distortion meter that needs a restore but I got it working and it indicates my Sine wave generator has 1.8% THD @ 2KHz 200mv , LabVIEW shows it at 1.4% Pretty close considering the Heathkit is about 50 years old.
Alan
Should I expect distortion in the 2% range , out of the first stage of a preamp tube with a u of 17 (12AU7)or so, given a low voltage supply like 30V at most? I know there is a ton of factors in distortion causes but should I expect to see ~ 2% at best for that low of supply?
Second question is how are most of you measuring distortion? And is anyone using LabVIEW for it? I have a Heathkit HD1 distortion meter that needs a restore but I got it working and it indicates my Sine wave generator has 1.8% THD @ 2KHz 200mv , LabVIEW shows it at 1.4% Pretty close considering the Heathkit is about 50 years old.
Alan
It will depend on tube biasing. For such low plate voltages your tube will need low bias to take reasonable currents. So the input level will be keep low enough not to cut off plate current in negative going signals nor draw grid current in positive hemicycles. Have in mind a reserve for transients in the signal input like an end at full orchestra, in example.
Hey, thanks for a quick reply! Let me rephrase the question a bit. If I do every thing exactly as perfect as can be done no expense spared , money is no object , Can I expect a distortion much under 2 % with only 30V supply.
Alan
Really not. There were few tubes specifically designed for 28 or 12VDC. I now remember ECH83 triode+hexode for mixer converters.
I sincerely can't quantify the amount of distortion and amplification, because I never done such a thing, but don't I imagine a good performance.
The question that arises is: why?. 30V for such a tube is too little, but sufficient enough for a JFET or MOSFET's devices.
I sincerely can't quantify the amount of distortion and amplification, because I never done such a thing, but don't I imagine a good performance.
The question that arises is: why?. 30V for such a tube is too little, but sufficient enough for a JFET or MOSFET's devices.
Thanks for such a quick reply! No I don't see a recommended operating voltage of 30V DC. This is my question: "Should I expect distortion in the 2% range , out of the first stage of a preamp tube with a u of 17 (12AU7)or so, given a low voltage supply like 30V at most? I know there is a ton of factors in distortion causes but should I expect to see ~ 2% at best for that low of supply?"
When using a tube outside of its normal design area of operation, the results are difficult to predict.
Even for simulation, the models (and tubes) are likely to vary widely in the low voltage area,
and vary more with amplitude than with a higher DC voltage. Hook it up and see what happens.
But a different tube will likely give different results.
Test equipment should be at least 5x better than the DUT being measured. Preferably more.
Even for simulation, the models (and tubes) are likely to vary widely in the low voltage area,
and vary more with amplitude than with a higher DC voltage. Hook it up and see what happens.
But a different tube will likely give different results.
Test equipment should be at least 5x better than the DUT being measured. Preferably more.
Last edited:
Maybe not a very useful reply but I have a 12AU7 at 12V with a MOSFET driver stage as a headphone amp and it's at least not noticeably distorted until the level gets pretty much to the max.
With a lower gain tube I think you might struggle at 12V but maybe you'll be alright at 30V. These are all just pure out of my *** guesses though as these tubes are not really ment for voltages this low 🙂
With a lower gain tube I think you might struggle at 12V but maybe you'll be alright at 30V. These are all just pure out of my *** guesses though as these tubes are not really ment for voltages this low 🙂
You should not expect low distortion, very likely a distortion level that is no good for a preamp. This is not just true for the ECC82 but also for the other ECC tubes except the ECC86.
You might have much better luck with the 6SN7 (or 12SN7). In fact the 12SX7 is a 12SN7 with twisted heater to get lower noise at low voltage operation with battery supply at 26.5V plate voltage. If you can run heaters with a quiet supply you should be fine in this department with the 12SN7 too.
Another choice that will be very linear at 30V plate voltage is the 6H30. The standard Sovtek will be just fine. With this tube you need to be aware of the fact that substantial grid current might start while the grid voltage is still negative, around -0.5V. Setting the anode current around 3 ma should be achieved with a bias of -1.7-1.8V. So you can input 1.2'-1.3V peak before getting into grid current territory. You might reduce the anode current to 2 mA with -2V bias and increase the plate resistor, if that makes a difference....
You might have much better luck with the 6SN7 (or 12SN7). In fact the 12SX7 is a 12SN7 with twisted heater to get lower noise at low voltage operation with battery supply at 26.5V plate voltage. If you can run heaters with a quiet supply you should be fine in this department with the 12SN7 too.
Another choice that will be very linear at 30V plate voltage is the 6H30. The standard Sovtek will be just fine. With this tube you need to be aware of the fact that substantial grid current might start while the grid voltage is still negative, around -0.5V. Setting the anode current around 3 ma should be achieved with a bias of -1.7-1.8V. So you can input 1.2'-1.3V peak before getting into grid current territory. You might reduce the anode current to 2 mA with -2V bias and increase the plate resistor, if that makes a difference....
Last edited:
WHAT IS THE SIGNAL LEVEL??
Down to around 30V supply on common tubes:
Peak unclipped output about 20% of B+ (so, 6V peak 4Vrms)
THD at peak output about 5% (assuming no NFB such as cathode resistor)
THD declines about as signal level. So we expect 2% to happen at 2/5 of the 5% point near 4Vrms, or 1.6Vrms
If bias is not optimized, I could imagine getting to this point with only 1Vrms. (Is that your test level?)
Down to around 30V supply on common tubes:
Peak unclipped output about 20% of B+ (so, 6V peak 4Vrms)
THD at peak output about 5% (assuming no NFB such as cathode resistor)
THD declines about as signal level. So we expect 2% to happen at 2/5 of the 5% point near 4Vrms, or 1.6Vrms
If bias is not optimized, I could imagine getting to this point with only 1Vrms. (Is that your test level?)
Thanks every one for their input! It is all very interesting and informative. To answer some suggestions and questions . Yes I did hook it up, and the current distortion is around 2%. The input conditions are 200mv Pk-Pk , out put is ~ 2V Pk-Pk. Signal is 2K sine. This is 1/2 of a 12AU7 , I figured this would have less distortion than a higher gain 12AX7, I also have a few of them on hand.
I agree completely on the Test Equipment comment. I am using an HP 3311A Function Generator. We used to use this at a Data Communications Mfg. Co. To test signal level detection, soft carrier turn off and such. Distortion did not matter, I was floored when I read a distortion of 1.4 ~1.8% at 2KHz ,So I found the manual and the spec is < 3%. I have a new one on the way with a spec of .1 % from 0 to 50Khz
A member here sells these boards with an almost unmeasurable level of distortion. I bought one a few years ago for $66 or so. You won’t find anything of this level until you get into Audio Precision territory. https://viccc42.wixsite.com/uld-audio