I have built a very simple inverting stage with a 12ax7 and mosfet buffer. 100k grid resistor, 47k anode resistor and 150 cathode resistor. Heater runs off 12 volts dc same as anode resistor. I am finding some 12ax7 valves wont conduct under these dc conditions. If I put a 470k pull up from grid to 12 volts the valves start to work. When the valves fail to work cathode voltage is zero.
I noticed some low voltage valve circuits work with 12au7, perhaps I should try that ? Left click on picture to resize it.
I noticed some low voltage valve circuits work with 12au7, perhaps I should try that ? Left click on picture to resize it.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
That's very similar to the bravo V2 headphone amp.. a 12au7 will be more likely to work, but see if you can run it off 24v instead
Vacuum is the perfect insulator.
You need significant voltage to pull electrons from the cloud around a red hot cathode towards a plate a few mm away.
Hence you need _significant_voltage_
Toying around unsuitable low voltages is just that, toying, having fun, whatever.
Yes, a tube will pass a little current with 12V supply.
Yes, a paper plane will fly a little too.
You need significant voltage to pull electrons from the cloud around a red hot cathode towards a plate a few mm away.
Hence you need _significant_voltage_
Toying around unsuitable low voltages is just that, toying, having fun, whatever.
Yes, a tube will pass a little current with 12V supply.
Yes, a paper plane will fly a little too.
Its for a head phone amp that can't have high voltages on it. I guess putting the two halves of 12ax7 in parallel didn't help. I have ordered a 12au7 to see if that works ok. Just seems a bit odd some 12ax7's work and others don't.
I've only ever built 2 valve circuits, but...
If running THAT LOW in volts surely anode resistor needs to be much smaller?
Just a (possibly useless) thought.
If running THAT LOW in volts surely anode resistor needs to be much smaller?
Just a (possibly useless) thought.
If you want to use low voltage on the anode then you need a low mu valve. This is why 12AU7/ECC82 works and 12AX7/ECC83 does not. In addition, you may need to use zero Vg-k bias and accept some grid current.
To a first approximation, the voltage actually seen at the cathode is Va/mu + Vg. This needs to be positive to get any significant anode current. 12V supply and mu=100 means that -0.12V is the grid cutoff - and this has to be modified by contact potential etc.
To a first approximation, the voltage actually seen at the cathode is Va/mu + Vg. This needs to be positive to get any significant anode current. 12V supply and mu=100 means that -0.12V is the grid cutoff - and this has to be modified by contact potential etc.
A much better tube for this is ECC86, specially made for 12V applications
https://frank.pocnet.net/sheets/030/e/ECC86.pdf
https://frank.pocnet.net/sheets/030/e/ECC86.pdf
Russian miniature double triodes (6N16) might work with 12 V, but surely this voltage is really a treshold value for tubes. For low level buffer the 6N16-s are working perfectly with 24V, but haven't tried them with lower voltages.
I have a bunch of rod pentodes which I might eventually build something with.
They are something like 60V on anodes in typical apps. I'm inclined to try them at 18 or 27V using 2 or 3 PP3 in series, whether they would work ok at 12V is something I dont know.
They are something like 60V on anodes in typical apps. I'm inclined to try them at 18 or 27V using 2 or 3 PP3 in series, whether they would work ok at 12V is something I dont know.
@Nigel, you may want to review the investigations and experiments by Ron Quan on running tubes at (very) low voltages down to 12VDC .
Jan
Jan
Attachments
Most guitar pedals based on ECC82, use a grounded cathode and between 4M7 and 6M8 as grid leak, They work well on 9volts.
I was just having a gander at the 12B4 transfer characteristics and plate curves and at first blush it might be a really good candidate for a low voltage preamp.
12AU7 was designed to work with low voltages, but 12AX7 was not designed to the same characteristics...
I bought a 12au7 and tried to get that working. It was just very distorted even when the grid voltage was set to give 1/2 vcc out. It also (as expected) wasnt as loud as the 12ax7. Looks like the 12ax7 will be ok so long as I use the 470k pull up from the grid and 100k to ground when that particular valve needs it.
The old secret-rule was that Mu should be less than B+.
This is similar to saying there is 1V uncertainty in actual grid-cathode bias due to "contact potential" (various processing variations).
The 12AU7 is not specifically low voltage (there is a 12U7 with low-voltage processing). But it comes from a line of low Mu tubes often used in 45V and 22V radios.
The 12AX7 is aimed at analog computers with over 250V supply, to give maximum voltage gain at low current. (Granted, it is a dual of an old radio 1st audio, from days when radios had 250V supply.) That's not what you want for 12V work. The "contact potential" variation (especially exploited in 12AX7 to jazz-up the bottom of the curves) means it is likely to saturate or cut-off from one tube to the next, or one month to the next on the same tube.
Yes, if you can tolerate a grid impedance of a few-K then taking a 12AX7 grid a few tenths positive "work", though it may not be stable over time.
This is similar to saying there is 1V uncertainty in actual grid-cathode bias due to "contact potential" (various processing variations).
The 12AU7 is not specifically low voltage (there is a 12U7 with low-voltage processing). But it comes from a line of low Mu tubes often used in 45V and 22V radios.
The 12AX7 is aimed at analog computers with over 250V supply, to give maximum voltage gain at low current. (Granted, it is a dual of an old radio 1st audio, from days when radios had 250V supply.) That's not what you want for 12V work. The "contact potential" variation (especially exploited in 12AX7 to jazz-up the bottom of the curves) means it is likely to saturate or cut-off from one tube to the next, or one month to the next on the same tube.
Yes, if you can tolerate a grid impedance of a few-K then taking a 12AX7 grid a few tenths positive "work", though it may not be stable over time.
Out of 3 12ax7's two need a 470k pull up resistor on the grid. For some reason the other one works without a pull up resistor. I couldn't get the 12au7 to work without severe distortion, even at 6 volts anode voltage. Been listening to one with a pull up resistor for a good few hours now and it sounds very good. Anode voltage seems stable at 6 volts.
I have simulated my circuit with ltspice and according to that both 12ax7 and 12au7 should work well. In practice 12ax7 works with a pull up resistor on grid to get anode to 6 volts, but 12au7 is very distorted even with anode at 6 volts. I can only guess I am on the limits of the valves working area and they need a bit of coaxing to work.
I have simulated my circuit with ltspice and according to that both 12ax7 and 12au7 should work well. In practice 12ax7 works with a pull up resistor on grid to get anode to 6 volts, but 12au7 is very distorted even with anode at 6 volts. I can only guess I am on the limits of the valves working area and they need a bit of coaxing to work.
So you are *selling* us a 14.5 Euro magazine edited by yourself? 😱
Is this allowed here? 😕
Is this Ethical? 😕
As a side note, the thread is about tubes fed 12V DC, period.
The unreadable (unless we pay that is) article you linked to states (in the preview):
🙄
Yes, this tube was part of an almost complete set of tubes that were designed to run directly off a car battery (ECC86-ECH83-EF97-EF98-EBF83). There was no power tube, of course. Car radios of that aera had germanium power devices.
Best regards!
More like working beyond the normal working area so some won't work at all.nigelwright7557 said:
Don't take any notice of LTspice. You are working in bias areas where the valve models are likely to be pure fiction. For example, some models don't have particularly good simulation of grid current because this is not an issue for most audio design. For you grid current is crucial.
- Status
- Not open for further replies.