I've got 12AX7 stand alone preamp, it's a straight anode to cap and to output. I understand that without the OT there is a mismatch but it sounds fine no matter what I connect it to even high load SS amp. Is the missmatch bad for some reason even tho I can't "hear" any problems? I want to build another preamp but don't really want to use cathode follower as an output. Maybe 6SN7 would be better choice for anode output or is the correct output load critical for triodes at all?
Last edited:
The only real difference you "might" hear is a low end rolloff on a SS amp that has a low impedance input, 10K-50K. ??? Maybe a lower volume level if you were to swap cables in a A/B sound comparison. What makes it a "preamp" as opposed to a driver in a full tube power amp? It's the input levels and the output gain. And the next stage input impedance.
The 12AX7/ECC83 will give distortion if used into a low impedance load. How much distortion depends on many things, including signal level and bias point. Some people like distortion.
If used into a high impedance load then the same valve will give far too much gain which will probably lead to clipping and hence distortion. Only in a guitar amp can this valve be used successfully without any thinking.
If used into a high impedance load then the same valve will give far too much gain which will probably lead to clipping and hence distortion. Only in a guitar amp can this valve be used successfully without any thinking.
So the mismatch doesn't sound like a big problem than. I don't think I should be getting much distortion if any, it is a single stage with only roughly 1:20 gain (for Baxandal tone control circuit) 2.4k on cathode without bypass cap. I might add a bypass cap switch in case I need more gain for something.
Mosfet sound like a good idea but I want to keep it SS free. I have 10k/4ohm OTs I could use but if the load mismatch is not generally big problem, the OTs might just degrade the sound since they're only from some old gramophone or something like that. Thanks.
Mosfet sound like a good idea but I want to keep it SS free. I have 10k/4ohm OTs I could use but if the load mismatch is not generally big problem, the OTs might just degrade the sound since they're only from some old gramophone or something like that. Thanks.
Last edited:
If you want a 12AX7 to not sound at all you need a larger anode resistor than 75k.stocktrader2000 said:
With an output impedance around 300k that is almost a transconductance stage, so the voltage gain will be roughly proportional to the load - including any stray capacitance. I believe the Baxandall tone control assumes a low impedance source, so 'flat' might not be flat.6V6dude said:
Hi 6v6 dude
In my designs (and those of many others) a pre-amp works chiefly as an audio buffer. I don't really need much gain, but the ability for it to drive signal through any kind of cable is a big plus.
I would suggest listing your desired characteristics like in this example:
1. wide bandwidth
2. very low distortion
3. input connections for enough sources
4. Decent volume control.
5. Low enough output impedance to drive any kind of cable or amp.
6. High enough input impedance to accommodate most any source, but without signal roll-off (think miller capacitance here)
7. reasonable cost (whatever that means to you)
8. make use of parts on hand (always dear to our DIY hearts)
9. Looks funky (or not)
I think you get the idea here...
I think that the *SN7 tube can do it. You can also consider using a follower circuit if you need more drive. Some users get exotic and use old directly heated triodes and swear by the sound. Other users like to go for solid state devices to regulate voltage and current...
There is a lot of possibilities beyond your 12ax7 differential amplifier, which likely has far more gain than you need and very high output impedance.
Ian
edit: fully agree with DF96's comment above too.
In my designs (and those of many others) a pre-amp works chiefly as an audio buffer. I don't really need much gain, but the ability for it to drive signal through any kind of cable is a big plus.
I would suggest listing your desired characteristics like in this example:
1. wide bandwidth
2. very low distortion
3. input connections for enough sources
4. Decent volume control.
5. Low enough output impedance to drive any kind of cable or amp.
6. High enough input impedance to accommodate most any source, but without signal roll-off (think miller capacitance here)
7. reasonable cost (whatever that means to you)
8. make use of parts on hand (always dear to our DIY hearts)
9. Looks funky (or not)
I think you get the idea here...
I think that the *SN7 tube can do it. You can also consider using a follower circuit if you need more drive. Some users get exotic and use old directly heated triodes and swear by the sound. Other users like to go for solid state devices to regulate voltage and current...
There is a lot of possibilities beyond your 12ax7 differential amplifier, which likely has far more gain than you need and very high output impedance.
Ian
edit: fully agree with DF96's comment above too.
Last edited:
I got the Baxandal output grounded to 470K plus what ever impedance is the other input it goes to. Doing A/B test with switch it seems to be pretty flat as far as Baxandal goes. I don't expect high precision, it is a tone control so I can tweak it and will have a bypass on it anyway. I'm only building it to go with my tube amp I use for PC because some of the sound quality from the Net is pretty bad and this way I can adjust it if I need to. Otherwise I don't use tone controls.
Yes I'm still considering the 6SN7, I'm gonna swap it for 12AX7 tomorrow and see how it goes. Reason I'm thinking 12AX7 right now is that it's more flexible if I want to add more gain or even RIAA phono preamp in the future.
the *SN7 will need more current than your 12ax7. Hope your power supply is ok for this and you can come up with a decent load line.
If you are not sure, post your proposed circuit.
We know far to little about the current power supply and the circuit. But yes, they will need 600mA instead of 300mA for the heaters of the 6SN7
Let's be honest though.. the poster's circuit with 12AX7 has probably an amplification factor of ~20. Why? Because the plate load is far too small for this triode and they are losing all the signal gain. We know why they can't just increase the load... if they do that, they will have to bias for less headroom. Look at the 12ax7 plate curves and you will see what I mean.
However if they are happy with this level of amplification and degree of headroom, then the *SN7 will be a good alternative. It will be far more linear and will be able to provide some drive.
I will assume that they have a decent high enough unloaded supply delivering a B+ around 300V. This would be getting them about 200V on the plate at roughly 1.25mA biased at -1.5V for their 12ax7
Their DC load line is definitely steep. The AC load line is even steeper. I won't even bother to go into much discussion here.. but such little current is not a good driver and there is tons of signal loss. Also the output impedance is really high.. others mentioned it so again no detail from me on this. Its frankly not good use of the 12ax7.
Using *SN7 instead, they could use a plate load of say 33k Ohm, drawing something conservative like 8mA from each triode, biased at 2V. B+ will be around 295V and the plate will be around 125V. Amplification factor will be about the same as their current set-up but it will be far more linear and output impedance will be lower.
I won't comment much further on this though. you could Band-Aid it with a cathode follower or source follower, but...
Maybe hpeter's solution is worth considering too. I'm not sure if he said that he used that mu follower on 12ax7 though. It might result in far more amplification than the original poster desires...
Ian
Hi Ian,
Yes, we're saying close to the same thing. Used properly, a 12AX7 is a great tube - for high voltage gain and high impedance loads. You want him to use a different tube properly. I'm trying to look at his current tube choice. If he wants more gain he could also look at the 6SL7. No matter which tube he selects, he should use a buffer to drive the outside world. That's another stage and hardly a band-aide approach. It's called a good engineering decision.
The choice of a single tube gain stage is rarely a good one unless it is being mounted inside the equipment and it's load is a known quantity.
Without knowing the power supply, there isn't any decision we can make except to observe what he has got - as you pointed out.
-Chris
Yes, we're saying close to the same thing. Used properly, a 12AX7 is a great tube - for high voltage gain and high impedance loads. You want him to use a different tube properly. I'm trying to look at his current tube choice. If he wants more gain he could also look at the 6SL7. No matter which tube he selects, he should use a buffer to drive the outside world. That's another stage and hardly a band-aide approach. It's called a good engineering decision.
The choice of a single tube gain stage is rarely a good one unless it is being mounted inside the equipment and it's load is a known quantity.
Without knowing the power supply, there isn't any decision we can make except to observe what he has got - as you pointed out.
-Chris
Hi Chris
The buffer is kinda optional. He's happy now using 12ax7 in kind of quick and dirty implementation and no buffer. So I think there is only upside.
Since he mentioned 6SN7, we can only assume he has one on hand.
I would almost prefer to do something like the MU Follower Hpeter posted, but using *SN7 instead. Maybe some of my parts would be different, but the general idea would be the same. And maybe I would use an even different triode...
But 6SN7 is really an excellent choice assuming the power supply will handle it and I am certain that it will be miles better than the current 12ax7 implementation.
Ian
The buffer is kinda optional. He's happy now using 12ax7 in kind of quick and dirty implementation and no buffer. So I think there is only upside.
Since he mentioned 6SN7, we can only assume he has one on hand.
I would almost prefer to do something like the MU Follower Hpeter posted, but using *SN7 instead. Maybe some of my parts would be different, but the general idea would be the same. And maybe I would use an even different triode...
But 6SN7 is really an excellent choice assuming the power supply will handle it and I am certain that it will be miles better than the current 12ax7 implementation.
Ian
Last edited:
Hi Ian,
Maybe yes, but only if the circuit was designed properly. The 12AX7 with a mu follower would be pretty good. If not for the design! So job #1 before anything else is considered might be to redesign the 12AX7 operating points while knowing the details of the power supply. Only then can he entertain additional tube types.
-Chris
Maybe yes, but only if the circuit was designed properly. The 12AX7 with a mu follower would be pretty good. If not for the design! So job #1 before anything else is considered might be to redesign the 12AX7 operating points while knowing the details of the power supply. Only then can he entertain additional tube types.
-Chris
Regarding the power supply that's not a problem. It can supply 4W, used to run 12AX7 with ECl86 and is with 6.3V. Maybe slight overkill but that's the smallest PT I have, about size of guitar amp 6V6 OT so not that big.
I didn't have much time today for proper test, but quickly tried the 6SN7. I couldn't really tell much difference, it's hard to tell running through a tone control.
But think I should somehow sort out that output load to roughly what it should be.
Right now, measuring the load from the 12AX7 output cap to the tone control, I get 570k. Connecting the output to SS amp (50K input), the 570K only drops by 20k to 550k.
Normally when 12AX7 is directly connected to say 6V6, there is a grid resistor of about 470k so the load is basically the same as the load I have on my preamp. This is what is confusing me. It's a same load so why is it wrong?
I didn't have much time today for proper test, but quickly tried the 6SN7. I couldn't really tell much difference, it's hard to tell running through a tone control.
But think I should somehow sort out that output load to roughly what it should be.
Right now, measuring the load from the 12AX7 output cap to the tone control, I get 570k. Connecting the output to SS amp (50K input), the 570K only drops by 20k to 550k.
Normally when 12AX7 is directly connected to say 6V6, there is a grid resistor of about 470k so the load is basically the same as the load I have on my preamp. This is what is confusing me. It's a same load so why is it wrong?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.