I've been doing several permutations of a 2-stage SE amp recently. Some are with 300b outputs, and some with EL12n outputs. The EL12n amp interests me a lot, since the mu of EL12n is 18, which allows me to use DHTs as drivers. I have so far tried 10Y, 26, 4P1L and 01A. I would dismiss the 4P1L as tonally inferior to the others. The 10Y is excellent and I can happily use it as a solution. The 26 struggles a bit at 7mA current, though the tone is good, on the warm side. The 01A is tantalising - the tonality is very fine, as we know, but at 3.4mA current it's asking for something in the circuit that would give it more drive.
So what are my options here? The aim is to preserve the tonality of the 01A intact but give it better drive into the EL12n outputs. I don't need more gain, just better drive. I'd be very grateful for any and all suggestions here - the tonality of the 01A is too good not to use it. I attach the schematic as it is currently.
So what are my options here? The aim is to preserve the tonality of the 01A intact but give it better drive into the EL12n outputs. I don't need more gain, just better drive. I'd be very grateful for any and all suggestions here - the tonality of the 01A is too good not to use it. I attach the schematic as it is currently.
Attachments
With a more constructive contribution here.
Andy, you know I love this valve. However, isn't expected to swing large volts. Depending on your output stage, with the 3mA you could run into slew rate distortion issues.
I'd go for a cathode/source follower at the output. DC-coupled to 01a anode. That's what worked best for me.
Alternatively, I was thinking on implementing a "super-triode" or assisting the 01a with some silicon. I know you're not a fan, but can keep the sonic signature of the 01a whilst providing current drive. See the attached
Challenge will be to calibrate the circuit on R5 as every MOSFET has wider parameters including VGS(th) and GFS. If you vary R5 by 100mR you can get a big variance on the drain current. Without a servo (or a trimming circuit) it may be hard to implement.
you can see that beyond 100Vpp, you are asking for trouble here with the 01a.
cheers
Ale
Andy, you know I love this valve. However, isn't expected to swing large volts. Depending on your output stage, with the 3mA you could run into slew rate distortion issues.
I'd go for a cathode/source follower at the output. DC-coupled to 01a anode. That's what worked best for me.
Alternatively, I was thinking on implementing a "super-triode" or assisting the 01a with some silicon. I know you're not a fan, but can keep the sonic signature of the 01a whilst providing current drive. See the attached
Challenge will be to calibrate the circuit on R5 as every MOSFET has wider parameters including VGS(th) and GFS. If you vary R5 by 100mR you can get a big variance on the drain current. Without a servo (or a trimming circuit) it may be hard to implement.
you can see that beyond 100Vpp, you are asking for trouble here with the 01a.
cheers
Ale
Attachments
@Andy,
BTW, if you want to experiment with a source follower, I can send you a pair of boards to test on your amp. Can bias them at 10 or 15mA and you can experiment easily. Let me know.
Ale
BTW, if you want to experiment with a source follower, I can send you a pair of boards to test on your amp. Can bias them at 10 or 15mA and you can experiment easily. Let me know.
Ale
Thanks Ale. You're a wizard with solid state, and that's not true in my case as you know, my knowledge base is valves. Before I try a solid state solution (which I could do), how would this compare with using a tube as a cathode follower? Direct coupled as per your suggestion? If so suggestions for what would make a good tube? I've never used cathode followers so would have to look up all the theory, but I have plenty of tubes I could use. I may have one of your boards already - I remember trying something out a few years ago. May be an older version.
Another thought that struck me is that I'm using a 100K grid leak on the EL12n, which is a left-over from my 300b output circuit. The data says maximum grid leak is 700K, so it looks like I could take that up to at least 470K and see if that helps.
I don't need more gain than I have at present using a plate choke. It's driven by my AK4490 DAC which outputs 2v. I could reduce the output resistance with a step-down interstage, but that would wipe out the gain and create all kinds of other issues.
Another thought that struck me is that I'm using a 100K grid leak on the EL12n, which is a left-over from my 300b output circuit. The data says maximum grid leak is 700K, so it looks like I could take that up to at least 470K and see if that helps.
I don't need more gain than I have at present using a plate choke. It's driven by my AK4490 DAC which outputs 2v. I could reduce the output resistance with a step-down interstage, but that would wipe out the gain and create all kinds of other issues.
Last edited:
Did you see on the sheet that typical class A operations expects from 5-9v signal input, as this tube has an amp factor of 8 and a gm 1/3 of a 12AU7?
Hi Andy, if you can increase the grid leak then is good for better linearity. What is the input capacitance of the EL12n?
A cathode follower is rather more complicated to implement and with poorer performance IMHO compared to a source follower.
I don't think you have my SF boards. I can send you a pair if you need.
What is the voltage swing needed in the EL12n grid?
I suspect that unless you have a good healthy pair of 01a with good element alignment, you may get a lot of distortion above 70Vpp. No sure, haven't used the 01a as a driver beyond 30vpp.
Ale
astouffer - Good question - that's what I'm trying to establish.......
May be a question of current, may help to reduce the output impedance of the 01A stage. I'm looking for suggestions.
Ale - here's the output stage. Don't know the input capacitance of the EL12n in triode. It's an 18W valve, S=18, Ri=1K approx.
May be a question of current, may help to reduce the output impedance of the 01A stage. I'm looking for suggestions.
Ale - here's the output stage. Don't know the input capacitance of the EL12n in triode. It's an 18W valve, S=18, Ri=1K approx.
Attachments
Last edited:
The output impedance of the 01a isn't the problem. The problem is how much current it can source/sink to charge/discharge the capacitance of the EL12n.
Slew rate distortion depends on voltage swing, frequency and capacitance, not the impedance of the driver
Ale
Slew rate distortion depends on voltage swing, frequency and capacitance, not the impedance of the driver
Ale
Sure Andy.
This post will give you the basics you need to get the head around it:
Slew Rate in Preamps – Bartola(R) Valves
This post will give you the basics you need to get the head around it:
Slew Rate in Preamps – Bartola(R) Valves
Thanks Ale - that's perfect. I'll have to study it well. The 01A is a favourite tube asking for some way to use it as a driver.
And thanks euro21. This is adding to the information needed. What's ayumi, incidentally?
And thanks euro21. This is adding to the information needed. What's ayumi, incidentally?
So the 2K plate load isn't going to be too low for a tube with 10K internal resistance? The following parallel grid resistor is only a small part of the input impedance the tube will push into. Why are you looking to increase the drive if 2v gives enough output signal? Looks like you only need a gain of 4 without using FB. What problem are you trying to overcome, exactly?
The plate load is a 136H choke.
The issue appears to be slew rate according to Ale, who knows the 01A tube very well and has worked with it for several years. Would you agree to that?
As I've said, the 10Y does the job very well, so I don't actually need the 01A to work. However, I would like it to work because I'm very fond of the tonal qualities of the 01A. It's a challenge, and hopefully one that has a solution. Ale has one example of a solution if you read his blog on Bartola Valves. Can you suggest other solutions? What I'm trying to achieve is to keep the tonal qualities of the 01A while giving it a more solid drive, which should be an audible improvement if the combination with the output stage is acting more efficiently.
As Radu says at the bottom of Ale's article, the 01A sounds good at lower volumes. In his words "at lower levels the 01A sound is gorgeous, but at higher levels it loses dynamics and the sound is thinner." That sounds close to what I'm hearing.
Last edited: