6C33C SE-OTL Project
#16 Re: 6C33C SE-OTL Project
Ok, got it. “Grid blocking’ shouldn’t be an issue with CCS but I’d rather be safe than sorry so the 100k instead of 470k is some compromise that doesn’t hurt. To use a coil/choke (I use chokes wherever I can because they add something wonderful to the sound of tubes IMO!) instead of a resistor could reduce the loading of the 6SN7 but we have enough gain for driving the 6C33. At 1V input we get 40V (pp) output from the 6SN7, the max before clipping is 2,3V input and 84V (pp) output but the 6C33 can only handle 1,2V input before starting to clip with about 1k for R(NFB).
So we don’t have to give up the better PSRR of the input stage and change to cathode bias with bypass cap for the second 6SN7 triode which would have been plan B.
The challenge with the 6C33 in this configuration was to get the cleanest neg half wave with Ip as high as possible at highest R(NFB) but without exceeding Pmax of 60W. So the best compromise for best transients and max P into 16 ohm with the 60W limit was B+ at 127V and Ip at 467mA with the tiniest bit of headroom left. The square waves with 16 ohm load look quite impressive (clean) even at 10k. I will post more pics soon.
Here is the the finalized circuit which I used:
So we don’t have to give up the better PSRR of the input stage and change to cathode bias with bypass cap for the second 6SN7 triode which would have been plan B.
The challenge with the 6C33 in this configuration was to get the cleanest neg half wave with Ip as high as possible at highest R(NFB) but without exceeding Pmax of 60W. So the best compromise for best transients and max P into 16 ohm with the 60W limit was B+ at 127V and Ip at 467mA with the tiniest bit of headroom left. The square waves with 16 ohm load look quite impressive (clean) even at 10k. I will post more pics soon.
Here is the the finalized circuit which I used:
#17 Re: 6C33C SE-OTL Project
You could get more adventurous, choke load the CF and use a negative supply and negative bias the grid you could remove the output coupling cap.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#19 Re: 6C33C SE-OTL Project
The results Wollfgang has achieved with his mono breadboard build are sufficient to make up my mind to go ahead with building my own project.
I'll start collecting the parts and will start work in earnest as soon as I finish my DSD decoder project. Updates in due course.
I'll start collecting the parts and will start work in earnest as soon as I finish my DSD decoder project. Updates in due course.
Sorry, I couldn't resist!
#20 Re: 6C33C SE-OTL Project
For matching the 6C33 OTL with the speakers I measured the CDF (critical damping factor ) for the Lowther DX4/15 ohm in my C-horns.
At 2,5 ohm it would be ideal but even within the range of up to 9ohm it look pretty good. And 6ohm output impedance I could already achieve with the test version by adding enough NFB. So overall it looks very good.
These results also corroborate perfectly with my listening experience with the original TS 300B OTLs which didn’t sound right at 0,8 ohm output impedance (completely overdamped speakers). I get the best sound with the C-horns if I set the NFB for an output impedance around 9 ohm. So we’ll see how this will finally go with the 6C33 OTL.
The measurements show the CDF at 9ohm,4,5 ohm, and 2,5ohm.
At 2,5 ohm it would be ideal but even within the range of up to 9ohm it look pretty good. And 6ohm output impedance I could already achieve with the test version by adding enough NFB. So overall it looks very good.
These results also corroborate perfectly with my listening experience with the original TS 300B OTLs which didn’t sound right at 0,8 ohm output impedance (completely overdamped speakers). I get the best sound with the C-horns if I set the NFB for an output impedance around 9 ohm. So we’ll see how this will finally go with the 6C33 OTL.
The measurements show the CDF at 9ohm,4,5 ohm, and 2,5ohm.
#21 Re: 6C33C SE-OTL Project
Just my tuppance, I know I'm a bit late to the party, but I have seen values supposedly held in check by current sources trying to run away, especially using high gm stuff like D3A, as Nick says the voltage just wanders off into the distance....and you get a nice glowing valveNick wrote: ↑Sat Apr 21, 2018 8:15 pmWhats to stop it. at least with a resistor instead of a current source there will be some degeneration to pull it down if it wants to float up, not so with a current source, The current may stay the same, but the voltage can go to wherever it wants.IslandPink wrote: ↑Sat Apr 21, 2018 6:48 pm
I can see there being some potential for distortion with grid current, but can it run away, with a current source in the cathode ?
Analogue, the lost world that lies between 0 and 1.
#22 Re: 6C33C SE-OTL Project
I think we need to be more specific if we have this interesting discussion. The only problem I can think of with Rg being too high is that the grid voltage can become positive (bias plus signal) and grid blocking would occur. The grid voltage would run away to positive and stay positive under these circumstances. This would indeed result in red plating of the tube with cathode/fixed bias. It is voltage not current that causes this and CCS couldn't stop this phenomenon at the voltage level.
But then again the result of grid blocking which can destroy the tube is because of too much current because of positive bias and this should not happen with CCS IMO. Maybe somebody could explain why CCS shouldn't be able to stop too much current as a consequence of positive grid voltage?
But then again the result of grid blocking which can destroy the tube is because of too much current because of positive bias and this should not happen with CCS IMO. Maybe somebody could explain why CCS shouldn't be able to stop too much current as a consequence of positive grid voltage?
#23 Re: 6C33C SE-OTL Project
Well, first, its not grid blocking I am talking about but what these valves do as they age. They start to pass a little grid current. That little grid current will pass the grid resistor (100k) the grid and the anode. The CCS wont see that current so it can't control it.
Remember in A2 the grid becomes a cathode.
It may be fine, but I would want to try sending some peak signals into the amp (with a load attached) and maybe a protective zener around the CCS, or use a CCS with a higher safe voltage limit.
Remember in A2 the grid becomes a cathode.
It may be fine, but I would want to try sending some peak signals into the amp (with a load attached) and maybe a protective zener around the CCS, or use a CCS with a higher safe voltage limit.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#24 Re: 6C33C SE-OTL Project
I thought we were talking about grid current that cannot sufficiently flow between grid and the cathode because of Rg (grid leakage resistor) being too high? Every valve passes this grid current, positive and negative, which evens out at some point and then the positive grid current is dominating. This could lead under certain conditions to grid blocking which could destroy the tube but that’s obviously not the problem right now and in this case.
It’s rather about giving the grid-current a sufficiently “easy” path to the cathode when it starts to become more and more positively charged compared to the cathode and starts acting like an anode? How could the grid become a cathode in this context? Maybe I am looking at the same situation but from a different point of view.With a cathode biased tube making the grid resistor smaller would clearly help so that more grid current could flow. With a voltage regulator instead of a cathode resistor it is a different situation altogether.
The cathode in this circuit is at about 22V with the anode at 127V. The TL783 is a high voltage regulator which can handle 125V at its input. I cannot see any good reason for a protective diode from output to input. B. Rosenblit used in the case of the 300B OTL and with the LM317 a zener with 36V but here the cathode was at 15V and the LM317 has an upper limit of 37V. But the sound was slightly dulled down so I took it out again after some listening and never had a problem.
But following Nick’s advice I will do some more testing with the finished amp.
It’s rather about giving the grid-current a sufficiently “easy” path to the cathode when it starts to become more and more positively charged compared to the cathode and starts acting like an anode? How could the grid become a cathode in this context? Maybe I am looking at the same situation but from a different point of view.With a cathode biased tube making the grid resistor smaller would clearly help so that more grid current could flow. With a voltage regulator instead of a cathode resistor it is a different situation altogether.
The cathode in this circuit is at about 22V with the anode at 127V. The TL783 is a high voltage regulator which can handle 125V at its input. I cannot see any good reason for a protective diode from output to input. B. Rosenblit used in the case of the 300B OTL and with the LM317 a zener with 36V but here the cathode was at 15V and the LM317 has an upper limit of 37V. But the sound was slightly dulled down so I took it out again after some listening and never had a problem.
But following Nick’s advice I will do some more testing with the finished amp.
#25 Re: 6C33C SE-OTL Project
Just remember, current doesn't flow from cathode to grid, but grid to anode. (or anode to grid if you want conventional).
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#26 Re: 6C33C SE-OTL Project
But its not a voltage regulator, its a current regulator, there is nothing to control the voltage across it other than the valve cathode. Once a valve grid goes +ve WRT its cathode two things happens, the first is it ceases to be a triode and becomes a forward biased diode, and second the grid becomes its own cathode.With a voltage regulator instead of a cathode resistor it is a different situation altogether.
With high gm valves like this, the distance between the cathode and grid is very small, over time some of the cathode material will be transferred to the grid, this is why the grid starts to pass current even when its -ve WRT the cathode. The grid starts to emit on its own. If the resistance in the grid circuit is too large then this grid current will cause a voltage to be developed across the grid resistor, as the voltage increases it will drag the cathode voltage up with it. The CCS in the cathode has no action to try and reduce the cathode voltage in the same way a cathode resistor would.
IMHO, a 6c33c is a very different beast to a 300b. The Almaro developed unstable bias with a resistor in the cathode as the valves aged, and I think that had a 220k grid resistance before if altered it.
It is only a suggestion, I will not in any way be offended if you chose not to follow it
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#27 Re: 6C33C SE-OTL Project
I simply try to follow the better argument, Nick. Especially your last explanation has helped me better understand your point of view.
Funny, that it was a misunderstanding of what I wanted to say that brought it up. Why I mentioned the voltage regulator was to point out that it doesn’t help to discharge the grid from potential difference to the cathode like a cathode resistor would. That’s exactly what you said: “The CCS in the cathode has no action to try and reduce the cathode voltage in the same way a cathode resistor would.” It’s clear that the VR regulates only the current.
My angle of looking at this situation was that when the grid voltage becomes positive and electrons are attracted to it, the grid may itself emit secondary electrons. This secondary electron current is in the opposite direction to the electron stream from cathode to grid- the grid would look rather like an anode from the position of the cathode. But of course we can also look at it like the grid has become its own cathode (now from the position of the anode) because this situation makes out of the tube at some point basically a diode.
How much this problem can affect valves depends on the valves. And obviously the 6C33 is effected very much by it as you could point out based on your practical experience. I use this tube for the first time so I have zero practical experience with it.
Reducing Rg even more (50k should be possible as we have enough headroom) and measuring periodically parameters of the 6C33C while it ages in order to get a better feel for this tube seems the right thing to do under those circumstances.
Funny, that it was a misunderstanding of what I wanted to say that brought it up. Why I mentioned the voltage regulator was to point out that it doesn’t help to discharge the grid from potential difference to the cathode like a cathode resistor would. That’s exactly what you said: “The CCS in the cathode has no action to try and reduce the cathode voltage in the same way a cathode resistor would.” It’s clear that the VR regulates only the current.
My angle of looking at this situation was that when the grid voltage becomes positive and electrons are attracted to it, the grid may itself emit secondary electrons. This secondary electron current is in the opposite direction to the electron stream from cathode to grid- the grid would look rather like an anode from the position of the cathode. But of course we can also look at it like the grid has become its own cathode (now from the position of the anode) because this situation makes out of the tube at some point basically a diode.
How much this problem can affect valves depends on the valves. And obviously the 6C33 is effected very much by it as you could point out based on your practical experience. I use this tube for the first time so I have zero practical experience with it.
Reducing Rg even more (50k should be possible as we have enough headroom) and measuring periodically parameters of the 6C33C while it ages in order to get a better feel for this tube seems the right thing to do under those circumstances.
#28 Re: 6C33C SE-OTL Project
Yep. I think just watching it will show you if its starting to get out of hand. Also how it behaves on peaks will help to reassure you it can keep the grid under control.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#29 Re: 6C33C SE-OTL Project
I finally could listen for the first time with the finished amp in my main system. After 20+ years of wasting a ton of money in the Highe-End scene and building my own equipment for more than 10 years I am still speechless and wondering what to say about this experience. What I can say is that this amp is as dead quiet as the very best of the solid state amps I have owned and it also makes music like I have never heard before from any DIY amp. Transcendent Sound products are of the past in my case.
I measured the grid voltage with and without signal both channels (33mA/55ma and 2V DC at max input)and I am considering adding a monitoring/safety LED/switch .
I measured the grid voltage with and without signal both channels (33mA/55ma and 2V DC at max input)and I am considering adding a monitoring/safety LED/switch .
- Cressy Snr
- Amstrad Tower of Power
- Posts: 10579
- Joined: Wed May 30, 2007 12:25 am
- Location: South Yorks.
#30 Re: 6C33C SE-OTL Project
Top result there Wolfgang
Sgt. Baker started talkin’ with a Bullhorn in his hand.