As we know Nick's had good results from the classic series pass valve reg with pentode error amp type documented by Steve Bench.
I have, with limited success, managed to get about 70mA 'usable current' out of a 'floated' LM317 this goes into a high speed op amp configured as voltage driver thus elimiating the need for decoupling between stages; each stage has its own "PSU". The power rails for the op amp are sourced from the B+ (which is designed to be 15volts more than I need) then, by using a voltage reference to divide out 30 volts I get plus and minus 15v for the op amp. I think I could push this design up the the limit of the op amps at about 50mA each, if need be. Two 317 power four op amps one for channel for each stage in the phono. The actually circuit, as I have said before, is detailed in MJ.
I am moderately pleased with the result but do have some humm to remove, I do not, at this stage, know if the hum is actually on the B+ or getting coupled in from elsewere.
So I made a meal out of building these, mostly due to my own stupidity and trying to get three to four times the current per unit as Morgan Jones does on his phono and not taking the extra power dissapation into account.
So I'm looking what to try next, better or worse etc.
I suspect the The Alan Weeks supereg design does not float so is not good for HT.
I noted a thread on PFM that might http://www.pinkfishmedia.net/forum/prin ... 6174&pp=15
float, what does the panel think?
There's also a design as used by Conrad Johnston that I need to study further, http://www.drtube.com/audioamp.htm#CJ, this is similar to one that is mooted to be used by AR.
It must be able to float to give HT and it must be able to deliver 40mA per stage if it is to be used for a high gm driver etc.
So the bottom line is I don't think these regs are quiet enough for an LCR phono with pentode input.
Any ideas?
cheers,
-- Andrew
Lets talk regulators.
#2
The PF link was one I found (and lost again), to my miund, it looks like a emitter follower with a stable base supply, and I think looks like it could work well, I don't understand the various names though, I don't see how its multiplying anycapacitance.
In my limited understanding this is what I thought of as a gyrator
http://en.wikipedia.org/wiki/Gyrator
I can see its worth experimenting with though.
I am sure that it should be possible to get a SS version of a valve rect, thats good for 200ma or so at 400v.
In my limited understanding this is what I thought of as a gyrator
http://en.wikipedia.org/wiki/Gyrator
I can see its worth experimenting with though.
I am sure that it should be possible to get a SS version of a valve rect, thats good for 200ma or so at 400v.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#3
Hi Nick,
Yes, I had the same thought, emitter follower, similar to the AR and CJ idea, I think, more study really required on my part.
I need some better test gear, to be sure, but I don't feel comfortable with the LM317 on its own. Reminds me, I must fix my scope.
Yes I agree 200mA should also be feasible from a Maida type circuit but I'm sure it well be noisy, to much so for a pre/phono anyway.
I think for a phono, at least, I need to follow up the 317 with something like this circuit then on into the voltage buffer.
The LCR is sounding very beguiling, very fluid, yet powerul, despite still needing a big shakedown, its not right yet, but it makes me want to pursue the solid state regs idea further. The LCR and the Zout of 250R to drive the TVC helps as well.
I can see any reason why this type of circuit would not float up at 200v or so? The reference needs lots of zeners but they're cheap enough, just noisy Why not even mix in a VR tube? Noisy too....
Somewhere I have a old schematic for a Harris bench PSU, these things were so good HP bought the company. I must look up how they did they're HT PSUs.
PS I need to read up on the Gyrator too.
cheers,
-- Andrewb
Yes, I had the same thought, emitter follower, similar to the AR and CJ idea, I think, more study really required on my part.
I need some better test gear, to be sure, but I don't feel comfortable with the LM317 on its own. Reminds me, I must fix my scope.
Yes I agree 200mA should also be feasible from a Maida type circuit but I'm sure it well be noisy, to much so for a pre/phono anyway.
I think for a phono, at least, I need to follow up the 317 with something like this circuit then on into the voltage buffer.
The LCR is sounding very beguiling, very fluid, yet powerul, despite still needing a big shakedown, its not right yet, but it makes me want to pursue the solid state regs idea further. The LCR and the Zout of 250R to drive the TVC helps as well.
I can see any reason why this type of circuit would not float up at 200v or so? The reference needs lots of zeners but they're cheap enough, just noisy Why not even mix in a VR tube? Noisy too....
Somewhere I have a old schematic for a Harris bench PSU, these things were so good HP bought the company. I must look up how they did they're HT PSUs.
PS I need to read up on the Gyrator too.
cheers,
-- Andrewb
#4
Hmm, as far as I can see, it requires a quiet, stable high voltage source.
I wonder if this cound be done with a common emiter HV TX, with its collector to B+, then a voltage divider from that down to op amp voltages, other op amp input from a good quiet voltage ref, than use the op amp output to drive the transistor.
Would that work?
I wonder if this cound be done with a common emiter HV TX, with its collector to B+, then a voltage divider from that down to op amp voltages, other op amp input from a good quiet voltage ref, than use the op amp output to drive the transistor.
Would that work?
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#6
No, sorry, I was just thinking out loud. Thinking about replacing that string of zeners with nice quiet DC reference.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#7
Ah right, still can't visualize it, too early in the morning perhaps.
The circuit I use makes B+ -15 and B+ -30 using a TL431, the limit for these beasts is 36 volts and, I suspect, given the size of cap MJ had across it, they're also quite noisy. MJ also has 20mA going through the ones on his circuit, again I suspect to keep the noise down. However, you could probably string these together.
The zener will also need >10mA to keep them quiet.
cheers,
-- Andrew
The circuit I use makes B+ -15 and B+ -30 using a TL431, the limit for these beasts is 36 volts and, I suspect, given the size of cap MJ had across it, they're also quite noisy. MJ also has 20mA going through the ones on his circuit, again I suspect to keep the noise down. However, you could probably string these together.
The zener will also need >10mA to keep them quiet.
cheers,
-- Andrew
#8
I notice the Pink Fish circuit just uses a voltage divider to create the DC reference, which is then passed through a CRC filter.
Does that sound right?
If it does, the combo of the CRC filter and the divider, which will drop noise as well, might well be enough to give a good enough DC ref and avoid the problem with the zeners? I do know the 317's are rock solid at varying mains load, no drift whatsoever, so a stable DC ref for the supply is not at issue, just a quiet one.
What am I missing?
So, let's assume the transistors were replaced by ones that can live at HT voltages. What if I then put this after the monlithic 317 HT reg and then finally the op amp voltage buffer?
My problem, as I think it stands, is that a 1k5 loaded pentode has very little in the way of PSRR compared to a typical grounded cathode.
As a result, any noise on the B+ just gets amplified up and, what's worse, its right bang in the region where the RIAA filter is of absolutely no help!
cheers,
-- Andrew
Does that sound right?
If it does, the combo of the CRC filter and the divider, which will drop noise as well, might well be enough to give a good enough DC ref and avoid the problem with the zeners? I do know the 317's are rock solid at varying mains load, no drift whatsoever, so a stable DC ref for the supply is not at issue, just a quiet one.
What am I missing?
So, let's assume the transistors were replaced by ones that can live at HT voltages. What if I then put this after the monlithic 317 HT reg and then finally the op amp voltage buffer?
My problem, as I think it stands, is that a 1k5 loaded pentode has very little in the way of PSRR compared to a typical grounded cathode.
As a result, any noise on the B+ just gets amplified up and, what's worse, its right bang in the region where the RIAA filter is of absolutely no help!
cheers,
-- Andrew
#9
Yes, good point, you are not interested in another voltage reference, you just need a quiet stable voltage to supply the darlington with.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#10
I just came across this
http://www.tubecad.com/2005/January/blog0031.htm
The last circuit on the page is similar to what I had in mind to create a stable HV tource for the emmiter follower, but replace the valve with a HV transistor
http://www.tubecad.com/2005/January/blog0031.htm
The last circuit on the page is similar to what I had in mind to create a stable HV tource for the emmiter follower, but replace the valve with a HV transistor
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#11
Now here is another thing to worry about.
Remember the d3a anode follower buffer stage I built?Well, I decided to tidy it up for Witham, at the time, the heater supply was a torroid, bridge, three pin reg, to the 6.3 heater. Very quiet, only noise was white noise.
However its a bit of a bodge (even by my standards), the heater tx and board is just hanging off the side, so before doing some more woodwork, I thought I would see how noisy it was with AC heaters (I went to DC to get the noise down, but the noise wasn't the heaters anyway). Not bad, but not perfect, still a 100hz buzz with ear next to speaker. "Bum,I need to do the woodwork anyway". But I couldn;t be bothered to do it then, so left it. Sat a listened to some music. "Hang on", it sounds much better, more open, less compressed and sat on.
Then I realised this was the same difference (If you see what I mean) I heard with the 300b driver stage replacing the DC supply with AC.
SO, there is something broken with the simple 6.3 reg supply, its too much of a coincidence for the same effect twice, and I don't think any part of my head was expecting that result.
Question is, whats doing the damage, could it be the regulator, or could it be the simple bridge rect. So things to try after Witham are, better diodes instead fo the bridge, and if that doesnt help, maybe a choke input supply, with and without the regulator, to see if its possible to get the good sound without the buzz,
Of course, its possible the buzz is creating the good sound, maybe acting like a HF bias in a tape deck, stopping the speakers having stiction.
Wouldn't it be good if this was simple
Remember the d3a anode follower buffer stage I built?Well, I decided to tidy it up for Witham, at the time, the heater supply was a torroid, bridge, three pin reg, to the 6.3 heater. Very quiet, only noise was white noise.
However its a bit of a bodge (even by my standards), the heater tx and board is just hanging off the side, so before doing some more woodwork, I thought I would see how noisy it was with AC heaters (I went to DC to get the noise down, but the noise wasn't the heaters anyway). Not bad, but not perfect, still a 100hz buzz with ear next to speaker. "Bum,I need to do the woodwork anyway". But I couldn;t be bothered to do it then, so left it. Sat a listened to some music. "Hang on", it sounds much better, more open, less compressed and sat on.
Then I realised this was the same difference (If you see what I mean) I heard with the 300b driver stage replacing the DC supply with AC.
SO, there is something broken with the simple 6.3 reg supply, its too much of a coincidence for the same effect twice, and I don't think any part of my head was expecting that result.
Question is, whats doing the damage, could it be the regulator, or could it be the simple bridge rect. So things to try after Witham are, better diodes instead fo the bridge, and if that doesnt help, maybe a choke input supply, with and without the regulator, to see if its possible to get the good sound without the buzz,
Of course, its possible the buzz is creating the good sound, maybe acting like a HF bias in a tape deck, stopping the speakers having stiction.
Wouldn't it be good if this was simple
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#12
Hi Nick,
We've just gotta try the scope on these little 3 legged blighters, as I said I think the line regulation is poor, lots of crap coming out of them. Now as to why that would kill the sound stone dead when its the heaters, well. I dunno.
Perhaps this is something for tomorrow, eh?
cheers,
-- Andrew
We've just gotta try the scope on these little 3 legged blighters, as I said I think the line regulation is poor, lots of crap coming out of them. Now as to why that would kill the sound stone dead when its the heaters, well. I dunno.
Perhaps this is something for tomorrow, eh?
cheers,
-- Andrew