Page 14 of 20
#196 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sat Dec 29, 2018 5:53 pm
by izzy wizzy
I'm going to have a play in ltspice to try and understand this. Just for fun.
#197 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sat Dec 29, 2018 6:20 pm
by RhythMick
izzy wizzy wrote: ↑Sat Dec 29, 2018 5:53 pm
I'm going to have a play in ltspice to try and understand this. Just for fun.
I would be very interested in your conclusions. However I'm knackered and will look into that buzz tomorrow.
#198 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sat Dec 29, 2018 6:21 pm
by Nick
3. Removed bias caps. Replaced 71A tail choke (LL1667, 2400DCR) with resistor (2k8 made from 2 x 5k6 5W - best I could do from what I have to hand). Whistle gone - and I'm sure I have more gain.
Is the centre tap connected to 0v or the top of the 2k8 resistor?
http://www.audio-talk.co.uk/phpBB3/down ... &mode=view
If its still like this where the centre tap of the interstage secondary is from a pot thats now on top of the resistor, then just remove the resistor, its still not doing anything helpful. If the centre tap is to the 0v on that stage (unlike the diagram) then you need the resistor. As Mike said just short the choke out with a clip lead would be the simple way.
#199 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sat Dec 29, 2018 6:25 pm
by RhythMick
Nick wrote: ↑Sat Dec 29, 2018 6:21 pm
3. Removed bias caps. Replaced 71A tail choke (LL1667, 2400DCR) with resistor (2k8 made from 2 x 5k6 5W - best I could do from what I have to hand). Whistle gone - and I'm sure I have more gain.
Is the centre tap connected to 0v or the top of the 2k8 resistor?
Centre tap connects to top of 2k8 resistor , which replaced the tail choke. There's another resistor below the 2k8
#200 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sat Dec 29, 2018 6:26 pm
by Nick
Then I ask the same question again. Whats that 2k8 resistor meant to be doing?
#201 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sat Dec 29, 2018 6:38 pm
by RhythMick
Dropping voltage.
So, going back to first principles which I followed for this design, namely ...
... A balanced design using a long tailed pair needs high impedance to get the best balance. This can be achieved by a large resistance, which in turn needs a high b+ to allow for the drop over the tail plus the Vak. The impedance can be improved using a constant current source (sink). Merlin Blencowe talks about using a pentode to give impedance in the megaohm region without needing huge voltages. I plan to try this at some point, also transistor CCS designs.
Around the same point I started to get my head around how chokes worked and realised that a choke on the tail should give a high AC impedance without a high DC drop.
Right now I've simply substituted the choke (DCR of 2k4) with a resistor. If I were to simply short out the choke I'd end up with too high a voltage over the valve and no doubt too high a bias to compensate.
#202 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sat Dec 29, 2018 6:55 pm
by izzy wizzy
RhythMick wrote: ↑Sat Dec 29, 2018 6:38 pm
Dropping voltage.
So, going back to first principles which I followed for this design, namely ...
... A balanced design using a long tailed pair needs high impedance to get the best balance. This can be achieved by a large resistance, which in turn needs a high b+ to allow for the drop over the tail plus the Vak. The impedance can be improved using a constant current source (sink). Merlin Blencowe talks about using a pentode to give impedance in the megaohm region without needing huge voltages. I plan to try this at some point, also transistor CCS designs.
Around the same point I started to get my head around how chokes worked and realised that a choke on the tail should give a high AC impedance without a high DC drop.
Right now I've simply substituted the choke (DCR of 2k4) with a resistor. If I were to simply short out the choke I'd end up with too high a voltage over the valve and no doubt too high a bias to compensate.
I can see the logic in this. And you got one amp to work. And I can see your voltage issues as well in changing from your original design.
The amp will balance after each stage anyway. With just resistance in the tail, you could bypass it with a big cap to see what happens if you fancied. Yes it would remove the advantage of a large tail but hey, this is fault finding.
Or you could stick RC decoupling in the ht and ground the grid and use a smaller resistance for bias. Again just for fault finding purposes.
Cheers,
Stephen
#203 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sat Dec 29, 2018 7:01 pm
by Nick
So put the resistor in the power supply before the last cap, it will drop the same voltage.
Again, I ask the question, how can the valve see the resistor? If the valve can't see the resistor how is it improving balance.
I agree, a choke in the tail of a resistively loaded LTP will improve AC balance in the same way a large resistor in the tail or a CCS will. But in this case with a inductive load all I can see it doing is altering the distortion spectra a bit.
With a resistive loaded pair then forcing the total current to be costabt will improve balance, if the current in one resistor increases by 1ma then the current in the other must drop by 1ma to keep the total constant, with a inductively loaded pair, I feel the two inductors will both try and maintain constant AC current, and I feel the result will at best be much the same as just the anode loads, at worst give all sorts of frequency dependent chaos.
#204 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sat Dec 29, 2018 7:01 pm
by Nick
but hey, this is fault finding.
Sorry, I disagree, I think its fixing a faulty design.
#205 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sat Dec 29, 2018 7:04 pm
by RhythMick
Yes I think I've mixed up my logic somewhere. Long tail with high impedance may be needed where the anode load is resistive, but I've kept that logic after moving to interstage coupling. Lots of re reading to do.
#206 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sat Dec 29, 2018 7:27 pm
by izzy wizzy
Nick wrote: ↑Sat Dec 29, 2018 7:01 pm
but hey, this is fault finding.
Sorry, I disagree, I think its fixing a faulty design.
I was trying to play nicely
It isn't what I would do but I can see the logic of it.
And it's not the only design like it. And sometimes fault finding can lead to a change in design thinking. I'm thinking of my own path in this with a not too dissimilar design. This can be a learning journey too and not a prescribed destination otherwise it ain't your amp but somebody else's.
Cheers,
Stephen
#207 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sat Dec 29, 2018 7:34 pm
by izzy wizzy
RhythMick wrote: ↑Sat Dec 29, 2018 7:04 pm
Yes I think I've mixed up my logic somewhere. Long tail with high impedance may be needed where the anode load is resistive, but I've kept that logic after moving to interstage coupling. Lots of re reading to do.
Partially true but there are a bunch of designs out there with high Z tail, ccs, and trans coupling.
Cheers,
Stephen
#208 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sat Dec 29, 2018 7:35 pm
by RhythMick
For me, at least, the learning is a bigger prize than the result.
#209 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sun Dec 30, 2018 9:02 am
by izzy wizzy
Nick wrote: ↑Sat Dec 29, 2018 2:11 pm
He floats the grid center tap on a choke to ground too.
Ok, any link I would be interested in seeing.
http://izzy-wizzy.com/audio/phono/Diffphono.gif
It's from a while ago I think.
Cheers,
Stephen
#210 Re: Finished and working - 12 valve all DHT PP 26-71a-45
Posted: Sun Dec 30, 2018 10:38 am
by Nick
Wow, is someone sponsored by Lundahl? Thats got the coupling from a interstage between the anodes, so not sure what you would call those stages. I assume the anode loads are coupled, and its not got a £300 transformer in each anode. However the 40ma rating possibly indicates a core gap, which would point to single ended use so the anodes are possibly not inductively coupled.
But to be curmudgeonly, just because something is found on the internet doesn't mean its right. I fully understand the logic from resistive to inductive loads leading to the choke in the tail looking like a good idea, and I don't see why RhythMick would be the only person to make that leap.
But thanks for the link anyway.
