D3a Phonostage reloaded Assistance required

[Dave the bass]

Kerrrrikey MrT, nice work here, you're way further down the Phono stage path than I am! Respektacles!

I've finally finished the WD Phono3, I know.....I know..... had to mod the heaters on it to run at 12.6V (by cutting PCB tacks and such like) as I wanted to use the 250V reg'd HT supply from my existing (now redundant) phono stage which has 12.6V DC reg'd heaters. Only thing that caught me out was are the heaters elevated on Phono2/3? If so why and how much? erks! I can't remember.

It does that airy top end that you got with your wooden box Phono, Like, plus, thumbs up etc etc etc.

[Nick]

With a smaller Rk the CCS has a less constant first C. 270R would give something like only 4k as the resistance seen by the cathode follower.

[Nick]

Only thing that caught me out was are the heaters elevated on Phono2/3? If so why and how much? erks! I can't remember

100v, something like that, to be nice to the final cathode follower.

[thomas]

Ah DTB congrats!
We'll have to do a compare/ contrast!
The cathode of your cathode follower looks to be at around 120v so raise those heaters for that one! 30 to 40v+ on the others is supposed to make a difference too (I read somewhere….)

Diet: Nick posted ahead of me there!

Edit too……. schematic for the WAD phono 3 in front of me, all heaters at +40v. Tad low for the CF I'd have thought….

[thomas]

Y'know thinking aloud here, obviously I will finish building this as is, but if I did bung a CF after the first D3a and before the network, I could have potentially scope for a, say, 2K ish impedance riaa, with rather convenient network values of around 300r, 1uf and 0.33uf….

[Dave the bass]

Edit too……. schematic for the WAD phono 3 in front of me, all heaters at +40v. Tad low for the CF I'd have thought….

Cheers MrT and MrG. Ah, I see what you mean, the CF is DC coupled to the previous stage. I just checked the data sheet and ECC83/12AX7 has a 200V DC max heater - cathode potential but it also stipulates no more than a 100V DC component when cathode is +Ve to heater, hence slight 40V lift yeah?

[IslandPink]

With a smaller Rk the CCS has a less constant first C. 270R would give something like only 4k as the resistance seen by the cathode follower.

Ah , right - so an Aikido relies on distortion cancelation against the first pair to get its extra-low overall distortion ?

[thomas]

Edit too……. schematic for the WAD phono 3 in front of me, all heaters at +40v. Tad low for the CF I'd have thought….

Cheers MrT and MrG. Ah, I see what you mean, the CF is DC coupled to the previous stage. I just checked the data sheet and ECC83/12AX7 has a 200V DC max heater - cathode potential but it also stipulates no more than a 100V DC component when cathode is +Ve to heater, hence slight 40V lift yeah?

Sounds spot on!

With a smaller Rk the CCS has a less constant first C. 270R would give something like only 4k as the resistance seen by the cathode follower.

Ah , right - so an Aikido relies on distortion cancelation against the first pair to get its extra-low overall distortion ?

I'd love to contribute to this but alas have hit the limits of my understanding of Rk loading ccs's here !!

[Dave the bass]

Good thread! Amazing what a bit of rubbish weather does to rekindle the forum spirit innits

[izzy wizzy]

You could connect the d3a as a pentode like I did if you wanted http://izzy-wizzy.com/audio/preampnew.html

cheers,

Stephen

[thomas]

You could connect the d3a as a pentode like I did if you wanted http://izzy-wizzy.com/audio/preampnew.html

cheers,

Stephen

Give me time…..I'll get there….
Actually I think I'll need to do a lot more reading before I'm confident using pentodes generally...
I was being half serious when I asked earlier about feedback, being pentode and all, the two have to go together don't they?!… like cheese and red wine...

[Nick]

[
I was being half serious when I asked earlier about feedback, being pentode and all, the two have to go together don't they?!… like cheese and red wine...

No, you are probably thinking about power output stages. Feedback is not needed for voltage amplification stages (though can be used). In fact, a good way of thinking about pentodes is to consider the feedback inherent in a triode. Imagine a triode with a load resistor. Send the grid positive, the current through the valve increases, so the voltage on the anode load increases, so the anode voltage drops. . Now because the voltage on the anode drops, there is less voltage across the valve so current drops. So the action of the anode moving is to counter the action of the grid going positive. Now imagine if you could somehow shield the anode so the cathode didn’t see the voltage on it changing. Grid goes positive, current increase, anode moves down, but the valve still sees the voltage on the screen grid, so the feedback is not happening, so gain is much higher. Thats what the screen grid goes on a pentode, it hides the motion of the anode from the cathode.

For the same reason miller capacitance is reduced. Miller is the effect of the anode moving against the grid. Stop the grid seeing the anode move, and miller capacitance is removed. So the addition of the screen removes miller capacitance increasing the frequency the valve can operate at, and also increases the gain.

The downside is the anode feedback is why triodes have low anode resistance, so pentodes have high ra. But for voltage amp use, thats not a problem.

Try and understand the way a triode works, in terms of a field of electrons and how they "see" the anode as a attractive force, and how the grid reduces the attractiveness of the anode to the electrons by hiding it in electrostatic terms. Once you can picture this add another grid, and see how this changes things and how the new grid hides the anode from both the cathode and its electrons and the first grid.

I am not being insulting when I say, reading is important, but understanding is more important.

[IslandPink]

Superb stuff, need to read through that again in case I missed something !
Meanwhile here's a sketch of my D3a phono as used at present - though there have only been minor component changes in the last 6 years.
Stephen - I did note your use of D3a as pentode driving the network, and nearly did it - but then got too heavily into thinking about a new-build LCR phono, then the speakers, etc etc. I will finally get to try this in the form of a Gm amp, in the new phono.

Anyway here's what I use and almost every component has been considered carefully or swapped at some stage. Unfortunately there are no proper substitutes for the Black Gates where they are used , though the 22uF could be swapped for an Ansar 400V PP . I didn't find LED's good enough for bias, on the occasion that I tried them over 10 years ago, but maybe that was a premature judgement and/or there may be a 'fix' .
I don't need very low Zout ( it's about 2.5k , I suppose ) as I use shortish interconnects and a 100k input at the power amp.

[thomas]

[
I was being half serious when I asked earlier about feedback, being pentode and all, the two have to go together don't they?!… like cheese and red wine...

No, you are probably thinking about power output stages. Feedback is not needed for voltage amplification stages (though can be used). In fact, a good way of thinking about pentodes is to consider the feedback inherent in a triode. Imagine a triode with a load resistor. Send the grid positive, the current through the valve increases, so the voltage on the anode load increases, so the anode voltage drops. . Now because the voltage on the anode drops, there is less voltage across the valve so current drops. So the action of the anode moving is to counter the action of the grid going positive. Now imagine if you could somehow shield the anode so the cathode didn’t see the voltage on it changing. Grid goes positive, current increase, anode moves down, but the valve still sees the voltage on the screen grid, so the feedback is not happening, so gain is much higher. Thats what the screen grid goes on a pentode, it hides the motion of the anode from the cathode.

For the same reason miller capacitance is reduced. Miller is the effect of the anode moving against the grid. Stop the grid seeing the anode move, and miller capacitance is removed. So the addition of the screen removes miller capacitance increasing the frequency the valve can operate at, and also increases the gain.

The downside is the anode feedback is why triodes have low anode resistance, so pentodes have high ra. But for voltage amp use, thats not a problem.

Try and understand the way a triode works, in terms of a field of electrons and how they "see" the anode as a attractive force, and how the grid reduces the attractiveness of the anode to the electrons by hiding it in electrostatic terms. Once you can picture this add another grid, and see how this changes things and how the new grid hides the anode from both the cathode and its electrons and the first grid.

I am not being insulting when I say, reading is important, but understanding is more important.

I've quoted this in full as its possibly the most comprehensible and succinct description of how a pentode works I think I'm almost getting it! But now it raises a lot more questions!
So small signal pentode voltage amplification is 'very good', but put a large voltage swing through one and current amplification is 'less good' and feedback is a good idea?

[izzy wizzy]

It's amazing how many of us have gone on a similar phono journey and ended up independantly with more or less with the same thing.

Mark, I have to ask, why battery bias the second stage and not the first? A 1.5V AAA does the trick in mine. Like you, I didn't like LED biasing no matter how many times I tried it.

I went for a pentode for a number of reasons. Hi gm valves have huge input capacitance which I don't think helps for an MC driving it via a stepup. I did some calcs once about HF resonant frequencies for the 103 in this kind of setup and IIRC, these hi gm input valves didn't look good. Wish I could remember how I did it now. A bit of peer review would be useful.

cheers,

Stephen