It's time I built a D3a -> 5687 phonostage

[Thermionic Idler]

Forget the number but there is a Russian high GM pentode that works as well. But I remember measuring some and finding two sets of curves as if there were two construction types.

Yeah, the 6J52P - I think I saw a post from Kevin of K&K Audio mentioning those, that he also found them all over the place, and there was another blog post I came across (forget where I saw it now) which showed an appalling 'spread' of measurements. Seems it's best avoided.

Well, theoretically I do have two NOS D3a's (even if they are in transit between Italy and here) and that's all I need cos it's 5687's everywhere else. And they are actually still reasonably accessible.

[Nick]

which showed an appalling 'spread' of measurements.

Testing them to give the pentode curves I found that there were two clear groupings of curves, I suspect if you measured these at a single point it might look like a spread.

IMHO, wide spread of measurements can be handled either by:

1. Design that copes with that spread (though that can go against the tube dogma of simplicity)
2. They are cheap and available in numbers were you can select matched sets (jfets fall into this category).

I feel that as the availability of a lot of these things tends to zero many of the traditional choices and designs will become impractical. When I was working on the ref design I originally used half of the 12ay7 as the cathode follower that drove the g2 of the d3a, but I found that only the EH gold ones were quiet enough to use that way, and then I found EH were stopping making them (I bought a box of them direct from EH, just because I found they were quieter even though the final design didn't need them).

To join this to a current PITA, a lot of modern parts are getting hard or impossible to find. I have stopped making the turntable supply now because LM3886's are unobtainable for at least the next year or so (if not forever), and can't be arsed redesigning to use something else. I have just bought 2k of a mosfet that is end of life that I use for the two box power supply. May buy some more while I can but I will eventually retire so such is life.

[IslandPink]

I just went checking on EBay, and it looks like D3a, C3g, C3m, C3o are all at least £40 each, or a lot more if you want to pay more. E280F's are not far behind. 6J52P's are available in big batches for as little as £1 each.

[Cressy Snr]

EF184 maybe?
https://www.langrex.co.uk/?s=ef184&post_type=product

https://frank.pocnet.net/sheets/010/e/EF184.pdf

[Nick]

EF184 maybe?
https://www.langrex.co.uk/?s=ef184&post_type=product

https://frank.pocnet.net/sheets/010/e/EF184.pdf

Its a long way off gm of 15ma/v as opposed to 35ma/v

[izzy wizzy]

IIRC, Nick's circuit is able to cope with a spread in the curves within reason which I think he is suggesting in some of his posts. That might be fine for the d3a.

I've accepted the current valve situation for my power amps. Many will use, say more exotic (now) unobtanium/expensice DHTs like 45/46 etc which I started with. I chose Russian/Ukranian DHPs and bought loads - 4P1L, 2P29L and select.

We might be just living in accelerated times for parts redundancy, have to accept it and design accordingly. Hold your own stock going forward.

[Thermionic Idler]

I feel that as the availability of a lot of these things tends to zero many of the traditional choices and designs will become impractical.

This is why I'm still periodically getting cold feet over this - I don't want to put a whole heap of effort and expense into something that I can't use after 5 years cos I can't get the damned valves for it. I'm starting to ponder how close one could get using discrete solid state + LCR instead of valves, I honestly don't care what tech it uses. Maybe that would open up the possibility of a 50R LCR network?

[Max N]

As someone already posted, d3a are rated 10,000 hours, so you wouldn’t need many to keep you going for a good few years. You’d need to get your cartridge re-tipped more often than replacing the tubes?

[IslandPink]

NOS premium E810F's are still at reasonable money, £30 each. Ideal for the front-end.
Otherwise, I still think there are loads of quality loctals around for £5 each which are really good. 7A4's for instance.

[Cressy Snr]

Its a long way off gm of 15ma/v as opposed to 35ma/v

I’ll get me coat.

[Thermionic Idler]

Not discounting any design options at this stage (cos I still effectively have a clean sheet) I'm currently studying a Linear Audio article I purchased a couple of years ago with a current input design for MC cartridges, where you get around the need for a step-up transformer by taking advantage of the MC cartridge's low impedance and applying almost a dead short to use it as a current source, rather than just giving it shit-tons of voltage amplification.

I have to say I like the concept - it seems very elegant to me. And it's a bit different.

I'm willing to bet nobody has ever tried partnering that with a following LCR stage - the article in question uses standard passive RC.

I wonder what that would be like.

[izzy wizzy]

Depends on the cartridge. It's not a totally universal solution and only for MCs. A bunch of commercial stages do this.

[Thermionic Idler]

Depends on the cartridge. It's not a totally universal solution and only for MCs. A bunch of commercial stages do this.

I think my current cartridge is about 16 ohms coil impedance, but the one I'm buying this year has a coil impedance of about 3 ohms. Seems ideal to me. Naturally I'm doing lots of reading...

[Thermionic Idler]

Okay, so after much pondering I've decided to take a chance and go down the current-input route with the Proteus circuit (Erno Borbely and Sigurd Ruschkowski, Linear Audio Vol 6, Sept 2013), but with LCR RIAA lovelyness in the middle.

From the research I've done, it has a very good chance of working well with my next cartridge (and may even work with the VdH). The first stage will drive a <600R load straight out of the gate, so the Sowter LCR will slot right in (or any of the other 600R off-the-shelf ones, for that matter). On a practical level, it doesn't call for unobtanium parts (granted LSK170's aren't cheap but at least they only have to be bought once).

Further practical considerations, I'm aware this thing will need a fair degree of bench testing, more so than my power amps, and solid state voltage levels mean I won't be dancing with death every time I do that. Things like the power supply become much less of a headache. Overall, I think that this is something I can handle. I decided I was in just a bit too deep with the D3a / 5687 and the supply situation for the former was also a factor in my decision.

There is of course no guarantee that the current input stage will work in practice, so my plan is to leave space for a more conventional voltage amplification stage and MC step-up transformers, which can interface to the same LCR and back-end, so I won't have to rebuild the whole thing if there's a disaster.

Sadly I don't think I can share the circuit because it's in a paid issue of Linear Audio and therefore subject to copyright. So I will have to be silent on the trials and tribulations of Spice modelling and PCB designing, but there should be a new thread coming when I start the actual build (probably some months away).

[Nick]

That's rather adventurous. The paradise was a bjt version of the same idea and certainly sounds good. I think PCB layout will be a big part of the work. I do note that the article suggests that the first stage can drive <600 but that's far from saying that it should. They then load it with a 20k network. But I guess that's partly driven by cap size limits they apply.