The perfect phono preamp.

[Brian Beck]

There will always be tradeoffs. I swear these RIAA stages are best left to philosipher's and theologian's.

Yes, RIAA stages are surely among the toughest audio design challenges. Hum, hiss, EQ distribution, overload margin, gain distribution, output Z, input Z. It's hard to balance everything. I agree that noise is a priority.

 

[Geek]

Heya,

I do have some qq regarding the decoupling of the Ck of the WCF and why a 10K tail end resistor is being used?

Personally, I'd ommit the decoupling cap, up the Rg for that stage to 1M and use a smaller output cap with a 100K tail instead. Unless I'm missing out on something fundamental perhaps?

The pre probably has enough output to drive an amp straight in, ZOut sure won't be a problem if it's a tube amp we're driving, so somewhere I'd like to see a volctrl. Probably best put right at the WCF's input, right?

Yeah, I forgot to explain this.....

The pre was actually designed to go into my "Ultimate Preamp" project. The mWCF is not really going to be DC coupled to the stage. There will be a cap and the selector switch and volume control in there. I just left it in to do the noise and THD tests as accurate as possible. (haven't settled on the line-gain tube yet, 6N6P or 12B4. Though I may DC couple that into the mWCF after all...)

For a standalone phonopreamp, just leave the 120R||100uF cap out completely.

The 10K tail end resistor's sole purpose is to bleed the static charge off the 4.7uF polycap.

Cheers!

 

[Burnedfingers]

Very interesting design. I think I will build it if I can scare up some 6AQ4 tubes.

Are you going to post your power supply design also? Should the power supply be built on another chassis?

 

[Geek]

I'm using a regulated bench supply now, but will be using my regulator circuit found here, just values changed for higher B+ and current to allow the decoupling resistors and caps: http://geek.scorpiorising.ca/Geek-0.html

You can build the supply on another chassis, or just make the cabinet big enough. I like the pre large enough so the TT can sit on it, so it'll be the physical size of a 3U rackmount, like my last one with nice wood covers

6AQ4's can be had off Ebay. I got my 10 for ~$15, NOS Mullard's.

 

[wrenchone]

Attached is a gain block I'm thinking of trying for a passive RIAA using the 6AM4. Constructive comments are welcome

 

[EC8010]

Hmmm. Interesting circuit. I assume that you've gone for the added complication of a folded cascode in order to get the DC voltages more nearly optimum for the cathode follower? 2.5mA seems rather a low current for the two input devices. In theory, driving a bipolar transistor with a current is a much better idea for linearity than driving it with a voltage, so your hybrid cascode should work well. However, there's nothing like a real measurement, and the exact distribution of the harmonics in the distortion will be important.

 

[Geek]

Hey SY,

I tried your idea, but a simplified version. Works cool

Don't want to retype everything, so here's the new schema and serious measurements:
http://geek.scorpiorising.ca/GeeK_ZonE/index.php?topic=2694.msg31339#msg31339

 

[wrenchone]

Regarding the folded cascode circuit I showed a couple of posts ago - lying in bed this morning, I realized that you can't close the loop as I've shown it because it's positive feedback (Doh!). Without the feedback, there's a lot of excess gain. I could keep the current arrangement sans feedback, beef up the top current source, and use a smaller resistor as load for the PNP folded cascode transistor. This would tone down the gain to sensible levels, but would require transistors with higher dissipation capability for the top current source cascode and the folded cascode transistor. I'm thinking a 2SA1535A would work ok. A lower noise JFET for the top current source is in order, too, as the noise gets passed right down into the folded cascode load resistor. Let's sling a 2SK170BL or V in there instead of a PN4303.

As for current through the input tube, I was going on earlier comments in this thread about the advisability for fairly low current in the tube. One could always run a couple more mA thropugh the 6AM8. At 100V plate voltage, dissipation is not a big problem.

Since the output folded cascode now will have a much lower value load resistor (I'm thinking ~ 5K), the possibility exists for just ditching the output cathode follower and driving the RIAA network directly off the folded cascode transistor. The cathode follower can be saved for the output gain block. With this type of arrangement, I could see a tube complement of (3) 6AM4, or maybe one 6AM4 at the input and a dual triode like a 5687 or 5965 for the output stage. Being contrary by nature, I would resist suggestions to use a 6DJ8 or 6922 as being too popular - this is all about lowly TV tubes, after all. Maybe a 6BQ7 would fill the bill.

 

[wrenchone]

A remark in passing - one of the advantages I see in the folded cascode approach is the relative freedom you have in biasing the tube, since tube bias current and the current in the load resistor are separate, while the current source supplies both. I'm mining this kind of approach for some sand state phono preamps, too. I'll find out sooner or later whether this approach sounds decent with tubes.

To Tubelab - Thanks for the heads up about the 6LU8 and 6LR8 triode/pentode vertical deflection tubes. I picked up a quad of the 6LR8s for about $2.50 apiece, and they are interesting looking in a "tubby the tube" sort of way. I never saw one of the later model tube TV sets, but a chassis full of compactrons must have been something strange to behold. I'll be salting the 6LR8s away for a rainy day project. Getting the little chubbettes to sing in SE ultralinear mode would be kinda neat.

 

[Tubelab_com]

Being contrary by nature, I would resist suggestions to use a 6DJ8 or 6922 as being too popular - this is all about lowly TV tubes, after all. Maybe a 6BQ7 would fill the bill.

I have used the 6BQ7 6BK7 tube as a LTP phase splitter in the input stage of a couple of amps. It works great there. I don't know if it is quiet enough for a phono preamp though.

To Tubelab - Thanks for the heads up about the 6LU8 and 6LR8 triode/pentode vertical deflection tubes.

I worked in a TV repair shop in 1969 - 1971. Philco TV's used them for vertical output in the 21 and 24 inch color TV's. I remembered them when I found some in my TV tube collection. I have a bunch of the 21 volt versions, which are sort of useless, so I "tested" a couple. I wired the pentode section of one into a pentode wired SE guitar amp and cranked on it pretty hard. With 325 volts on the plate and screen, the tube just rocked, no glow, no tosted parts. It sounded like a 6V6 in this application. I think that they would work good in UL SE, but I have not tried it yet. On my list of things to try. Looks like a good match for the $18 Edcor transformer for a low cost amp.

There was a P-P amp based on these tubes described in the following thread. I think that more power would have been available with more screen voltage, although it might not have sounded as good.

http://www.diyaudio.com/forums/showthread.php?s=&threadid=60346&highlight=6lu8+6lr8

 

[wrenchone]

I think the 6BQ7 would be quiet enough for the second stage and output buffer of a phono preamp. I'm still looking at using the 6AM8 for front end duty. BTW, the 6J4 is another one of those RF triodes that looks interesting.

 

[lndm]

How would a 417A do as the upper device in a cascode?

Is there any point to using one as the lower device? Would that be a poor choice?

 

[EC8010]

How would a 417A do as the upper device in a cascode?

It would do nicely as the lower device. A 6J5 would be fine up top.

 

[lndm]

Ta.

Now I realise (brain fart) I meant to ask, would there be any benefit to one as the top device? (though you may have already answered that )

 

[EC8010]

... would there be any benefit to one as the top device?

Not really, although it's common to use the same device top and bottom. The lower valve needs lots of mutual conductance (and possibly additional current to achieve it), whereas the upper valve needs linearity at large voltage swings.

 

[lndm]

The 6J5 would be a good choice for linearity. I have been led to believe that I should run a 417A at 25mA. I could run it at less but don't want to defeat gm (unless there is another good point at which to operate it). I could use both halves of a 6SN7 or something else altogether. Any idea where I should go from here?

 

[EC8010]

Mutual conductance is always proportional to anode current. I'd be a bit wary of running a 417A at 25mA, I know it's within ratings, but you can expect a shortened life. As you hint, even 15mA would be too much for a 6J5, so perhaps you might want to look at some of the Russian valves (6H30?).

 

[Brian Beck]

Another option would be to run the 417A somewhere in the range of 15mA to 20mA, achieving a pretty decent gm, and then use both sections of a 6SN7 in parallel up top. Each 6SN7 section should be quite happy at 8 or 9 mA. I've done something similar to this in a driver circuit before.

 

[lndm]

I'll be glad to bring the current down. Let's say I went with the 6SN7 (I conveniently have them). At first looks with 150V across the 417A, and I want 180V on the 6SN7 and with a 20k plate load, I think I'll need a 650V supply.

That's more that 10W from a phono stage that'll be a record for me.

 

[Tubelab_com]

Since my wife is out of town, I have dragged several amplifiers into the living foom for some loud listening. My 300Beast doesn't work any more . It looks like one of the power supply caps has shorted. No big deal since I have 6 working tube amps.

The big problem is that my phono stage is also dead. It has been in the long slow death spiral for a long time. I haven't used it in a while, and now one channel is all hiss, and the other goes snap, crackle, pop. It is no longer usable. It was a crappy 12AX7 design anyway. I could replace all of the cheap resistors, and get it going again, but I think its time for a new one. Which one?

Very early in this thread SY talked about a hybrid design that used mystery tubes and DN2540's. Well I bought a bunch of DN2540's. Probably got the Mystery Tubes too, i just don't know it yet. Any progress on your design? I should have next week off of work, and Sherri is out of town. This means I work on the hot rod during the day, and tube stuff by night.

Schematic?