EF86 Phono pre-amp or Mcintosh MC2200?

[TV7]

Hi all, does anyone have any experience with EF86 phono pre-amps? what do they sound like, should I build the leak point 1 clone?

I have built several before around the 12AX7, mainly because the Mcintosh C22 and Marantz 7C both just use those tubes, however I have several NOS Phillips EF86 and am keen to find a use for them. (The gold pin variety)

Here is a schematic pilfered from the Mcintosh MC2200 preamp, giving a tidy and simple phono pre-amp with 12AX7 and 12AT7 tubes.

It seems RIAA feedback is via the 12AT7 cathode, I was thinking of eliminating one side of this tube per channel to simplify things, one big question is how will this affect the feedback? Anyone have any ideas?

If successful a nice phono pre made with only 3 tubes.

Open to any and all suggestions, including biasing for 3 tubes....

Which way should I go? thanks.

[SY]

As a former owner of a Leak Point 1, I can suggest that nearly anything would be an improvement. Noisy, poor headroom, poor RIAA conformance. I ended up gutting it and building a decent phono stage into the chassis.

[TV7]

Leak is out of the question then. Is there any hope for the EF86, or are they destined to live in the spare parts box for ever.....

[SY]

I think a decent EF86 preamp could be designed and built- but not with one tube per channel!

[TV7]

Thanks for the help, Ive decided to stick with the basic Mcintosh circuit as shown above, nice and simple, no input transformers.... That schematic was from the C2200, looks to use feedback via the 12AT7 cathode. My only problem with this is, if the input impedance varies between power amps, would this cause the feedback to change? I guess this is why they use both sections of the 12AT7 to give it plenty of clout. Otherwise I could use a variation of the C22 commemorative edition, where the feedback is via the anode of the second section of the first 12AX7, this makes plenty of sense to me to keep the cathode follower output independent. Having said this I have only ever had the good fortune to listen to a C2200 and not the C22 through my MC275 so who knows what the sound difference is like. I should have a C22 commemorative edition sitting next to the MC275 but ebay scoundrels took me to the cleaners for the sum of $3500 years ago. But thats another story....

[Eli Duttman]

I agree with SY that the EF86 could be very useful in a phono preamp. IMO, the "gotcha" is noise factor. So, don't use the tube at the I/P. A 6922 section at the I/P will provide enough low noise gain to feed a passive EQ network and overcome the EF86's noise limitations. The pentode will provide the bulk of the circuitry's overall gain. A buffer between the pentode and the O/P is necessary. For that job, I "nominate" a DC coupled ZV0545A source follower.

[TV7]

The 6DJ8 seems to be getting popular, I have a few of these also lying around that I thought I'd never use..... Better than a 12Ax7 I/P by any definition? Maybe I'll revise my ideas. Thanks.

[Eli Duttman]

That should have been ZVN0545A for the buffer.

The 6DJ8 family is quieter and exhibits lower C_Miller than the 12AX7. All well and good, but it's not a slam dunk, as stage gain is "compromised" by a μ of 33. You have EF86s in the "stash", which are more than capable of making up the difference. IMO, a combination of a 6DJ8 family tube section in the 1st gain block and an EF86 in the 2nd gain block will be highly synergistic. Watch out for parasitic oscillation.

[kevinkr]

A cascode connected input stage using 6DJ8 can work extremely well, sound good, and provide lots of gain. (Up to ~40dB or more is possible)

Enough gain can be realized in a 2 stage passively EQ'd phono stage to make direct use with medium output moving coil cartridges practical, although depending on operating point they may not tolerate high output vintage MM types well. (Say much > 5mV @5cm) All mostly just a matter of knowing what the design goal is about..

[TV7]

All very interesting, looks like I'll have to bread board something together and see what the performance is before I go ahead and build one from scratch. I suppose input transformers would get around the voltage gain problem of a 6dj8. It might even then be viable to use a 12AT7 cathode buffer after that as a nice combination instead of the EF86, but I'll have to play......of course finding the balance between performance and price is the restraining factor.