Changes Modifications to paramour IIs for 6J5 input tubes?

[Jim R.]

I recall that a number of folks converted their paramours and/or paramounts to use the 6j5 as an input tube  and wonder if anybody has the necessary changes to bias resistors, C4S changes, etc. that one would need to make in order to use this tube?

The first paramour IIs I heard had the 6j5s in them and they sounded quite good and still had plenty of gain for my system.

I just found a pair of teflon octal sockets today that I didn't know I had, and as long as I have to replace the tube socket on one amp, might as well just go directly to the 6j5.

Thanks,

Jim

[Paul Joppa]

Book standard 6J5 operating point is 250v, 9mA. That's not enough compliance on the C4S to be comfortable; I'd shoot for 200 volts at the plate, 6mA current - change R1 on the C4S board to 150 ohms.

Bias will be about 6.7v. The easy way is to use a 1100 ohm resistor, with a bypass cap of 270uF or greater. You might consider a 2000 ohm pot or trimpot, to set the plate voltage exactly.

I have an old 2A3 amp that uses the 6J5, and I liked it a lot (gathering dust for the last decade because it's series feed, and parallel feed just sounds better to me).

You could use the 431 chip shunt regulator for bias - there is unused space on the board. The board is an older version, and would take some archaeology to figure out which connections to use.

[Jim R.]

Paul,

Thanks much, sounds pretty straightforward.  Though I'm not really sure what you are saying about the 2k trim pot and exactly where that should go and/or be wired.

[Jim R.]

Ok, somehow was just thinking in terms of the ccs, but now understand that you are just talking about putting the pot between B+ and ground and with the wiper going to the plate connection.

Sorry for the brain fart -- had my tea and breakfast now :-).

-- Jim

[Paul Joppa]

Umm ... no, I was talking about making the cathode resistor variable instead of a fixed 1100 ohms. Sorry for the lack of clarity!

With a C4S plate load, the cathode bias resistor drops a fixed voltage - it no longer functions as a DC feedback. Since the bias is fixed, and the current is fixed, the plate voltage is set by the tube parameters, which can vary +/-20% or more between tubes and will vary as the tube ages. The most effective place to adjust the plate voltage is to vary the bias voltage. In the "soft-start" board we have finally made allowance for a variable resistance in the 431 shunt regulator circuit, so that it can be used as an adjustable Zener-equivalent. The advantage of this is that the shunt-reg chip has a very low impedance and need not be bypassed.

You can substitute a variable resistor for the adjustable Zener-equivalent, but to maintain low plate resistance you have to bypass it with a capacitor. Thinking further about this today, I calculated that if the 1100 ohms is not bypassed, the effective plate resistance feeding the 2A3 grid is a bit over 30K ohms. The 2A3 Miller capacitance at the grid is about 70pF, which gives -3dB at 75kHz. That's well beyond audibility, so I would leave off the bypass cap - it will probably sound better that way.

[Jim R.]

Ah, ok, now it all makes sense -- thank you!  And yes, I always try to keep the cathode circuit as simple as possible and with good parts as this seems one place where it counts the most, and if I can leave off a bypass cap, all the better.

What I don't know yet, is if this is as critical in a pf amp as it has been in the series feed amps I've worked on.

That's why I'll be doing some experimentation with these things once the amps are working.

I'm having a bottlehead weekend and just got back from the shop.  Tomorrow is painting day.

-- Jim

[Jim R.]

Paul,

Another question: -- what will this do to the plate current through the choke?  Will I still be ok with the bpc-15 50 mA?

And do you mean that the pot should be wired in parallel or series with the existing 1.1 k cathode resistor?

Thanks,

Jim

[Paul Joppa]

Paramour is cap coupled, so the driver cathode bias has no effect on the 2A3 current. (You knew that!)

I meant to replace the 1100 ohm cathode resistor in the driver stage with a 2K pot wired as a variable resistor. Then you can adjust the pot to get the desired plate voltage of 200 volts.

If you just use a 1100 ohm fixed resistor, I expect the plate voltage will be in the range of 150 to 250 volts. That still leaves plenty of headroom for the driver. My guess is that the plate voltage is not that critical. Running 6.5mA, the 6J5 saturates at 70 volts and the B+ is 320 volts in a Paramour II, so a swing of +/-60v will still leave clearance, positive and negative.

The more I think about it, the more inclined I would be to use a fixed resistor and match the tubes. That way both channels have closer to the same driver impedance, which might help keep the imaging stable. Plus, you can get a really good resistor - without the bypass cap, this is probably an important one for sonics.

Sorry for the moving target. As Dan says, sometimes you have to shoot the engineer and pry the design out of his cold dead hands, because otherwise they will just keep tinkering with it.  :^)

[Jim R.]

Paul,

Ok, I knew that, but thanks for reminding me :-).  Seriously, being a global learner has it's ups and downs -- when learning a new subject, or re-learning from a very different perspective, knowledge is like this giant amorphous cloud of factoids until one day some tiny little thing is realized and then the whole thing crystalizes and often in most dramatic fashion.  Still trying to find that little nugget on tube amp design that will make me seem a bit less stupid :-).  Doesn't help either thant I'm a visual learner and well... :-)  I'll get there!  And thanks for being patient with me in the mean time -- it's very much appreciated.

Getting tubes matched is no problem either.  Just get them by the lot and my buddy has a nice tester that can match things very nicely.

-- Jim