Bottlehead C4S as a differential pair constant current sink?

[Paul Joppa]

Ah! I get it! Thanks, Mikey.

You can remove the current source triode, the 401K resistor, the two Zeners, the 499K resistor, the cap (is it 47uF or .47uF?) - and the triode. Replace them all with a C4S. You can keep the triode and replace the MJE340 with it if you prefer, grid is base, cathode is emitter, and plate is collector.

[Mikey]

Hi Paul,

Thank you for the help!
I'm glad you were able to open the second attachment...

It seems as though I've got one last decision to make:

- Should I use the MJE340, or use the existing vacuum tube triode (in this case, a 6SN7).

The rework to the existing circuit board will be about the same in either case so that's not an issue.
However, if at a later date I decided to really 'ramp up' the current through this circuit it seems the vacuum
tube might hold an advantage over the transistor in terms of dissipation.  From the 6SN7GTA datasheet,
each plate can dissipate 3.75 watts (when both are operating).  That seems like it might be better than
a scathing hot transisitor with a HUGE heatsink!

So my questions are:

- In your opinion, does one option hold a technical advantage over the other?
- Would you expect one option to sound better than the other?

Mike

[Paul Joppa]

In my opinion there is not going to be any noticeable sonic difference; the device is largely controlled by the small transistor. But the tube will be more robust - less chance of blowing something if anything goes wrong - and it will get a heat source up on top and out of the chassis insides, which is always a good thing. Downsides are of course heater power and real estate.

[Mikey]

Hi Paul,

I will try the 'hybrid' C4S first, using the vacuum tube in place of the MJE340.

(https://forum.bottlehead.com/proxy.php?request=http%3A%2F%2Fi3.photobucket.com%2Falbums%2Fy80%2Fmpaschetto%2FAtma-Sphere%2520MP-1%2FDSC07452.jpg&hash=5eab5b2852ffb3a44c66c1f51d56d448432c9bd1)

You can see the existing 6SN7 constant current sink in the above photo.
I'll splice in nearby, and leave the C4S board supported by the wires for
the time being....photos to follow!

Mike

[Mikey]

Hi Paul,

Well, at the last minute, I decided to install the C4S 'as is',
without utilizing the existing triode in place of the MJE340.

(https://forum.bottlehead.com/proxy.php?request=http%3A%2F%2Fi3.photobucket.com%2Falbums%2Fy80%2Fmpaschetto%2FAtma-Sphere%2520MP-1%2FDSC07508.jpg&hash=5defbfef5241ccceaad3fbfbe8df8c6b8e479b4d)

One channel is up and running with the C4S sink, I'll convert the
second channel tonight....

Mike

[Grainger49]

Mike, is that battery on a cathode?

[Mikey]

Mike, is that battery on a cathode?

Hi Grainger,

The input stage of this phono preamp is a cascode, and I replaced the bottom
tube (2X 12AT7's in parallel) with a MAT12 transistor.  With the transistor in
the circuit, I can use the battery to bias the top tube in the cascode...

Mike

[Mikey]

I retrofitted the second channel last night, so I will
be able to give the new setup a listen this weekend.

Stay tuned...

Mike

[Grainger49]

Hi Grainger,

The input stage of this phono preamp is a cascode, and I replaced the bottom
tube (2X 12AT7's in parallel) with a MAT12 transistor.  With the transistor in
the circuit, I can use the battery to bias the top tube in the cascode...

Mike

Got'cha!  There was a mod of the Paramour where you biased the driver, changed to a 76, with a 9V rechargeable.  Once it was charged the circuit kept it charged.

Should this be quieter?  

[Mikey]

Should this be quieter? 

Grainger,

I honestly don't know, it was the lazy way out....
This way, I didn't have to construct a voltage divider.

Mike

[Mikey]

In my opinion there is not going to be any noticeable sonic difference; the device is largely controlled by the small transistor. But the tube will be more robust - less chance of blowing something if anything goes wrong - and it will get a heat source up on top and out of the chassis insides, which is always a good thing. Downsides are of course heater power and real estate.

Hey PJ,

One additional thought:

Could an IRF820 (see attachment) be substituted for the MJE340 without any issues?
It looks like a pretty beefy transistor, and I happen to have a few of them on hand...

Mike

[Paul Joppa]

...
Could an IRF820 (see attachment) be substituted for the MJE340 without any issues?
It looks like a pretty beefy transistor,...

It's not a transistor, it's  FET - MOSFET to be specific. It's enhancement more so it needs several volts of positive bias, more than can be supplied by an LED. I don't know to what extend the cascode arrangement will reduce the potential for distortion from the  high and nonlinear capacitances. We never experimented with FETs, though I know Tucker and the late John "Buddha" Camille tried several circuits a decade or more ago and rejected them as inferior sounding.

[Mikey]

Hi Paul,

I got some time on the newly C4S'ed preamp, and I liked what I heard!  In fact, I liked it enough
to tackle the next iteration...

In a previous post, I mentioned the desire to try a 'hybrid' C4S, using the existing 6SN7 triodes in
place of the MJE340s.  Well, I perfomed the modification this evening, and it works like a charm!

Here's a photo:

(https://forum.bottlehead.com/proxy.php?request=http%3A%2F%2Fi3.photobucket.com%2Falbums%2Fy80%2Fmpaschetto%2FAtma-Sphere%2520MP-1%2FDSC07513.jpg&hash=95650386a4a39d467580b1df9641d9bb67ab47de)

The swap required a bunch of 'cut-and-splice' to the circuit board, but it was worth it.  Now the
vacuum tube can dissipate the heat when I bump up the current thru the circuit.  Although the
C4S board is dangling from the wires at the moment, I'll be able to mount it on the underside of
the existing circuit board...without the MJE340's, it is a very low profile assembly!

Thanks again for all of your 'hand holding' throughout this project, I couldn't have done it without
your assistance!

Mike

[Mikey]

Hi PJ,

I wanted to wrap up this thread with a few photos of the completed installation,
and also ask one additional question...

I mounted the C4S boards to the underside of the existing circuit boards.  To bolt
them in place, I drilled a couple of holes in the existing board, and used 1/4" long
nylon spacers with some #4-40 hardware.  Jumper wires were made to splice the
C4S into the existing curcuit:

(https://forum.bottlehead.com/proxy.php?request=http%3A%2F%2Fi3.photobucket.com%2Falbums%2Fy80%2Fmpaschetto%2FAtma-Sphere%2520MP-1%2FDSC07528.jpg&hash=6ea878da969b27c818b8b5da93b020a9228c576e)

(https://forum.bottlehead.com/proxy.php?request=http%3A%2F%2Fi3.photobucket.com%2Falbums%2Fy80%2Fmpaschetto%2FAtma-Sphere%2520MP-1%2FDSC07525.jpg&hash=0083f5d03c58117fa579f4765a14b841c2da4f2f)

(https://forum.bottlehead.com/proxy.php?request=http%3A%2F%2Fi3.photobucket.com%2Falbums%2Fy80%2Fmpaschetto%2FAtma-Sphere%2520MP-1%2FDSC07529.jpg&hash=f1750050a11ae5518c6cc8b4cd5cade031164eea)

The C4S certainly helped reject any last vestige of power supply noise from this
circuit....al that remains is a bit of tube rush.


With the 6SN7 in place of the MJE340, I think I'm going to be limited in the amount
of current I can draw through this arrangement.  With both triodes (of the 6SN7) in
use, the RCA data sheet specifies the maximum dissipation of the tube is 7.5 watts,
or 3.75 watts per triode.  By my reckoning, that equates to a current of 15mA
through the C4S.

Does that sound about right?  Is 15mA the upper limit of this scheme?

Mike

[Paul Joppa]

With a 250v drop, then yes, 15mA is your maximum without over-heating the 6SN7. You can draw up to 20mA, the 6SN7 maximum rating, if the low voltage is reduced to maintain the maximum heating.

To be careful, check the 6SN7 bias voltage and compute the dissipation in the small transistor. There are clip-on heat sinks for those tiny metal cans, and I recommend them if you exceed 1/2 of the free-air dissipation limit.