Pass mosfet citation 12 clone

[Tekko]

I just got my mosfet citation 12 clone working after some trouble with dc offet prolly caused by the wrong transistors.

2N3904/6 were the closest i had to the MPSL01/51, hfe pretty much the same.

Here are a few pics:

And no that opamp is not part of the circuit, i just dident bother with desoldering it before using the board.

 

[Frank Berry]

I've got the parts to convert my Citation 12 DeLuxe to the MOSFET Citation 12 but haven't done the conversion yet. I may do as you have done and build the MOSFET version and leave the old Citation 12 as-is.

 

[Tekko]

Theres only one problem, when turned off and trhe caps drain and the voltage drops low enuf, this thing bottoms out the speaker with a loud "KASCHUNK" as the voice coil hits the back of the pole piece.

 

[Frank Berry]

You may need to add a relay to open the connection to the speakers when the amplifier is powered down.

 

[Frank Berry]

I just realized. The amplifier might be going into oscillation as it powers down.
Have you looked at the output with a scope while the amp is shutting down?

 

[Tekko]

Yes i have, it gos to the positive rail with a resulting bang from the speaker.

Schematic:

Heres a vid of this amplifier in action: YouTube - MosFet Citation 12 by Nelson Pass

 

[Tekko]

I have a feelin that Q4 should be attached to the heatsink, bias current is slowly creeping up, i st it to 130mA and now a few songs later its up at 180mA and climbing.

I just need to dig out some ribbon cable and stick it to the heatsink to counter thermal runaway.

 

[Frank Berry]

Probably correct. Let us know the results after you mount Q4 to the heatsink.

 

[Tekko]

Q4 mounted to heatsink, quiescent current lingering around 110mA and doesent seem to be rising, atleast not yet.

 

[Tekko]

There is another thing i am considering, replacing the 18k resistor on the input stage with a current source. Not sure if that would make a difference tho.

If it posed a significant difference in performance, i think pass would have included it in shcematic, unless he just did the mods the board allowed for.

What do you guys think, would a current source in place of the 18k resistor do any significant difference or would it just be a waste of time ?

I don't have the equipment to measure the difference if any, but i may try it just for fun.

 

[juma]

There are benefits in using a CCS to bias the input LTP, but some people prefer the resistor.
Your ears will give you the answer...

 

[Tekko]

So a ccs is on the list, its easy, just a transistor, a led and two resistors.

 

[Tekko]

Heres my proposal:

An externally hosted image should be here but it was not working when we last tested it.

I've also ordered 4 of these fets to test out:
http://jm.plantefeve.pagesperso-orange.fr/acd100nsd.pdf

 

[Tekko]

Adding that current source was a bad idea, it worked for a while, until i turned up the input signal, then the bias jumped up to 500mA and a few seconds later it blew up, so back to the good old resistor.

 

[juma]

If your PS voltage is +/-35V (as the sch. shows) then you have a bit less than 1mA through each LTP transistor when you use 18k to supply the LTP.
With CCS (as you drew it in post #13 - red LED and 330R) there will be about 1.5mA through each transistor in LTP and that changes everything. Put a trim-pot across the R8 (330R in your sch.) in order to adjust the current through LTP and that will settle working points in the amp.

 

[Tekko]

Im not gonna mess with it anymore, the 18k resistor is back and its working flawlessly, had it up to clipping with no prollem into a 3.9ohm dummyload.

 

[juma]

Im not gonna mess with it anymore, ...

That's a good strategy: run away quickly from everything that doesn't work first time !

 

[Tekko]

There prolly is a reason why pass dident use current sources.

This thing was gonna be a bass amplifier anyways so it doesent have to be perfect.

 

[Nelson Pass]

I didn't use current sources because I wanted to minimize
the changes to the original, making it easier to do the
conversion.

In general, there's no reason that current sources wouldn't
work fine.

 

[Tekko]

The question is why it failed on medium volume into a 4ohm speaker with the current source but survived up to clipping into a 1.95ohm load with the 18k resistor.

Now i gonna wait till the laterals arrive and test those out.