L'Amp: A simple SIT Amp

Member
Joined 2009
Paid Member
ha, ha, thanks for the applause.....

but of cause I do not know why it is as it is.....:p


and I still do not know why Papa puts so many caps near to the Vishay resistors in SIT-1...... nobody any idea?

maybe he loves caps of tea?:D
 
Last edited:
Official Court Jester
Joined 2003
Paid Member
and I still do not know why Papa puts so many caps near to the Vishay resistors in SIT-1...... nobody any idea?

maybe he loves caps of tea?:D

convenient place for mucho PSU caps (and possibly just few output ones) - sounds logicall ?

2 reasons for that - poor PSRR of circuit itself , empty pcb if you leave just resistors there

edited - few important words swallowed , probably .... usual ZM's brainfart
 
Last edited:
Member
Joined 2009
Paid Member
yes Zen Mod and ra7,

that is the most logical and simplest answer.....:)

and I do not see CRC in the PSU so more caps are useful....!

but my intention is not to discuss Papas SIT here primarily, I was more interested if the picture gives us an advice to "better" our thick film resistor solution......sounding not so dark....I repeat, I understand now the people doing "Bulb Rolling" on the search of the right sound mixture....:idea:


but I confess...... I am always looking at the pictures of Papas Amps at 6moons like a junkie....on the search of hidden riddles and treasures....I confess...:nownow:
 
Last edited:
Yes, the 47K. I've seen SITs with 100uA or more gate leakage current. With 100uA, you'd be dropping 4.7V across a 47K resistor. It's worth a look anyway. Being aware of it is half the solution.
I did check with some and yes the one with 11.5 vgs had a high leakage that accelerated once the sit reached a certain temp. With a 12k there was almost none.However the variation, testing with the 12 k now, is still very wide as I described so no real change there although it seems now the two sits with the highest vgs are more closely matched now with the 12 k. So out of 8 sits I have about three pairs that seem reasonably matched.
 
Pass DIY Apprentice
Joined 2001
Paid Member
Good work, protos.

It's very temperature dependent, so you may find things will behave a little better still once they're gooped-up and bolted to a big heatsink. You'll have to use the negative bias supply version to generate the correct bias voltage. You could also do as Generg does and use a battery and pot for testing. You just have to be careful the battery doesn't drain and cause a meltdown.
 

ra7

Member
Joined 2009
Paid Member
People experienced in reading Papa's schematics and PCB layouts will note that he likes to put some filtering capacitance near the main circuit. It probably also helps with transient demands. Of course, this could be wrong and the cap might be serving some other purpose in the SIT circuit.
 
Drat:mad:! My first signal tests failed.
I am using neg bias supply for a Vds of around 18V and my lab 30V ps configured as a 2amp ccs. The bias (around 5v) Vds 18v and everything seems to be functioning as expected in steady state. However I was only able to get something like .4V AC rms on a test speaker at 1khz before gross distortion. I have input and output cap as expected and I tested even with music which sounds normal on the speaker at a low level of course. I have not put the scope in yet to actually check exact level of clipping but still it is obvious something is off.
 
Member
Joined 2009
Paid Member
People experienced in reading Papa's schematics and PCB layouts will note that he likes to put some filtering capacitance near the main circuit. It probably also helps with transient demands. Of course, this could be wrong and the cap might be serving some other purpose in the SIT circuit.


Thank you all for your reflections!

In the SIT user manual papa says that 8 resistors bias the circuit, so I suppose these 8 resistors (each 100W) form the drain resistor. 800W for the drain resistor, not bad, really overkill, he keeps them in a safe area....:D

Not strained resistors seem to sound better....
 
Member
Joined 2009
Paid Member
my last try in the CCS....

you see in the picture that I am a bit old.....

I do not why, but with the 500R pot I manage to get a good mixture of the K2 and k3. Either in series with the cap or parallel to the resistors like now...

without the pot it was much more difficult to get the right mix only by the SIT bias pot and the SS bias pot at the recommended values....

I hope I write no nonsense.....

Of course I would be happy if anybody on his journey could test the 500R pot idea too and possibly confirm the use of it.:D
 

Attachments

  • SIT MU2 b.jpg
    SIT MU2 b.jpg
    33.9 KB · Views: 612
diyAudio Chief Moderator
Joined 2002
Paid Member
Drat:mad:! My first signal tests failed.
I am using neg bias supply for a Vds of around 18V and my lab 30V ps configured as a 2amp ccs. The bias (around 5v) Vds 18v and everything seems to be functioning as expected in steady state. However I was only able to get something like .4V AC rms on a test speaker at 1khz before gross distortion. I have input and output cap as expected and I tested even with music which sounds normal on the speaker at a low level of course. I have not put the scope in yet to actually check exact level of clipping but still it is obvious something is off.

Is it a real ccs or they just mean the adj. current limit on the lab supply?
 
Official Court Jester
Joined 2003
Paid Member
the idea to handle the pot in this way was the idea of young German Ralph.....:D:D

previous variation (10k pot/variable resistor along with 10k resistor to bias string) gave you ~6db less of max modulation , and softer drive

this variation is giving you broader range , and firmer drive of Mu gate

what's more to you liking , let ears decide
 
Yes it is not a "real" ccs. But for simplicity's sake I am using the bench PS in place of the ccs in the schematic. Basically it is a 30v 2amp ps with adj voltage and current. I max the voltage adjustment so it can swing to 30V and of course limit the current to just under 2amps. At steady state it is settling at about 18v which is the Vds of the sit. So theoretically it behaves like a ccs keeping current constant but swinging the voltage as needed to keep the current at 2A.
I should get at least 7v ac rms swing with this set up I think.
 
Last edited: