USSA-5 Build with Review

Taccodude,

37c is not even close enough for proper bias unless you have active cooling. I can make fajitas on my chassis... around 50c is when you have proper bias in a 3U chassis. Yes, you may be cooler because of heatsink area with a bigger chassis or ambient conditions, but you should never be that cool.
 
My USSA5 was working fine, and i thought of fine tuning the bias up to 1.35A.

A stupid mistake costs me dearly, magic smoke came out!
But not the output fets shorted, it is one of the Thermister that touches the heatsink all these time and today the insulation decides to fail!

It's back to repair again !!!

Pls, don't make the same mistake as me, leave a small gap between, if you have to let the thermistor touches the heatsink, pls put a mica sheet in between.
 
Today I checked some more and realised that the Transistors seem to have different Temperatur,so I maybe better check point 8.7now .Before doing something wrong,I got some Questions.Do I better set Bias back to say 50mV ? In Manual it says Connect V+V-and PGND(the one close to V-)On PCB is PGND close to V+ ,what is right ?So hook up Zero to PGND and TP10+11 to GND ? I got a Bridge between both Ground at the Moment.
Regards DT
 

fab

Member
Joined 2004
Paid Member
Hi meanie
Sorry for your thermistor failing. This thermistor is rated 150C and/or 425mw. Because it is in series with another resistor which network is in parallel with a pot and another resistor only few mw is dissipated even when thermistor is at lowest value (high temperature). It is impossible you got such a high temperature for heatsink. I also recommend thermal switch on main heatsink as per the manual. The lateral mosfet which is the main heat contributor are self protected at higher bias because of their temperature coefficient.
I see 2 possibilities for the thermistor to fail:
1) contact with other conductive parts. It happened to me once.
2) premature failure due to specific batch or parts bad manufacturing and Aggravated with higher temperature. It is possible that thermistor fabrication is not mastered as it is for simple resistor...

A small space (even 1mm) between thermistor and heatsink is good practice. I did mention it in the USSS-1 manual but maybe not in the USSA5 manual...

I suppose you have checked that the thermistor is the only part damaged.
Fab
 

fab

Member
Joined 2004
Paid Member
Taccodude
26mv is ok but 23mv is too low for the intent and both must be equal within 1mv.
This means your jfet have very very low idss. Change R9/R10 from 470 to 330 ohms or even 220 ohms depending on what you have. You can also install a parallel resistor with R9 and R10 to get the esquivant value. After that you readjust P3/P4 for 28mv. You can also uses 26mv for one side and change resistor on the other side to get 26mv.
Fab
 
Last edited: