Babelfish XA252 / Babelfish XA252 SIT / Babelfish XA252 SET

Well, still having problems and at a dead end. I've gone over everything on this board for placement, value, operation, solder joints, etc.. and haven't found anything. I replaced Q1,2,T6 even though they tested fine. Powering up with light bulb, everything looks ok. No longer having any rail voltage offset. Switched out light bulb, and looked good on startup, went for a few seconds, until fuse blew on negative rail.

Is there anything I can probe with it powered by light bulb, or variac for troubleshooting?


Official Court Jester
Joined 2003
Paid Member
did you tested all surrounding components, what I wrote in #1936?

what value of fuses in rails? say that 2A5 is more than adequate

now, you can test FE itself, removing pcb from pucks (so no jeopardy of big nasty currents), first dialing Iq pot all the way CCW and shorting G to S for both pucks on pcb, to establish proper NFB loop for FE

but - first check ZD3 and ZD4 are they both alive
Official Court Jester
Joined 2003
Paid Member

here it is - testing only FE
steps are - removing pucks from game , so no nasty surprises
establishing NFB loo ( broken without pucks in place) so FE is having all what's necessary to behave


  • Babelfish XA252 (straight MOS) FE only TESTING.png
    Babelfish XA252 (straight MOS) FE only TESTING.png
    108.8 KB · Views: 61
  • Thank You
Reactions: 1 user
Official Court Jester
Joined 2003
Paid Member
so, it seems that everything is OK with FE now

while setup as is - can you check voltages across small smd resistors in biasing circ, practically voltage between positive rail and emiters of Q3, Q5, same as voltages between negative rail and emiters of Q4, Q8

that will tell is biasing circ operational

when that done, do not mount pucks yet, remove bridges you soldered form G to S of pucks , then with circ powered Off, check all resistors in between/around puck gates, meaning 10K R32,R33 and gate resistors R51, R52

it'll be cunning to dial Iq pot all the way CCW, preparing for full circuit testing afterwards

important note - pucks mounting - side with slit for screw goes looking at 0R56 group on pcb, side with hole for screw goes to down edge of pcb, speaker trace
Same voltage across Q3,5,4,8: 0v793

In circuit, R32+33 start at 6k and climb towards 10kohms. I checked them out of circuit earlier and they exhibited same, and were 10k when removed.

Yup, I know correct puck orientation, which was heeded originally.

So, next is to start on the startup procedure as you detailed pg 1. I will await your ok.

Thank You!!!!
Official Court Jester
Joined 2003
Paid Member
what you have populated for R** and R***?

sum value needs to be in range of 121K (270K//220K), or one single 120K resistor in place of any of those

so, LT3092 works as 10uA through said resistor ( group)

now, if 10uA goes through 120K, gives 1V20

divided with (R28 + P2) (when P2 is all the way CCW, so 500R) goes to (1K5+500R)

goes to 1V20/2K=600uA

through R29 and R30 (both 470R) gives 282mV

you have 0V793, saying that you have way too much programmed current thorough LT3092, in range of 0V793/40R=1.68mA

if powered like that, it'll give way too high OS Iq