• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

SA-8100 protection circuit issue

Thanks for that explanation, that was very helpful. The function of that part of the protection (that we've been discussing so much) now makes sense to me. I figured Q5 was essential, but I couldn't see why there were two paths (q3 & q4) to it. It also now makes sense why I should try lifting Q4 C, then measure the voltage at the solder pool for Q4 C. If Q4 is faulty, the voltage should come up to equal Q3 C, right?
 
Yes, the voltage on the Q4 solder pad should be equal to the supply voltage since there will no longer be any current flowing through R14. But, I would expect that the Q3 collector voltage will drop since the collector current will increase when the Q4 collector current is interrupted.
 
OK, will try that in a little bit.

A friend who fixes electronics had NTE123AP transistors on hand that he said is equal or better specifications to the original 2SC945 in Q3 and Q4. Would you concur? If so, I could try those once I confirm the voltages comes up on Q4 C with the collector lifted.
 
All connections in except Q4 C out.

E, C, B
Q3: +0.3 (variable), +34, 0
Q4: +0.1 (variable), +9 (variable), 0
amp stayed in protection; however after 2 or 3 minutes while making measurements, it moved in and out of protection. Got some readings during this time:

Q3: -0.6, 17, 0
Q4: 0.5, 34, 0

Should I try putting the NTE123AP.
 
OK, put Q4 C back in and pulled Q3 C

E, C, B
Q3: -0.8, 34.2, -0.2
Q4: -0.6, 6, -0.1

(Collector for Q4 is measured at the solder pad.)
 
interesting .. a new pair should hopefully fix the issue . if your meter has hfe test i would be interested what hfe is for q3 q4 . after pulling them of course .
 
Put in the NTE123AP transistors for Q3 & Q4:

E, C, B
Q3: -0.6, +18.6, 0
Q4: -0.6, +18.6, 0
Q5: 16.4, -1.2, 19.1

Haleluiah!! Amp comes out of protection as it should. Q5 C should be -0.7, but I assume that reading is ok?
 
petehall347 said:
interesting .. a new pair should hopefully fix the issue . if your meter has hfe test i would be interested what hfe is for q3 q4 . after pulling them of course .
Click to expand...
My DMM does not have functions for testing transistors. I can use the diode test, but that's it. I think one lesson here is that 1845 can't be used to replace a 945. I did look at the spec sheets before I put the 123s in, and those are a much better match for a 945 than the 1845.

Q5 C might settle down? Those were just initial readings. I was too excited to wait longer! :banana:
 
Actually in post 174, LesE's simulation came out with -1.5 V for Q5 C. His values for Q3 & 4 base also line up. Mine were negative but in the single digit milivolts, so I called that zero. He has -8.7 and -8.3 mV.
 
Great that the transisitor changed sorted it, but I would steer clear of NTE parts. Most people would avoid them on here... best to get some modern 945's from somewhere like mouser. KSC945CGTA.

945's can sometimes be changed to 1845's but it does depend as they have different "sensitivities" as maybe we've seen here in the protection circuit. And protection circuits are very sensitive anyway.

The 945 issue plagued me quite a bit until I compared the datasheets...
 
I guess the NTE story is something from the past/happening in the old past?

A reputable company like Newark supplies them also.
 
JURB said:
Well, IMO NTE transistors are not for in the audio path. Reason being is they take any transistors that will do the job and rebadge them. There are variances, and no detailed spec sheet available.

They are fine for certain applications but I would have to be between a rock and a hard place to use them in the audio path. They do not make transistors, they buy rejects. One batch could be just fine and another batch be ****. You never know. But if you need a pass transistor on a regulator board they are fine. It's just that they aren't real picky or specific on certain other parameters.
Click to expand...

I couldn't find anything that MTF or Echowars have said in the past quickly, but installing NTE is never a good idea. Best to get transistors from a good and reputable company like Fairchild/OnSemi for one. No affiliation.
 
OK, thanks for that. The protection circuit/amp is working at this point, but I will put some KSC945CGTA's on my next mouser order and replace the NTEs. Interestingly, the NTEs are 5 -10 times more expensive than Fairchild.
 
KeithD said:
Interestingly, the NTEs are 5 -10 times more expensive than Fairchild.
Click to expand...

Better quality AND cheaper! :rflmao:

Just check the data sheet for the 945CGTA's when you put them in. I think the "C" signifies that the center leg is the Collector. so they are ECB rather than the usual 945's which are EBC. IIRC...
 
You must log in or register to reply here.
Back
Top Bottom