SA-9100 Protection light on and off

All transistors and caps are on order for awr-032, awh-023, & AWM-025-C. Will replace q2 and q4 on awr-032 and report results. Do you have the complete spreadsheet for sa 9100 markthefixer? KSC1845 are on back order.
 
512-KSA733CYTA or 512-KSA733CGBU ??
512-KSC945CYTA or 512-KSC945CGTA ??

Watch that 'C' suffix. It means center collector.
 
so these are the same transistors with a different pin layout? What's the difference between CYTA vs CGBU or CYTA vs CGTA??? thank you
 
C means center collector.
L / Y / G is the Hfe rank.
TA = tape package
BU= bulk packaged.
Look at the data sheets. They tell you this information.
 
I avoid the a733 and c945 BECAUSE these particular damn transistors are available with two different pinouts.
And people get visual dyslexia from the parts numbers.
This causes more errors and confusion, headaches and lots and lots of typing for me.

the C says center collector (ecb) in both
checking the data sheet shows G and Y to be gain codes. DOWNLOAD the data sheets!!!
then TA versus BU : BU is bulk transistors, TA means they are mounted on a machine feeder tape.

don't worry (as much) about the "obsolete", usually that means they have changed the letters in the suffix, after the number (733, 945 and so forth)
just search on : 512-KSA733 and 512-KSC945.

I don't have time to dig them out, just where the heck are they ALL in a sa-9100?
The 945 in the protect is a problem waiting to reoccur, the the 2383.
the power supply can use the 2383 in q3 IF necessary.


q2 is blown, q4 probably as well. When they are out, check for burned resistors around them electronically. (not necessarily physically).


edit: redk beat me by a hair.
 
So again, that leads me to original question... What's the best replacement for KSA733CLTA & 512-KSC945CLTA
 
I avoid the a733 and c945 BECAUSE these particular damn transistors are available with two different pinouts.
What the hell were they thinking when they came up with this numbering scheme. The original Japanese transistors were 2SA733 and 2SC945. Then the Koreans (?) thought it would be a good idea to clone these as KSA733 and KSC945 but add the suffix "C" to have a center collector like the Japanese transistors. And why not just leave the first two characters off of the P/N. What could possibly go wrong? Oh Boy!
 
Changed q2 Ksa940 and q4 ksa1013 new voltages

1) 0V
2) -0.750mV
3) -1.95V
4) -3.2V
5) -6.1V
6) -59.5V
7) 59.03V
8) 42.8V
9) 35.34V
10) 25.09V
11) 16.51V
 
Q2
E -10.5V
B -9.9v
C -59.75

Q4
E 0V
B -0.525V
C -10.7V

I cannot figure out what is pulling Q2 E and Q4 C from -42.6v up to -10.7V. C2? R4 and R2 test good.
 
What the hell were they thinking when they came up with this numbering scheme. The original Japanese transistors were 2SA733 and 2SC945. Then the Koreans (?) thought it would be a good idea to clone these as KSA733 and KSC945 but add the suffix "C" to have a center collector like the Japanese transistors. And why not just leave the first two characters off of the P/N. What could possibly go wrong? Oh Boy!

The transistors concerned were originally NEC fabricated epitaxials.

KSC (Korean Semi Corp?) was simply a manufacturer code (a subsiduary of Silicon Transistor Corp - STC). Now it has gone on to mean of 'Fairchild' origin- maybe Fairchild bought STC way back and manufactured in South Korea? (what exactly happened with NEC's divestment of its semiconductor business relating to epitaxial transistors is anyone's guess and where Fairchild fitted in is interesting, considering the issues they had in the 1970s with their business arrangement with NEC) The 2SC945 had the NEC pinout and Sanyo(2sc4641), Toshiba (2sc1815), Mitsubishi (2sc2320) and Hitachi (2sc2308) had their own JIS numbers (idenitical) transistors some with their own pinouts.

ECB across the face was never the standard for Japanese T092s, it was just one of many- the most common.

The 'c' suffix was never needed because all the details on the pinouts were tabulated in Japanese Data Manuals that everyone had on their bench and you learned which manufacturers used what pinouts on what transistors. The 2SC945 has to be the most common T092 ever used.

There's nothing hard about it. :)
 
Last edited:
So again, that leads me to original question... What's the best replacement for KSA733CLTA & 512-KSC945CLTA
And again I say, somewhat differently: In WHAT circuit positions? Each circuit position does different things, and must be analyzed for best fit parameters.
And I don't have enough time to LOOK for all of them in there and analyze each circuit position.

Maximum voltage, maximum collector current, power dissipation, gain, speed = they all figure in.
And I HAVE found both faults in Pioneer's math and selections, as well as where they cut it too close on some ratings, and now years later those effects are becoming apparent. No less how they might survive faults.
 
check r9, r11 for 22,000 ohms resistance, as well as R10 at 910 ohms.. Then what is the status of C7? Is it original? Or new? Or "leaking" electronically?

The circuit is regulating. Just not where you want it. It is a mirror regulator, and it wants to keep the base of Q4 at about -0.6 volts.
A voltage divider of 3 resistors does this.
R9 is connected to the +42v, R11 needs to be connected to -42v to make -0.6v at the junction of R11 and R10.
IF for the sake of an example R10 is zero ohms, that -0.6v point would be zero volts. Balanced exactly half way between +42 and -42 volts.
Now we add resistance at R10 to shift that 0v to -0.6v where the transistor is just starting to turn on.
That means that the +voltage and the -voltage has to be the same magnitude, opposite sign. Thus R9 = R11 and any voltage imbalance is a difference in those resistances relative to each other.

This is why I spoke of C7 and "leakage" which can shift the conductance / resistance.
 
Will get resistors at work tomorrow, caps come in tomorrow or Wednesday. All transistors and caps are original. All be damned if I don't figure this out
 
Last edited:
One of your (markthefixer) previous post on another thread for awm-062 rebuild. That why I asked about the 2 transistors.

Q3, Q4 SSC945-Q
512-KSC945CLTA to-92 eCb 50v .15a .25w 300mhz 40-700hfe $0.05 ea

Q5 2SA733-Q
512-KSA733CLTA to-92eCb 50v .15a .25w 180mhz 40-700hfe $0.05 ea
 
The transistors concerned were originally NEC fabricated epitaxials.

KSC (Korean Semi Corp?) was simply a manufacturer code (a subsiduary of Silicon Transistor Corp - STC). Now it has gone on to mean of 'Fairchild' origin- maybe Fairchild bought STC way back and manufactured in South Korea? (what exactly happened with NEC's divestment of its semiconductor business relating to epitaxial transistors is anyone's guess and where Fairchild fitted in is interesting, considering the issues they had in the 1970s with their business arrangement with NEC) The 2SC945 had the NEC pinout and Sanyo(2sc4641), Toshiba (2sc1815), Mitsubishi (2sc2320) and Hitachi (2sc2308) had their own JIS numbers (idenitical) transistors some with their own pinouts.

ECB across the face was never the standard for Japanese T092s, it was just one of many- the most common.

The 'c' suffix was never needed because all the details on the pinouts were tabulated in Japanese Data Manuals that everyone had on their bench and you learned which manufacturers used what pinouts on what transistors. The 2SC945 has to be the most common T092 ever used.

There's nothing hard about it. :)
The thing that I thought was out of hand is that since the transistors do not have 2S or KS printed on them, you do not know if you are looking at a KSC945 or 2SC945. A 2SC945 subs to KSC945C to get the same pin out. Not exactly brilliant in my opinion.
 
One of your (markthefixer) previous post on another thread for awm-062 rebuild. That why I asked about the 2 transistors.

Q3, Q4 SSC945-Q
512-KSC945CLTA to-92 eCb 50v .15a .25w 300mhz 40-700hfe $0.05 ea

Q5 2SA733-Q
512-KSA733CLTA to-92eCb 50v .15a .25w 180mhz 40-700hfe $0.05 ea
Ok, here's the "dirty secret":

I usually need a link to the post where I say such things. I have to evaluate where in the timeline since 2005 that post was done, as I gained more information and experience by doing all the reverse engineering and interacting with members - like for example my building dislike for the confusion generated by the 733 and 945 transistors and their variations and numbers - and all the problems traceable to that confusion that were simply solved by stopping mentioning those transistors. Yet they STILL plague me now - just not as much...
for example
You may find earlier posts mentioning ksd1616 transistors, (mostly protect circuit charged timing capacitor's discharge current dump transistors) that were no better than the originals that were destined to fail by their marginal "safe operating area" that was changed to the ksc2383, which has a larger die and thus larger "safe operating area".

you start to see why the circuit location where a particular part number is used is important.

I took notes while doing that work, and they still exist - some on paper and some online.
I don't toss worksheets like scratch paper - well - I toss them, but into a file folder, not the trash.

So every once in a while (like post #20) if it looks good (no proscribed parts numbers), I just grab a section of my notes and post them.
Then if I have doubts if the numbers are fresh enough, I make a test project at Mouser to see what it doesn't like - before posting a flawed list.

Thus how the capacitor lists were started.
 
...since the transistors do not have 2S or KS printed on them, you do not know if you are looking at a KSC945 or 2SC945. A 2SC945 subs to KSC945C to get the same pin out. Not exactly brilliant in my opinion.

Totally agree. IMO C945 should be an NEC 2SC945- nothing else, with an ECB across the front.

I have to admit, the day I got a automatic transistor pinout/hfe tester way back in 1990 on my DMM, I thought it was the best thing since sliced bread.

If it was me, I'd be using brand new Toshiba stock, 2SC1815/2SA1015. They are available in the tens of thousands, in all Hfe ratings, with the ECB config.
 
Today I got caps and resistors from Mouser and started with R9, R10, R11. Original values R9 - 21.3Kohms, R10 909ohms, R11 21.64Kohms, R24 181ohms, R1 5.59Kohms. I replaced R9, R10, and R11 anyway. Decided to move onto C7 and replaced it. Turned amp on and it came right out of protection mode. I absolutely amazed by your ability to troubleshoot these circuits.
Voltages after C7, R9, R10, & R11 replacement:
AWR-032
1) 0V
2) -15.89V
3) -24.60V
4) -12.93V
5) -42.65V
6) -59.25V
7) 58.40V
8) 42.66V
9) 34.93V
10) 24.42V
11) 15.75V

AWM-025C
1) 0.016V
2) 12.99V
3) 83mV
4) 0V
5) -15mV
6) 48mV
7) 100mV
8) 66mV
9) 8.54V
10) 35.12V
11) 35.12V
12) 0V

AWH-023 RIght
1) 24.5V
2) 0V
3) 0V
4) 0V
5) 0V
6) -42.72V
7) 0V
8) 42.78
9) 24mV
10) 0V
11) -44.53V
12) 8mV
13) 9mV
14) 44.53V

AWH-023 Left
1) 24.51V
2) 0V
3) 0V
4) 0V
5) 0V
6) -42.73V
7) 0V
8) 42.79
9) 24mV
10) 0V
11) -44.66V
12) 7mV
13) -37mV
14) 44.63V
 
Voltages after electrolytic replacement of the Elna junk caps C1, C5, C6, C7, C8, C9, C12, C13, C14, C15, C20 and Q2 Ksa940 and Q4 ksa1013:

AWR-032
1) 0V
2) -15.95V
3) -24.67V
4) -12.95V
5) -42.73V
6) -59.34V
7) 58.50V
8) 42.76V
9) 35.04V
10) 24.48V
11) 15.66V

AWM-025C
1) 6mV
2) 13.0V
3) 7mV
4) 0V
5) -38mV
6) 26mV
7) 24mV
8) -10mV
9) 8.53V
10) 35.15V
11) 35.15V
12) 0V

AWH-023 RIght
1) 24.49V
2) 0V
3) 0V
4) 0V
5) 1mV
6) -42.70V
7) 0V
8) 42.77
9) 24mV
10) 0V
11) -44.38V
12) 7mV
13) -10mV
14) 44.39V

AWH-023 Left
1) 24.49V
2) 0V
3) 0V
4) 0V
5) 0V
6) -42.72V
7) 0V
8) 42.77
9) 25mV
10) 0V
11) -44.45V
12) -7mV
13) -38mV
14) 44.48V

It's been playing for hours. No issues.
 

Attachments

  • 20171114_213111.jpg
    20171114_213111.jpg
    54.9 KB · Views: 37
  • 20171114_213222.jpg
    20171114_213222.jpg
    20.6 KB · Views: 37
Last edited:
Back
Top Bottom