Keyboard Shortcuts
ctrl + shift + ? :
Показати всі комбінації клавіш
ctrl + g :
Не доступний для безкоштовних груп.
ctrl + shift + f :
Знайти
ctrl + / :
Сповіщення
esc to dismiss
Лайки
Пошук
SC503 -12V power supply
|
Greetings for Louisiana!
I have an SC503 that has an issue with the -12V Power Supply. When I started troubleshooting the PS I came across a bad transistor, Q8052, plus several tantalums. At that time the voltage was about -3 volts. After I changed the transistor the voltage jumped to -33.4 volts. I course this helped identify 2 other weak tantalums. I have put sockets for all the transistors in the -12V Supply having checked each transistor using my Atlas DCA Pro. I believe I have also checked all of the diodes and all of the resistors. The voltages of Q6041 is as follows E=0.703, B=-1.273, C=0.01. the schematic shows that E=+0.62 B=0.0 and Collector=-10.3Volts. Is there anyone in the group that can offer suggestions? Thanks! Barry |
|
On Sat, Dec 21, 2024 at 10:01 AM, Barry Breaux wrote:
..Since you have all the transistors in a socket: - Does removing Q7042 and Q7051 make output go ~ +0.6V, if not power transistor is bad/leaky. When all transistors are in the socket, what are the voltages at: - Base of Q8051 with respect to its emitter (-33.5V line). - Base of Q7042 wrt its emitter - Base of Q7051 wrt to ground (not emitter) - Voltage across R9041 (=Vbe of power transistor) - Voltage across R9051, needs to be measured at mV level. These voltages should tell what is wrong. Since your supply board is out only +'ve current to -12V line comes from R6041. Q7042, Q7051, and the power transistor are possible paths to -33V. They should only pass enough current to set output to ~ -12V. In your setup Q6041 is on (although its Vbe is quite high), that should pull base of Q8051 higher through Q7054, which should short across base of Q7042 and turn it off, making voltage across R9041 ~ 0V and turn off power transistor. Q7051 base should be ~ -10V so it should be off with -33V at the output. However, one of these conditions is wrong in your board. Ozan |
|
Barry
Testing transistors out-of-circuit on a Tester is NOT a fail-safe diagnosis. I also have an Peak Atlas DCA, and many's the time that it showed a transistor OK, but under load in the circuit it was bad. You have at least TWO ways to check a transistor IN-CIRCUIT. 1. Scope the transistors pins - but you need to know what the correct waveform is, which isn't always possible. 2. Freeze-spray the suspect transistor, and measure the pins' voltage behaviors - usually massive fluctuations will identify the faulty component - much easier to accomplish, and that's what I normally do. If you don't have freeze-spray, get a can - it's an indispensable tool on your workbench. Menahem |
|
Ozan,
Many thanks for your help in this matter. Glad to learn as much as I can! I replaced all the transistors (except the pass transistor) and I still have -32.5 at the output. The voltages you question are below. On Sat, Dec 21, 2024 at 05:28 PM, Ozan wrote: Thanks again for your help! Barry |
|
Hi Barry,
переключити цитоване повідомлення
Показати цитований текст
By removing transistors and seeing +0.6V we confirmed power transistor is not responsible for the fault (at least not the -33V output issue). Voltage across R9041 is 0.672V, this turns on the power transistor. However, in this state (output exceeding target voltage), Q7042 should be off and voltage across R9041 should be zero. This means there is current through Q7042. It could either be faulty Q7042 or faulty Q8051. Transistor Q8051 has a good base drive, it should be able to short across Q7042 and turn it off. Is C-E terminals of Q7042 correct, i.e. not swapped? If you short across C-E of Q8051 (with or without Q8051 in place is OK, you can do it at the socket) does output go to ~ +0.6V? If it does Q8051 is bad. If it doesn't Q7042 is bad or C-E swapped. Is Q7042 base voltage of -29.3V wrt to its emitter or ground? Vbe of -29.3V is unlikely and would damage the transistor. Ozan On Sun, Dec 22, 2024 at 12:20 PM, Barry Breaux wrote:
|
|
Ozan,
Is C-E terminals of Q7042 correct, i.e. not swapped?I verified the terminals are NOT swapped! If you short across C-E of Q8051 (with or without Q8051 in place is OK, youWhen I short across C-E of Q8051 the output voltage goes to -15.1. Q7042: E=-34.2 B=-34.2 C=-32.5 Is it possible that the replacement transistors I chose are causing this issue? I choose 2N5401 for 2N4249 and 2N5551 for 2N3565. Thanks again! Barry Barry |
|
Barry,
Q7042 crosses to a 2N2219A. This is a 3W transistor. A bit of power is needed to drive the series pass. Your list of parts are small signal, not enough power in this case. That could be your problem or a problem. On the 2N2219A, the case is directly connected to the collector. Check to see if resistors are bad in some way, open, burned, etc.. Mark |
|
On Sun, Dec 22, 2024 at 08:03 PM, Mark Vincent wrote:
Mark, I have checked all the resistors. Thanks for your comments! I have a lot to consider at this point. Barry |
|
Hi Barry,
My comments are inline below. Ozan On Sun, Dec 22, 2024 at 05:38 PM, Barry Breaux wrote: Noted.Is C-E terminals of Q7042 correct, i.e. not swapped?I verified the terminals are NOT swapped! At -15.1V this test doesn't say much about Q8051 (output is not completely cut off even if Q8051 was fully on) but see the note below for Q7042.If you short across C-E of Q8051 (with or without Q8051 in place is OK, youWhen I short across C-E of Q8051 the output voltage goes to -15.1. Q7042: E=-34.2 B=-34.2 C=-32.5This is interesting because B & E are at the same voltage (i.e. same as if they are shorted) and output is ~ -33V. However, if you physically short across C-E of Q8051 output goes to -15.1V. R8052 is only difference between two cases but it is a small 27-ohm resistor and shouldn't matter for this experiment. Looks like Q7042 is not fully turning off. To rule out bad Q7042, do you have the original transistor or another spare? Is it possible that the replacement transistors I chose are causing thisData for the original transistors 151-0103/0341/342 is hard to compare to the ones you replaced with. One possibility is their fT is too high/low and supply is oscillating. If you look at base voltage of Q7042 is there any oscillation? Is there any ripple at +12V and -33.5V filtered DC lines? One problem with this theory is shorting across B-E of Q7042 should turn off Q7042 regardless of oscillations. Transistors Q6041 and Q8051 have compensation networks around them and their fT ~ 40MHz. Transistor Q7042 doesn't have a compensation network but its fT is higher ( 100MHz min). These three transistors are important for stability of the loop in addition to the power transistor (I assume it wasn't replaced). Rest of them are not critical. Q7054 is a cascode Xtr, Q8052/Q7043/Q7051 are short circuit protection. If you still have the three original transistors you could try them. Ozan |
|
Ozan,
I found a broken trace which I repaired. I also swapped out a few transistors. Today I am measuring -0.48 volts when I short across C-E of Q8051. I understand that 2N2222A is a lower wattage equivalent to the 2N2219. I have ordered additional 2N2219. The 2N2219 was/is shorted. There is a .3VPP oscillation at the base of Q7042. a 50mv ripple on the +12V and a .2V ripple on the -33.5V. I am still plugging away at it. Thank you! |
|
Hi Barry,
переключити цитоване повідомлення
Показати цитований текст
I think you fixed the root cause of why shorting across C-E of Q8051 didn't work. I was expecting to see the same voltage as if Q7042 was removed (+0.6V) but at this point the regulator should be able to control the voltage. Do you still see ~ -33V at the output? Do you have the original transistors to try or was Q7042 the shorted one? Q7042 (151-0103-00) cross references to 2n2219A, plain 2n2219 has a Vceo of 30V and lower fT than 2n2219A. This device will see Vce~ 33V in normal operation. page 6-3 of: https://w140.com/tekwiki/images/5/5e/Tektronix_Xref_sm.pdf Ozan On Mon, Dec 23, 2024 at 01:47 PM, Barry Breaux wrote:
|
|
Ozan,
I see -32.5V across at the output Q7042 was found to be bad when I started several days ago. I did not have one so I used a 2222A temoprarly. I have a couple "2N2219 PBFREE Central Semiconductor 60Vcbo 30Vceo 5.0Veb / Bipolar Transistors - BJT" on on order from Mouser. I expect them on the 26th. I also have "5x 2N2219 MOTOROLA"on order from eBay. I didn't read into the significance of your comment. Are you saying I need the 2N2219A part instead of the 2N2219 part? Regards, Barry |
|
Hi Barry,
переключити цитоване повідомлення
Показати цитований текст
My comment about 2N2219A vs 2N2219 was: - "A" version is in the Tek replacement recommendation - "A" version fT matches to original Tek part - "A" version has higher Vceo of 40V. Plain 2N2219 has a Vceo of 30V and in my opinion it is too low for this application which will see 33V or above across CE. I recommend 2N2219A. Mouser/Digikey has 2N2219A's. I don't use them often but Jameco also has them at a cheaper price although manufacturer is not reported so there is some risk. Ozan On Tue, Dec 24, 2024 at 01:23 PM, Barry Breaux wrote:
|
|
I finally got to the bottom of this! A great learning for me. The bottom line is that the Q8051 transistor needs a lot of gain to drive 2n2219A. The original transistor was a PN3565, 151-0341. The original was bad so I couldn't measure it. I had only 2 of those on hand and the one with a gain of 465 worked.
Onan, Thanks for all the help you provided me. I am sorry if some of the information I provided was a bit unclear. In the end I learned a lot. Many thanks! Barry |
|
Hi Barry,
переключити цитоване повідомлення
Показати цитований текст
Good news. We did try shorting C-E of Q8051 and it didn't change the output, it should have eliminated any effect of Q8051 as the only root cause, i.e. there must have been something in addition to possibly bad Q8051. Perhaps the test was done before you fixed the broken traces and/or swapping 2N2219A. Congratulations for the repair, it was an interesting puzzle. Ozan On Sat, Dec 28, 2024 at 07:25 AM, Barry Breaux wrote:
|
Jean-Paul
Повідомлення
Більше
Додаткові параметри
Більше
to navigate to use esc to dismiss