Author Topic: looking for smd diode reference : marking = T3Y7 (Read 3456 times)

kripton2035 · « **on:** February 16, 2019, 08:30:11 pm »

just found this diode in a PFC power supply. between the mains and the pfc controller Vcc. (it's an NCP1605D)
my googlefu doesn't give any information on this marking. also tried Y7T3
don't know if it's a diode or a zener. no helping text on the pcb either.
It *should* be a 75v zener diode ?
anyone has an idea about a ref ?
thanks.

mr.fabe · « **Reply #1 on:** February 16, 2019, 09:07:13 pm »

Could it be SDZ6V8Q ?

http://pdf1.alldatasheet.com/datasheet-pdf/view/425353/KODENSHI/SDZ6V8Q.html

kripton2035 · « **Reply #2 on:** February 16, 2019, 09:45:51 pm »

really fits the marking, but I don't see how a 7v zener could fit at this place in the circuit ...

fzabkar · « **Reply #3 on:** February 16, 2019, 11:19:25 pm »

I would think that "Y7" would be a date or batch code. Perhaps it might help if the unknown diode could be compared against known diodes elsewhere in the circuit. Designers will often source their parts from one particular manufacturer.

A circuit diagram or photo (of the circuit) might help ...

shakalnokturn · « **Reply #4 on:** February 17, 2019, 01:12:35 am »

BAV21W ?
What is the package type?

kripton2035 · « **Reply #5 on:** February 17, 2019, 07:37:16 am »

Quote from: shakalnokturn on February 17, 2019, 01:12:35 am

BAV21W ?
What is the package type?

from july 2011 would fit the markings too. see the picture for the package. SOD123 ?

kripton2035 · « **Reply #6 on:** February 17, 2019, 11:57:48 am »

this is the approximate schematic, reversed from pcb. I don't have anything else.
in the red rectangles are the measured values on the bad power supply.
the vcc is oscillating between 8 and 12 volts.
there should be 10-20V on the pfc controller Vcc, but a fixed value. from reference designs there should be 17V dc.
I presume the (zener?) diode is bad ?
what do you think ?

shakalnokturn · « **Reply #7 on:** February 17, 2019, 01:49:47 pm »

What makes you think that it should be a 75V zener?
Vcc oscillating (how fast?) between 8 and 12V looks like you are in the UVLO hysteresis range.
If anywhere, I'd be looking where the NCP1605 gets it's Vcc from.
Start circuit with dropper from mains? Probably OK from the symptoms.
Is there a dedicated winding from the main PSU transformer to take over from the start-up supply? Look for open resistors or small bad electrolytic.

kripton2035 · « **Reply #8 on:** February 17, 2019, 03:16:06 pm »

Quote

What makes you think that it should be a 75V zener?

because it would substract tha 90v and get some 20v to the ncp1605 ?

kripton2035 · « **Reply #9 on:** February 17, 2019, 03:18:27 pm »

Quote

If anywhere, I'd be looking where the NCP1605 gets it's Vcc from.

the vcc circuit is the above circuit I uploaded (except any mistake I could make)
the pcb is black, and you can't see the tracks on it, only hte big ones.
I had to trace the circuit with a low-ohm meter, but I may forget some tracks (although I did it several times...)

kripton2035 · « **Reply #10 on:** February 17, 2019, 03:23:36 pm »

Quote

Vcc oscillating (how fast?) between 8 and 12V looks like you are in the UVLO hysteresis range.

around 1Hz. mistake the 2-4V is in fact steady 2V

kripton2035 · « **Reply #11 on:** February 17, 2019, 03:24:55 pm »

Quote

Is there a dedicated winding from the main PSU transformer to take over from the start-up supply?

did not find one.

kripton2035 · « **Reply #12 on:** February 17, 2019, 03:25:52 pm »

also the NCP1605 gets very hot.

kripton2035 · « **Reply #13 on:** February 17, 2019, 03:34:12 pm »

pcb details

shakalnokturn · « **Reply #14 on:** February 17, 2019, 05:12:51 pm »

Ouch! That PCB is bloody awful... The last time I was in something that ugly it was a Hannspree monitor.

I sort of guessed that's where you got your 75V zener from, not sure about your schematic, I don't see any way that is going to power the NCP1605 in any case.
If the NCP1605 is "very" hot it's bad news for a start, not sure it's even worth troubleshooting in the current state as a fried IC may be causing the main PSU to hiccup...

You could test your mystery diode out of circuit, if it isn't open or shorted you'll soon find out if it has a zener voltage.
How about the larger black SMD's with a white stripe are those diodes?

Maybe have a look of a couple of app notes or get hold of a schematic for a similar power supply to get some picture of how the NCP1605 is usually used.
https://pdfs.semanticscholar.org/5832/d4c4211f05483841941e39ce3639e95a44e5.pdf
https://static5.arrow.com/pdfs/2012/11/13/0/28/53/955/ons_/manual/evbum2152-d.pdf

kripton2035 · « **Reply #15 on:** February 17, 2019, 05:38:55 pm »

Quote

If the NCP1605 is "very" hot it's bad news for a start, not sure it's even worth troubleshooting in the current state as a fried IC may be causing the main PSU to hiccup...

it oscillates at 1Hz between 60°C and 70°C. rest of the circuit is cold.

kripton2035 · « **Reply #16 on:** February 17, 2019, 05:41:18 pm »

Quote

Ouch! That PCB is bloody awful... The last time I was in something that ugly it was a Hannspree monitor.

it's from an iMac 27" model 2011. people are selling these power supplies refurbished, but I would like to fix this one just for the "pleasure" !
will try to change the ncp1605 it is a good candidate too.

kripton2035 · « **Reply #17 on:** February 17, 2019, 05:57:20 pm »

Quote

You could test your mystery diode out of circuit, if it isn't open or shorted you'll soon find out if it has a zener voltage.

just tested it, it's a simple diode, so certainly a BAV21, and it's perfectly working.
going to find a ncp1605

shakalnokturn · « **Reply #18 on:** February 17, 2019, 08:43:51 pm »

Crap, I didn't have a very high opinion of Apple even before you told me that...

Well depending on the voltage applied to test the diode it will show as a simple diode even if it is a zener.
The trick is to have a high enough test voltage to reveal the zener but with low enough current to not destroy anything by excessive heat/current in the case of a zener or excessive avalanche current in the case of a reversed biased (beyond blocking voltage) signal diode.

Anyway if your NCP1605 is cooked, that usually doesn't happen alone.

kripton2035 · « **Reply #19 on:** February 17, 2019, 09:34:18 pm »

Quote

Well depending on the voltage applied to test the diode it will show as a simple diode even if it is a zener.
The trick is to have a high enough test voltage to reveal the zener but with low enough current to not destroy anything by excessive heat/current in the case of a zener or excessive avalanche current in the case of a reversed biased (beyond blocking voltage) signal diode.

I tested it with the multifunction transistor tester, one model has a zener diode test, that goes up to 100V IIRC.

kripton2035 · « **Reply #20 on:** February 17, 2019, 09:37:55 pm »

Quote

Anyway if your NCP1605 is cooked, that usually doesn't happen alone.

not sure for this time, because I tested almost all semiconductors and capacitors, none of them are shorted or high esr.
the fuse is not blown. I'm confident that an 8 years old pfc chip can fail alone.
ordered 5 pieces at rs-particuliers, they should arrive in some days. will see.

kripton2035 · « **Reply #21 on:** February 17, 2019, 09:40:40 pm »

it's also quite bad that you can find almost any schematic of apple motherboards online without big efforts,
but absolute no power supply, or inverters . and they often fail first (when you don't have a defective graphic chip !)

kripton2035 · « **Reply #22 on:** February 20, 2019, 06:29:36 pm »

well I changed the ncp1605 and everything is the same. 60 to 70°C oscillating at 1Hz, and Vcc of the chip oscillating between 8 and 12v where it should have steady 20v max.

kripton2035 · « **Reply #23 on:** February 20, 2019, 07:27:49 pm »

replaced the diode too, with an smd 1n4007 and same symptoms ...

SilverSolder · « **Reply #24 on:** February 20, 2019, 09:27:38 pm »

Is there no hope of teasing more of the diagram out of it - will a strong light shine through it or anything like that?


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: looking for smd diode reference : marking = T3Y7 (Read 3456 times)

Share me