EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: spilihps on March 06, 2018, 12:46:58 pm
-
Hello!
So i am trying to repair a ecu, that has a fault that prevents it from activiating a relay.
I did find what i think is a SOT-89 transistor (marked with DK QR)
Now, i found that it has a dead short between base and collector. So i did remove it, and it is still shorted, so i guess it is shot.
The question is, now since it just have those markings, what replacement would be best?
-
Hello spilihps,
Take a look at this Anachip datasheet for AP131. The marking chart on page five shows DK as the marking for AP131-25Y, a 2.5 Volt LDO in a SOT89-5L package.
The QR part doesn't match up with their date code system of digit+letter.
Can you post a close-up photo of the device? Does it only have three leads?
RF+ Tech
-
Hello and thank you for your response.
It have three leads (plus flange)
googling the first response is this:
http://www.s-manuals.com/smd/dk (http://www.s-manuals.com/smd/dk) And it shows a PNP resistor under DK. first in the list.
I dont think a LDO would make any sense, since this is in charge of switching ground to a relay-circuit.
So using that first pnp datasheet, it make sense.
It have base connected (via some discretes) to the main microcontroller, then it have collector connected to the relay pin, and emitter is directly to ground.
Would that make any sense?
-
That connection to the microcontroller and the relay coil makes sense, but only if the transistor is NPN and the car is negative earth. Assuming that it is a negative earth car, the emitter connection to ground means that the transistor must be an NPN type.
-
Thanks!
Yes that made me confused, but the designation DK only seem to apply to an PNP?
-
This device looks closer - MCC 2SC4672-Q, SOT89 NPN, marking DKQ.
No mention of R,
RF+ Tech
-
Hello!
please have a look at the attached image and see the markings
(https://preview.ibb.co/dAKwq7/h_transist.jpg) (https://ibb.co/kzGy3S)
-
Hello spilihps,
Go to https://images.google.com/ (https://images.google.com/) and search for 2SC4672
;D :-+
RF+ Tech
-
Yep that seem to be the right one!
My main source dont seem to have a 2amp 60v one though..
The closest that comes up is a BCX-54 and its just 1amp and 45v. Guess that is a bit to far off spec?
-
While it is not clear in your original post, it seems to infer this transistor is used to switch a relay coil. Ask yourself "how much current would the relay coil require?"
I doubt it will be 2 Amps, or even 1 Amp, if the relay is located on the same board. Something on the order of several hundred milliamps. If the relay is located external, then the coil resistance might be lower than what I suspect. Still, I would expect the coil current to be less than 1 Amp.
Measure the coil resistance, apply Ohm's law and make an educated decision.
RF+ Tech
-
Hello!
An update,
I did actually measure the currentdraw of the relay coil, and it was continious 800mA.
First i thought that there were something wrong with it, however, we had another one on hand and it also read that high.
So the 1A transistors i got hold of is probably wrong?
-
I agree a 1 Amp rated transistor sinking 800 mA is on the edge for reliability. It should work long enough until the proper replacement transistor is found.
Digikey shows 276 pieces of Rohm 2SC4672Q in stock for $0.71 each. I note too, that this part is marked "Not for new designs", so that suggests supplies globally are dwindling.
Your country flag is not set, I'm not sure in what country you live. It may be necessary to look further abroad for another stocking distributor.
RF+ Tech
-
Again thanks for your help!
Digikey is a last resort for me,
the best for me is www.elfadistrelec.com (http://www.elfadistrelec.com) or maybe www.farnell.com (http://www.farnell.com)
-
So i did find a retailer carrying this: 2SC5994-TD-E
Do you think its a suitable replacement?
-
The 2SC5994 is an interesting find. Comparing the DC Safe Operating Area curves of both devices, the 2SC5994 looks nearly an order of magnitude better than the 2SC4672.
Look at the 800 mA line on the left axis, follow this horizontally to the DC curve intersection, then down to the C-E voltage axis, for both devices. These graphs reveal the 2SC4672 really needs high beta and sufficient base current to keep the C-E junction well saturated, at 800 mA of collector current. Insufficient base current or beta can result in higher C-E saturation voltage, which pushes collector dissipation closer toward the DC SOA limit line.
The product of collector current and C-E saturation voltage is what determines the power dissipated, when the relay coil is pulled in continuously. There isn't much headroom for the 2SC4672. Note how much more margin exists for the 2SC5994, at the same 800 mA collector current line. This is about 1 kilometer wide, relatively speaking. Good choice.
RF+ Tech
-
The 2SC5994 is an interesting find. Comparing the DC Safe Operating Area curves of both devices, the 2SC5994 looks nearly an order of magnitude better than the 2SC4672.
Look at the 800 mA line on the left axis, follow this horizontally to the DC curve intersection, then down to the C-E voltage axis, for both devices. These graphs reveal the 2SC4672 really needs high beta and sufficient base current to keep the C-E junction well saturated, at 800 mA of collector current. Insufficient base current or beta can result in higher C-E saturation voltage, which pushes collector dissipation closer toward the DC SOA limit line.
The product of collector current and C-E saturation voltage is what determines the power dissipated, when the relay coil is pulled in continuously. There isn't much headroom for the 2SC4672. Note how much more margin exists for the 2SC5994, at the same 800 mA collector current line. This is about 1 kilometer wide, relatively speaking. Good choice.
RF+ Tech
First of all, Thank you very much with your help through this repair. It have been very valueable for me!
It was interesting to read about the curve comparision, i did just choose the one that was close or "better" on the quick specifications.
I can also tell you that i have tested the ECU live without the transistor in place, because i was worried that it may have had some other components between the microcontroller and this transistor base.
But it was all OK on the scope.
So, chances are that we can get this up and running again. Will report back when the part arrives and we have tested it.
Thanks
-
Hello,
Good news, the repair was a success.
The unit works great!
Thanks for your help. :-+
-
Hello,
Thanks for the report of success.
And you are most welcome, glad to lend some assistance.
Could you edit the subject line and insert Solved, so others may know?
RF+ Tech