Deltran Battery Tender

[TheRuler8510]

RE: Deltran Battery Tender

Has anybody successfully repaired a "Plus" model or a "junior" model?

What typically goes wrong on these units?

I opened a dead "Plus"model, and the main MOSFET seems good. So I don't know how much effort is needed to troubleshoot further.

Thanks,

[DAVEISGREAT]

Same problem hear, deadsky.  Looks like at least a blown voltage regulator:

Here is my YouTube "video" describing the problem:


Link in description shows a blown-up high rez. photo of burned component.

Anyone have access to Deltran's internal documents or
can someone post a clearer picture of component???

[floobydust]

These products are really cheap, the electrolytic capacitors fail. Doesn't matter if Deltran, CTEK, Schumacher, etc. they don't last. Sorry Jay Leno, you got ripped off.

U4 looks like a linear regulator providing power to the MCU. Like LM1117. The MCU p/n would tell you if 5V or 3.3V part. If an ohmmeter reads short from in-out then the whole board is toast I think, as 12V to the MCU would damage it.

All the youtube vids are fanboys with camera focused at the table, not the board so you're doomed to see a better picture.

[DAVEISGREAT]

Thanks for info on cap failure problem.

In another repair forum I believe dead U4 was partially ID'd as ( the S-814___ Series) the last few digits got obliterated.:
https://digchip.com/datasheets/parts/datasheet/432/S-814-pdf.php

The MCU is STM8S003K3T6C
https://www.mouser.com/ProductDetail/STMicroelectronics/STM8S003K3T6C?qs=vV0pr1uq7fDCy7ClrSvsdA%3D%3D

with Operating Supply Voltage:   2.95 V to 5.5 V

Forgive my ignorance but if I get say a 5V output reading on MCU does that mean U4 would also be the 5V option?

Note: S814 series of cmos voltage regulators has a 3 pin type like U4 with SOT-89-3 case, but...
they range in output from 2v up to 6v with only a tenth of a volt seperating each
so choosing the correct voltage regulator without knowing last digits of U4 is very difficult, no ?

also, i was not able to souce a 3pin case option and the other case options seamed very different?

Front and back of board pictures attached:

[floobydust]

That Seiko stuff is never used by anyone, pinout is OUT, GND, IN which I don't think matches the pcb? What is the pinout needed GND, OUT, IN ?
I see the burnt IC U4 has the same logo as U3 AS324 by Diodes Inc. It looks like that to me, so worth seeing what LDO's they have.
I would look at AZ1117C SOT89-3 but four different pinouts offered  SMD marking code G27M for 5V part.

edit: what is Q4?

[DAVEISGREAT]

Yes, pin 1=Vout 2=ground 3=Vin

So you are thinking it may be a diode?

[floobydust]

No, what I am saying is the burnt U4 is made by Diodes Inc. and out of around 7 possible regulator IC's my guess is a 1117.
But the pinout we can't agree on, does 1 or 2 connect to the GND pour? I have 3 as OUT. Pinout SOT89-3 or TO-243AA package, you figure it's OUT, GND, IN?
This is a jelly bean part, dozens of manufacturers make it but they offer different pinouts.

I would test Q4 it likely got damaged and seems to pre-regulate and feed power to U4.

[dc101]

Q4 looks like it might connect to the tab which is vout according to the datasheet below, it also shows pin 3 as input. It looks like U4 is being fed from U2 (top side).

https://docs.rs-online.com/9bfb/0900766b81519f6c.pdf

[DAVEISGREAT]

So what to do if can't find right package, solder some wire to pins then solder to board, and maybe silcone it so it is not flopping around? Actually, this may be easier, soldering one pin at a time especially since I have never worked with SMD before..

Guess it's time to take some voltage measurements and check that transistor.  Thank you both for the great advice!

Update:  I HATE SMD!  The trasistor tested bad but honestly not sure if it was beforehand or my butchering made it bad.  Regardless it gets replaced.  2SC1008 NPN for those keeping score. Had to remove cap right next to it to extricate Q4 -wish I had done that first.  The 50V 100uf cap tested at 89uf, borderline but will replaced.

With a blown regulator, and missing transistor + cap do I dare power it up to take voltage measurements, and what if any VR do you suggest ordering?

[dc101]

I would desolder U4, that should help you a lot with determining the correct pinout. If you don't have a hot air gun, you might be able to get away with using some small wire snipers and cutting the 3 pins, then use a soldering iron to heat up the tab and remove the device, and go back over each pin individually and remove them.

I suppose worst-case, buy both pinouts and use whichever one doesn't blow up

[dc101]

So what to do if can't find right package, solder some wire to pins then solder to board, and maybe silcone it so it is not flopping around? Actually, this may be easier, soldering one pin at a time especially since I have never worked with SMD before..

Make sure the tab is soldered to the board, that is the heat sink for this part. It really needs to be soldered to the board so the pcb can dissipate the heat from the LDO.

[floobydust]

U4 smd marking code I can see a cooked letter G___ but not the next letters, they're burnt but maybe you have a loupe or try wet the top and see if there is a shadow left. This tells us the voltage. I'm guessing its 5V for the MCU.
So it would be an AZ1117CR2-5.0 which is marking G33N (GND, IN, OUT). I think the out goes to the opto LED through large 3.3k resistor R22. You can use an ohmmeter to follow the out pin to the the MCU power pin.

What is Q4 part number? Test it. It looks like a pre-regulator for U4.

[dc101]

So it would be an AZ1117CR2-5.0 which is marking G33N (GND, IN, OUT). I think the out goes to the opto LED through large 3.3k resistor R22. You can use an ohmmeter to follow the out pin to the the MCU power pin.

Is the transformer not the source of input power to the device? U2 seems to route to a tap on the transformer. I'm not following why the output of the LDO would be routed back to the source of input power? To me it appears that the general flow of this board (looking at the top of the PCB) is from left to right.

[DAVEISGREAT]

Update:  I HATE SMD!  The transistor tested bad, but honestly not sure if it was beforehand or my butchering made it bad.  Regardless it gets replaced (2SC1008 NPN for those keeping score. Had to remove cap right next to it to extricate Q4 -wish I had done that first.  The 50V 100uf cap tested at 89uf, borderline but will replace.

With a blown regulator, and missing transistor + cap do I dare power it up to take some voltage measurements, and what if any VR do you suggest ordering? btw, Seiko data sheet confirmed pinout 3=Vin

[dc101]

I would go for the 3.3V first, if that doesn't work try the 5V? They're pretty cheap, so might as well order a few of each in case some don't survive the transplant. You could probably get away with soldering wires like you mentioned for a short test run just to make sure it works, but long term the LDO really needs to be soldered to the board.

https://www.digikey.com/en/products/detail/diodes-incorporated/AZ1117IH-3-3TRG1/5699672
https://www.digikey.com/en/products/detail/diodes-incorporated/AZ1117IH-5-0TRG1/5699673

Hard to say about the transistor, semiconductors are usually more heat sensitive than other components. I burned a couple up during my very first soldering project (a wireless microphone) when I was in school.

[dc101]

To me, pin 1 on this LDO has to be ground. There is a bypass cap connected to pin 3 and the cap is also connected to a flood fill that pin 1 is connected to. Pin 2 does not appear connected to anything. The sure fire way to answer your question is to remove the LDO, and probe the pads.

None of that matches the datasheet for the Seiko, as floobydust also mentioned previously.

That Seiko stuff is never used by anyone, pinout is OUT, GND, IN which I don't think matches the pcb?

Is there another datasheet for a Seiko which has a different pin layout?

[DAVEISGREAT]

Extricated U4 and tried cleaning it up -have a look at pin 3's tab, it got blown out and is connected to pin 2?

[dc101]

I should have clarified earlier, I don't believe that pin 2 is not connected to anything, but rather it is most likely connected to the tab which connects to Q4.

I think it's just excess solder that is obscuring a better view. Do you have some solder wick you can use to remove the excess solder? If not, you can try just soaking it up with your iron and wiping the excess off on a sponge, although a wick is much more efficient. You can fix the lifted trace for pin 3 easily enough when you solder the new component on. Unfortunately that is very common when components explode.

[DAVEISGREAT]

Before lifting U4 I did take some other pics but they were not very decisive either.  The best one "looked like" S814JDA.  If you'd like to take a look, I posted them in another forum, but no one there could make heads or tails of the later blasted digits.

Here is link and it's post #12 with 5 pics:
https://www.electronicspoint.com/forums/threads/battery-tender-charger-model-021-0128-blown-component.295125/

unfortunately, you may have to register to see them, but again, problaby not worth looking at them.

Will order parts tomorrow -thanks again!!!!!!!!!!!!!

[DAVEISGREAT]

There is continuity between the SMD cap you circled to all three pins on U4.  I will try to remove some more solder but I could swear that middle trace near pin 2 connects all 3 pins even though it make no sense.  Also is continuity between 2 of 3 pins on vacant Q4 and vacant U4 traces. 
Any other tests I can do to determine Vin Vout & ground, and if U4 replacement should be 3.3, 5 or some other voltage?

[dc101]

If after you remove all the solder on the board and there is still continuity between pin 1 and pin 2,  then I would suspect another component has failed and is causing a short. Both pin 2 and pin 3 should not be a direct short to ground (pin 1).