The day I killed my Atten PPS 3005

[Hole]

So I bought myself a present and it was a Atten PPS 3005S with USB from ebay. First thing before actually using it was to modify the fan. I tested a 7805 with the build in fan and found that it works. Stupid me than decided to optimize the solution. I found a 45° C NO switch and added it across the 7805 as a security solution.

Will testing it I touched the heat sink of those two power transistors with (I think) ground (of the switch or 7805). It resulted in a lot of smoke (really!) and a dysfunctional Atten.

Visual inspection led to the ULN2003A (see picture 1). I repaired that (see picture 2). But that seems not to be the only problem.

Current status is: when switched on, the display shows all 8's, then sometimes clears, I can see the numbers scrolling away. This lets me think that the Atmega is not dead. The fan does not run (I checked it, it works).

So finally the big question: how to search for the error? I'm not very experienced in that topic, so any guideline and ideas would be a great help.

(I found the schematic of the analog part in the forum.)

Any help is welcome!

[mariush]

Start by checking voltages.. input to regulators , output of regulators, voltage going into opamps, check for shorts on transistors and mosfets (continuity mode on meter and check gate-source , gate-drain , source-drain etc) ...

[dannyf]

how to search for the error?

I would take the schematic and figure out its key functional blocks and go from there. Starting from the rectified ac first.

[David_AVD]

I found a 45° C NO switch and added it across the 7805 as a security solution.

Is that a typo?  how exactly did you wire that switch in?

[Hole]

--A----[7805]--->|---B---

From A to B.

[Thilo78]

Hmm... I'm not quite sure, but did you notice that solder bridge giant solder blob on the two pins of the ULN2003A ?

I don't think that this would be original or even work in any way. 

[senso]

Those two shorted pins appear to have a nice big square pad of cooper under them, at least, appears like that.

So, when your switch closed the idea was to bypass the 7805 and give the fan the full 12v?
I don't think the 7805 will be very happy with the same voltage at input AND output.

[Thilo78]

Those two shorted pins appear to have a nice big square pad of cooper under them, at least, appears like that.

Looks very strange. Never seen something like that before. 

So, when your switch closed the idea was to bypass the 7805 and give the fan the full 12v?
I don't think the 7805 will be very happy with the same voltage at input AND output.

I might think that equal voltage may be okay for the 7805. Not sure about 12 V, though.

Pretty sure it will die with 12 V on the output and 5 V on the input. (tried this once...)
Maybe it would be better to completely bypass it instead of feeding the 12 V back to the output...

[mariush]

If that's how the pins on the pcb were made, it's perfectly fine.

ULN2003A is like I said an array of darlington transistors.. on the bottom side of the chip (below the text) you have the base pin of each transistor (7 of them), on the top you have the collectors for each transistor. The emitter is common to all and it's the eight pin at the bottom.




It's quite common to short two commons or even several of them when there's high current requirements .. each collector can do maximum 500mA, you can join them to do up to about 2.5A for short periods of time .. see figure 14 and figure 15 in datasheet (page 10) : https://www.sparkfun.com/datasheets/IC/ULN2003A.pdf

-------

Did you measure the voltages like I said?

Get some pictures of the whole board so that we can give you some tips (what to measure, what may look faulty/dubious to us etc)

[David_AVD]

--A----[7805]--->|---B---

From A to B.

That's not really any clearer.  Do you mean you shorted the 7805 input to output with the thermo switch?  I doubt that's a good idea.

[Fraser]

Visual forensics.......

In a case like this I always search the area where the 'event' occurred with a strong magnifier. I am looking for the evidence of a spark burn mark on items that could have potentially been shorted. If you find such, you need to look at the schematic and consider what a short circuit from that point to 0V would actually do. If there is no evidence of a contact point, but you have a suspicion a heat sink was involved, work out the effect of a short to those points from 0V. A lot can be learnt from physical evidence and considering the effects of such. As you have the schematic you are very lucky. If you want to post the schematic and detail the suspected short circuit points I will be happy to consider the impact on the components.

[Hole]

The fan solution: atten03.jpg shows the circuit I use. Perhaps the diode is a Schottky, I don't remember.

The solder blob: Due to the hole in the IC to identify it I used a photo in one of the Atten-forum threads. And those two legs sit on one copper pad. So I soldered them together.

Images and schematics come from https://www.eevblog.com/forum/reviews/playing-with-the-atten-pps3005s-dc-power-supply/

mariush, I will start checking voltage next.

Next step I thought was to replace the 595 just below.

[PA0PBZ]

The fan does not run (I checked it, it works).

Is the 7805 still in place? Measure the input and output voltage.

[Hole]

The fan circuit works well. I tested it with 12 V from my other power supply, it does what I think it should do. Fan runs with 4,something volts, heating the switch it turns into full speed. The 7805 seems to like it.

I see nothing wrong with it. On the other hand: the pps 3005 worked without the fan. So this should not be the problem at all, I think.