Pickit 3 repairable?

[dr pepper]

So working on a project I think I put 9v accross one or more of the pins on my pickit.
I didnt thinbk it'd do any harm, but its dead.
The software comes up with overcurrent every time I plug the pickit into the usb port, even with no pic chip connected to the pickt just an open 6 way connector.
Going into the settings screen you can turn on the 5v power, but whether its turned on or off if you go into the status screen it says theres 4v present.
The end of the pickit case gets warm so summats up.
Is there a common repair for this, or should I just bin my programmer and get another.

 

[Mickster]

Troubleshooting boils down to time vs cost...

For around 28 quid, less than the hourly rate for pretty much anything these days, buying a new one is most likely the better option but you don't necessarily have to bin it.

You could save it as a side project, for when you have less important things to do.

 

[dr pepper]

I see what you mean, I think I'll limit an hour looking at it just in case its easy.
The voltage reg is turning on the power led so it looks like the programmable reg is working, looking at the schem u6 switches power to the pic connector, I'm thinking maybe the body diode in the fet in u6 has gone short due to a reverse polarity feed back up the power o/p.
If I turn off the power in the settings section and try and power the pickit from a seperate 5v supply it pulls loads of current, so I'll pull u6 and try it again, if it doesnt pull current then thats it, I'd then be able to program a device that has seperate power, maybe I could replace the smd with a couple of vn10km fets I have.

The other way of course is to feed 5v back up the pickit with the internal power set to off and see where the smoke comes from, if its from the main chip then dustbin.

 

[Nigel Goodwin]

The other way of course is to feed 5v back up the pickit with the internal power set to off and see where the smoke comes from, if its from the main chip then dustbin.

Don't dustbin it! - I couldn't bear to think of it.

If you can't mend it, I'll pay you to post it to me and I'll have a play with it, I've already got a PICKit3, so could presumably program a new PIC for it if it was dead.

Or find another (more deserving) candidate, don't just bin it.

 

[Mickster]

I wouldn't even attempt to backfeed it and look for the magic smoke.

For the price of a new unit and knowing how ridiculously easy Microchip made allowances for the case to be opened, I'd spend a little time taking measurements between the two.

You already have the freely available schematics.

 

[Mr RB]

9v applied to an input pin or to Vdd generally won't blow a PIC.

It has probably blown the 5v analogue switching IC between the internal PIC and the output connector.

 

[Wp100]

I didnt thinbk it'd do any harm, but its dead.

Hi,

Did you buy it from one of the big suppliers ? - if so speak to their technical dept.

If you are seen as a trade buyer they will probably just exchange it for a new one.

They did it for me when my Pk2 failed; their / Microchips attitude was they do not want the Pickits failure to stop your development /production .


Wonder why RS have reduced the price to £28 + vat ? - something new on the horizon ? - even ebay ones are dearer .

 

[Mike - K8LH]

The two or three times I've seen anyone mention a blown PK2 or PK3, they've also mentioned getting a free replacement after talking to someone at Microchip...

 

[Mickster]

MC do have a pretty generous RMA system, they also provided an enviable service with regard to sampling.

Anything is open to exploit.

A dead unit which is likely attributed to user error does not warrant replacement by the supplier.

In the same shoes, I'd buy a new unit and chalk it up to experience.

 

[dr pepper]

Well home now so had chance to have a closer look.
I dont think its a power issue, not as far as the tartget pic is concerned, it looks like the bus transceivers for dat and clk despite having zeners and thermistors protecting them have fried, and they are pulling down the supply rail withion the pickit.
My usual suppliers dont list the 74lvct45 but RS do, so I'll order some when I'm back to work next week.
The pickit still connects to mplab, you can upload firmare to it and the status leds still work, so the internal pic still seems to be behaving.
If new xceivers dont work then it'll be on its way over to you Nige.
T'other thing is, when 5v for the pic came from the pickit it didnt run properly, with 5v from my relic bench supply it works fine, I think its a ground issue with the 9vac supply.

 

[be80be]

Yep Microchip will give you a new one. They done that for a bunch of people. That's if you have a real one from them.

 

[dr pepper]

Yes its a realun, heard some of the copies didnt work so I bought a proper one.
I wouldnt expect microchip to replace something blown up by my own stupidity, and its a couple of years old.
Looks like you were right roman with the xceivers.

I did recently get a zif socket that connects to the pickit from malaysia, and it was quite good, even has a 3m zif socket, so I might just get their version of the pickit for a tener.

 

[dr pepper]

Well taking the xceivers off the board made a difference, I can now set the o/p voltage to what I want and it reads back correct, as well as showing correct on my meter, so I'm fairly sure its the drivers.

BUT I being old and going blind I messed up and pulled one of the pads off on each of the xceivers, ones come off alltoghether and the other is still stuck to its track.

I've spent enough time on this now and I'm only gonna make it worse, so nige do you still want it, I'm not to a stamp pm me and I'll send it.

 

[large_ghostman]

i messed my PK3 up and sent microchip a message explaining i was at fault, they asked for the serial number and 2 days later i got a new one from them!!

 

[dr pepper]

2 days, that was quick.
I blew mine up through an act of laziness and stupidity so I spose it serves me right.

 

[dr pepper]

Dropped it in the postbox last nite Nige, in a jiffy with 3 second class stamps, not sure what class it'll go but it'll probably be there sometime this week.

Let us know how you get on, probably working in 30 minutes or something.

I've been using my kitsrus programmer, and realised just how convenient the pickit is.
I've ordered a pickit 2 from malaysia, should have got a 2 the first time.

 

[Nigel Goodwin]

Dropped it in the postbox last nite Nige, in a jiffy with 3 second class stamps, not sure what class it'll go but it'll probably be there sometime this week.

Cheers.

Let us know how you get on, probably working in 30 minutes or something.

Sounds unlikely, presumably I'll have to order the SM chip

I've been using my kitsrus programmer, and realised just how convenient the pickit is.
I've ordered a pickit 2 from malaysia, should have got a 2 the first time.

I've got both, there's not much to choose between them - I got the three because the two didn't support the chip I was using.

 

[jpanhalt]

The PK3 can be pretty finicky, and there were problems with some of the early units. That and the fact that Microchip realizes the value of people having working PK3's is probably why its replacement policy is so generous. I suspect it might even give them away, except for the potential for abuse.

Here's a piece of advice I picked up last week. If you are powering the target from the PK3 using 5V, and the PK3 returns an error for the wrong chip ID, try lowering the voltage to 4.75 or even lower. The 12F683 is known for having that issue. There may be others.

John

 

[dr pepper]

The pk3 I just sent is the one with the clear case, the early ones had a opaque case.
Oh yeah, pin 1 on the missing 2 xceiver devices (rxd and txd) is the same orientation as the remaining one (vpp), Ie all the píns 1's face the same way, rs comps do the xceiver chips.
When you set the vcc to 5v on the pk3 I had the v returned was 4.88, that'll be why then jp, and is probably why the copies also have trouble as they probably chuck out 5v.
The pk3 I used wasnt brill with 10f200's, after a couple of dozen writes the pickit would come back with missing osc calibration data, and the calibration word would be corrupted, I beleive this isnt an issue with the pk2.

 

[jpanhalt]

When you set the vcc to 5v on the pk3 I had the v returned was 4.88, that'll be why then jp, and is probably why the copies also have trouble as they probably chuck out 5v.

I am not quite sure what you are saying. Why would 4.88 Vcc (actual) only affect some chips and not others?

John