Home built Pickit 2 not behaving normally.

[mikejp56]

Hi All,
I am having a strange problem with a homebuilt PickIt2. It is wired just as MicroChip designed it, but I substituted a few parts to facilitate thru hole construction. I subbed a BS250P P channel MOSFET for the IRLML6402 and the Q2 part of U5, and a TN0702N3 for the Q1 part of U5. I also subbed a MCP6041 op amp for the MCP6001 op amp. The board is recognized by the PickIt2 v2.61 software. The problems are as follows:

1) When first plugged I get the error message bo that says "Pickit 2 VDD voltage level error. Check target and retry operation." I get this error message whether a PIC is connected or not.

2) When I hit OK the PickIt 2 Programmer software runs and shows that a PickIt 2 was found and connected. Sometimes it finds a connected device, other times it doesn't.

3) If it finds a device, pressing Erase or Blank Check brings up the VDD error box again, and then the software reports "No device detected".

When I run the troubleshooter, the following occurs:

Step 1 passes but I always have about .2V difference between what I set and what I read in the results box and on my DMM.
Step 2 Pressing "Test VPP" the first time gives me a "Test Failed: VDD Short Detected" error. I can see with my scope that the level is high, and I can measure around +12V with the DMM. If I press "Test VPP" a second time it reports that the test passed with a voltage reading close to what I get on my DMM. When I press "/MCLR On" the VPP level drops to 0.
Step 3 I can control the PGC and PGD lines during step 3 testing.

I feel that Step 2 is the main problem, but tracing my connections has shown no miswires. Plus if there was a miswire, just running the step a second time would not pass the test.
Did one of my substitutions break something?
Any ideas on how to get this working would be greatly appreciated.

Regards,
Mike

[Bruce Abbott]

BS250P Rdson is 14 Ohms max, compared to 0.08 Ohms for the IRLML6402. This is at -10V Gate drive and the threshold is 3.5V max, so it's probably even higher with 5V Gate drive.

Perhaps the extra resistance is slowing down +V_TGT rise time enough for the test to fail the first time. Without knowing what is acceptable it might be better to just use the correct parts (or close equivalents).

[mikejp56]

Hi Bruce,
Thanks for the quick response. I was trying to build this with all thru hole parts. Maybe I need to use both thru hole and surface mount?
Thanks for your time!
Regards,
Mike

[picandmix]

These 2 well proven pk2 circuits using common through hole parts might help you locate your problem.

They do not use the variable Vdd, but all you need do is run the whole circuit at 3v3 is you want to program 3v3 chips.

[mikejp56]

Hi picandmix,
Thanks for those 2 schematics. I will investigate them.
Regards,
Mike

[picandmix]

Hi picandmix,
Thanks for those 2 schematics. I will investigate them.
Regards,
Mike

You are welcome,  they are 100% proven circuits, that second one being mine that I built as a backup to my original Microchip Pk2 in 2008, still works fine today 

[KL27x]

Mike nailed it. The PICKIT2 checks for short circuit. If the expected target VDD is not met within a very short time, probably on order of 1 mS, it will take this as a user error of short circuit between Vdd and ground rail and turn off the power to prevent damage.

U5 also must be responsive. This part is even faster than the IRLML402. This is a dual trench power FET with sub 100mohm RDSon at only 4.5V and only 10nS turn on times. So you need to pay attention to all 3 of these FETs, but particularly to the 2 PFETs if you want to make this VDD error go away. But slower FETs will also slow the feedback loop and might potentially allow poorer regulation or oscillation under certain conditions.


I subbed all these FETs, on my own clones. But I used faster and cheaper FETs.

Maybe one of the biggest waste of money on the original parts list is the tantalum FET for Vpp generator. I dunno why that is there. It costs like 2 dollars. I use a cheap ceramic. My understanding is Vpp has a very wide range of acceptable voltage and does not need to be particularly accurate or stable, and the draw on it is very low.

[mikejp56]

Hi KL27x,
Thanks for your response. I subbed thru hole parts for the SMT parts hoping that it would not degrade performance much. Anyway I was way wrong! So I have ordered the correct parts along with SOT23 to DIP and SSOT6 to DIP adapters. Damn adapters have a 6 week lead time.
I will report back to this forum once I receive and install the correct parts.
Thanks again for reading and lending your experience!
Regards,
Mike

[KL27x]

I just searched for comparable thru hole transistors, thinking there would be some. Wow, who knew thru hole has been dead this long?

[mikejp56]

Hi KL27x,
Yeah I didn't think it would be such a problem.
Oh well!
Regards,
Mike