Really nasty JDM issues

[Guest Da5id]

OK, my SID is done (finally!) except for programming the PIC.

And here I'm running into some problems.

My JDM fails all of the voltage tests.

I get around 0.23 volts where I should see 14v, and the like.

Any ideas? I've triple-checked the diode polarities, and quadruple-checked the transistor polarities. And I've tried two sets of transistors.

I have as of yet only tried it on one computer, one I set up specifically for this purpose, running Win98. It's an old DEC server I had kicking around.

So where should I start with the troubleshooting on this?

[ThomasT]

Are you sure your com-cable is correct?

I had similar problems until I realized that I interchanged the plugs. I used female where I should use mal or so. I maybe a crosscable should work.

But I got voltages in the expected range, I did not realize it immediately. But clock and data was interchanged and the voltage were a little bit too low, AFAIR.  

[Guest Da5id]

Well, I don't think the cable's the problem, I made it myself and it tests fine.

Just to make sure, it's wired straight through, with shield attached to the connector housing at both ends, right?

I still haven't tried another system yet, that's probably the next step. I was having a hard time getting the PIC software to work on my other windows box (running XP), but I'll dig up a harddisk somewhere and do a temporary install of 98 on the XP system. Crap, I don't think I'll have time for that until Thursday... Grrr...

[ThomasT]

>Just to make sure, it's wired straight through

I don't know. In principle yes. But it can make a difference if you using a female connector on the JDM PCB and a male plug or vice versa.

I don't know how to explain. It's like looking from the frontside or backside.

I don't remember what is right. I only remember that I interchanged the sides. (mirrored)

[Guest Da5id]

OK, well, if I built the JDM exactly as the schematic and board layout show, I would use a normal straight-through serial cable, right?

And the pins on those connectors ARE numbered, and I wired 1 to 1, etc.

Computer serial ports are male, one on the JDM is female, which is in keeping with a straight-through cable.

God, I hope I get this resolved soon!

I'm going out of town tomorrow, not getting back until Sunday evening, and early Monday I have school, and school all week... Eek!!!

[Guest Da5id]

So, I still can't get my JDM working, here's what I've done since last postig:

Set up a different computer, running Win98se just for this purpose.

jdm is connected directly to one of the serial ports, so I've eliminated the possibility of a funky cable.

thuroughly checked the pcb - it's fine.

swapped out the transistors, soldered in new ones VERY carefully, with my iron turned down real low, and allowing lots of cooling time between soldering each pin.

Where it's at presently, is still failing the voltage tests! I have followed the test procedure to the letter, and I get the same results every time.

Here goes:

With all signals disabled, the initial voltages are

12-40:   4.09v

12-39:  -0.64v

12-1:   10.81v

12-11:   4.54v

enable MCLR:    12-1:    10.81v   (no change)

enable Vcc:         12-11:    4.54v   (no change)

disable Vcc:        12-11:    5.04v

                          12-39:    5.84v

enable clock:       12-39:  -0.64v

disable clock:      12-39:  -0.64v

enable dataout:    12-40:  -7.63v

I'd really appreciate some insights here, I'm pretty clueless at this point!

Da5id

[TK.]

Did you ever try the external power supply like suggested in the MBHP_JDM troubleshooting section?

Since I own a new computer with a ASUS mainboard, I've the same problem: I cannot program a PIC18F452 without this external voltage source anymore...

Best Regards, Thorsten.

[Guest Da5id]

No, I haven't... I didn't realize that I would be seeing behaviour this strange from something like that, I'll give that a shot, though...

Thanks!

UPDATE:

Very minor change, the voltages are a little higher, like around .18 volts or so, but it's still totally misbehaving....

[TK.]

Hm - no idea - then it must be a wiring problem.

A while ago two users sent me their non-working JDM for analysis, and in both cases I found bad or missing solderings (and additionally: in one case the transistor was damaged due to a wrong connection...). So the only hint that I can give you: check the informations given at the MBHP_JDM page again and again. Sooner or later you will find the error - and you will know the circuit better than anybody else ;-)

Best Regards, Thorsten.

[Guest Da5id]

OK, after going over my damned JDM about 8 times, I deided to "give up" and buy one of the kits from SmashTV. The parts arrived a few days ago, and last night I finally had time to solder it up.

Exact same results!

So I tried a third computer, and now things are working! (well, almost, my MCLR is only 12.9v, but I think there's enough info in the forum for me to figure it out).

Thanks for the suggestions!