Prop seems normal, yet pins are inoperative.

xanadu

I have a Propeller board that is programmable, and will return text over the PST no problem. The IO pins seem to be toasted and I'm trying to figure out why. Currently pins 30&31 work fine. Pin 29 is stuck high, and the rest of the IO pins are +80mv and uncontrollable. I was using 12 IO pins, and not drawing current from any of them. The 12 pins I used were in the lower range, 0-15 skipping pins 6-8. It's not the code or anything like that as I have swapped out different boards to verify everything else is okay.

I am carefully going over my circuit that was attached and there is nothing I can find wrong with it. It all works fine connected to another Prop. Does this sound more like I abused some IO pins, or 3.3v Prop power supply abuse?

I could believe the power supply issue, it was cheap. But would overpowering a Propeller cause what I describe?

Thanks.

SRLM

What about adding a pullup to a pin, and then driving it to 0v? Or adding a pull down and driving to 1v? Does input work?

xanadu

I tried and nothing. I used the PST to show the state of each pin, made a little wiring rig and tested a bunch of pins. They all show 0 even when pulled high. I swapped out the board to ensure the code and wiring for the test worked and it did. The PST is beginning to act funky now. Sometimes the data takes up to a couple of minutes to display.

I guess all I can do is post a schematic, but I maintain there is little chance it was the circuit. I was reading some encoders and driving a few transistors. Stuff I've been doing for years without issue. I guess I'll replace the power supply and run it for a week and see if it burns out again. If it doesn't I will have to assume it was the power supply or somehow something shorted, but we're talking 5v input power and resistors on the IO pins soo.... arrrg...

Phil Pilgrim (PhiPi)

Which Prop chip are you using (Q44 or D40), how many bypass caps have you installed, what values are they, and where are they located?

-Phil

xanadu

I'm using my favorite M44D40 by WBA. That's the M44. It has 3 on-board bypass caps. I'm not sure if I can test them in circuit. I was able to compare results with known good board on various test points and so far everything checks out but the IO pins (except 30&31).

Maybe I should be asking if there is something that would protect pins 30&31 from damage if a certain condition exists, and what that condition might be?

At failure time: It worked and was purposely powered off. Upon the next power up it was dead. It didn't have a meltdown or act strange when powered up.

Thanks.

Phil Pilgrim (PhiPi)

The Prop chip on that board has only one bypass cap that I can see. That's a bit on the scanty side. With inadequate bypassing, the Prop has demonstrated a failure mode wherein the PLL does not operate. Try running a pin-test program in one of the non-PLL clock modes and see what happens.

-Phil

xanadu

Phil,

That works. I am able to toggle pins in non-PLL. The documentation shows 3 bypass caps, and they look like caps. I will replace and test. Do you think a power supply spike could cause a cap(s) to fail, leaving the Prop itself undamaged?

I have a VIN of ~8.6v (2S lipo).

Thanks.

Phil Pilgrim (PhiPi)

xanadu wrote:

The documentation shows 3 bypass caps ...

The board layout appears to differ. I see only one applied to the Prop itself, and that's not enough.

Do you think a power supply spike could cause a cap(s) to fail, leaving the Prop itself undamaged?

No. Given your symptoms, I'm virtually certain that the Prop chip itself has been compromised.

Also, I should add that, anytime you transfer power from one board to another, you also need a filter cap (10 uF+ tantalum or multi-layer ceramic) at the power-input point, which the board you're using seems to lack.

-Phil

Publison

I went through Andrew's other threads, and one picture show two additional caps on the bottom side of the board.

Phil Pilgrim (PhiPi)

Publison wrote:

I went through Andrew's other threads, and one picture show two additional caps on the bottom side of the board.

Ah so; that explains the discrepancy, then. Can you provide a link? I'd like to see where they're positioned and how they're connected.

In any event, the loss of PLL is consistent with poor bypassing but, in this case, may not be caused by it.

-Phil

xanadu

There are two caps on the bottom, one on top. I can get a pic this evening. It sounds like there's a chance the caps are fine, and I did something to an IO pin to cause the failure. I'll pull the caps and test them to be sure.

Publison

Phil Pilgrim (PhiPi) wrote: »

Publison wrote:

I went through Andrew's other threads, and one picture show two additional caps on the bottom side of the board.

Ah so; that explains the discrepancy, then. Can you provide a link? I'd like to see where they're positioned and how they're connected.

In any event, the loss of PLL is consistent with poor bypassing but, in this case, may not be caused by it.

-Phil

Could be in here:
http://forums.parallax.com/discussion/131867/m44d40-propeller-basic-circuit-module-in-dip40-form-factor

or here:http://forums.parallax.com/discussion/123776/program-to-fully-test-the-propeller-for-new-qfn-dip40-module/p1

I had to download the photo an zoom to see the caps. The vias on the top side seen to match the power pins.

Phil Pilgrim (PhiPi)

I see maybe one SMT component on the bottom, but it's hard to tell what it is.

-Phil

xanadu

There are two on the bottom identical to the one on top. The color and size look appropriate for a ceramic capacitor.

Phil Pilgrim (PhiPi)

xanadu,

Whatever the case, I think the Prop chip is toast, beyond repair, since this is a well-documented failure mode. You need to move on and get a replacement module.

-Phil

xanadu

Phil,

Agreed, replacements are on the way.