PDA

View Full Version : A short on the Prop Proto USB Board?



Timothy D. Swieter
07-04-2009, 03:12 PM
I am with the Prop Proto USB Board. I am creating two identical devices by soldering the components to the prototyping area of the Proto Board. The first board I created, I completely soldered up. When I first powered it on there was some missing parts and pieces, as to be expected, but no shorts. As I added the parts and pieces, interspersed with downloading test code to test LED and button functionality, I noticed the power LED glowing dimmer than usual. I quickly figured out a short had occurred! Now, I thought that in my late night soldering I someway criss crossed some wires and made a short. I spent considerable time the following evening (all fresh) looking for the problem but couldn't find it.

So....I set out to build the second identical device. With this device I add each section of components and tested them as I went along. Constantly looking for shorts, adding the test code for the Propeller, etc. For the past hour or so I have only been working on the software for LED, buttons and a couple transistors. My serial port terminal from the Prop Proto USB Board started to hang. Again I noticed the glowing LED and the overheating regulator.

I checked this second board and it now has a short. The board has been working fine, but in the middle of software testing it developed a short. This sounds very odd. My work space is clean. I have added any items. The code was running and while then - over heating.

So....I am looking for some advice on narrowing down this problem. What could cause this? The device currently has two LEDs, three buttons and a couple Prop I/Os to the base of NPN transistors (with a series resistor). I am thinking this is a system design problem, since the problem now appears on both boards.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com (http://www.brilldea.com) - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com (http://www.tdswieter.com)

StefanL38
07-04-2009, 04:24 PM
Hello Timothy,

I think it would help us to analyse it
if you post a schematic of your circuit

best regards

Stefan

Timothy D. Swieter
07-04-2009, 10:57 PM
True - it is on;y hand sketches at the moment. Tomorrow morning I will scan it and post it.

On one of the shorted boards I started removing the wires and connections on the 3.3V/Vdd, that is the bus with the short. I have removed everything in the circuit that I added to the board, but there is still a short. So I did some resistance testing and trace cutting and it appears the short is in from the center portion where the Prop is. Although I am not sure I cut all traces for Vdd to the center - is this design two layer or four layer? So, if there is a short in the center, maybe it is something with the Propeller IC that the external circuits caused to fail.

Tomorrow a schematic.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com (http://www.brilldea.com) - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com (http://www.tdswieter.com)

Timothy D. Swieter
07-05-2009, 05:50 PM
I attached a hand drawn schematic and a couple photos of the front side of the two assemblies. I am trying to find a short, a short that developed during testing of the board.

To repeat, I soldered up fully Board 1 while during continuity checks. The board checked out OK and I started doing firmware. I then noticed the regulators hot and the LED glowing dimly. There was a short in the 3.3V line. The testing was being done with no external equipment connected.

Board 2 was started after the above incident. So, as each section was soldered on I checked for continuity and grounding and power. I have only added the LEDs, the buttons and some of the transistors. I was writing software for a over an hour. Things were going fine and then the short occurred again. So, I am thinking the fault is in the circuitry and is somehow shorting the Propeller or an I/O pin, but it is a delayed short in that it showed after being powered on for a while. Again without any external equipment attached other than a power cord and the USB programming cable.

Is there a potential for ground loops or shorts to develop through the USB bus?

I hope to find a resolution to this problem soon and before I swap out the Propeller IC - if the short is there.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com (http://www.brilldea.com) - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com (http://www.tdswieter.com)

shanghai_fool
07-05-2009, 06:50 PM
Timothy, I can't tell about board2 but on board1, it looks like a transistor pin is connected to 3v3 near P23. It's on the second row of pads that connect to the prop.

Timothy D. Swieter
07-05-2009, 07:19 PM
Thanks for check Shanghai_fool. I believe the transistor is OK, it is just outside of the double row of pins that connect to the prop. I am thinking this because there isn't a silk screen line between the pads that you see soldered by P23. When I first put together Board 1 I did have this error, but caught it quickly and fixed it before I added the remaining transistors.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com (http://www.brilldea.com) - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com (http://www.tdswieter.com)

Whit
07-06-2009, 09:07 PM
Timothy,

How about pics of the backsides of the boards? (could the shorting element be there?)

Also, it occured to me (because of the description of the problem above), could an I/O pin (in software) be being used for conflicting purposes in seperate places in code? Could that create such a problem? I honestly do know if this is possible, but you said the problem showed itself during software tests.

On board 1, what do I see above the Vdd and Vss that·are between P7 and P8? Something appears to run diagonally from the outter Vss (near P8) the the pad 45 degrees above it to the right and and from there, then 45 degrees down to the right.

Hope you get your problem solved!

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Whit+


"We keep moving forward, opening new doors, and doing new things, because we're curious and curiosity keeps leading us down new paths." - Walt Disney

Post Edited (Whit) : 7/6/2009 1:58:26 PM GMT

Leon
07-06-2009, 09:37 PM
This might be useful:

www.edn.com/article/CA408390.html (http://www.edn.com/article/CA408390.html)

Low resistance meters are made, but they are expensive.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

Timothy D. Swieter
07-06-2009, 09:53 PM
I am positive this is not a software problem - as the short exists when the board is unpowered. However that isn't to say that the I/O didn't fail some how because I wired something to an I/O pin that once toggles caused the pin to fail.

On board 1 what you see above P8 are cut traces of the two 3.3V traces to the center area. I was trying to segregate where the short may be on the board so I started cutting traces. I think these are the only two traces to the center area and with the cuts in place the short is showing up on the VSS/VDD pins around the propeller.

I am thinking of removing the propeller IC, however I am not quick to slap another one on till I understand why both boards failed in the same way.

Thanks for the link Leon - I will certainly read through and study that article tomorrow. I may have to cook up such a tool.

Keep the suggestions and questions coming as they are helping me to think through the problem and to double check my work.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com (http://www.brilldea.com) - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com (http://www.tdswieter.com)

Leon
07-06-2009, 10:11 PM
Many years ago I bought a gadget like a logic probe for finding shorts on PCBs. It had a tone output on a little piezo speaker and the tone changed as it got nearer to the short. I found it quite useful at the time, as I was selling a lot of transputer systems on four-layer boards and I had shorts on a few of them. It was probably just a couple of op amps for measuring low resistance values and a CMOS tone generator; it was quite cheap.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

Rayman
07-07-2009, 01:06 AM
You might try measuring resistance from your Prop I/O pins to ground (with the thing off)... I think maybe I fried a chip once and they read low impedance...
I just checked on my working demo board (no power) and the pins read in the mega-Ohm range.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My Prop Info&Apps: ·http://www.rayslogic.com/propeller/propeller.htm

Pliers
07-07-2009, 02:35 AM
A capacitor in backwards can work for a long time before failing.
Double check the polarities.

Some caps go poof right away some don't.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
11100011100011101101010100001111100011100011100101 01011

Timothy D. Swieter
07-07-2009, 07:51 AM
Getting closer, but no "Ah-ha" yet.

Rayman I did as you suggested and found the I/O pins to be in the mega-ohm range as well. At first the three I/O to the switches was telling me 81.6K or around there and I quickly realized that this was because the 3.3V bus is shorted and the current was taking the path of least resistance which was through the pull-up resistors. I desoldered that I/O briefly and confirm the pins are in the high 2 mega-ohm or low three.

Pliers - good though on backwards capacitors. In this design all capacitors were already insert and assembled by Parallax. The ceramic caps aren't polarity sensitive, however the large electrolytic cap for the servo power and the small tantalum cap for the 5V regulator hiss are polarity sensitive. From what I can tell both of these caps are inserted correctly, plus the short is on the 3.3V bus and not the 5V bus. Recently I had a design where the tantalum caps were installed backwards and you are right, it worked for an hour or two of time and then smoked.

So, if the I/O checks out, then perhaps the Propeller power is some how wrong? The other 3.3V circuit would be the FTDI USB circuit. Any ideas on how to check that circuit out? Could there have been some sort of ground loop between my laptop's USB and the circuit powering the PCB that caused the FTDI IC to fail? I will check the IC and serial comms a little later today to test this and report back.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com (http://www.brilldea.com) - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com (http://www.tdswieter.com)

Timothy D. Swieter
07-21-2009, 10:01 PM
Alight - here is an update, I hope someone can shed some ideas into my abyss of trying to solve this problem.

One Board 1, I cut out the propeller and soldered in a new IC. I suspected that perhaps somehow the Prop IC got fried. With the new IC in, I double checked for shorts before powering up and there isn't any. I powered up (only applied power), voltages look fine. Powered down, still looking fine. I plug the USB cable into the PCB (remember it is a USB Proto board, so just the usb cable) and things are fine. As soon as I plug the USB cable into the PC a short appears. This isn't making sense to me. It is happening on both boards.

The USB cable that I am using is the exact same cable that I use with my Prop Plug. This USB cable has been in service for a good year or more and has worked fine thus far. In fact, earlier today I was using the USB cable with the Prop Plug for another product.

So, what is it about the power/intialization of the USB portion of the prop proto board that would cause it to short VSS/VDD? I am going batty trying to find this problem. I am thinking I may need to rebuild the device on the standard Prop Proto Board because I know my prop plug is OK.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com (http://www.brilldea.com) - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com (http://www.tdswieter.com)

WBA Consulting
07-22-2009, 12:28 AM
From the added explanation, it sounds like your FTDI chip/circuit has an issue. It is only powered up when connected to a USB port, so that would explain why the problem only appears then.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Andrew Williams
WBA Consulting
IT / Web / PCB / Audio

jazzed
07-22-2009, 02:41 AM
Timothy, I found in one or more situations that having the PropPlug connected to a power hungry board without the external power connected (switch in off position) caused damage to the PropPlug. I don't know what improvements the USB proto-board has over PropPlug. Your board doesn't look very power hungry, but without a schematic it's hard to tell. I never connect a PropPlug without power to the propeller board these days after seeing one get beat up.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
--Steve


Propalyzer: Propeller PC Logic Analyzer (http://www.brouhaha.com/~sdenson/Propalyzer)
http://forums.parallax.com/showthread.php?p=788230 (http://forums.parallax.com/showthread.php?p=788230)

Phil Pilgrim (PhiPi)
07-22-2009, 03:06 AM
Timothy,

It sounds like you might be outputing to the serial I/O pin that's supposed to be an input. You won't get a bus conflict, though, unless the FTDI chip is powered up from your PC and also driving that pin. All you have to do to cause a problem is to set dira[31] high. This will drive the pin low, while the FTDI chip is driving it high (normal idle state).

-Phil

Rayman
07-22-2009, 09:06 PM
I bet Phil's right...· Jet another example where having some series resistors between FTDI and Prop can prevent common issues...

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My Prop Info&Apps: ·http://www.rayslogic.com/propeller/propeller.htm

Timothy D. Swieter
07-23-2009, 07:21 PM
Thank you again gang for the feedback, unfortunately I still haven't found the problem.

I checked the FTDI circuit and what Phil suggested. I have nothing else tied to P30/P31. I also double checked the software to ensure I wasn't doing something screwy and setting the pins to an output by accident. No problems found. I am using the Parallax Terminal Object and prior to the board failing I was programming away and getting serial/terminal data back. The slow/latent failure mode is what keeps me wondering what happened.

I cut out one of the Propellers and soldered in a new one. The Propeller isn't getting recognized, but the USB Virtual Com port is getting initialized. Voltages look OK. Solder job looks OK. Orientation looks OK. I am afraid I may be getting multiple problems stacking up on me and that the time windows that I have to resolve the problem are too far apart or too small and so I am rushing.

This weekend, I will build a new board, this time on the non-USB proto board. Just like board two, I will methodically build it, test it, ring it out as I go along. Third time is the charm, right? I will be using the schematic that I posted up on July 5th.

I'd like to think I don't give up easily, but this problem(s) is really making me scratch my head. Hopefully I will find the problem and then the head smack will probably be heard around the world. http://forums.parallax.com/images/smilies/wink.gif

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com (http://www.brilldea.com) - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com (http://www.tdswieter.com)

Rayman
07-23-2009, 08:21 PM
I just took a look at your schematic... I see some transistors with no obvious resistors on the collector... Just a guess, but perhaps you're turning on a transistor that directly connects Vdd to GND?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My Prop Info&Apps: ·http://www.rayslogic.com/propeller/propeller.htm

Timothy D. Swieter
07-23-2009, 09:54 PM
Thanks for looking at the schematic Rayman. The MP3 Player interface and the LED interface are just headers at this point with no external connections yet. I have checked the boards here to see if I got the transistors backwards or solder bridged or anything wrong but don't see it. Of course that could be the problem, I don't see it - a flaw in my understanding.

Do you recommend normally putting some lower value resistor on the collector to limit current? for this particular setup, it might be nice if I did that since I don't have full control of the external connections.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com (http://www.brilldea.com) - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com (http://www.tdswieter.com)

Rayman
07-23-2009, 10:07 PM
Ok, if there's nothing yet connected there, it can't be the problem...

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My Prop Info&Apps: ·http://www.rayslogic.com/propeller/propeller.htm

Rayman
07-24-2009, 01:09 AM
This one area on the first board looks very close to shorting Vdd to Vss...· But, it's hard to say from the picture...

Probably not the real problem since it happens only when USB is connected...

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My Prop Info&Apps: ·http://www.rayslogic.com/propeller/propeller.htm

Timothy D. Swieter
07-24-2009, 08:05 AM
WOW - good eyes. Those are actually cut marks where I disconnected Vdd from the center area with the Propeller. I was trying to isolate where the short may be so I cut the two small traces of VDD that feed the Propeller and EEPROM.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com (http://www.brilldea.com) - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com (http://www.tdswieter.com)

Timothy D. Swieter
08-06-2009, 02:57 PM
Over the past week I rebuilt the circuits on Propeller Proto Boards, the Prop Plug variety. So far with a few days of testing with and without the devices the boards have been running fine. Same circuit, same component types, same firmware, only difference was a Prop Plug versus the Prop Proto Board USB. I am suspicious of ground issues between a laptop top, the Prop Proto Board, etc. Perhaps something occurred that took out the FTDI IC and causes the 3.3V and VSS to short.

For now I am back on track to get the project done, hopefully this weekend. The fried boards though are stashed away as a curiosity to solve at a later date.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com (http://www.brilldea.com) - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com (http://www.tdswieter.com)