Blown PDB . . . with Fire
Bill Chennault
Posts: 1,198
All--
While working on Ugly Buster's new brains, I came to the point wherein I needed to do some additional testing of the PDB off my 12 volt batteries. I had previously done so successfully, but had taken it apart.
Well, you can guess what I did. I reversed the polarity and sent 12 volts at 24 amps straight into the PDB power connector. A nice fire ensued.
In the following picture, you will see the damage. I have some questions.
1. WERE those the two 10uF capacitors that you see as charred embers?
2. Is it worth trying to fix the PDB by replacing these two capacitors?
3. Along the lines of #2, I wonder how far into the PDB the reverse polarity reached. Just to the·LM1084 or further?·(I had a BS2-OEM in the breadboard area . . . it is also dead now.)
The OSC1 pin was merely charred by the blast affects and ensuing fire.
Thanks!
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
While working on Ugly Buster's new brains, I came to the point wherein I needed to do some additional testing of the PDB off my 12 volt batteries. I had previously done so successfully, but had taken it apart.
Well, you can guess what I did. I reversed the polarity and sent 12 volts at 24 amps straight into the PDB power connector. A nice fire ensued.
In the following picture, you will see the damage. I have some questions.
1. WERE those the two 10uF capacitors that you see as charred embers?
2. Is it worth trying to fix the PDB by replacing these two capacitors?
3. Along the lines of #2, I wonder how far into the PDB the reverse polarity reached. Just to the·LM1084 or further?·(I had a BS2-OEM in the breadboard area . . . it is also dead now.)
The OSC1 pin was merely charred by the blast affects and ensuing fire.
Thanks!
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
Comments
Often, it's some "Power" or "Ground" trace that opens from the excess current.
Having said that, I did once have a power supply fail, and in that failure bump +5 to +12. It did blow out all the TTL logic I had on that board. So over-voltage CAN destroy multiple devices.
So, what you have to ask yourself is, do you feel lucky? Is it cheaper to replace the board, or spend the time replacing components on the blown board in the hope that's all that's wrong with it?
If you've got the time, it might be faster to 'fix' the board. Personally I find it's faster to get a new, un-blown, assumably 'more reliable' board. And if the new board is $20 or so, then that's the way to go.
Oh, and you have my HUGE sympathies for this.· What a pain in the butt this sort of thing can be.
I thought most linear regulators protected themselves against hooking up power backwards, though. Check the data sheet.
Yes. I am quite proud of it. What an explosion! And, Man! You should have seen the fire! It not only scorched the nearest OSC1 pin, it even got the side of the power switch closest to the blast!
But, wait! There's more! (Don't delay! Call now! Operators are standing by! Actually, they are sitting down, but that's not important right now.)
I had a BS2-OEM plugged into the breadboard area running off of Vdd. Although it shows no damage, it is DEAD, also! Can you top that?
If I don't get an answer concerning how likely it is that the PDB can be repaired by replacing the caps and the regulator I will buy a kit BS2-OEM and use the parts to fix the blown BS2-OEM. I am fairly confident that the same thing is wrong with it as is with the PDB. I will probably just go ahead and get another PDB while I am ordering stuff, but I sure would love the satisfaction of fixing this one.
After the horrible detonation and resulting all-consuming fire, I did something more fit to my capabilities: I went out in the garage and painted a chair with an aereosol spray paint. (I never mastered the brush.)
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
Post Edited (Bill Chennault) : 4/12/2008 3:00:45 PM GMT
I suggest that you remove the regulator and those caps and try running it (the PDB and/or OEM) with an external 5V source that you might (??)·have.
Here is how I would check it-
First, remove the charred remains of the caps and also the 5v regulator.
If you have a meter, ohm from the regulator pin 2 hole (Vout) to ground. + lead on pin 2. If you get anything other than a dead short (or open) you might be ok.
If you have a 5vdc source, connect it to the pin 2 hole. That should power everything on the board except the stamp sockets (Vin). No smoke? Proceed!
Short Vin to Vout (pin 3 to 2)- that should give you power to the stamp sockets at 5vdc
If you don't have a 5vdc source then run to the shack and get a 7805 5vdc regulator (don't sweat the current rating- just for testing). The pinout is NOT the same as the 1084 though, so you will have to bend the leads in an unnatural way. The 7805 pin out is: 1- Vin, 2- Ground, 3- Vout, compared to the 1084 which is 1- Ground, 2- Vout, 3- Vin. You don't need the caps for the board to work. Apply power and check it.
If it doesn't work take it out and spray paint it!
Once you get is back together you might want to add a diode inline with the power. That way if the polarity is reversed it just does nothing. On a DC powered piece of test gear I made up for checking motors/encoders I added a bridge at the connector where the power comes in. That way it works no mater how the power leads are connected.
Robert
http://www.parallax.com/Portals/0/Downloads/docs/prod/schem/PDB-RevD-Schematic.pdf
jmalaysia, thank you for the detailed instructions! I have a schematic from Parallax. It tells me the two caps are 10uF. I will take a look at the link you provided.
PJ, how can I say this sheepishly? The leads were already marked. The REAL problem was that I did not have my "robotics" glasses on. I don't need them for reading, but I do in the gun shop and in my machine shop and OBVIOUSLY at the robotics workstation. (Somehow, I get the feeling you sensed that already. Cyber-sense, perhaps?)
Robert, will you recommend a diode? RS is just a few miles from here. I suppose I can figure out which way to use it . . . connect it the wrong way and it doesn't work? Connect it the right way and it does?
I have a bunch of BS2-OEM Stamps, another PDB, and a variable, regulated desktop DC power supply. I will do the 5 volt test with the latter.
Thank you!
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
My schematic says 10uF. Your schematic says 47uF. However, my schematic is older (1/30/07) than yours (5/1/07). So, I will test use the 47uF from the latest schematic. (I wonder if it makes any difference?)
Thanks!
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
Regarding the two 47uF capacitors in the PDB power supply circuit.
Forgive my ignorance, but are those two capacitors polarized, electrolytic devices?
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
Thanks! I will get some. I guess that since they are tantalum, I won't be getting them at RS. But, I'll check there first.
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Think Inside the box first and if that doesn't work..
Re-arrange what's inside the box then...
Think outside the BOX!
Regulator- AP1084T50L-UDI-ND $1.91 EA
Caps- 495-1559-1-ND $0.98 EA
metron9, this entire thing was an el-stupido on my part. There are many things I should have done differently. One of them would have been to put my glasses on so I could see how the wires were marked. I will stick a fuse inline. Parallax should consider a slight redesign of the PDB to incorporate a fuse to protect that excellent product from dummies like me. While they are at it, polarity-reversal protection would be nice, as well.
jmalasia, thank you for doing that bit of research for me. I had some strange malady that prevented from doing ANYTHING today, but I think I am recovering. I printed your test instructions out and will perform them prior to getting parts. I may have questions! (Here is a question right now: Was there a reason you selected Digikey over Jameco?)
Thanks to both of you over and over again.
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
There's nothing like a new idea and a warm soldering iron.
The cap p/n I gave you was for a 7mm overall length cap. I think the originals might be 6mm but I'm sure you could use either one. Digikey had no listing for the 6mm. I estimated the cap size relative to the regulator width, which is 10mm.
As for the polarity reversal protection a lot of people are recommending, it's a pretty good idea, but I'm not crazy about putting a diode in-line with the power lead. The correct way to do it is to put diode on the load side of the connection, reverse biased between hot and ground. When you hook it up backwards it forward biases the diode and blows the fuse. It also provides you with voltage surge protection if you selct the appropriate breakdown voltage for the diode. I don't use any of that stuff- I wear my glasses instead.
Yes and a note for those younger ones that dont need em yet. Safety glasses are a very good thing to have as well, hot solder balls and flying shrapnel from tants can become a problem as well. Marking wires is all well and good but I typically have dozens of wires all over the place, jmalaysia has it right using the diode and fuse, remember a ptc fuse will reset when it cools back down. I use this method on all my circuits as the public connects the batteries.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Think Inside the box first and if that doesn't work..
Re-arrange what's inside the box then...
Think outside the BOX!
One thing that i do is for the Power Leads to an OEM board
·I always try to ·use ( Vin RED +)·and (Vss BLACK - )·wires for the Power Supply and differnet wire colors
·for every thing else this help me alot when putting Projects together
And if I use Vdd off the board then I use ether Yellow or Red
And for inputs to the Basic Stamps board White or Green or Blue
When·I have not done this it has cause me problems in the past
I do not know if this will help you or not but I also have glasses to read any thing
this why I had to start doing using differnet colors for different thing
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
··Thanks for any·
·
·
·
·
Sam
Sorry to hear about your catastrophe. The fact that the output capacitor was blown too worries me. Some of the circuits that could be damaged are the Pulse Generator, Audio Amplifier, RTC and MAX232 driver. I guess the only way to be sure would be to first do as suggested…Supply 5V directly to the board via VSS and VDD and see if these circuits still work (as well as programming a BASIC Stamp). Then go from there. Take care.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
It will be a great learning experience. The first thing I am going to do will be to clean it up a bit and perform the tests that jmalasia recommended. If they fail, THEN I will follow his SECOND recommendation!
Thank all of you for helping me. If I bring it back to life--looks doubtful [noparse]:)[/noparse] --I will let you know!
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.