Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

ESD problems?

Status
Not open for further replies.

hjl4

Member
Hi,

Ok so my dilemna is that I have a track machine, that operates in a rough environment, and from time to time things like structural brackets break.
I have 2 encoders on 2 arms of this machine, so we can read position they are in.
Before the end of day, in a file (let's call it temp), the Pic stores final value from quadrature encoder, and the machine is shutdown for servicing. The welder comes in, does the repairs on the track frame, and when I start up the machine again, the Pic has lost all values in temp file, and displays "00000"
How can I protect this from happening. I'm thinking of having Pic operate on a separate power supply or isolated from the machine's batteries. Would that help??
Also ESD chips?Are they any good for this??
I just need a solution.
Steer me in the right direction, as I have tried a few things except those mentioned above.
 
hjl4 said:
Hi,
...
Before the end of day, in a file (let's call it temp), the Pic stores final value from quadrature encoder, and the machine is shutdown for servicing....

What do you mean by file. Does the PIC have mass storage? Does it store the "file" to EEPROM?
Are you disconnecting the battery prior to welding? The welder should be doing that even if there was no processor.
Can you disconnect the PIC board from the rest of the wiring prior to welding?
If all else fails can you pull the PIC board and reinstall it after welding.

Since the failure happens when the machine is off a separate supply for the pic would not be of much use.

3v0
 
No, Pic does not write to EEPROM. ( would it help??).

Why is welding without disconnecting batteries so bad??
Any input here will be appreciated.

Also, there are a few ways to isolate circuit board, (relay, opto isolators, etc...), but if disconnecting batteries is only option, then maybe that's what we will have to do.

Please just explain why this welding procedure of disconnecting batteries is necessary. I have pretty good idea why, we probably were taught this in college, but that was years ago(18yrs to be exact), and I haven't had to think about this.

I know the batteries can explode, but only if hydrogen gas is present.

Help me out, or maybe I will dig out the old engineering book of mine.

Thanks for any help.
 
hjl4 said:
No, Pic does not write to EEPROM. ( would it help??).

Why is welding without disconnecting batteries so bad??
Any input here will be appreciated.

Also, there are a few ways to isolate circuit board, (relay, opto isolators, etc...), but if disconnecting batteries is only option, then maybe that's what we will have to do.

Please just explain why this welding procedure of disconnecting batteries is necessary. I have pretty good idea why, we probably were taught this in college, but that was years ago(18yrs to be exact), and I haven't had to think about this.

I know the batteries can explode, but only if hydrogen gas is present.

Help me out, or maybe I will dig out the old engineering book of mine.

Thanks for any help.

Well if the only storage is in RAM the welder might well be disconnecting it and that would be the whole thing... As I recall the PICs have nonvolatile storage available on chip, but there is a balance to be stuck between saving your times and wear and tear on the FLASH if that is what it is stored in. NVRAM is better as it typically has 100,000 write cycles as opposed to most FLASH only dealing with 10,000.

D.
 
Some more info would be good - what in the world is a "track machine", and how is the power hooked up? - is the PIC just storing the values in ram (i.e. the register file?!)

Putting a lithium battery into the circuit and setting it up so that it can hold the PIC above the memory retention voltage(something like 2V) is an option, but it typically requires adding some additional software checks to handle startup/reset and corrupted ram properly.

However, you've got an arc welder nearby, you'll have to do some pretty drastic things to make it work...

BTW, real NVRAM is just battery backed SRAM which has an unlimited number of cycles, but only lasts ~10 years (when the battery dies). EEPROM usually is in the 100K - millon's of cycles, and flash can range from 100's of cycles to 10K's of cycles depending on how it's designed.
 
hjl4 said:
No, Pic does not write to EEPROM. ( would it help??).

Yes, because it will keep the readings with the power off - shut down the power to it when welding. When it's powered back up, read the vlaues back out fo the data EEPROM.

Why is welding without disconnecting batteries so bad??
Any input here will be appreciated.

Welding is EXTREMELY noisey electrically - it's liable to upset electronics, unless you take extraordinary precautions.
 
Cadstarsucks may may have figured it out. The first thing a welder should do is disconnect the battery. That was standard practice 20 years ago.

Again, how is the lost data being stored?
When you switch off the machine does it remove power from the PIC?
Which PIC are you using?
Did you program it?
If not where did the board with the PIC on it come from?
If not can you get the person/company who did to fix the problem?

If you programed the PIC save the values to EEPROM prior to power down.

If someone else designed the system have them fix it.

As a last resort I would setup the PIC with its own battery and switch to disconnect it from the machines electrical system.
 
hi ijl,

Just a couple of questions.

Does the track machine normally run 24/7, ie; never powered down except for servicing.?

Are you sure that the battery is not being disconnected during welding ?

The reason for asking are the PIC and its arm encoders powered ON, even when the machine is switched OFF,
[does the PIC system have its own ON/OFF switch]

From your description of the system it appears that the encoders do not output a absolute position code
value [grey or binary], but are relative [counting from a datum point], thats why I expect you have to save the 'positions' to flash.

The fact that you get '0000' from the indicators after the weld operation, suggests the PIC is performing
a power up reset and clearing the last flash saved position to zero.

I would expect on some occasions, if the problem was caused by EMI you would get corrupted data displayed
and not always zero's on either one of the displays.

Lets know, like to help.


Eric
 
Last edited:
ericgibbs said:
hi ijl,

I would expect on some occasions, if the problem was caused by EMI you would get corrupted data displayed and not always zero's on either one of the displays.

Eric

Except for example a Brown-out Reset (BOR < improper isolation, bulk, bypass and decuopling>) condition where VDD below VBOR is greater than TBOR, or any reset condition caused by similar factors
 
Last edited:
hi luis,
I agree, from the 'zero' condition at power up, it sounds like some form of init Reset is taking place.

Regards
Eric
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top