Dallas Semiconductor DS1213C

[cncjerry]

Hello,  I have the subject smart socket in a Keithley 213.  I assume there are batteries embedded in the epoxy as this is used to keep alive NVRAM.  They batteries have died, I've ordered another, but was wondering if I could swap out the LIPOs by digging in the epoxy for a near term fix.  Any ideas?  I guess I have nothing to lose by trying.

Thanks

Jerry

[gibbled]

Different dallas chip but there's this...

http://www.vk2hmc.net/blog/?p=245

[JacquesBBB]

I had the problem with a DS1220  on a HP 54601a oscilloscope that would not start.

I  dig the chip and solder a cr2032 socket that I glued on the top. I also put a socket on the chip for easy
replacement if needed.

[cncjerry]

Different dallas chip but there's this...

http://www.vk2hmc.net/blog/?p=245

Thanks, Dremel to the rescue.  I was going to put it on the small milling machine.   this one looks like it has two batteries.

[amyk]

Has been done multiple times:

http://www.mcamafia.de/mcapage0/dsrework.htm
http://www.classic-computers.org.nz/blog/2009-10-10-renovating-a-dallas-battery-chip.htm

[cncjerry]

I started to grind out the old batteries, got distracted, and by the time I got back to it the new one arrived.  I put it into the Keithley 213 and the error at power on was gone after the first POR when I cleared it with software. Taking it out and reinstalling it gives a power on error the first time and once cleared, you can power it on and off without error.  With the old DS1213c the unit wouldn't pass POST so it looks like it is working, maybe?

The reason I say 'maybe' is that after pressing the calibrate button and writing new gain and offset values (without error) they reset to default after powering off and on.  No errors as when I remove it, just default values as if the calibrate button wasn't pressed before writing the values.  I checked the button with an ohm meter and all is well.  This isn't a huge deal since you can't operate it except from software and I can send the offset and gain setting for the reference voltage I need with each command.  I don't know if I will be able to build step waveforms in memory as they are voltage only, I think, without offset and gain.  I actually doubt I would ever use it in that mode anyway but the calibration failure bugs me.  It could still be operator error or maybe the DS1213c isn't keeping the voltage up when powered off?

The 213 is a +/- 0 to 10V reference source at 10ma.  It seems really stable, as good as any 6.5 digit reference source I have, and once warmed up I get little if any flicker on the lowest digit of my HP 3457a. 

Thanks

[cncjerry]

I put a meter on it and the DS1213c is holding the voltage at around 3.77V which is more than enough, I believe.  So I guess the RAM is bad?  No checksum error at POST though, it is almost like it isn't even trying to write to the RAM.

Ideas?

[cncjerry]

I found an 'S' command that writes the settings and calibration constants to NVRAM.  Funny it wasn't mentioned in the calibration procedures which implied it was done automatically by setting the calibration switch in the back of the unit. 

So I guess it is working.  The guy I bought it from refunded all my money and told me to keep it since when I bought it, it was hanging on POST and useless.  I couldn't send or receive commands, basically a veg.  I have about $60 in parts now plus about 10 hours of debugging.  I told him I would send him the net, if any, after repairs.  At minimum wage I think I can safely judge I am under water on the deal but will send him some cash anyway. 

I still haven't figured out a good use for it.  I planned to use it initially for reference voltages for a meter I am building. 

Enough said.

Jerry