EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: TRN on February 26, 2022, 04:38:04 pm
-
I purchased a defect SMU200A on eBay, and when the unit arrived it had several issues.
First of all the fans were not working, which caused the unit to switch off every time shortly after powering up. This fault was caused by a blown fuse of a 5.2V supply rail on the motherboard. After replacing the fuse, the unit refused to power up anymore, and after some troubleshooting this was traced back to a faulty Vega 450 PSU. Swapping the primary side of this PSU, with another one resolved the issue, the unit now boots, and remains powered up, but with the message that the synthesis module "SSYN" is missing. This module with Part number 1141.4208.02, however is actually installed. and the reason that the SMU200 thinks that it is missing, is caused by the fact that the Atmel AT25160AN serial eeprom on this board has a missing Vcc pin. Replacing this serial eeprom should not be a problem, however I do not have the eeprom dump file, to re-program the new one, and due to the fact that the Vcc pin is missing, I won't be able to dump the contents of the existing eeprom.
Does anybody have any suggestions how I can temporary feed power to the damaged chip, so I can dump the contents, or else does anybody who owns a SMU200A provide me the eeprom dump?
Thks
-
Is it possible to cut into the chip to gain access to the Vcc connection?
-
Assuming it's not possible to grind a away a bit of the packaging and solder a wire to the stub that used to be the Vcc pin, I would try powering it through the ESD protection diodes that are at most digital inputs (that's where the Vcc + 0.5V Vih(max) spec (http://ww1.microchip.com/downloads/en/DeviceDoc/doc3347.pdf) comes from). The problem is that the supply current with Vcc=5V at 4MHz is specified as 4 mA typical, while the protection diode might be specified for max 500 uA continuous (at least for the old PICs (https://ww1.microchip.com/downloads/en/AppNotes/00521c.pdf)). So I'd reduce the supply voltage to say 2.5V (about a diode drop above 1.8V) and reduce the frequency as much as possible.
I think you could supply power to /WP and /HOLD, and hopefully that would be enough to be able to read it. Depending on the voltage drop you might have to increase the voltage slightly. Make sure whatever you use to interface with (Arduino?) it supports this voltage.
-
Is it possible to cut into the chip to gain access to the Vcc connection?
thanks for the suggestion, and I actually tried to do this by chipping the mold compound a little bit with a side cutter, but the wire-bond is so tiny that I won't be able to solder anything to it.
-
Assuming it's not possible to grind a away a bit of the packaging and solder a wire to the stub that used to be the Vcc pin, I would try powering it through the ESD protection diodes that are at most digital inputs (that's where the Vcc + 0.5V Vih(max) spec (http://ww1.microchip.com/downloads/en/DeviceDoc/doc3347.pdf) comes from). The problem is that the supply current with Vcc=5V at 4MHz is specified as 4 mA typical, while the protection diode might be specified for max 500 uA continuous (at least for the old PICs (https://ww1.microchip.com/downloads/en/AppNotes/00521c.pdf)). So I'd reduce the supply voltage to say 2.5V (about a diode drop above 1.8V) and reduce the frequency as much as possible.
I think you could supply power to /WP and /HOLD, and hopefully that would be enough to be able to read it. Depending on the voltage drop you might have to increase the voltage slightly. Make sure whatever you use to interface with (Arduino?) it supports this voltage.
Thanks, that is a suggestion which is definitely worthwhile looking into. :-+
Just for your info, I am using a Conitec Datasystems GALEP-5 programmer for reading and programming
-
Another member that formerly repaired these had a bunch of parts and stuff for sale, it might be worth asking if he has the part or the file.
https://www.eevblog.com/forum/buysellwanted/fs-rohde-schwarz-cmu200-tested-parts-and-modules/msg3925028/#msg3925028 (https://www.eevblog.com/forum/buysellwanted/fs-rohde-schwarz-cmu200-tested-parts-and-modules/msg3925028/#msg3925028)
-
Another member that formerly repaired these had a bunch of parts and stuff for sale,
SMU200A != CMU200
-
SMU200A != CMU200
Oops!
-
Since my last suggestion wasn't very helpful, let me offer another that I've gotten to work in sort-of similar circumstances. If the bond wire is exposed at all, get some silver conductive ink and blob it on there in layers until you have a nice spot. Then you can use a sharp test probe to supply your power by holding it against the blob by hand while it is in the programmer being read.