HP-41CV / CX Memory Transplant

[rocket.scientist]

The bottom line here is that D/S ICs can be transplanted between CV and CX calculators, at least in this circumstance. The recipient CX now passes all of the 5061-7221 diagnostics. I have had the CX running on the bench for a few days now and it appears to be without issue. If anyone knows a good stress test, I'd be happy to have it and report the results here.

On to the details.

This post is a follow-on to my earlier HP-41CX Missing Clock thread. I used a NoV-64d module with the CV/CX diagnostic software loaded to page 4. The repaired CX passed all diagnostics but one - D/S 3 was reported as having issues. Ugh.

Not wanting to unduly increase the cost of my cheaply-had HP-41CX, I turned my gaze to an inop HP-41CV I had handy. The CV diagnostics reported only an OS ROM problem - all the D/S chips checked in okay. I did some research and found scholarly opinions both ways on how freely ICs can be transplanted between models. In my case, the CX had a failing RAM, and the ROMS reported okay. The CV had a failing ROM, and the RAMs reported okay. After due consideration, I decided no further harm could be done by trying a transplant.

The CV board is the 00041-60104 "HP-41C/CV Common Board" and the CX is the 00041-60154 "HP-41C/CV/CX Common Board". I used a professional desoldering station to remove all ICs from the CV, and only the U6/13, U7/12 and U8/9 stacks from the CX. After unstacking the dual layer ICs, I did a considered swapping of RAMS from the CV to the CX. I did use low-temp solder paste on the CX, just in case I need to remove them later. Note the CV remains a parts donor. The remaining ICs removed from the CV are stashed safely away.

Here are the IC layouts, both before and after the surgery.

Code:

Device    CV(before)    CV(after)        CX(before)    CX(after)
D/S 0    1LA7-0001=>    Storage        1LA7-0001    1LA7-0001
D/S 1    1LE7-0001=>    Storage        1LH5-0001    1LH5-0001(*)
D/S 2    1LE7-0002=>    Gone            1LH5-0002=>    1LE7-0002
D/S 3    1LE7-0003=>    Gone            1LH5-0003=>    1LE7-0003
D/S 4    1LE7-0004=>    Storage        1LH5-0004=>    1LH5-0004
D/S 5    1LE7-0005=>    Storage        1LH5-0005=>    1LH5-0005
D/S 6    N/A            N/A            1LH5-0006=>    1LH5-0006
D/S 7    N/A            N/A            1LH5-0007=>    1LH5-0007

Of note above:
1) "=>" means the IC was removed from the board. ICs ILH5-0004 thru -0007 were desoldered and later replaced in the CX, but not in the same place.
2) D/S 0 (1LA7-0001) was found to be identical between the two models
3) RAM modules (1LE7 and 1LH5) can be freely intermixed on the CX. I did not test mixing RAM on the CV.
4) (*) I did not disturb the dual layer ICs at U4/5 and U3/10 on the CX. The marking on D/S 1 are presumed.
5) Be aware the orientation of pin 1 of the IC stacks are reversed between the two models. I almost messed up big time here. Fortunately I caught my error before applying power.
6) It turns out the pinouts for almost all of the RAM and ROM modules are identical, so placement is not critical. Only the devices at U3/10 appear to be different at pin 8, perhaps due to a clock battery hold-up.

I'm happy to have this as a project. I have learned a lot about the internal workings of these marvelous machines, and have gained a great respect for some of the greybeards on this list, and on others. Thank you for your assistance.

n.b.- I plan on converting the CV into a CL once (if?) the next batch of Monte's awesome boards become available.

R/

-Pat (KG5YPQ)

[/code]