The Museum of HP Calculators

HP Forum Archive 19

[ Return to Index | Top of Index ]

HP 41C data storage integrated circuit swapping: Maybe Eric could answer this ;-)
Message #1 Posted by Geoff Quickfall on 17 Mar 2010, 3:44 a.m.

Hello experts,

I posted this many months ago and cannot find it in the archives. I know Eric Smith had many good answers including swapping the memory with a 41CV. I just can't find the info so here is the question again

Here is a photo of an early HP 41C with all the bugs and also with a defunct D/S integrated circuit (RAM) (ser# 1931Axxxxx). It is the one at the bottom right labelled:

1LA705

When SIZE 000 is executed the program registers available equal 47 out of the required 63. All programs execute and the data registers used for STO when SIZE 047 is executed are accessable. This indicates that (my logic) the last D/S chip is unserviceable and must be replaced:

1LA705:

My quandry is that I have a newer HP 41C board (1981) with the following chips. They have the following codes from first to last:

1LA701, 
1LE701, 
1LB738, 
ILB739,
1LB740 (this chip is in the same position as the chip on the 
        early board)

My assumption is that the last chip; 1LB740 would be the one to use. My question is, is it swappable with 1LA705.

I also have a two CV's with similar blocks of 5 D/S chips:

1LA701
1LE701
1LE702
1LE703
1LE704  this chip is in the same position as the chip on the 
        early board)

AND

1LG901 1LA701 1LH502 1LH503 1LH505 (this chip is in the same position as the chip on the early board)

My questions are as follows:

1.  is my assumption correct and the last D/S chip is the 
    defunct chip on the photo above; chip 1LA705?
2.  Which chip in the 3 lists above are swappable; 
    1LB740, 1LE704, 1LH505 or any of the others?

The reason for doing this, to keep the early with bugs board intact with its bugs. That precludes me just swapping out the board for a newer one which I have. That is I would like to use a donar chip from the three options above to replace the bad chip that I hopefully have identified correctly.

Thanks in advance, Geoff

Edited: 17 Mar 2010, 3:51 a.m.

      
Re: HP 41C data storage integrated circuit swapping: Maybe Eric could answer this ;-)
Message #2 Posted by Randy on 17 Mar 2010, 11:25 a.m.,
in response to message #1 by Geoff Quickfall

Quote:
My assumption is that the last chip; 1LB740 would be the one to use. My question is, is it swappable with 1LA705.

1LB738/39/40 are the roms so no, it is not swappable.

Quote:
My questions are as follows:

1. is my assumption correct and the last D/S chip is the defunct chip on the photo above; chip 1LA705? 2. Which chip in the 3 lists above are swappable; 1LB740, 1LE704, 1LH505 or any of the others?


1 - Not a valid assumption, you can only identify the faulty d/s with the service rom.

2 - AFAIK, none of the three can be interchanged.

      
Re: HP 41C data storage integrated circuit swapping: Maybe Eric could answer this ;-)
Message #3 Posted by Jim Horn on 17 Mar 2010, 12:06 p.m.,
in response to message #1 by Geoff Quickfall

If it helps, Geoff, when the '41CV came out a few of us who worked at HP Santa Rosa got the new CV RAMs directly from Corvallis and replaced the ones in our '41C machines, turning them into CVs but with all the first generation bugs. As I recall, the CV had one more RAM than the C did, so one was piggybacked onto the board to make it fit. One of the reasons for doing that was that in-house orders for the CV had something like a six month wait list whereas the ICs were available immediately.

Last time I used it it still worked fine but, like most of my things, it's in long term storage. Otherwise I'd pop it apart to verify the details including IC numbers.

Jim Horn - my 41C has 319 registers, power contacts in the battery access tunnel, etc.

      
Thanks Jim and Randy, further info
Message #4 Posted by geoff quickfall on 17 Mar 2010, 2:20 p.m.,
in response to message #1 by Geoff Quickfall

after running the diagnostic with the service module the calculator passes all tests except for the D/S which doesn't surprise me.

The failure message is:

D/S   1   BAD

If I am doing this correctly and according to the 41C service manual that would imply, refering to the photo, that 1LA7 002 (the second D/S chip on the top right) is bad.

The first chip 1LA701 would be used internally for the X,Y,Z,T LASTX and ALPHA which all work correctly.

Knowing that would the 1LE701 from the 1981 HP 41C be usuable as a replacement, it too is the D/S 1 chip?

Failing that, Jim, the HP 41CV conversion sounds interesting as I have the ability to swap out all the c chips for working cv chips. More info on that would be interesting, converting this to a tall keys with gold balls 41C also!

Cheers, Geoff

Edited: 17 Mar 2010, 2:25 p.m.

            
Re: Thanks Jim and Randy, further info
Message #5 Posted by Ignazio Cara (Italy) on 17 Mar 2010, 3:19 p.m.,
in response to message #4 by geoff quickfall

Hello friends, you can find all information you need at the following link , about all the HP-41C released (Initial, two transistor and final) with the tabs for the replaceables parts. I hope this could be a little help for Geoff and Randy.

Regards

Ignazio

                  
Thanks Ignazio
Message #6 Posted by Geoff Quickfall on 17 Mar 2010, 6:57 p.m.,
in response to message #5 by Ignazio Cara (Italy)

I currently am using the HP 41C service manual but unfortunately only have access to the 41CV and 41CX diagnostics rom.

It looks like I will be swapping the four HP 41C D/S rams with four HP 41CV D/S rams as described by Jim Horn and Eric Smith.

This should give me an HP 41C "with all the bugs" in a tall keys 1979 case but with HP 41CV memory (319) registers.

Cheers, Geoff

                  
Re: Thanks Jim and Randy, further info
Message #7 Posted by John Robinson on 17 Mar 2010, 9:18 p.m.,
in response to message #5 by Ignazio Cara (Italy)

Hi Ignazio,

I am curious where this information came from ?, and whether similar information is available for the fullnut CV and CX boards ?

Thanks John

                        
Re: Thanks Jim and Randy, further info
Message #8 Posted by Ignazio Cara (Italy) on 18 Mar 2010, 3:45 a.m.,
in response to message #7 by John Robinson

Hi John, the info I have posted came from the 'HP-41C/CV/CX Service manual", 00041-90472 integrated by the addendum 00041-90542 related to the half-nut release of the three calculator. I'm trying to send this manual and other (Card reader, wand and printer 82143 service manuals) to Dave via ftp, to integrate further editions of MoHPC DVDs. Regards

Ignazio

                              
Re: Thanks Jim and Randy, further info
Message #9 Posted by John Robinson on 19 Mar 2010, 1:27 a.m.,
in response to message #8 by Ignazio Cara (Italy)

Thanks for the info Ignazio

Cheers, John

            
Re: Thanks Jim and Randy, further info
Message #10 Posted by Eric Smith on 17 Mar 2010, 5:13 p.m.,
in response to message #4 by geoff quickfall

Which diagnostic ROM did you use? If it's reporting a bad D/S 1, it probably is a diag for a 41CV/CX.

With 48 registers working, I'd think that the 1LA7-0005 is the bad one. When the memory is sized, they start from address 0C0 and count upwards until they find a non-existent register. If a RAM in the middle fails, the good RAMs past it are not found. (They can still be accessed as data registers, though. However if the 41C D/S 1 (1LA7-0002) fails, it won't be able to find any registers, and won't be able to create the permanent ".END.". I'd expect the calculator to work very poorly if at all.

From memory, which might be mistaken, I believe the mapping is:

000-00F  D/S 0  1LA7-0001  status registers
010-03F  gap
040-07F  X FCN memory (41C/CV),  D/S 6 1LE7-0006 or 1LH5-0006 (41CX)
080-0BF  X FCN memory (41C/CV),  D/S 7 1LE7-0007 or 1LH5-0007 (41CX)
0C0-0CF  D/S 1  1LA7-0002 (41C) \  
0D0-0DF  D/S 2  1LA7-0003 (41C)  \ D/S 1 1LE7-0001 or 1LH5-0001 (41CV/CX)
0E0-0EF  D/S 3  1LA7-0004 (41C)  /
0F0-0FF  D/S 4  1LA7-0005 (41C) /
100-13F  mem module 1 or quad mem (41C), D/S 2 1LE7-0002 or 1LH5-0002 (41CV/CX)
140-17F  mem module 2 or quad mem (41C), D/S 3 1LE7-0003 or 1LH5-0003 (41CV/CX)
180-1BF  mem module 3 or quad mem (41C), D/S 4 1LE7-0004 or 1LH5-0004 (41CV/CX)
1C0-1FF  mem module 4 or quad mem (41C), D/S 5 1LE7-0005 or 1LH5-0005 (41CV/CX)
200      gap
201-2EF  X MEM 1
2F0-300  gap
301-3EF  X MEM 2
3F0-3FF  gap

You could replace all four chips 1LA7-0002 through 1LA7-0005 with a single 1LE7-0001 or 1LH5-0001. You'd still have the equivalent memory of a 41C if you did that without adding 1LE7-0002 through 1LE7-0005 or 1LH5-0002 through 1LH5-0005.

                  
Eric, YEAH!!
Message #11 Posted by geoff quickfall on 17 Mar 2010, 6:35 p.m.,
in response to message #10 by Eric Smith

From memory I remember you mentioning this. I was thinking also that if a bad memory chip was found the program registers would stop and not continue, which is not the case here.

You are correct, the diagnostic is the CV and CX version, I mistakenly thought it would diagnose the chips in order. But intuitively it did not seem correct that the first chip would be the bad one considering the calculator acts pefectly up to 46 registers in the program mode.

Quote:
You could replace all four chips 1LA7-0002 through 1LA7-0005 with a single 1LE7-0001 (do have a spare 1LE7-001) or 1LH5-0001 (don't have a spare 1LH5-001)

It would seem that this is the way to go as I have 3 spare working cv boards.

To clarify, I would remove 1LA7-002 through 1LA7-005 and replace with a single 1LE7-0001 at the 1LA7-002 position. Thus, obviously, leaving the other 3 successive chip solder points empty.

Sorry to be such a neophyte with this, if it was my calculator I wouldn't hesitate to experiment but trying to keep this buggy and with the correct memory registers is the intent.

Cheers, and thanks, Geoff

p.s. to get fancy and to increase the memory to a CV with the C bugs would the chip swap below result in the HP-41C with bugs having the same 319 registers as the CV the chips came from. I mention this as Jim Horns posting sounds as though this is possible:

ORIGINAL 41C SWAP ORIGINAL 41CV CHIPS CHIPS

1LA7-002 <--> 1LE701 1LA7-003 <--> 1LE702 1LA7-004 <--> 1LE703 1LA7-005 <--> 1LE704

I think Gene would like a 41C with all the bugs plus the memory of a CV in a tall keys case.

Wouldn't you Gene????????

                        
Re: Eric, YEAH!!
Message #12 Posted by Eric Smith on 17 Mar 2010, 7:21 p.m.,
in response to message #11 by geoff quickfall

If you want the full CV memory, you'd have to put in all five of the 1LE7 chips from the CV, and keep the 1LA7-0001, so you'd have to piggyback one of them.

The RAMs have identical pinout and wiring, so it doesn't matter which physical location they're installed into.

                              
Piggy backing
Message #13 Posted by Geoff Quickfall on 17 Mar 2010, 7:29 p.m.,
in response to message #12 by Eric Smith

Physically what does this entail. find a location for the extra chip and use jumper wires to the previous chip. Is there a crossover required such as in the double-x-memory mod.

Actually, is this explained in the archives or journal?

Thanks for you help.

Of course the easiest thing is to just remove the four and swap for the one.

Cheers, Geoff

Edited: 17 Mar 2010, 7:43 p.m.

                                    
Re: Piggy backing
Message #14 Posted by Eric Smith on 17 Mar 2010, 8:06 p.m.,
in response to message #13 by Geoff Quickfall

No crossovers, the pinout is identical and there is no chip select. Piggybacking is just soldering one chip onto the legs of another, avoiding use of jumper wires. HP did this themselves in the 41CX (fullnut only), in order to get two 12Kword ROMS and eight RAMs to fit in five 8-pin DIP footprints.

                                          
Re: Piggy backing
Message #15 Posted by Randy on 17 Mar 2010, 8:13 p.m.,
in response to message #14 by Eric Smith

Piggybacking was also done in CV's when using the common C/CV/CX PC board 5081-5564.

If you want to do it in the original 41C board, use only the top most position next to the toroidal transformer as any lower and the thickness will be too much for the case back to fit properly.

                                                
Re: Piggy backing, good suggestion and noted.
Message #16 Posted by Geoff Quickfall on 17 Mar 2010, 9:00 p.m.,
in response to message #15 by Randy

don't want to have to do any milling of the plastic thats for sure.

Thanks Randy,

Geoff

                                                
Re: Piggy backing
Message #17 Posted by Jim Horn on 18 Mar 2010, 11:56 a.m.,
in response to message #15 by Randy

Geoff, Eric and Randy hit the nail on the head. That's exactly how I did mine, including the piggybacking and which chips went where. Thanks, team!

By the way, piggybacking was not uncommon back then. Many of the first Amiga owners (model 1000) piggybacked two more sets of DRAMs on those in it, giving (with the optional 256k plug in) a total of 1M of RAM - a staggering amount back then. I remember having 24 active programs crowding the screen with memory to spare. Now my HP-50g has far more memory, CPU horsepower, etc. (as does my cell phone, my PDA, and so on). Moore's Law is our friend!

I won't even describe what we did to my HP-110 Portable...

                                          
Thanks Eric
Message #18 Posted by Geoff Quickfall on 17 Mar 2010, 8:13 p.m.,
in response to message #14 by Eric Smith

Well that is that then, might throw this in an appendix for the heck of it (if it works ;-)

Cheers, Geoff

p.s. can you send my your address via email. Then I can send you the HP-01 circuit board.

                                          
one final question ;-(
Message #19 Posted by geoff quickfall on 18 Mar 2010, 12:21 a.m.,
in response to message #14 by Eric Smith

the donar chip is labelled:

1LE701

is that equivalent to your:

1LE7-0001

===============================================================

That would be all but as an extra I am going to take pictures of the 3 CV boards I have.

two of these have piggy backed D/S and one does not.

                                                
Re: one final question ;-(
Message #20 Posted by Eric Smith on 18 Mar 2010, 5:18 p.m.,
in response to message #19 by geoff quickfall

Yes. They usually leave off the first two zeros of the suffix printed on the 8-pin parts, but the official part number is still 1LE7-0001.


[ Return to Index | Top of Index ]

Go back to the main exhibit hall