Found these inside a non-working 71B

[Paul Berger (Canada)]

It looks like both in the U9 position are ROMs and the original 4K module was removed.

Paul.

[MikeSD]

Dave,

Yes i'll post label and back side of F/A, if that is what it is. The chip p/n above U9 is shown above.

Paul,

it's out. that chip above U9 seems soldered ok. Just thin wires so it moves. Buzzes ok top to bottom.

Now to the cleanup. Step by step.





My original problem: Calc wouldn't power on. No LCD content displayed at all.

My assumption was it might have had something to do with broken wires in these new chips. Assuming that the broken wires didnt touch something else it shouldnt, can not being able to configure these chips break the chain so that nothing would show on display?

I'm now wondering if these mods might not have anything to do with tbe lack of LCD.

The repair label said the original problem in 86 was CPU lockup. That appears to have been repaired. If the 71B doesnt run after removing my RAM and FORTH, I might switch 71B.

Question on replacing flex cable. There are two pins connected together on one end on bottom board, but on the display side the outside pin doesnt have a trace or via to it. But does have a trace in the flex for it. Why is that? and/or what is that pin for?

[Paul Berger (Canada)]

(03-11-2018 04:19 PM)MikeSD Wrote:  Here is a closeup of that topndevice. It doesn't appear to have the 3-7 lines like on RAM. Could be under but why? Anyway it does have part number that might help identify what it is.

The visible part number would appear to be the number for the raw circuit card and does not seem to be related to what is on it. All four RAM modules and the ROM module of the board I am looking at have the 5180-2912 part number on them as does a 4K plugin I have. Those signal lines are probably on the other side of the circuit card, different part number different layout I would guess. What does the back of the "FORTH" module look like does it look similar?

[MikeSD]

I see. The 5180-2912 is the bare PCB number and the 5180-2918 is a different bare PCB. I guess that makes sense. Ill lift that chip and get a pic of the other side. Probably another app chip. Maybe something good

I'll also clean off the goop on the back of forth and see whats there.

Wow lucky.

[MikeSD]

Wow, lucky. The comp above U9 looks like a math module. Part number 5061-7226 [typo corrected. it is 7226]. I'll post pics soon.

Math?

[Paul Berger (Canada)]

(03-11-2018 06:09 PM)MikeSD Wrote:  Dave,

Yes i'll post label and back side of F/A, if that is what it is. The chip p/n above U9 is shown above.

Paul,

it's out. that chip above U9 seems soldered ok. Just thin wires so it moves. Buzzes ok top to bottom.

Now to the cleanup. Step by step.

My original problem: Calc wouldn't power on. No LCD content displayed at all.

My assumption was it might have had something to do with broken wires in these new chips. Assuming that the broken wires didnt touch something else it shouldnt, can not being able to configure these chips break the chain so that nothing would show on display?

The way that is wired it should not it looks loke they broke into the chain at the end so the only thing that would not work is the HP-IL. You can check the chain U6 Din is connected to +5V, U6 Dout to U7 Din U7 Dout to U8 Din U8 Dout to U9 Din (bottom of two Roms ) U9 Dout to Din of top ROM. your picture showing the edge on vie confirms that Dout from the top ROM is where the RAM and "FORTH" modules are inserted. The OS ROM is hard addressed and is not part of a chain.

(03-11-2018 06:09 PM)MikeSD Wrote:  I'm now wondering if these mods might not have anything to do with the lack of LCD.

The repair label said the original problem in 86 was CPU lockup. That appears to have been repaired. If the 71B doesnt run after removing my RAM and FORTH, I might switch 71B.

It might be useful to test them in a working 71B what you could do is parallel the first 7 pins and ground with one of the ROMs and with no HPIL module installed connect Din to the end of the chain there is at least one via in the signal path to the HPIL module or you can find it on pin 8 of the empty top connector. There is a via that might be easy to pick up the signal from just under a 10K resister just below the 00071-80002 BD. NO., I think that would be the easiest spot to pick up the signal.

I had made up a plug in to bring out those 7 signals plus ground to the front to monitor the bus with a logic analyzer, but now it is coming in handy when I am testing some RAM modules.

(03-11-2018 06:09 PM)MikeSD Wrote:  Question on replacing flex cable. There are two pins connected together on one end on bottom board, but on the display side the outside pin doesnt have a trace or via to it. But does have a trace in the flex for it. Why is that? and/or what is that pin for?

That is the ground pin and on my spare it looks like there was the same number of pads at both ends (Some pads damaged at the CPU end) but it does not look like the outermost pad was connected but they may have been solder bridged.

[rprosperi]

(03-11-2018 07:15 PM)MikeSD Wrote:  Wow, lucky. The comp above U9 looks like a math module. Part number 5061-7236. I'll post pics soon.

The Math Module is part number 5061-7226, not 36. Typo, or maybe some numbers aren't clear on this chip too?

The Math and Forth modules were by far the most popular among enthusiasts, so it makes sense these 2 were installed. Do you have any info on where this came from, or just another eBay special?

[Paul Berger (Canada)]

(03-11-2018 07:15 PM)MikeSD Wrote:  Wow, lucky. The comp above U9 looks like a math module. Part number 5061-7236. I'll post pics soon.

Not a huge surprise the math module was a very popular module. That looks like pretty nasty stuff they used to hold the FORTH module in place, double sided tape work nice for that.

Now that I see wider pictures I have my 32K modules in pretty much the same place, I wonder why they felt it was necessary to relocate some of the buzzer components.

You mentioned earlier that you had a number of non-working 71Bs would yo be interested in selling one or two as is? If so we can discuss via private message or email.

[MikeSD]

(03-11-2018 07:35 PM)rprosperi Wrote:  

(03-11-2018 07:15 PM)MikeSD Wrote:  Wow, lucky. The comp above U9 looks like a math module. Part number 5061-7236. I'll post pics soon.

The Math Module is part number 5061-7226, not 36. Typo, or maybe some numbers aren't clear on this chip too?

Nah, just a typo on my part. i have better images on my camera. Just using phone for this thread. will replace with better photos later.

I think i will test these chips in another 71b thats known working. Math, FORTH and 64K RAM would be a good find, in a calc i marked junk a few years when it didnt work when I turned it on, assuming they are working.

[rprosperi]

(03-11-2018 07:53 PM)MikeSD Wrote:  I think i will test these chips in another 71b thats known working. Math, FORTH and 64K RAM would be a good find, in a calc i marked junk a few years when it didnt work when I turned it on, assuming they are working.

Keep in mind the FORTH ROM Module actually contained 2 ROMs; one is fixed address at 0xE0000 and the other is loadable anywhere, placed at the 'right' address by the OS' config routines.

Perhaps the board with the very different layout is a board that holds the 32KB ROM that is fixed at 0xE0000 ? Makes sense that it would have fewer leads, I think...

[Paul Berger (Canada)]

(03-11-2018 08:15 PM)rprosperi Wrote:  

(03-11-2018 07:53 PM)MikeSD Wrote:  I think i will test these chips in another 71b thats known working. Math, FORTH and 64K RAM would be a good find, in a calc i marked junk a few years when it didnt work when I turned it on, assuming they are working.

Keep in mind the FORTH ROM Module actually contained 2 ROMs; one is fixed address at 0xE0000 and the other is loadable anywhere, placed at the 'right' address by the OS' config routines.

Perhaps the board with the very different layout is a board that holds the 32KB ROM that is fixed at 0xE0000 ? Makes sense that it would have fewer leads, I think...

It appears to have all the connections it needs there are 3 pins next to +5 that are of no use to a RAM or ROM module they are there because these modules are designed to be put into front side plugins The fourth missing connection is Dout which is not brought back to the board because it goes on to connect to the other added RAM and ROM modules.

Paul.

[rprosperi]

(03-11-2018 08:27 PM)Paul Berger (Canada) Wrote:  It appears to have all the connections it needs there are 3 pins next to +5 that are of no use to a RAM or ROM module they are there because these modules are designed to be put into front side plugins The fourth missing connection is Dout which is not brought back to the board because it goes on to connect to the other added RAM and ROM modules.

Thanks for 'splaining that. I've never opened a Forth module, but I always assumed it contains 2 separate boards, one for each ROM, but that could simply be naïve. Assuming that, I was trying to figure out why that one board (which looks like it is mounted under the labeled Forth ROM) would be different from all the other RAM/ROM boards. Anyhow, hopefully we'll soon know. Thx.

[MikeSD]

(03-11-2018 08:27 PM)Paul Berger (Canada) Wrote:  [quote='rprosperi' pid='92797' dateline='1520799314']

It appears to have all the connections it needs there are 3 pins next to +5 that are of no use to a RAM or ROM module they are there because these modules are designed to be put into front side plugins The fourth missing connection is Dout which is not brought back to the board because it goes on to connect to the other added RAM and ROM modules.

Paul.

I'm not sure about that. The chip above U9 definitely has part number of MATH. Any chance a 2nd FORTH module might be under the same goop with the first? As far as I can tell those are the only installed points. So if it's not there, it never was

I also think the FORTH does go back to HPIL Din. I think it had a wire going from FORTH pin back to HPIL. You can barely see it in this pic. First page photos 5 and 6 show wire going from FORTH to HPIL. Pin9 goes behind resistor that picks up the HPIL chain.

I was looking for an easy way to break into the chain on a working 71B and I'm not sure I have to. Looks love HP might have planned for this. U9-pin goes two places. See pic. I might just be able to tack the RAM to that single alone chip, on another calc and just connect pin 9 of it and run it to RAM and don't connect an HPIL



But I'm confused now. You have been talking about a U9 chip. I have been under the impression that is the normal chip in an unmodified 71B. And the chip above it is the odd chip. Shown top left, in this pic. I thought that was a MATH chip, based on the part number. I don't see any place for a 2nd ROM for forth. The top left I'm calling a Math chip. I also suppose I might not have two 32k RAMs. It might be the missing FORTH ROM and something else. I've been assuming it's all RAM but I could be wrong. Never looked at it very closely. I was thinking I might replace those wires, so I will be able to get to the bottom of that.

But, using this pic above which are RAM and which are ROM. i thought the 4 to lect were 4K RAMs, and the 5th on bottom left was a RAM also (or a ROM). and that the only suspect is the Math, top left.

BTW, found something interesting in shaf I thought aas 64K RAM. Photo coming. Last 2 photos.

[MikeSD]

Might have found the missing 2nd FORTH ROM. I was assuming I had a 64K RAM config. It might bea 32K, and the 2nd FORTH.

one device is marked HM?256LFP (assume 32k RAM) but the other is unmarked. Might be ROM.

So the modification might add 32K RAM, FORTH (2 ROMs), and a Math ROM. Now, if i can just get these tested and the 71B working. Spent the whole day tearing apart an documenting. Need a break to digest all this.

[Paul Berger (Canada)]

(03-11-2018 09:51 PM)MikeSD Wrote:  [quote='Paul Berger (Canada)' pid='92799' dateline='1520800020']

It appears to have all the connections it needs there are 3 pins next to +5 that are of no use to a RAM or ROM module they are there because these modules are designed to be put into front side plugins The fourth missing connection is Dout which is not brought back to the board because it goes on to connect to the other added RAM and ROM modules.

Paul.

(03-11-2018 09:51 PM)MikeSD Wrote:  I'm not sure about that. The chip above U9 definitely has part number of MATH. Any chance a 2nd FORTH module might be under the same goop with the first? As far as I can tell those are the only installed points. So if it's not there, it never was

I also think the FORTH does go back to HPIL Din. I think it had a wire going from FORTH pin back to HPIL. You can barely see it in this pic. First page photos 5 and 6 show wire going from FORTH to HPIL. Pin9 goes behind resistor that picks up the HPIL chain.

Both the hard configured part of the FORTH and the soft part of the ROM would be on the same module, there isn't room inside the shell of a front plugin for more than one module. The dies for the ROMs are wire bonded directly to the circuit card and then encapsulated with epoxy and it would be possible to have more than one die on the circuit card of a module, in fact they refer to the base RAM and ROM modules as "hybrids" which usually implies there are multiple components involved. The 4K modules used for the base RAM and the 4K plugins are composed of 4 1K dies and the ROM is also 4 dies.

(03-11-2018 09:51 PM)MikeSD Wrote:  I was looking for an easy way to break into the chain on a working 71B and I'm not sure I have to. Looks love HP might have planned for this. U9-pin goes two places. See pic. I might just be able to tack the RAM to that single alone chip, on another calc and just connect pin 9 of it and run it to RAM and don't connect an HPIL

Yes I think that is what I suggested in one of my last posts, you can tack it onto end of the chain under that resistor where you found it connected on the one you are repairing. You would not want to have a HPIL module plugged in though.

(03-11-2018 09:51 PM)MikeSD Wrote:  But I'm confused now. You have been talking about a U9 chip. I have been under the impression that is the normal chip in an unmodified 71B. And the chip above it is the odd chip. Shown top left, in this pic. I thought that was a MATH chip, based on the part number. I don't see any place for a 2nd ROM for forth. The top left I'm calling a Math chip. I also suppose I might not have two 32k RAMs. It might be the missing FORTH ROM and something else. I've been assuming it's all RAM but I could be wrong. Never looked at it very closely. I was thinking I might replace those wires, so I will be able to get to the bottom of that

Sorry for the confusion I was just using the to reference the position of the chip, under what you discovered is the math chip, bottom left, and yes U9 would be a normal 4K module if the original module was still there but in your case that position appears to be occupied by something else since circuit card looks different from the other 4K modules.

The two modules that are bonded together and hanging on the end of wires look like CMT 32K modules that is why I have been referring to them as such and the SOIC package chip on the is a 32KB SRAM which makes it even more likely.

I assure you the entire FORTH module would be a single module.

Paul.

[Paul Berger (Canada)]

(03-11-2018 11:11 PM)MikeSD Wrote:  Might have found the missing 2nd FORTH ROM. I was assuming I had a 64K RAM config. It might bea 32K, and the 2nd FORTH.

one device is marked HM?256LFP (assume 32k RAM) but the other is unmarked. Might be ROM.

So the modification might add 32K RAM, FORTH (2 ROMs), and a Math ROM. Now, if i can just get these tested and the 71B working. Spent the whole day tearing apart an documenting. Need a break to digest all this.

I still think they are both RAM and likely 32K the whole of the FORTH would be on a single module.

Look for CMT somewhere on the card I am pretty sure they are their modules.

Paul.

[MikeSD]

ok, that mystery solved. When you said 2 modules i thought you meant two devices. got it now.

I wonder if my 32K RAM is actually two for 64, or one32K and a different ROM. odd one would be marked and the other not.

One last question befor calling it quits for tonight. You said the chip under Math looked different. i dont see any difference. There are 5 other chips besides the Math. All appear to be the same, except for the center bottom. it has a different solder jumper, than the other 4. its the same on.my other 71B as well.

[Paul Berger (Canada)]

(03-11-2018 09:51 PM)MikeSD Wrote:  

Its hard to say why it is this way, perhaps when they where originally planning the 71B they thought they might need a 6th module inside and laid out the board to accommodate that, but later changed their minds and instead of removing all the pads just joined the the two pads so that if in the future the did need the top left it would be a simple cut rather than having to layout and produce a new board. I have run into odd little things like this before and I try not to get too wound up with why, I am just happy that they work. They may have used pin 10 because it was easier to route the connections and 10, 11 and 12 are all not connected on RAM / ROM modules. The connection from Dout of U8 actually goes though pin 11 of top left, pin 11 of top left and bottom left are joined and 11 of bottom left goes to pin 8 on the top side of the board under the module I am sure this was done for ease of routing the signal lines.

[Paul Berger (Canada)]

(03-11-2018 11:21 PM)MikeSD Wrote:  ok, that mystery solved. When you said 2 modules i thought you meant two devices. got it now.

I wonder if my 32K RAM is actually two for 64, or one32K and a different ROM. odd one would be marked and the other not.

I am pretty confident that they are both RAM.

(03-11-2018 11:21 PM)MikeSD Wrote:  One last question befor calling it quits for tonight. You said the chip under Math looked different. i dont see any difference. There are 5 other chips besides the Math. All appear to be the same, except for the center bottom. it has a different solder jumper, than the other 4. its the same on.my other 71B as well.

Right you are I thought I saw something in one of the earlier pictures that made it look different.

Paul.

[rprosperi]

I don't know if this helps or makes things more complicated, but do note that some 71's were modified by replacing (some or all of) the original 4K RAM chips on the mobo with 32K chips, and then (usually) also had additional modules wired-in as well.

Perhaps you guys are already well past considering this (I can't tell for sure) but thought I'd throw it out there in case it helps to decipher things.