The Museum of HP Calculators

HP Forum Archive 20

[ Return to Index | Top of Index ]

Another approach - shucking the clamshell, further. Book?
Message #1 Posted by Wanderley Navarro on 3 June 2011, 2:13 p.m.

After swapping several good HP-18C bodies with good HP-19BII & HP-28S CPUs I find now that dealing with the Clamshells is not too dificult. Not true for patience however...

For part of this demonstration was used the body of an 19BII without the CPU that was previously removed prying the upper part of the front since both front and back was already broken. This is why you see that the keyboard stakes are intact. The labels are from the 18C body donor.

Warm up well before attempting to remove the labels.

Initiate the removal lifting one corner with the help of a sharp blade and keep blowing hot hair.

This is what you will find. Notice that the stakes near the latch area was already broken at its base and came off entirely along with the label (as pointed out by Randy on the previous thread).

I use a Dremel with vertical suport and a drill bit of 3/32" to remove the mushroomed heads of the stakes. Remember that there are stakes under the label located above the display too.

To set up the maximum down movement of the drill: First - lower the drill just enough to touch the stake head with the drill bit.

Second - with the drill bit touching the stake head adjust the lever on the suport to have the vertical movement stoped at this point.

Third - release the drill to its upper position and insert a sheet (+/- 1 mm) of cardboard below the calculator that will be lifted a bit. This way the cardboard thickness will define the deep of the drill.

Notice (between the keys 5-6-8-9) that after drilling the head a small portion of the stake must remain available to be mushroomed in the reassembling process.

The stake's head that was removed.

So, let's take a look at the most common key issues that affect the models of this series:

1 - Broken threads on the keyboard flex PCB;

The most common area to search for broken threads are located on the lower center corners where the flex PCB folds. Look at the other areas where the PCB folds also.

Use the keyboard schematic (provided below at the end) to find the faulty thread and fix it using something like a conductive silver mix for heat sensitive surfaces.

2 - Bad connection between the keyboard flex PCB and the main PCB that can be due to:

2.1 - Decomposing compression foam;

To solve this problem I go in another direction instead of replacing the foam, with the use of the silver mix on the contact points of the two PCBs. Depositing a good amount to form a small mound will fill the gap of contact between the two PCBs.

On the flex PCB points of contact...

Notice the small mound...

On the main PCB contacts...

Small mounds also...

2.2 - Broken stakes near the display area;

2.2.1 - You saved the original stakes. Better if they are still glued on the keyboard overlay in its original position. This is relevant since the stakes broke at variable angles and lenghts.

2.2.2 - The original stakes are missing. So a request from a donor is needed.

Cut flush to the back of the donor and sand any eventual remaining piece of stake on the receiver. See that the model on the picture below has already donated almost all its stakes.

With the stakes at hand some tools will be needed to accomplish the task of reinsert them. Super glue, a sharp pointing tool, a needle alike tool and a small piece of plastic to serve as the glue base.

Press the needle at the top of the stake and you will have a good way to work with it.

Drop a small amount of glue on the plastic base and use the pointing tool to put only the amount needed on the place where the stake will be glued.

Reposition the empty keyboard structure back in place and insert the stake on its hole. Allow a bit for the glue to dry. Remove the keyboard structure and repeat for the next stake. After all the glued stakes in place allow some time (I use a day) to work on the next step.

Now in the process of reassembling you will mushroom the stakes to form a head. Use a not too hard tool to apply pressure near the stake that will be mushroomed.

The pressure must be applied and maintained while the stake is mushroomed. This is maybe the most important moment since pressure between the halves is a key factor. So avoiding any damage apply how much you can. Wait the tool a second in place for the plastic cool down.

This is how it will look after the melting process. The new head must be entirely inside the hole. With the above mentioned pressure the result is a calculator as sturdy as the original.

For good mushroomed heads try to use a not too sharp iron tip.

Color of body, labels, keys and CPUs aside the five models of the series are all the same. But one thing to do when swapping is to reposition the beeper since the 18C have it on the middle and the others on the upper left. The front HP logo can be swapped too. Just heat it a bit and remove with the aid of

a sharp blade. Another diference between the 18C and the others is the lenght of the springs inside batteries's chamber. On the 18C they are a bit short and that is the reason you do not see too much 18Cs with broken batteries door area. It seems that at some point HP changed the project using longer

springs that resulted in more pressure on the door. So, this is why changing batteries on a 18C is a much more easier task than on the others. Not sure now but I think the 28C has the short springs too.

I was able to map the keyboard schematic as follows. In the absence of a more appropriate design tool the good old manual method was used. I'm sure several of you will be able to do a better job translating it with a proper sofware.

      
Re: Another approach - shucking the clamshell, further. Book?
Message #2 Posted by Geoff Quickfall on 3 June 2011, 3:52 p.m.,
in response to message #1 by Wanderley Navarro

Lots of useful information here, great job. The restaking looks fantastic and allows for an unblemished outer case. Your dremel looks to be alot cleaner then mine ;-). I like the dremel as it allows a quick drill job with perfect depth control.

some questions;

1.  How is the long term survival of the posts once glued?  I ask
 as the glue option does not seem as strong as the original one
 piece design.  This is not a critism at all, as your repair is
 extremely elegant.  I ask as this is an option I now have to
 include in the text of the book, (with your permission). 
 I elected to continue throught the back with new stakes as the
 glue weld option has a larger surface area to work with for
 strength.

2. I see the strip contacts on the ribbon were reinforced with the silver conducter paint. They appear to be worn or scratched probably due to the housing flexing as the stakes broke. How did you address a compressed foam, or did you replace it along with addressing the worn stips on the ribbon.

Thanks for the posting also!

Cheers, Geoff

Edited: 3 June 2011, 3:53 p.m.

            
Re: Another approach - shucking the clamshell, further. Book?
Message #3 Posted by Wanderley Navarro on 3 June 2011, 5:08 p.m.,
in response to message #2 by Geoff Quickfall

Hello Geoff,

First, I want to thank you for all your repair tips. Has been invaluable to me. Second, congrats for the book initiative... and put my name on the list.

Quote:
1. How is the long term survival of the posts once glued?

Dificult to determine or even know since after reassembled the calc may still work well after the fail of some glued stakes due to the use of the silver conductive mix to "enlarge" the thickness of the contacts.

Quote:
I see the strip contacts on the ribbon were reinforced with...

Two diferent keyboard PCBs was used in the demonstration.

On the picture where the conductive silver pen is pointing: the worn on the ribbon contacts are due I used this keyboard PCB to map the schematic. So during this work the multmeter leads scratched the contacts.

On the picture where the keyboard schematic is presented: there are some contact points that received the silver mix that I used to test the conductivity of the product when I bought it.

Quote:
How did you address a compressed foam, or did you replace it...

As I mentioned on the post I do not replace the foam but I use an amount of the silver mix to form a small mound to compensate the lack of pressure the foam should make.

Cheers. Wanderley

                  
Re: Another approach - shucking the clamshell, further. Book?
Message #4 Posted by Dave Shaffer (Arizona) on 3 June 2011, 10:30 p.m.,
in response to message #3 by Wanderley Navarro

I have not worked on any of these calculators, but a question about your drills:

I assume you want as much of the original heat stake to remain as possible. So, instead of using standard drills (which leave a v-shaped depresion at the top of the heat stake), how about removing the tip of the drill (with a grinder?) so that the tip of the drill is flat (looking more like a router bit), which would leave a flat top on the "drilled" out heat stake? i.e. More like it was milled away.

                        
Re: Another approach - shucking the clamshell, further. Book?
Message #5 Posted by Wanderley Navarro on 3 June 2011, 11:45 p.m.,
in response to message #4 by Dave Shaffer (Arizona)

I think it's a good idea. Just not sure how better is to work with since the tip helps to determine and keep the drill centered.

                              
Re: Another approach - shucking the clamshell, further. Book?
Message #6 Posted by Dave Shaffer (Arizona) on 4 June 2011, 12:25 a.m.,
in response to message #5 by Wanderley Navarro

If you can make the end really flat, then how well you center the Dremel bit will determine where you take off the plastic (and, I think, the faster you spin your Dremel, the better). I have a battery-powered, hand-held Dremel tool, and I find the faster I spin it, the better, for almost everything I (try to) do with it.

      
Re: Another approach - shucking the clamshell, further. Book?
Message #7 Posted by Mike Morrow on 4 June 2011, 1:01 a.m.,
in response to message #1 by Wanderley Navarro

Absolutely stunning and beautiful work!

Absolutely insane HP design, an example of HP's disposable calculator ethos that began with the clamshells and pioneers.

            
Re: Another approach - shucking the clamshell, further. Book?
Message #8 Posted by Victor Bilandzic on 6 June 2011, 5:04 p.m.,
in response to message #7 by Mike Morrow

Fascinating work and fix!

Probably one of the earliest applications of flex circuits around...

- Vic
San Francisco Circuits

                  
Re: Another approach - shucking the clamshell, further. Book?
Message #9 Posted by Randy on 7 June 2011, 7:23 a.m.,
in response to message #8 by Victor Bilandzic

Quote:
Probably one of the earliest applications of flex circuits around...
Not even close as the clamshell appeared in 1986.

Print heads in the 67/97/19C (1976) where possibly first followed by the HP41C Battery & I/O block (1979). These were followed by the original Voyager design flex circuit (1981).


[ Return to Index | Top of Index ]

Go back to the main exhibit hall