HP Articles Forum
[Return to the Index ]
[ Previous | Next ]
Resurrecting a shattered HP 19B
Posted by Neil Hamilton (Ottawa) on 30 Jan 2012, 2:14 p.m.
I recently acquired a truly devastated HP 19B. When it arrived every heat stake in the lower section was shattered. The keyboard was lifting off the body to such an extent that the keys in the lower rows were in danger of falling out. This is in addition to the missing battery door. What a wretched life this poor machine must have had!
However, when I hooked it up to power, there was life! I had to pinch the body just below the display (and in several other places as well -- some photos show a big paper clip holding the bottom of the keyboard together -- though this was equally to keep the keys from falling out) but it was most definitely capable of life!
I am sorry I did not think to take a photo of the before state...
Having read Geoff Quickfall's excellent thread back in June ("HP 28S: Shucking the Clamshell" ), I covered the LCD with a block of blue Styrofoam and heated the upper and lower labels with a blow drier to remove them. As with Geoff's experience, I had a few heat stake heads come away with the label. As I discovered later, the ones that were left in their little wells were shattered also. In fact, the only ones which were intact were the ones above the battery area.
There are 2 sizes of heat stake used in this calculator. The majority are 2.4mm (3/32") while there are 6 which are 1.6mm (1/16"). These smaller heat stakes are the 4 middle ones above the top row of keys and the 2 middle ones just below and to the right of the ENTER key (oops! I meant INPUT key). After taking out all the shattered remnants of the heat stake heads, I drilled through every one using a drill press to maintain alignment. I used 0.060" and 0.0935 (ie: 3/32") bits to ensure a good, tight fit for the replacement styrene rods.
The battery compartment heat stakes where shaved conventionally (if not a bit over enthusiastically) with a brad-point bit.
Any crud left over from the drilling process had to be removed from the body cavity so the main portion of the calculator had to be fully disassembled at this point. When taking the front shell off the chassis, most of the keys fell out or drifted out of their assigned holes. The thin rubber membrane did essentially nothing to hold them in place.
It turned out that the tight fit for the styrene that I was hoping for was a little too tight. I had to ream each hole out by hand using the drill bits as tiny rasps. This did the trick and allowed the styrene rods to fit sufficiently well. (Note that, by mistake, I purchased 3/32" styrene tubing instead of rods. I don't think this made any structural difference but it would complicate matters if you wished to paint the back after you were done.)
For obvious reasons, the keyboard wants to be key-side down when the reassembly takes place. However, it turns out to be a non-trivial exercise getting the keys back in and aligning all the keyboard sandwich layers together while trying to poke the styrene replacement rivets into place. As others have mentioned, my state of evolution has not yet advanced to having the required 3rd (or 4th) arm necessary for such a task. None the less, I persevered and repeated the step more than a few times before I finally got everything to stick in the right place when the 2 halves were joined.
With a minimal number of the styrene rods holding the keyboard (and keys!) in place, I was able to flip the calculator over and test the connectivity and integrity of the keyboard. All keys worked acceptably except that there was very little snap. I have some suspicion that I may have applied too much heat to the shell in order to remove the labels. This may have compromised the snap domes somewhat (only thing I can think of).
With the keyboard integrity proven, I inserted the remainder of the styrene rods.
With the still reasonably tight fit I had created for myself, I discovered that the easiest way to get the rods through was to feed the styrene in through the back and push the whole calculator down against the table to drive the rods up through the sandwich and out the keyboard side.
I had purchased a solvent cement that worked for both the styrene and the original ABS. This was a methylene chloride based solvent and due to its low viscosity was capable of wicking down between the tightly fitted styrene rods and the holes in the back of the calculator. I bonded both the back to the styrene rods and the battery compartment heat stakes as well -- since I drilled them a little too much in the first place!
Because of the solvent's low viscosity, I did a nasty job making sure it stayed on my tiny paint brush without dripping onto the back side of the calculator. (It drips pretty easily!!) Unfortunately (or fortunately?), none of my photos does my sloppy job justice.
I next trimmed the keyboard area rods down to just above the keyboard surface, compressed the keyboard, and melted the styrene with a soldering iron.
This left a final testing of the assembly, the replacement of the labels, and some work to get a suitable replacement for the battery door.
For the latter, I recalled seeing a link somewhere for a "replacement" Clamshell battery door. The following pictures show quite a reasonable, if not pretty solution to the issue.
Mine is made from an old PC expansion card bracket and is nowhere near as well made as the one depicted in the links but will do until I improve my metal bashing skills.
It will not win any prizes for beauty but it is now functional and ready for action once again!
- I would use a bit less heat to remove the labels as I think this might have softened the plastic snap domes a bit.
- I would look into cutting down the battery springs somewhat before I reassembled everything so they wouldn't impart so much force to the battery door area through the batteries.
- I would buy styrene rods instead of tubes in case I wanted to paint the back afterward.
[ Return to the Message Index ]
Go back to the main exhibit hall