25, or 4 to 6 [Not Chicago Content]

[Les Bell]

I have an HP-25 (non-C), for which I've rebuilt the battery pack and added a 6.3V zener across the battery contacts (not that I've even plugged in the charger, so far - in fact, it's still running on the initial charge in the new NiMH batteries).

It's got a keyboard quirk, though; sometimes pressing the '4' key places a '6' in the display. Repeated keypresses will give strings of 6's and sometimes 4's, and on occasion it will actually kill the display. A bit of jiggling sometimes resolves the problem, and right now it seems to be behaving itself, but the problem comes back after a few days of disuse.

A bit of Googling revealed a post in the old forum from someone with the same problem, but no fix was suggested.

Any suggestions? Without having stripped the calculator down and performed more comprehensive testing and cleaning, I tend to suspect dust under the key, but would appreciate opinions from those with more 25 experience.

[Geoff Quickfall]

Try this Les

DEOXIT gold introduced through the pinhole associated with the switch. Hopefully you have the keyboard design with the pinholes. If it were mine and it did not have the pinholes I might consider drilling a tiny hole BY HAND (tiny drill bit between thumb and forefinger or held in a pin vice) over the centre point of the offending button and introduce the DEOXIT that way.

Still no joy, then the extreme, remove the tops of the heat stakes, minimal amount leaving enough in place to remelt after cleaning the board. The last method is the most extreme and difficult so try drilling the tiny hole first.


See this posting!

http://www.hpmuseum.org/cgi-sys/cgiwrap/...869#217869

Geoff

[PANAMATIK]

(01-29-2015 02:17 AM)Les Bell Wrote:  I have an HP-25 (non-C), for which I've rebuilt the battery pack and added a 6.3V zener across the battery contacts (not that I've even plugged in the charger, so far - in fact, it's still running on the initial charge in the new NiMH batteries).

It's got a keyboard quirk, though; sometimes pressing the '4' key places a '6' in the display. Repeated keypresses will give strings of 6's and sometimes 4's, and on occasion it will actually kill the display. A bit of jiggling sometimes resolves the problem, and right now it seems to be behaving itself, but the problem comes back after a few days of disuse.

A bit of Googling revealed a post in the old forum from someone with the same problem, but no fix was suggested.

Any suggestions? Without having stripped the calculator down and performed more comprehensive testing and cleaning, I tend to suspect dust under the key, but would appreciate opinions from those with more 25 experience.

Getting 6's instead of 4's is strange because these keys are in different keyboard columns, there might be some contact or smear, that has to be removed between the columns.

Each row, these are the horizontal five keys from left to right, is scanned every 308 microseconds. After 7 rows, that is after 2,1 ms, all keys are read. If a key is pressed, the voltage of the column goes from 6.4V either to 0V or to a midrange voltage about 3,5V , if the LED display segment is on, that is displayed with the same row as the keyboard. In my ACT chip I read the keyboard columns by its analog values, because they can have three different levels.

What do you mean with "kill the display"?

If you have an HP-21, HP-22 or another HP-25 or HP-29C, they all have they same keyboard, you can exchange them and check whether it is only your keyboard that fails.

I had a HP-21 keyboard with sticky 5 and it seemed to get better when moving the key from right to left several times, while still being pressed.

I have many (about twenty) HP-25/HP-21 etc. keyboards. I never had wrong key detection (4 to 6), always the key was sticky and needed more pressure.

Goeff is right to recommend cleaning!

Bernhard

[Les Bell]

(01-29-2015 03:39 PM)Geoff Quickfall Wrote:  DEOXIT gold introduced through the pinhole associated with the switch. Hopefully you have the keyboard design with the pinholes. If it were mine and it did not have the pinholes I might consider drilling a tiny hole BY HAND (tiny drill bit between thumb and forefinger or held in a pin vice) over the centre point of the offending button and introduce the DEOXIT that way.

Thanks, Geoff - I think this is the way to go; it just "feels" like a mechanical/corrosion problem to me, from the way that jiggling the key or pressing it from different corners seems to affect the number it actually enters. I'll order a tube of DEOXIT Gold this morning.

(01-29-2015 03:39 PM)Geoff Quickfall Wrote:  Still no joy, then the extreme, remove the tops of the heat stakes, minimal amount leaving enough in place to remelt after cleaning the board. The last method is the most extreme and difficult so try drilling the tiny hole first.

Let's hope it doesn't come to that! There's a special place in you-know-where reserved for the evil genius who invented heat stakes.

[Les Bell]

(01-29-2015 06:06 PM)PANAMATIK Wrote:  Getting 6's instead of 4's is strange because these keys are in different keyboard columns, there might be some contact or smear, that has to be removed between the columns.

Good point, Bernhard - when I take it to bits to clean it, I'll take a close look at that row.

(01-29-2015 06:06 PM)PANAMATIK Wrote:  Each row, these are the horizontal five keys from left to right, is scanned every 308 microseconds. After 7 rows, that is after 2,1 ms, all keys are read. If a key is pressed, the voltage of the column goes from 6.4V either to 0V or to a midrange voltage about 3,5V , if the LED display segment is on, that is displayed with the same row as the keyboard. In my ACT chip I read the keyboard columns by its analog values, because they can have three different levels.

What do you mean with "kill the display"?

Thanks for that - although I don't have much test equipment myself these days, I can always borrow a scope in one of the labs at work. By "kill the display" I mean that the display goes dark, which suggests to me that keyboard/display scanning has stopped.

I'll give it a clean with DEOXIT Gold - chances are, that will fix it (fingers crossed!).

[Bruce Larrabee]

Les,
My experience is that everything they said is right on target. Don't know if that's helpful but for what it is worth...
Bruce

[PANAMATIK]

(01-29-2015 11:06 PM)Les Bell Wrote:  By "kill the display" I mean that the display goes dark, which suggests to me that keyboard/display scanning has stopped.

I'll give it a clean with DEOXIT Gold - chances are, that will fix it (fingers crossed!).

If the display gets dark because you just pressed a key or you get another key than expected, it is unlikely only a debouncing problem. But give it a chance, I will cross my fingers too for you.

Bernhard

[Les Bell]

(01-30-2015 12:25 AM)Bruce Larrabee Wrote:  My experience is that everything they said is right on target. Don't know if that's helpful but for what it is worth...

Don't worry, Bruce - I know a guru when I see one.

[Les Bell]

(01-30-2015 02:27 AM)PANAMATIK Wrote:  If the display gets dark because you just pressed a key or you get another key than expected, it is unlikely only a debouncing problem. But give it a chance, I will cross my fingers too for you.

In this case, when pressing the '4' key repeatedly to test it, the symptom is to get a few '6's and then the calculator goes dead - i.e. the display is dark and no other keys work. Switching off and then on again brings it back to life, though.

But you could be right about debouncing - I just picked the calc up to see what it would do - it had been behaving itself - and this time I got two '6's and then it died. On repeated testing, I'm not seeing two - sometimes three? - '4's or '6's being entered in response to one keypress. And the problem seems to have spread to the '5' key as well, only not as bad.

Sigh. No rush; I'll wait for the DEOXIT to arrive, and then see how it responds to that.

[Geoff Quickfall]

Mel's,

If just a debouncing issue then the DEOXIT will work. This sounds a little bit more complicated.

Personally I start with the "KISS" principal or at least the medical tenant to "do no harm". I would start with the pinholes for the 4 key and the 6 key and introduce a tiny (bigger then a pinhead) drop of DEOXIT gold by placing the tiny end of the tube against the offending pinhole and squeezing, then working the button. Clean up the excess.

If this does not work I would disassemble the calc and clean all the pin contacts for the PCA and display. Reassemble and check.

Next check the PCB for corrosion and cold solder joints maybe rinsing in a white vinegar bath followed by a water rinse and thorough dry. This bath does not include the LED BLOCK. Reassemble and check.

I have actually washed the keyboard but only with water and a mild liquid dish soap solution. No nail brush but maybe a natural bristle paint brush. Do not use hot water, in some cases this may cause a bloom on the black background of the keyboard facia. This is due to the lacquer breakdown by UV light. It happens rarely and can be fixed by allowing the keyboard to dry thoroughly. If still present, some conditioners like Armoral can be applied with a Q-tip to the facia of the keyboard. Do not use alcohol or solvents as the silk screened characters will vanish.

Lastly, swear an oath to the inventor of the heat stake and then removing the caps sparingly leaving some residue for re melting. Not to be attempted by the faint of heart!

Or find a dead 25 on your favourite auction site and harvest the keyboard from it. The dead ACTS 25s are plentiful!

Cheers

[ElectroDuende]

(01-30-2015 08:16 AM)Geoff Quickfall Wrote:  I have actually washed the keyboard but only with water and a mild liquid dish soap solution. No nail brush but maybe a natural bristle paint brush. Do not use hot water, in some cases this may cause a bloom on the black background of the keyboard facia. This is due to the lacquer breakdown by UV light. It happens rarely and can be fixed by allowing the keyboard to dry thoroughly. If still present, some conditioners like Armoral can be applied with a Q-tip to the facia of the keyboard. Do not use alcohol or solvents as the silk screened characters will vanish.

Sorry to take the thread... but I have something that meay be related to this. I have an HP22S where the keyboard plate, where the blue silkscreened symbols are, has developed really tiny "blobs", like the surface of an orange. In the begining I thought them to be stains, so I used isopropyl alcohol and rubbed... bad idea! Now I have a couple of spots where bare metal is shown...

As it is a second hand unit, I don't know its history, maybe worked in harash enviroment or strong sunlight for long time... but no other of my pionneers (and I have quite a good number) has ever developed a single defect in its frontplate (by itself; yes, some have scratches).

What type of lacker would you recommend? I'm not very confident on the results... I have even considered the option of making a vinyl replacement (like the one done for WP34s).

[PANAMATIK]

(01-30-2015 08:16 AM)Geoff Quickfall Wrote:  Or find a dead 25 on your favourite auction site and harvest the keyboard from it. The dead ACTS 25s are plentiful!

Cheers

All the dead ACts 25s can be repaired now. So you will have two working units. If all cleaning fails, then you can still exchange the ACT and there will no longer be 6s for 4s and a dark display.

The first three new ACTs will be sent in a few days to some members and I hope they will give some feedback here.

Bernhard

[Geoff Quickfall]

Hello Bernhard,

Looking forward to the ACT. Of course the dilemma is that dead ACT woodstocks are no longer keyboard donors!

Electrodude,

Check out this posting:

http://www.hpmuseum.org/cgi-sys/cgiwrap/...ead=158000

Geoff

[Les Bell]

(01-30-2015 08:16 AM)Geoff Quickfall Wrote:  If just a debouncing issue then the DEOXIT will work. This sounds a little bit more complicated.

Thanks, Geoff. I'll have to sit tight for a few days, until the DEOXIT arrives; I'll know more after that.

Thanks to all for the info - I've bookmarked the thread and will return!