An HP-42S fix success, and a contrast problem


I bought a "dead" 42s from TAS for $20 USD with shipping, so I was happy with that. After putting batteries in it, I got a constant on and off beeping. I looked that up and most opinions pointed to a bad RAM chip. I've now done a few upgrades on these, so I decided to upgrade this one. After taking it apart, I put the batteries in it one last time and it now wouldn't even beep. I finished the upgrade and it now works! Very happy with that. (Side note: It had a Toshiba RAM chip in it, the only one of those I've seen so far)

Now I notice that I can only see the display if I'm looking at the screen from a very low angle. If you look at it straight on, it literally looks off. No problem I thought, but when I went to adjust the contrast, nothing. The contrast will not change at all, and seems to be stuck in low :(

This is a version A with the flush display, S/N 2942S30206 and it passes it's self test.

Has anyone else run into this, or have any ideas?



Hi Dave

LCD drive voltage is made by on-board step-up converter and it should be 7.0V(light) ~ 8.3V(dark) normally.

How much is the voltage at the 10u/16V tantalum capacitor near the crystal resonator?

My first guess is a broken solid-tantalum capacitor.

Regards, Lyuka


Thanks, I'll check that when I get back from work.

I neglected to mention that batteries had leaked in it at one point. It doesn't look too bad as only a few copper traces were green. I cleaned those some, but could probably do a better job. I've only taken the back off of the calculator, so I don't know if there could be some hidden damage on the other side of the board due to that.



I'm getting 4.52V across the 10uF capacitor. That looks to be quite a bit lower than the "light" voltage of 7.0 and that seems about right as you can barely see the display.

Replace the 10uF cap then?



Hi Dave

It seems that 4.52V is just below the battery voltage. It implies that

1) the inductor is not in open-mode failure.

2) the Ta capacitor is not in short-mode but may be in high-leakage state that is just before short-mode failure.

3) the SiSB diode may be in short-mode failure.

4) the FET switch in the processor may not working or in open-mode failure.

Replacing both capacitor and diode would be better.

Regards, Lyuka



Thank you so much for your help. I have already ordered some of the capacitors, but am having a little trouble getting the right diode. Does this look like one that would work?:

LL4148 Fast Switching Diode, SMD 1N4148



Hi Dave

1S4148 is a Si-switching diode and is not suitable for this purpose.

I recommend to use one of small signal Si-Schottky diode with main characteristics of

    Vrrm >= 30V, If >= 0.1A, Vf <= 0.4V (@If=10mA, Tj=25C), SOT-23 package
such as, for instance, STMicroelectronics's BAR42FILM or BAR43FILM.
(You may be able to purchase it at Digi-Key.)



I think I found what you are referring to here. Unless I'm looking at the wrong component on the board, this doesn't look anything like the diode just under the 10uF Ta capacitor. But I see a place where something like this could fit, just to the left of the diode, still under that capacitor. There is currently nothing there now. You're suggesting that I remove the other diode and replace it with this sot-23 package in that other spot?




You're suggesting that I remove the other diode and replace it with this sot-23 package in that other spot?

Just as I've replaced both capacitor and diode of one of my HP-42S.



Thank you so much. Actually, I've been referring to one of your photos while talking to you back and forth because my 42s is at home. I didn't realize that was your site until you sent the link.

Your website has been a great help to me and I'm sure many others. Thank you for taking the time to document what you're doing, and with nice, clear photos!



My pleasure.




I have just replaced both the cap and the diode without any effect on the display. I get 4.49V across the cap, similar to before.

Maybe there is something else I could try, but I'm wondering if it's possible to change the value of the cap or other component so that at least the display is "stuck" in a usable contrast setting.




That's too bad.

There seems to be a problem of the switching regulator in the processor, or PCB traces between the diode and the processor.

The last hope is that the factor of the problem is outside of the processor. Let us check the PCB again.

Remove the battery and make both contacts connected for several seconds and remove the connection, then check the continuity by an ohmmeter of a DMM.

 1) D1(-), U1-116(Vcc)(+) should be near the value of 27 Ohm (series resistance of the inductor L1)
2) D1-A(-), U1-117(+) should be less than 0.1 Ohm
3) D1-K(-), U1-119(+) should be less than 0.1 Ohm
Note: (+) : positive measurement voltage side, generally red lead of the DMM,
(-) : negative measurement voltage side, generally black lead of the DMM

1) possible problem is short circuit between U1-116 and U1-117 which will make the voltage stacked to Vcc.

2) possible problem is breakage of the trace which will make the voltage stacked to just below the Vcc by loss of switching.

3) possible problem is braakage of the trace which may cause loss of voltage control but I'm not sure.

Good luck.



Hello David and Lyuka,

I have a similar problem, I just upgraded the Toshiba RAM to 32 K on an HP42S #3012Sxxxxx with the flush display and version A.

The contrast does work but the display itself is weak. Unlike yours David, the contrast works but when it is backed off to allow the digits to display the display is noticeably lighter then my newer models.

Was this a function of the older displays (less inherent contrast between digits and background) or could this also be remedied by the diode and capacitor change?

My calculator also resided in the case and was in pristine condition and fully functional (no bad foam contact).

Cheers, Geoff

P.S. Lyuka, could you provide the part number for the capacitor that you used. Thanks

some added specs:

the capacitor: 

- calculator off gives 4.6 volts
- calculator on with low contrast 6.3 volts
- calculator on with high contrast 7.6 volts

Edited: 3 Dec 2009, 1:20 p.m.


Here is the link to the capacitor that I ordered. I believe (hope) this is the right one, but maybe Lyuka could confirm that.

I'll report my results here when I've replaced the cap and diode. I think I'll get them in 3-4 days.




just sent you an email and you posted the answer here!

so I have ordered the TA and the diodes also.

Cheers, Geoff



The contrast of the LCD of my HP-42S #3005Sxxxxx with flash display and version A is relatively weak than that of #3242Sxxxxx with matte display.

Voltage at the 10uF capacitor at Ta=25C, Vbattery=4.70V

  Serial#/revision  : light  usual  dark
#3005Sxxxxx/Rev.A : 6.27 7.40 7.71 [V]
#3242Sxxxxx/Rev.C : 6.39 7.52 7.58 [V]

The capacitor that I used was TDK C3225JB1C106M (EIA X5R characteristic), but
TDK C3225X7R1C106M that Dave mentioned has wider temperature range and is good enough.

Please take care not to make short circuit by the shield plate at the back of the housing, though there is a bumper near the diode and capacitor,




Mine sounds very close to yours vis a vis the voltage. This is probably typical for the early displays.

It is readable, just not as contrasting. So the upgrade will probably not help. Still I will order the parts for backup.

I agree with David, your site is fantastic, as are your descriptions.

Many thanks


Well I ordered the following parts from DigiKey and recieved them the next day (yesterday).

So, I swapped out the diode and the capacitor for the following as per Lyukas instructions. Although the 42S was in specs for the voltage across the original capacitor the swap out did improve the contrast between the alpha/numeric display and the background and boosted the overall ability for the display to contrast to dark.

I did this to include it in the book with photos as I am trying everything I see posted.

Here are the parts I ordered including the capacitor using the X5R characteristic:

SCHED B: 854110 ECCN: EAR99

CAP CER 10UF 16V X5R 20% 1210
SCHED B: 853224 ECCN: EAR99

Thanks Lyuka, and Dave (hope your trace tracking works out for you)

Cheers, Geoff


Hi Geoff,

That's good :-)

The swap out you did also improved the long term reliability and reduced the power supply current, I think.

In my case I_idle reduced from 440uA to 380uA, I_standby reduced from 14.0uA to 13.5uA. (Ta=24, V_battery=4.73V, f_osc=65.536kHz, Serial#=3005S0xxxx)


The recommended parts for this replacement (which Geoff used) are

  SiSB diode : STMicroelectronics, BAR42FILM
10uF MLCC : TDK, C3225X5R1C106M
It's also recommended to add 1uF MLCC (TDK C3216X7R1E105K) as a bypassing capacitor, on to the pad between the RAM and the processor, to improve the stability.

Cheers, Lyuka

Possibly Related Threads...
Thread Author Replies Views Last Post
  SCIOVR and ENGOVR in FIX mode Andrew Nikitin 5 1,366 07-02-2013, 04:46 AM
Last Post: Victor Koechli
  PILBox on Mac OS X success Håkan Thörngren 5 1,499 03-25-2013, 11:45 AM
Last Post: Egan Ford
  Contrast in HP42s flush display , another one.... aurelio 0 631 09-01-2012, 08:47 AM
Last Post: aurelio
  A temporary "fix", the Casio Prism (fx-CG10) LHH 5 1,519 08-25-2012, 07:39 PM
Last Post: LHH
  Craigslist find and fix - HP 19C Matt Fegenbush 12 2,854 08-13-2012, 02:57 PM
Last Post: Matt Fegenbush
  Keyboard fix for the HP 15C LE Jeff Raines 2 870 03-31-2012, 12:49 AM
Last Post: Jeff Raines
  Just a thought I had--a 35S Trig fix Matt Agajanian 62 8,343 03-26-2012, 10:40 AM
Last Post: bill platt
  15C-LE Overdue for a fix? Matt Agajanian 7 1,674 03-24-2012, 01:46 PM
Last Post: nina scholz
  Keyboard fix for the HP 15C LE? Jeff Raines 4 1,226 03-23-2012, 10:54 AM
Last Post: Jeff Raines
  Success: Flashing a 30b on Windows 7/x64 Ingo Paschke 2 933 03-01-2012, 12:31 PM
Last Post: Walter B

Forum Jump: