The Museum of HP Calculators

HP Forum Archive 08

[ Return to Index | Top of Index ]

HP-41CV power supply
Message #1 Posted by Rob Stone on 8 Aug 2002, 4:43 p.m.

I have recently acquired a hp-41cv and while its in use it constantly turns off while idle for about 15 seconds. Is this normal or is it a problem? Thanks for the info.

HP-41CV power supply is innocent
Message #2 Posted by Andrés C. Rodríguez (Argentina) on 8 Aug 2002, 6:51 p.m.,
in response to message #1 by Rob Stone

Eventually, the display, not the power supply, is the main suspect... The display is the only active part of an idling HP41C/CV.

If I recall correctly, a PowerOff signal is generated by the display after the idle timeout. Perhaps Tony Duell or others may confirm.

Re: HP-41CV power supply is innocent
Message #3 Posted by Ed Martin on 8 Aug 2002, 7:11 p.m.,
in response to message #2 by Andrés C. Rodríguez (Argentina)

I have exactly the same problem on my CX, and I have investigated the problem. I thought it was such an odd problem that it must only have been me.

Andres is correct. The 41 actually has three power modes - Run, Standby, and Sleep. The 41 actually stays in Run mode for a very short period of time (to conserve battery power).

The power moding is controlled by two signals. PWO (controlled by the processor) is only high when the 41 is in Run mode. When the processor sets PWO low, it also instructs the display to take over power moding. The display keeps showing whatever is in its data storage registers while its internal clock counts down. When the countdown is complete (which is SUPPOSED to take 10 minutes), the display sets DPWO low, which puts the 41 into Sleep mode.

Despite understanding all of this, I failed in fixing my 41. At first, I thought I might have some bad connections, as I also suffered from poor display quality on the LCD (which I could fix by pressing above the display). I subsequently rebuilt the machine (checked for bad solder joints, cleaned contacts & fixed the screw posts and a cracked back cover). My LCD now works great - good as new - but the machine still turns itself off after 15 seconds.

About the only workaround I have found is to assign the ON function to a USER key - I have found Shift-ENTER to be handy. Then, after turning on the 41, I hit USER ... Shift-ENTER ... USER. I just have to remember to turn it off.

BTW, the 41 will stay on as long as a program is running.

If someone else has a solution to this problem, I would love to know. I've thought about getting a good display driver from a dead 41, but I'm not sure I want to take that risk.

- Ed

Re: HP-41CV power supply is innocent
Message #4 Posted by Steve (Australia) on 8 Aug 2002, 8:25 p.m.,
in response to message #3 by Ed Martin

How did you check the display contacts?

I had a problem with these contacts on an early HP41 that could only be seen under magnification.

It would be interesting to know *how* the display calculates the 10 minutes (is it affected by clock speed for instance).

Re: HP-41CV power supply is innocent
Message #5 Posted by David Smith on 9 Aug 2002, 4:45 p.m.,
in response to message #4 by Steve (Australia)

I have seen LOTS of HP41 machines with bad display contacts. They are particularly prone to come undone from the LCD module side, although they also pop loose from the keyboard. When I suspect an LCD problem (or have a dead machine that a replacement CPU will not fix) I go down the line of contacts and lever them gently with a dental probe and see which ones pop loose. If one does, usually more than one will. Be very careful soldering to the LCD module board.

Anothe problem that seems to occur quite a bit is solder flux or invisible grunge on the CPU card, LCD module circut board, or keyboard (in that order of probability). This is a primary cause of screens that go blank when a key is not pressed. Clean with 91% isopropyl alcohol. Dont let the alcohol drip down on the LCD window. It can cause clouding. If that happens either live with it or remove the LCD module and re-polish it.

Re: HP-41CV power supply is innocent
Message #6 Posted by Vieira, Luiz C. (Brazil) on 8 Aug 2002, 10:45 p.m.,
in response to message #3 by Ed Martin


I have three 41`s, two of them being halfnut and one (the grandma) being fullnut.

Does the problem you are dealing with appear in both models?

When you removed the LCD, you remove it from the keyboard's PCB or you took all of the assy together?

I'm somewhat curious about it. If the display generate this signal, does it have a separate line for it? If so, can the connection be simply removed? This way, the ON function would be no longer needed, as the LCD's circuitry will not send Poewr-Down signal (unless it is multiplexed in one of the lines in the display; I do not have the schematics in hand, now).

I'd like to know how to find more info about this particular problem, so I could be of more help. If any of you feel O.K. doing this, send me a detailled e-mail; I have one fullnut spread around my tech stuff and I'll try verifying as far as possible.


Re: HP-41CV power supply is innocent
Message #7 Posted by Rob Stone on 9 Aug 2002, 6:33 a.m.,
in response to message #3 by Ed Martin

Ed tried your workaround was unsuccessful can you describe the keystroke order again I'm not sure if I'm performing it properly. Thanks

Three comments
Message #8 Posted by Andrés C. Rodríguez (Argentina) on 9 Aug 2002, 7:41 a.m.,
in response to message #3 by Ed Martin

Three comments:

a) To Ed: The CX may be somehow different because of TIME functions; I don't have any experience with the CX. You description of PWO and DPWO is accurate and helped to clarify the issue!

b) To Rob: The ON function prevents the machine from turning off because of inactivity timeout, the standadrd manner to invoke it is [XEQ] [ALPHA] "ON" [ALPHA].

As this may be cumbersome for everyday use, Ed´s solution uses a redefined or "assigned" key. To this end, you need first (only once) to press [shift][ASN][ALPHA] "ON" [ALPHA] and then your choice of key; as per Ed suggestion you might press [shift][ENTER].

Then, every time you turn your calculator on, you just press [USER][shift][ENTER] to call the ON function, and then [USER] again to exit the USER mode. Just check this doesn´t conflict with other key assignments you may will to use.

c) To Luiz: As far as I can remember, DPWO is a physical signal, so there should be a physical trace which can be opened in the PCB. It may be needed to solder a pull-up or pull-down resistor (I suggest something around 200 Kohm) to ensure that the now opened trace has the proper level on the CPU side...

I ADVISE ANYONE NOT PROFICIENT WITH THIS KIND OF HARDWARE MODIFICATION JUST TO STAY WITH THE SOFTWARE "PATCH", SO TO AVOID RISKING A VALUABLE CALCULATOR. But if you know what you are doing and accept the risk, the thrill of a successful hardwired modification is a nice sensation!

Re: Three comments
Message #9 Posted by Ed Martin on 9 Aug 2002, 5:47 p.m.,
in response to message #8 by Andrés C. Rodríguez (Argentina)


The posting by Andres describes the software workaround I intended to give. Andres just did a much better job in describing it than I did.

If you execute the "ON" function, it sets flag 44 (Continuous ON) so that the machine will not turn itself off. I assume that this causes it to stay in RUN mode and therefore have a higher battery drain, but I'm not sure.

Note that you can't set flag 44 directly via the keyboard or under the control of a normal (non-synthetic) program, as it is in the protected area.

Assigning "ON" to a USER key is only a shortcut to save time on startup.

- Ed

A little clarification about power modes
Message #10 Posted by Andrés C. Rodríguez (Argentina) on 10 Aug 2002, 6:12 p.m.,
in response to message #9 by Ed Martin

The HP 41C/CV has 3 power modes:

1-RUN: While executing a program or a function (for instance, for less than a second when you press [SIN], or all the time needed to finish a program since you pressed [R/S]). In this mode, the calculator draws the most power (peripheral operation like Wand or Card reading may take even more power). It is supposed that the calculator spends not a large perecentage of time in RUN mode.

2-STANDBY or IDLE: The calculator is doing "nothing", except detecting eventual keyboard activity. The display keeps showing the last contents it had. This mode uses little power, and is key to the long battery duration, since the calculator seems to be active when it is really doing very little.

As the CPU is not running while IDLE, the display, which is the only component "working" in that moment, is the only place in which a timer could be counting. After 10 minutes in such mode, the display intends to shut off the calculator by means of the DPWO signal (Display PoWer Off).

Any RUN mode activity (i.e.: keypresses) resets the timer.

If the calculator has Flag 44 on (continuous mode on), it will ignore the display timeout, but in normal cases it will shut off.

So, after executing ON, the calculator may remain for a long time in IDLE mode, even without any RUN mode event.

[[Note: I am not sure if ON prevents the timer to generate the DPWO signal; or if flag 44 causes the CPU software to ignore DPWO, which seems more probable. A nice challenge for hardware experimenters...]]

3-SLEEP. In SLEEP mode the calculator appears to be off, it is just feeding a little current from the batteries to the RAM chips in order to keep their contents. Apart from detecting activity on the ON key, nothing is being done, and the power consumption is very low.

[[A sleeping calculator may be turned on by some peripheral activity like Wand or Card reading, or by the Time module]]

Re: HP-41CV power supply is innocent
Message #11 Posted by Tony Duell (UK) on 10 Aug 2002, 6:44 p.m.,
in response to message #2 by Andrés C. Rodríguez (Argentina)

One thing I've just thought of. On the fullnut machines (separate CPU/logic board), there's a capacitor on the CPU board that's connected (via the CPU board -> keyboard connector) to the display hybrid. This capacitor is used as part of the timer circuit for the auto-power-off function. IIRC, there were 2 versions of the display driver hybrid, which used different value capacitors. So there are 2 things to think ahout. Firstly, do clean and re-clean the contacts. An open-circuit here could cause some interesting problems. Secondly, has this machine ever been repaired? Has the CPU board or display driver hybrid ever been replaced? If so, is it possible that the capacitor is the wrong value. This does not apply to halfnut machines as far as I know. I can't find an equivalent capacitor on that machine.

Re: HP-41CV power supply is innocent
Message #12 Posted by Ed Martin on 11 Aug 2002, 4:09 p.m.,
in response to message #11 by Tony Duell (UK)

Ahh, the capacitor connection is intriguing. By any chance, do you know what kind of capacitor it is, and if there is any way it may experience current leakage? If the capacitor is leaky, that could explain the premature timeout.

What bothers me about this problem is that the calculator * does * wait a while before turning itself off, but it is so far shorter than 10 minutes. I cannot imagine a scenario where a logic problem or broken connection would lead to this condition, but a shift in an analog component .... hmmmm.

- Ed

Re: HP-41CV power supply is innocent
Message #13 Posted by Vieira, Luiz C. (Brazil) on 12 Aug 2002, 12:54 a.m.,
in response to message #12 by Ed Martin

Hi, Ed;

I am not sure about it, but I read a post about batteries leakage being cleaned; I cannot ensure it was related to this specific subject, but one thing occurred to me when reading your post: you are right. If the display is the one who controls DPOW (is it correct?) signal, it counts (down?) from a reference. The only way to accelerate it, if not by changing the starting countdown value (HW fixed?) is by increasing count speed. And that can be accomplished only by increasing clock reference.

Wow! Where are those Tecnnical Data, Maintenance Guides and the like about the HP41? They must have something about it...

Re: HP-41CV power supply is innocent
Message #14 Posted by David Smith on 12 Aug 2002, 12:54 p.m.,
in response to message #13 by Vieira, Luiz C. (Brazil)

The HP41 service manual is on the HPMUSEUM CDROM set.

Re: HP-41CV power supply is innocent
Message #15 Posted by Tony Duell (UK) on 12 Aug 2002, 6:51 p.m.,
in response to message #13 by Vieira, Luiz C. (Brazil)

As far as I know (see my other post) the display driver hybrid does control DPWO (at least in fullnut machines). And when the CPU is idle, the display runs off its own internal oscillator, controlled by the capacitor I mentioned. So either the dispaly driver hybrid has failed (I guess it's possible for some of the logic to fail so that it no longer counts correclty) or the oscillator is running far too fast. I haev no idea how fast this thing will run with no capacitor connected, but maybe an open-circuit connection in the logic-keyboard connector will cause this. Or maybe the capacitor is faulty.

Re: HP-41CV power supply is innocent
Message #16 Posted by Tony Duell (UK) on 12 Aug 2002, 6:38 p.m.,
in response to message #12 by Ed Martin

The capacitor you want is on the logic (CPU) board. It moves around a bit (there are at least 6 versions of the HP41 logic board, of which 3 were used in 41CV machines). But it's generally close to the left end (looking at the component side of the PCB with the front edge towards you) of the CPU chip -- the largest chip on the board. It is connected between the front left pad that connects to the keyboard PCB and logic ground. Originally (Version F display hybrid, with a plasic cap over the chips) it was a 0.01uF capacitor. With the more common revision G display hybrid (epoxy blobs over the chips) it's 470pF. The front left pad of the keyboard - logic connecto is connected to the 4th pin from the left (looking at the back of the keyboard PCB) of the display hybrid pins. It would appear that the display time-put is handled by the display driver hybrid -- when the PWO signal goes low (i.e. the CPU is sleeping) then the displey runs off its own internal oscillator (rather then from the CPU bus signals). Then after 10 minutes (or so), the display times out, turns itself off, and brings the DPWO signal low as well (which is an input to the CPU). As far as I know, this only applies to the older fullnut models. The halfnut machines don't seem to have an equivalent capacitor, so I guess they work somwhat differently.

Re: HP-41CV power supply is innocent
Message #17 Posted by Ed Martin on 25 Aug 2002, 12:40 p.m.,
in response to message #16 by Tony Duell (UK)


I finally got around to replacing the suspect capacitor. It was exactly where you said it was, just to the left of the CPU (a 470pF cap, since I have the revision G hybrid with the epoxy blobs).

I had high hopes, since the epoxy coating on the old capacitor was cracked. Unfortunately, my 41 still does the same thing - turns itself off after ~14 seconds. This is identical to the failure mode reported by the person who originally posted this problem.

Any other ideas? Is it possible that I have a bad timer circuit in the display itself?

BTW, my calc is a fullnut 41CX made in 1984 week 21.

Any ideas you have would be greatly appreciated.

- Ed

Re: HP-41CV power supply is innocent
Message #18 Posted by Tony Duell (UK) on 25 Aug 2002, 3:42 p.m.,
in response to message #17 by Ed Martin

YEs, it is possible that there's a fault in the DDH (Display Driver Hybrid). The good news, though, is that this part is the same on all Fullnut 41s (C, CV, CX) -- well, apart from there being the 2 revisions that I mentioned in a previous message. So you don't have to hunt around for a CX to get the part. Somebody must have a beat-up CV (by far the most common HP41 over here anyway) that you could get cheaply. However, before doing that, do check the contacts between the DDH and the capacitor (it goes via one pin on the zebrastrip connector between the keyboard and logic board, for example. You might be lucky!

Re: HP-41CV power supply is innocent
Message #19 Posted by Ed Martin on 25 Aug 2002, 3:50 p.m.,
in response to message #18 by Tony Duell (UK)


Thanks! I've checked out the zebra connector, and everything appears to be OK there. It looks like the problem lies in the DDH, so I'll scrounge for a dead fullnut 41 to see if I can fix it that way.

- Ed

[ Return to Index | Top of Index ]

Go back to the main exhibit hall