HP-97 battery pack substitute?

[Mike McCauley]

I have a HP-97 that I purchased some time ago off eBay. I use it almost daily, and it works great, save for one problem that pops up periodically. Every few years, it gobbles up a battery pack.

I use the machine on AC power only, never off the battery pack. As I think of it I unplug the wall wart and turn the machine on to run the battery down, but most times it is plugged in and turned off. As is well known, in time this trashes the batteries.

I chopped the battery pack open some time ago, and I rebuild it myself as required. This is not big deal, but the sub-C batteries aren't cheap, and NiCds are getting harder to find all the time.

After reading a number of posts on the site about how the charger/PS circuit works, I decided that I could substitute an electrolytic cap and a zener for the battery pack, the cap to provide the filtering to eliminate the "fluttering digit" problem, and the zener to insure that the machine would not be exposed to a high voltage spike resulting from the cap charging to the peak voltage of the wall wart.

I paralleled a 2200 uF 10V cap and a 5.1V 5W zener and attached it to the battery pack spring clips inside the battery compartment with clip leads. This works very well... for the most part.

When the machine is on, the wall wart/cap combo voltage gets pulled down to about 4.89V and the zener dissipates essentially zero power. When I turn the machine off, the cap voltage is clamped at 5.1V, and turning the machine on and off produces no obvious problems. All in all, the idea works great, except for one problem.

With the machine off, the zener heats up, quickly and to a relatively high temperature. Obviously, this indicates that the current limiting resistors on the output side of the unregulated supply in the machine are sized to allow too much current to flow though the zener.

I could use a higher voltage zener, but I'm concerned that a turn-on spike from the cap would harm the machine.

I could use a simple zener-transistor shunt regulator circuit with a small heat sink to spread the heat out. But the power being dissipated is still the same, and I am concerned about burning out the series resistors in the calculator

The idea behind all of this is to simply eliminate the "low power" LED being on when the next set of batteries gives out, and thus the buying of one set after another of batteries in the first place, batteries that, in my situation, are acting as no more than short-life filter caps.

Higher voltage zener? Shunt regulator? Bad idea, won't work? Something else?

Thanks in advance for your comments,

Mike

[Patrick Rendulic]

Hello. At the moment I have no answer to your question, but I just want to tell you how I "power" my HP-97.

I have rebuilt the battery pack with Ni-MH cells from Sanyo, 3.700 mAh. To charge them, I put the switched off machine on AC power for a whole day. I then remove the AC power adapter and use the machine on battery power.

I don't use it as often as you, so in my case the charge is sufficient for several months of use. In your case the charge would certainly be sufficient for several weeks.

Up to now, no problems with the Ni-MH cells.

[marais]

I use mine as often as Mike and wear out cells at an alarming rate. For some time I was wondering if a stabilized 4.8V (or 5V, for that matter) supply with 2A would not be a better solution. Does anyone know the max. current the 97 draws?

Andreas

[Eric Smith]

My bench power supply shows peaks over 3A, but its display update rate is fairly slow, so the actual peak could be somewhat higher.

[Randy]

I have modified a number of units to use an external power supply. A Digikey T989-P5P wall wart switcher works well, 6V, 2.5A, 5% regulation for $17.00. It's a simple matter to move the 97's charger plug out of the way, let it dangle inside and route the PS wires through the holes in the 97 and solder to the pcb areas marked + and -. The battery connects to the other side of the pins which you should pull off and tape up just in case anybody puts a battery into the unit. Simple and working well for many years now.

Edited: 3 Oct 2007, 10:24 a.m.

[Mike McCauley]

Thanks for the tip. Given my situation, this sounds like the best option for me.

Given what you say regarding connecting the switcher PS, am I correct in thinking that where you have it connected is at the same points in the circuit as the existing battery pack?

Also, depending on design, some switchers don't work properly under zero load. Using the PS you specify, is turning the machine off a problem, or must it be left on all the time?

Thanks again for the info!

Mike

[Randy]

Quote:

---

am I correct in thinking that where you have it connected is at the same points in the circuit as the existing battery pack?

---

Yes, it connects to the battery input.

Quote:

---

Also, depending on design, some switchers don't work properly under zero load. Using the PS you specify, is turning the machine off a problem, or must it be left on all the time?

---

It's a wall wart consumer power supply. Zero load is an acceptable condition.

[Mike McCauley]

Understood. Given all of what's been thrown out, for me, I think this is the best option.

Thanks again,

Mike

[Mike McCauley]

Thanks to all who responded. I appreciate the help and show of interest.

Mike

[Dia C. Tran]

When I rebuilt my battery pack I didn't glue them back together but simply taped them. It makes it easy to replace the cells again. I used NiMH instead of NiCad so the pack last at least 3 times longer than the original pack.

[Mike McCauley]

I, too, use only tape. Works fine.

Did you have to make any circuit changes in order to use NiMH?

I'm sure they would last much longer, and they would also be far easier and cheaper to obtain than NiCd. In my area (Dallas, Texas, USA), sub-C NiCd cells are tough to obtain. And on the Internet, with shipping and sometimes minimum order charges, it's not a cheap prospect.

Every 2-3 years, the cost of the cells is more than the cost of a brand new TI scientific calculator. But then, that's not an HP...

;<)

[Dia C. Tran]

No it just takes longer to charge. About 2 to 3 days to fully charge the battery with the stock walwart. Besides the walwart is charging the battery so slowly, it's almost safe to keep it pluged in all the time.

Edited: 3 Oct 2007, 4:14 p.m.

[Geoff Quickfall]

I have rebuilt 5 battery packs with Nimh and have yet to have a problem. I do however, recharge them in the accessory battery charger that HP sold at the time. I also do this with my hp67's, which I use in the 'office' recharging the batteries on the 82004a instead of the calculator.

I do this for a two reasons, the heat generated in the calculator cannot be good after 30 years of use and the stress on the older ciruits in the calculator is just not required if you can use the external chargers. Number 2, the external chargers are easy to fix and do not involve dis-assembly of the units to get to the recharging circuitry.

I may be a bit paranoid here but my HP 97, HP 92 and the 4 67's I have (I use the 97 and one of the 67's still) keep their charges for very long times.

Cheers, Geoff