HP Forums

Full Version: Half & Half Recharging
You're currently viewing a stripped down version of our content. View the full version with proper formatting.

Hello all.

As I've recited that I'm the new parent of a 32E I got from Romania. I started out at 4:00PM yesterday recharging the calculator. Not knowing when I'd be home today so that I could stop the session in a timely manner, I decided to disconnect and unplug the charger from the 32E. I had thought to finish up the full session today. So, I reconnected the charger this morning to finish the cycle. Although I would not go through this two-step on a regular basis, is it okay to do for this time?

Edited: 7 Apr 2012, 4:25 p.m.

The charging circuitry is essentially just a series resistor, so there is no significant difference between charging eight hours continuously, or four hours today and four hours tomorrow.

This is true for all HP calculators that use NiCd cells. It is NOT true for any device that uses lithium ion or lithium polymer cells.

Edited: 7 Apr 2012, 4:30 p.m.

By the way--on the subject of memory effect, does that apply to the NiCd batteries that are in the 29?

Yes. That is a concern for all NiCd batteries.

Most of the research I have done on memory effect in NiCd batteries argues that true memory effect is not the culprit of a perceived shortened lifespan. Memory effect can occur in (for example) a satellite battery that discharges and charges to precisely the same level over and over, but is unlikely to occur in consumer products because the discharge from one use the the next is never exactly the same. I can probably dig out a few sources of papers explaining this issue better than I ever could - if you are interested.

Edited: 7 Apr 2012, 6:12 p.m.

This is absolutely correct from all the reading that I have done. What you are actually experiencing (and it gets thrown in under the heading of memory effect) is Voltage Depression. Here is an article that explains it. I haven't read the full thing so if you need more info search for "NiCd Memory Effect Voltage Depression".

Hope this helps,



Edited: 7 Apr 2012, 7:06 p.m.

Now that the relevant information has been given and the mystical "memory effect" got demystified, let me add a few thoughts.

First, I do not see any advantage in using old NiCd batteries in the Spice series. I do not know about the situation in the US, but at least over here in Europe NiCds have been banned many years ago for environmental reasons. This means, except some special batteries for even more special applications, you are not able to get new, fresh ones. NiCd has been replaced by NiMH. And that's exactly what I use in my 34C.

I use two Sanyo Eneloop batteries instead of the original NiCd pack. They fit nicely into the battery compartment, the two terminals on the right hand side are conneted with a thin metal strip (I also tried some folded aluminium foil - works as well). While these two batteries are in use, two other fully charged ones sit in a drawer. Since Eneloops show virtually no self-discharge, they may be stored for months or even more than a year without major capacity loss. They offer about four times (!) the operation time of the original NiCds, so they only see the charger once in a while. Of course they are charged outside the calculator, in a dedicated and processor-controlled charger that operates reasonably fast and, most important, prevents overcharge.

And this is what will ruin your batteries. Overcharge is one of the worst things you can do to your them. So, if you really want to use the HP-charger with the batteries inside the calculator you should at least check the charging current. The right time can then be calculated: 1,2 * current / capacity. Of course this is the time for a fully depleted battery - partially used ones need less charging time. But how do you want to know how much capacity is left inside? That's another reason for a processor-controlled charger that stops at exactly 100% charge.

Does anyone have any data regaring the Spices' charging current? A simple multimeter will do. :-)


Absolutely true. I have also replaced my NiCd batteries with NiMH cells and they work very well. While I do have several high end micro processor controlled chargers I generally don't bother to charge out of the calculator (I don't use Eneloop cells however for this purpose).

Note that, as Dieter stated, overcharging is a battery killer. But in particular it is (relatively) rapid overcharge leading to heat buildup and venting. Some high end chargers will even leave batteries on a very low trickle charge after fast charge has completed with little detrimental effect. Trickle charge for a NiMH is approximately 0.05C. NiMH are one of the chemistries *LESS* tolerant of trickle charging--less tolerant than NiCd cells at any rate.



Spices duplicate the charging circuit (aka el destructo) of the Woodstocks: a 10 vac supply, half wave rectifier, 8.2 ohm resistor in series. It starts out at ~200ma at full discharge, goes down to maybe 20 ma at full charge. When the calculator power switch is on, the resistor is shorted, increasing the current to the battery. I guess this was to offset the operating current. The behavior to note here is that you cannot easily calculate charge times for NiMH as the circuit does not provide constant current.

Overall, it is a horrible design, even calling it a design is giving it more credit than it deserves. It is bad for NiCds, even worse for NiMH. Personally, I would never charge a NiMH pack using an HP charger, it will die a very quick death.

Hi all, Randy.

Just to make things a little more, say, 'spicy', let's get back to the time these calculators were designed and, surely, their chargers. What else could designers add that was available/known at that time? I do completely agree with the fact that those chargers are really bad compared to what we have today, but what about 'back then'? I really remember reading about charging batteries and charging cycles and all of them demanded human supervision. Automatic control circuits in portable, relatively low-budget equipment (compared to other HP devices at that time, Woodstocks were low-budget) were simply not available. In fact, whatever mixed analog to digital at that time were very expensive.

Today we have all sorts of microcontrolled chargers, new chemical technology, battery models... Compare these 70's techs to today's is not that fair.

Flames? Blames?


Luiz (Brazil)

Edited: 8 Apr 2012, 2:27 p.m.

Thanks for the link to the article Marwan, interesting reading.



Luiz wrote:

Compare these 70's techs to today's is not that fair.

Well, it's the late 1970s, not 1870s. ;-) A decade earlier man made it to the moon. It's not rocket science we are talkting about.

Even a low-tech design could have been better than what was actually built into the Spices according to Randy. A simple constant current design like the one in the '41's battery pack 82120A, maybe slightly modified to a lower charging voltage so that the current automatically decreases as the battery is fully charged - down to a level that will not hurt the cells if applied for a longer time, such as 20 mA (= 0,04 C). Or, the current could have been set to 0,1 C for a regular 12-hour-charge, and a battery type might have been chosen that easily tolerates such a rate even for twice or three times that (e.g. special fast-charge NiCds), so that overcharge will not harm.

Even if a simple 78xx voltage regulator (as used in the 82120A) is considered high tech there are ways to control the charging current: use a simple light bulb instead of a resistor. If you ever used a Metz 45-series flashgun and its NiCd-pack you will know how well this works. Yes, Metz eventually switched to NiMH-packs and a controlled charger, but it took them 20 years since the introduction of the first 45CT-1 in 1976. ;-)

Using the charger as a power supply for the calculator should not have been a problem either: If the power switch is on, simply cut off the battery and the charging circuitry. If I understand Randy's description right, this may also be the key to using the available hardware without too much risk for the batteries: simply use the charger only for charging, and turn the calculator off while you do so. An initial charging current of 200 mA will do no harm to a current 2000 mA NiMH cell, and 20 mA is way below its trickle charge rate (50...80 mA). So the major problem is the time required to give such a battery a full charge. ;-)


Hallo, Dieter.

First of all, thank you for pointing out so many technological aspects and facts, but I ask you to forgive me when I say that I suspect we have a bit drift on the main subject .

I agree with all discussions about technology, mostly the constant-current supply circuit (a single transistor with a couple resistors would suffice, I know, but then it would be another faulty point), but considering that the HP21 was introduced in 1975 and some of its technology was borrowed from previous HP designs (whichever) while designers had their attention called to the new firmware (ACT concept, for instance), I guess that the battery charging circuit was not necessarily that important at the moment, so the choice of a resistor + diode may have been fair enough. I know that the Classics offered a constant current circuit for charging the batteries with a separate power line for the circuits, but their charger was somehow more complex, with transistors and two power supply lines, also much more expensive (I guess this complexity led designers to try the Woodstocks simplicity). Spices have a better charging circuit and were introduced three years later.

I also agree with you when you mention the HP41 rechargeable pack having a better charging circuit, but remember that it was an optional and that the HP41 (introduced in 1979) offers no way to charge batteries inside of it. Also, I am not sure about when the 78Lxx voltage regulators were available, but would the 100mA 78Lxx be capable of holding the power needs for the Woodstock while charging and being used? And how much would this enhancement add to final cost? The HP41 rechargeable battery pack was somehow expensive.

You see, I agree with all of what you wrote, hands down, but my point was only comparing available technology at that time the Woodstocks were introduced with the existing one when they were designed, given design priorities, not the technology that arose and was used in later models.


Luiz (Brazil)

Edited: 8 Apr 2012, 5:08 p.m.

Back in the early 70's I remember those flashlights that you would plug into the wall. The idea was that they would always be ready for use. The darn things never lasted very long in my hands--batteries tended to get destroyed inside of 6 months.

Off topic but this thread brought them to mind.



Makes sense to me...

As it happens in most scientific fields, not all factory tests neither all specs predict final product full spread of operation. Now we know much more about NiCad's than their own manufacturer/suppliers knew back then.


Luiz (Brazil)

Edited: 8 Apr 2012, 9:27 p.m.

Actually, the half & half approach is a safe method. Isidor Buchman has an excellent battery theory book called "Batteries in a Portable World". In it he says, "High temperature during charge and standby kills batteries. Overcharging occurs when the charger keeps the battery at a temperature that is warm to touch (body temperature) while in ready condition .... A temperature peak is reached when the battery approaches full charge". So multiple, short charge cycles that don't allow the temperature to rise much are safe for the battery, and there's no real danger of overcharging in this scenario. He also states "Any prolonged trickle charging will damage the battery". Hope this helps.

Sounds like an interesting book. Can you confirm from your reading that this is the case for *all* chargers? It sounds reasonable for smart chargers that use voltage drop detection (dV/dt) to terminate charging. However, for a charger such as those supplied with HP calculators wouldn't one risk pushing the battery into overcharge and thereby heating and venting even with a relatively short charge if the battery is nearly full when you start charging?

I know that at least a couple of my numerous chargers also detect temperature rise. If the cells begin to heat up charge is terminated. I am not sure how often (if ever) this kicks in. One of my chargers I believe actually will report charge termination due to overheating the others simply stop charging and report charge complete.



The NiCd/NiMH chargers I know have temperature sensors only for safety reasons - if the battery gets too hot, they simply stop charging. However, the usual delta-V method that detects a voltage drop (dU/dt gets negative) relies on on a temperature rise as well: as soon as the cell is fully charged, the charging current no longer can be transformed into battery capacity (the terminal voltage no longer rises), so the battery simply gets warmer. This in turn causes a voltage drop (the higher the temperature, the lower the battery voltage and vice versa). The decreasing voltage then is the cut-off signal for the charging circuit.

However, this applies to charging currents significantly higher than 0,1 C. Simply because the battery will not warm up sufficiently when charged with lower currents. So if a 500 mAh battery gets hot during a 50 mA charge either the battery or the charger have a problem. ;-)


Yes, exactly. But it sounds like the HP chargers were charging at above .1C. Certainly with the lower capacity batteries of the time. Or am I mistaken? I know that my calculators get warm during charging although I have not specifically verified that the cells themselves are warm.



I'm no battery expert, so I won't speak to ALL chargers, but from the little I do know, battery chemistry (NiCad, NiMH, LiON, etc) dictates the actual charging current curve. NiCads are tough, and can be charged quickly at high current with no damage. Not so for the others. LiON is particularly sensitive to temperature rise, and any approach to quick-charging them has to be super-tightly monitored/controlled, since temp can get pretty high at the end of the cycle. Maha Energy used to market a couple of cross-chemistry smart chargers. I have the last one they made (MH-C77Plus-II), which disappeared off the market after consumer complaints that they were ruining batteries. The unit had two settings: "Ni", which charged at 800ma, and "Li", which charged at 400ma. I always "baby-sit" a battery pack I'm charging, and almost always use the 400ma charge rate. I check the battery every 15 minutes, and pull it off to cool if it feels warm. Best I can tell from the book, you'll never overcharge a battery if it never gets warm. That's been my approach to all my charging - even AA cells for my digital camera. I never let them get hot. Just my two cents worth, in case it's helpful to anyone.

True. Due to my other interest (well one of several others) batteries and battery chargers are high on my list of interests. That other interest is RC aircraft and specifically *electric* RC flight. While you are mostly correct on the capability of Lithium based chemistries to take a fast charge modern LiPo cells used in flight packs have come a long way in the past 2-3 years. So have the chargers. There are several batteries on the market that can take up to a 5C charge and some very sophisticated chargers that can push that number even higher safely. LiPo batteries should *NEVER* get hot (EDIT: or even warm) when charging and I have seen others charge their cells at higher than 5C rate with no heat buildup at all. I, personally, charge even 5C rated cells at no higher than 2C rate and often don't even exceed 1C. Mind you, with a 4,000 mAh pack you have to have one heck of a charger to handle 5C. I have one charger that can pull that off but I certainly won't try it off my car battery!



Edited: 10 Apr 2012, 3:42 p.m.

The right time can then be calculated: 1,2 * current / capacity. Of course this is the time for a fully depleted battery - partially used ones need less charging time.

At 0.1*C, charging efficiency is around 60% to 70% for NiMH and NiCd, this results in approx. 70% - 80% charge.

However, this is exactly what I do. I currently use 1800mAH NiMH batteries. I use the battery until is shows signs of discharge (erratic behaviour, dimmed display etc.) - never until it is completely dead. I use a very simple constant current circuit (1 transistor, 2 diodes, 2 resistors) in the order of 0.1*Capacity for 12 hours (on a one-shot timer! lest I forget to switch it off). Thus the cells should not go down to 0V (or worse, an under-performing cell in a pack going reverse voltage), and reducing the risk of over-charge (although it doesn't immediately destroy the battery, can reduce its life).