Speaking of chargers (HP41 Nicad Pack)

[steve]

I'm rebuilding an HP41 nicad pack. The nicads were not particularly hard to get (if a little expensive), and the pack itself was not too hard to open.

However...

The circuit seems to allow the charger to power the calculator while it is plugged in (the charger that it).

To do this it has connections to either end of the battery, but also a seperate lead that powers the calculator.

This is where I need a little help. Corrosion has damaged the flexible PCB material that contains this seperate connection.

What I need is confirmation that it DOES NOT make contact with the battery.

I will probably trace out the circuit while I have the thing open. If anyone has already done this then I'd love to know, otherwise, is anyone interested in the circuit for the charger?

[Walt Devaney]

Since you have taken the pack apart, I assume you are talking about the connections to the battery stack, not the connector to the transformer. Some early packs did have a third pin on that connector. I have taken apart several of the battery packs (boards marked 82120-60001 and 82120-60005) and none have the third connection to the batteries you describe. In the ones I have looked at the charger board is connected directly to the battery stack using strips (brass? copper?)welded to the batteries and then there are spring clips from the battery ends that contact the calculator terminals. The batteries were always across the calculator inputs, as a precaution I assume. The charger circuit changed depending on the age of the pack, HP never did get it right. I have schematics for two different vintages of board and would be interested in trading for what you see in yours. Hope this helps.

[steve]

That's really interesting.

Mine appears to have a charging circuit that will deliver power to the calculator via a different circuit when the charger is connected than when it is absent.

The brief inkling I have is that it operates in 3 modes:

1) normal (no charger attached). The nicads provide power to the calculator

2) charging (charger attached and powered up). The charger charges the batteries and provides power to the calculator seperately.

3) failsafe (charger attached but not powered up). The batteries power the calculator via a diode. This may also allow the batteries to provide any high peak currents required (by the card reader for example) while charging that cannot be provided by the charger.

This provides be with more incentive to fully trace the circuit :-)

And by the way, I have found a local (Australian) source for the nicads, but they're not cheap... $A4.50 (about $US3) each. What's the price in the rest of the world?

The nicads appear to be 1/3 AAA. Well, 1/3 AAA is a little smaller than the ones in my pack, but the 1/3 AA are a little too large (albeit at 110 mAh, the 1/3 AA would be worth trying to squeeze in in place of the 50 mAh 1/3 AAA)

[steve]

I have traced out the charger circuit diagram. It seems to contain a small three terminal regulator.

A very small BMP of the circuit diagram is here...

http://www.wantree.com.au/~shodges/HP41-CHG.BMP

C1 is 100 uF 25V C2 is a small (Tantalum?) capacitor with no markings and a ferrite bead on the leg nearest the output of the regulator. R1 is a 580 ohm resistor

The switch is open when the charger is plugged in.

Interesting notes:

The unloaded voltage across the main filter capacitor (C1) is approximatly 15.5V, so allowing for 4.8V across the nicads, the charge current is 17 mA which gives an approx 3 hour charge, and is WAY over the recommended charge current (for safe (?) overcharging) of about 5 mA.

Refer to the following site for information about the nicads (this is a Canadian site) the N-50AAA are the one's I'm getting.

http://209.167.73.164/indust/batt/cadm.htm

This site gives some info on recommended charge currents

(5 mA for 14 hours or 50 mA for 1.5 hours)

And an article on nicad charging. With an indication that the rate that the HP 41 charger charges at is too high for continuous charge.

http://www.greenapple.com/~tailspinners/nicads.htm

My estimation that a 1.8 K resistor would be better in this circuit (to give a charge current of 5.5 mA)

I'd appreciate it if you could email me the circuit diagrams you have.

BTW, the battery pack I have is:

SN: 2132S60515 (mid 1981 manufacture) made in Singapore

[Walt Devaney]

Steve - you appear to have a very early 82120-60001 charger board. After your description I went and looked at the 60001 I have and it does indeed have a place on the connector for the switch you describe. However the switch has been eliminated and the calculator is always connected directly to the battery terminals. Anyways, the unit you describe appears to be an attempt to build a genuine battery eliminator that HP quickly abandoned. HP claimed later that the 82120 was a charger and if you left it plugged in you would cook the batteries. This is still true with the arrangement you describe, unless you reduce the charging current to inconveniently low levels. Real NiCad chargers switch off. I will send you the circuit diagram for the later 82120-60005 units. It will take a few days. You may want to modify yours and eliminate the extra circuitry.

[steve]

Yep, the underside of the board reads 82120-60001 B-1951

It might be practical to fit a switch to it (externally) to enable either the current moderately fast charge, or a trickle charge.

I have no problems leaving the charger on overnight, and it beats replacing the nicads :-)

[Barrie Fletcher]

Steve,

The circuit of mine was as follows: The transformer conects to 4 diodes arranged as a bridge rectifier, followed by a 100uF capacitor, then a 78L06 voltage regulator. A 10 ohm resistor, a diode and a switch were in series, bypassing the regulator when the switch was closed.

The board you referred to has three pins - the two outer ones connect to the charger, and the centre one is the switch contact connected to the battery positive. The other switch contact is connected to the diode's cathode.

Perhaps the battery pack was designed to work with more than one type of charger.

Barrie

[steve]

That's interesting, but I can't quite figure out how it's all connected. It seems to suggest a charge current of 60 mA, which is way too high (so I've obviously got something wrong)

Or is the regulated output delivered directly across the nicads? (Horror!)

Can you sketch out the diagram as a bitmap or something and email it to me?

My unit also has a 78L06A but the output appears to be only a little above 4.1v. I'm not sure if it's related to the dead nicads or if the regulator is fried.

I'm going to replace the nicads with a 1K resistor to provide a suitable load and measure the output again.

The unknown capacitor is marked K1M 103, which seems to suggest 0.01 uF

Also interesting is that two if the diodes in the bridge rectifier in mine are different from the others, and appear to have a much higher forward voltage drop. (Interesting but not useful)

OH! I just discovered a fault in the design of this charger. If you partially insert the charger, then the regulated output is connected directly to the batteries. This could result in a rather rapid charge. I'm not sure if the 7806 is current limited (although in common with the 78xx series regulators it probably has a thermal shutdown).

Hmmm some information about the 78L06 and its application in this circuit.

Input voltage 15v, output 6v (so 9 volts dropped across regulator.

Max junction temp 125C (thermal shutdown)

Junction to ambient thermal resistance for TO92 package 230C/W

Assuming 25C ambient then the device can dissipate about 0.43 of a watt continuously without going into shutdown. (and practically a whole lot less because the temperature in the charger will not remain at 25C for long.

That means that it can deliver 47mA. And I'd derate that by 50%. So the charger can deliver about 25mA.

That's probably OK if you're not pulling cards through the card reader continuously :-)

Curiouser and curiouser. I get an 9v output! (and that's with an 9mA load)

Aha! Without the nicads the charger circuit raises the output voltage. Nasty. The regulator was actually putting out 6.3 volts.

This means that this model nicad pack could deliver voltages to the calculator that are dangerously (?) high if the nicads deteriorate to the point that one goes open circuit (or high resistance). So it may not be a good idea to use this model as a battery eliminator if it contains dead nicads.

Yes, with the (dead) batteries replaced the output voltage is about 8.7v I'm glad I didn't use this in my HP41 :-)

I'm not sure where the reading of 4.1v I measured earlier came from. It may have something to do with the regulator being open circuit???

Gee, this is long winded isn't it? Back to real work...

[Barrie Fletcher]

Steve,

The circuit you mentioned at http://www.wantree.com.au/~shodges/HP41-CHG.BMP is the same as the original circuit of my battery pack. The batteries in mine failed in the mid eighties, and were discarded after they corroded one of the calculator's battery contacts.

Tinned hookup wire bridges the gap left by the corrosion. I now use Sanyo N150N NiCads (150mAh) in the battery holder which was supplied with the HP41, and charge them in an Arlec NiCad charger.

The rechargable battery pack is now used as a battery eliminator, with R1 removed to protect the HP41.

The battey pack's circuit can be broken into three sections:

(1) The bridge rectifier and electrolytic capacitor, which provide 13 to 15V to the other sections. The bridge contains two types of diode. The diodes connected to the negative terminal are grey with a black band. The others are black with a white band indicating the cathode.

(2) The charging circuit, consisting of R1, two diodes and the switch. The switch is to allow full battery voltage to the HP41 when no charger is connected. The diode across the switch is to ensure that the regulator's output does not contribute to battery charging. The diode connected to R1 is to prevent C1's being charged by the battery when the charger is switched off.

(3) The regulator circuit consists of the 78L06, C2 and the diode between C2 and the positive terminal. The 78L06 is rated to provide 100mA, but the TO92 case is only rated at 0.75W, giving 83mA maximum load at 15V input. Its output voltage should be between 5.75V and 6.25V. C2 is marked as 0.01uF in mine. The diode is probably to isolate the 78L06 from the battery voltage during normal operation, but it could also indicate that more than 6V may be present during charging, as you have found. This section's purpose appears to be to provide power for the calculator while the batteries are being charged, but the voltage across them is still too low to operate the calculator.

If you wish to rebuild the battery pack, you may wish to add a zener diode between R1 and negative, to limit the voltage that may be presented to the calculator if one of the batteries goes open circuit. The current drawn by this diode will reduce the charging current, but you may also wish to increase the value of R1.

I shall be using the space freed by the batteries to install a regulator rated at 1A and a larger capacitor in parallel with C1, so I can run the card reader from the battery eliminator. I believe the card reader draws 200 to 250mA.

[steve]

most of the information in points 1, 2, and 3 are fairly obvious once you have the circuit diagram.

TO-92 cases are not rated for a particular power however. The 78L series regulators will supply up to approx 100mA until the temperature of the chip reaches approx 125C at which point they go into thermal shutdown. My previous note suggests at what continuous current this may occur (but only after some time).

Last night I placed a 6.8V zener between the resistor and ground to limit the voltage to about 6.2 volts. I disconnected one of the nicad leads (+ve) so as not to stress these poor dead nicads further. About 15mA flows through the resistor/zener combunation leading to a continuous total dissapation of 250mW which is well within their ratings.

Contrary to what you say, this will result in NO decrease in charging current with nicads attached as the nicad voltage will be below 6.2 volts (the zener knee voltage less the forward biassed diode's voltage drop). The zener will only begin to conduct when the battery is well overcharged, or goes open circuit.

I also found that it wouldn't power the card reader (no surprise there)

The diode in series with the output of the 78L06 is to protect the device from power flowing back into it when the charger is not powered up. These regulators have a problem with having a voltage present on the output terminal which is higher than that at the input terminal. In many circuits you'll see a diode connected in parallel to the regulator to effectively shunt current around the regulator in this case, but this is only a protective measure for power supplies that have relatively large output capacitors. In the case of batteries this would cause a continuous current drain which would be unacceptable.

I too was thinking about making the unit into a battery eliminator, but I considered placing some very small nicads in the pack to provide power for memory backup. At the moment I just replace the pack with the normal batteries when I'm finished.

A question exists still as to why the diodes in the bridge are not matched. The best guess I have at the moment is that 2 of them may be zener diodes to act as a form of spike/surge protection (and this was prompted by someone who said they "blew" in such a situation)