I have some old Texas, HP and Novus calculators (with red LED displays) exactly like that and I have done these NiCd to NiMH conversions to work perfectly fine.
These are the points of attention IMO:
Q: Would that be a case where NiCd chemistry would behave differently from NiMH?
A: No problem for the device itself - pocket calculator being powered by 2 rechargeable batteries, as either NiCd and NiMH have nominal voltage of 1.2 V.
Charging NiMH is different story, but quite feasible to address.
Spoiler: To use with the original charger, your circuit should be adapted to reduce the trickle current and increase the charging time.
Slow charging and Trickle charging:
Charging processes of NiCd or NiMH are made controlling the charging current, which categorize if the charging is considered fast, slow or trickle.
Fast chargers need to monitor battery temperature or the rate of voltage change along the time (dV/dT). These features are quite complex and should be done with dedicated circuits/products.
Let’s talk briefly about slow and trickle charging.
NiCd batteries support slow charging as 10% (more usual) to 20% (rare and aggressive) of its rated current capacity. Most old-times AA NiCd batteries were ~400 mAh, said as C = 400 mAh. Typical slow charging of NiMH is C/10; in the above case, designed to be 40 mA average, or ~80 mA thru a half-wave rectifier (1 diode as shown).
Then, being charged as pulsed 80mA (half wave) for 14 hours, would charge a NiCd completely.
Assuming your wall-charger is a conventional transformer that delivers 10VAC with negligible voltage drop (worst case), for an average charging current of 80mA, resistor R4 should be:
i_charge = 0.08 A = 50% [(10-0.7-2.8)/R4] ;
R4_NiCd = 40.6Ω
Probable cause of calculator’s problem:
As R4 value in your circuit (8.2R) is too low, the peak current might be ~ 400mA with the resistor in series, and even higher if the bypass switch connects directly diode and battery.
Even assuming both Eneloops are good, during a higher-current charging they will have higher termination voltages added to the higher ripple voltages.
And these overcurrent spikes might be the major reason why the calculator’s voltage & operation is erratic, as these calculators’ circuits expect to operate at 2.2 V to 2.8V.
R4 should be much higher (see ahead) and never use that bypass switch, especially with the calculator connected.
Charging and Overcharging:
Near the fully charged state, Ni-based batteries will warm-up as the most perceptible symptom, as internal pressure and dV/dT would require additional equipment for their identification.
In general when batteries become warm, the state of charge is above 70%, soon will reach 100% and then will be overcharging - and probably being damaged.
The “solution” becomes to do trickle charging to limit the overcharging and allow the recombination chemistry to do its work.
NiCd accepts overcharging up to C/10.
NiMH doesn’t accept well the same overcharging - some manufacturers indicate that can withstand C/30 up to C/20.
NiMH-mod I have done:
Trickle charging at C/30 is the method I have used to charge these *NiMH-modified calculators.
As the internal resistance of NiMH is much smaller than R4, the voltage ripple is negligible and the calculator works as originally intended.
Replacing old NiCd by 2000 mAh NiMH:
Your calculator’s NiMH will be 2000 mAh, so if C/30 is assumed for trickle charging:
I_trickle = 2000/30 = 67mA.
R4 = (10 - 0.7 - 2.8)/(0.067/2) = 194R.
The “2” is due to the half-wave rectification. Divide by “1” in case the circuit is modified to be full-wave rectified.
Commercial value: R4_NiMH = 180R.
And remove that bypass switch.
Charging time and final comments:
As the capacity of your battery is ~500% of original, once fully charged you will have a long working time, with negligible self-discharge too.
Just remember to allow the slow & trickle charging do its job, as it might take almost 2 days to fully charge (140% x 30h = 42h) the exhausted NiMHs.
If you Recharge partially-discharged NiMH batteries, charging time will be less and NiMHs also do not have “memory effect”.
Finally, in case you feel the battery-pack warmer, you will have reached +70% of a full charge and more than the originally designed capacity. In this case, if it was my calculator, I would terminate the charge, but would not be overly concerned if it gets warm for some extra hours, as the charging is sized as C/30.
