Topic: HP-35/45/55 etc. Charger Circuit

Hi Jacques,

I have been experimenting with the charger circuit. I found that when the charger is coupled to a calculator, the output voltage drops considerably, to around 3.5V. When the charger logic supply is connected to a multimeter, I get more like 4.5V. These numbers come from a couple of chargers I repaired by replacing the Zener diode with a 4.7V 1W Zener.

I had been assuming that the output voltage would be fixed at the Vz less the .6V drop across the NPN darlington. However it appears the output voltage is not fixed at all.

To confirm I created a PSpice model. I found a model for the MJE800 Darlington at the On Semi site. Added a 470 Ohm resistor, the Zener, and a constant voltage source at 15V. When I simulate a load across the output voltage, the voltage varies considerably depending on the load. It routinely goes down to 3.5V with a load resulting in a current of about 150mA, which is about what we would see from the calculator.

Has anyone else done this? If this is correct, the 4.7V Zener should probably be more like 5.1V, since a true output voltage of 3.5V is just barely above the low voltage indicator of the calculator, which is at 3.4V, and the calculator stops working about 3.3V. Most chargers I see have output voltages close to 5.0 volts (when connected to a multimeter which has a very high impedance). I had always thought this was too high. Perhaps not. Comments?

Re: HP-35/45/55 etc. Charger Circuit

Hi Dan,

I think the Zener in this 'transistor series regulator' should be 5.1 or something (or 5.6 maybe).
You'll have to redesign the circuit since the HP 1854-0581 that should be hard to find is an equivalence not a clone!
Particularly, VBE could be a bit different.

To calculate the circuit:

VL (load voltage) = VZ - VBE = 5.1 - 0.7 = 4.4V
VS is the unregulated voltage from the transformer ; I measured 16V
VCE = VS  -  VL = around 12V

current thru the 470 ohms IR= (VS-VZ)/470 = 23ma

current thru the base IB = IL/beta (load resistor calculated from the measured load current IL = VL/RL) = 200ua
taking beta from the specs at 750 (but it depends on IC).

current thru the zener IZ = IR  - IB around 23ma, 120 mw (Vishay Zener 1N4733 should do it).

I'd try with a 5.1 or a 5.6 Zener.

Keep me posted on this issue.
I will order the 3 replacement BJT to reconstruct the whole pack.

Cheers

Re: HP-35/45/55 etc. Charger Circuit

Hi Dan,

I have done measures on a charger 82002A.

1) voltage regulator

The voltage at the collector of Q3 (HP 1854-0581) is 16V unloaded and 11.5 V loaded (calculator on).
Due to the drop in the 470 ohms and action of the zener, VB of Q3 is 5.3.

But the drop in the Darlington pair is more than I thought (in fact it adds) ; I measure VBE = 1.15

VL (load voltage) = VZ - VBE
I measure 4.15 loaded and 5.85 unloaded.

That gives : 4.15 = 5.3 - 1.15 (5.3 for the zener)

VCE = 11.5 - 4.15 = 7.35

2) Constant current source
Unloaded voltage = 16v
Loaded = 4.15
Current 50.9 ma

Cheers.

Re: HP-35/45/55 etc. Charger Circuit

Hi Jacques,

Yes I figured the drop across the Darlington pair would be probably around 2x the drop across one NPN. This particular device has a couple of resistors and a diode, in addition to the pair of NPN's.

One possibility is that for the Zener to have a voltage drop of the Zener voltage, it needs a minimum current, Iz. For most 500mW Zeners, that current is about 20 to 25 mA. For 1W Zeners, it can be much higher, like 200 mA. The current you and I calculate is 21 to 23 mA. It could be right on the edge thus not regulating all that well.

I might remove the Zener on a working PSU and make up a test circuit to see what the output really is. I am not all that confident in the PSpice model - which uses a regular diode reverse biased with a breakdown voltage equal to the Zener voltage. Nevertheless the PSpice model seems to be about right. I simulated a load using a resistor of about 15 to 20 Ohms to generate a current of 150 mA or so.

Regardless, I have had success replacing the Zener with a 1W 5.1V Zener and I'll probably order some 500mW devices to see how they work.

Also, I have discovered that it is possible to replace the entire Logic supply with an active regulator like the LM317T. It can be done without changing the circuit board. In fact I built one charger with 2 LM317T devices, one to supply constant current and the other for constant voltage. It is much better regulated.

-- Dan

Re: HP-35/45/55 etc. Charger Circuit

Hi Dan,
Good job! wink

The total BE of the Darlington pair is double that of a convention NPN and, you're right there are internal base-emitter resistors (but not always).
In the MJE800 yes, but in the HP 1854-0581,  I don't know.
If there are no BE resistors the current gain is maximized, but it lengthens switching time and increases thermal variations in leakage current.

I will do my reconstruction with the MJE800 so we can compare.

Your use of the LM317 is very elegant. But look at the internal schematic diagram of the 317, it's much much complicated!
Another epoch!

A temptation would be to boost the current source above 50ma to charge modern big NiMH.
I tried it last year on a HP-25.
But NiMH are less tolerant than NiCads to trickle charge rates and repeatedly recharging a full battery will damage it and reduce its life time.
Safe trickle charging rate should be around C/250.

I'll look in National catalog for a circuit handling the trickle rate.

Cheers
Jacques