HP-35 Charger/Adapter Repair (82002C)
Message #1 Posted by LHH on 19 May 2012, 1:21 a.m.
My new NiMH battery pack is sweet but now I have discovered I have the familiar "charges the battery but won't run the calculator" problem. The problem has actually existed for some time prior and probably started when I was running the 35 with dead and shorted battery packs. Anyway, it's time to get this fixed.
Once inside the charger the problem boiled down to the power transistor routed to the unmarked connector pin. Some measurements indicated it was still working but in a strange kind of way. The V(be) measured about 1.25 volts so at first I was thinking it might be a Darlington. But even with a 50mA load the voltage at the emitter was only about 3.2-3.3 volts ( Vz ~ 4.7V) which seemed suspicious. At that point I removed it from the PCB for further study and found it behaved much more like an SCR with a high trigger voltage. Once the base current was sufficient to raise the base voltage to about 1.2V the collector current ramped quickly from nearly 0 to full saturation and the base no longer controlled the collector current! If the base current was removed (or greatly reduced) the collector current would fall to nearly zero (unlike an SCR) again and the cycle could be repeated. So in a way the transistor is still functioning but just in an unusual and undesireable way, at least for this application.
So it appears this transistor needs to be replaced but the more important question is why did it fail? I have noticed that the connector switch in the calculator "probably" remains closed for a short period after the actual connector contacts have mated. If the battery pack was completely shorted by defective cells this could cause a momentary high current through this transistor. The other most likely factor could just be heat since the adapter case is sealed from any airflow. The transformer itself gets fairly warm and this transistor could be dissipating 1-2W under normal operating conditions so I've been wondering about adding some ventilation holes to the charger cover (good or bad idea from a "collectable" point of view?). I do want to give this some thought before I just replace the transistor and call it done. At the very least I may add some small resistance between the bulk supply and the collector of this transistor. I'm planning to measure the nominal and peak operating currents of the 35 to determine an appropriate value.
One other thought is to add a switch to the (functioning) battery charging circuit. The switch would select between the normal charging current of 55-60mA or a small trickle current (maybe 5-10mA) so I could keep the calculator plugged in most of the time, not overcharge the new NiMH batteries and still have a full charge at all times. Of course I'll then need to remember to set the switch correctly! An earlier though was to design a "smart" charger circuit but I've decided that's too much trouble to go to for the time being.
I did find a few threads about this in some archives but many seemed to relate this problem to something within the calculator itself. In this case I'm pretty sure the problem lies in the adapter.
Any thoughts or suggestions welcomed!
LHH
| 
Go back to the main exhibit hall