Hi All,

I have a Classic charger (no voltage switch) that doesn't supply the higher charging voltage. With information I gleened from looking back through the archives here, I checked about the fuse, and mine has none, so I am going to replace the 400uF cap. Now, my local electronics shop can supply a 330uF or a 470uF cap, but not a 400uF. Can anyone tell me how critical the value is ?, and which value of the two mentioned I should use as a substitute. - Thanks, John Robinson

have you checked the cable? sometimes the wires near the plug on the calculator end become diconnected from the metal connectors.

**vp

Yep, checked the cable, I have continuity all the way from the plug to the electronics inside the charger. - John

Go for the 470ìF -- these things have a -20+50% tolerance anyway. Make sure you connect it with the proper polarity, and that you get a capacitor rated for the same or higher voltage.

/ji

To answer the last question first, either will work, but I'd use the 470uF. It's always best to go slightly higher on smoothing caps.
However, I am not convinced this is the problem. The transformer, rectifier, and smoothing cap are common to both outputs, so a problem with any of those normally causes problems with both outputs.
I'd look at the constant current source circuitry. If yo look at the PCB on the 'output' side (near the cable), then on one side there's a single transistor with a 'heatsink' (a piece of metal rod screwed to the transistor), on the other side there are 2 transistors. Check the latter 2 transsitors, and the resistors near them.

It could still be the filter cap... but probably not. The low current charge circuit noes not even need the filter. The high current operting supply does. I have seen a couple of chargers that charge just fine but to not operate the calculator. Both were bad filter caps. But usually a bad filter causes a highly erratic flashing in the display.

The original fault was a problem with the high voltage charger output. Which is the constant current supply to charge the NiCds. I read that implying that the lower (constant) voltage logic supply was correct and that the constant current
output was missing. Now, the constant voltage output comes from a zener+emitter follower circuit run off the smoothed DC from the 400uF capacitor. It is impossible for this
output to be higher than the voltage across that cap. And thus, if the fault is the filter cap and we're seeing something on the constant voltage output, we'll also see something on the constant current output.
My guess is the capacitor is fine, and the fault is in the constant current regulator (2 transistors and a few resistors).

Thanks for all the postings guys. Well, I tried a new filter cap (the 470uF), and it hasn't changed a thing. So does anyone have any suggestions on how to fix the constant current source electronics ?, anyone got a circuit diagram ?, are replacement/equivalent parts available ? Should I just perhaps just remove the lot and replace it with a modern day constant current source (chip ? if they exist). Any suggestions welcome. Thanks - John

I have an ascii-art schematic, which I have (hopefully) privately e-mailed to you. I can't think of a sane way to include it in this post.
The constant current circuit is 4 components. 2 resistors and 2 transistors. The idea is that when the load current increases, the voltage drop across the 13R (Ohm) resistor increases, thus turning the small-signal
transistor on harder. This reduces the bias on the power transistor (normally applied by the 4k7 resistor to ground), thus reducing the load current again.
It's not going to be hard to debug. Check the transistors (remove them, and at least check them as a pair of diodes with a multimeter). Most likely the power transistor is open-circuit on at least one junction. I don't think these
transistors have exact modern equivalents, but they're not exactly critcal, and any PNP small power transistor would work. Similarly, for the small-signal transistor, anything will probably do. I'd try a 2N3906 first, as I've got those by the dozen. But I haven't _actually_ tried it.