The database errors are preventing me from maintaining the original version of this post, which is one of the ones I refer to myself most often. My apologies to readers who may be sick of seeing it.
Warning: if the mains input is connected to actual mains, the "GND" mentioned here is at lethal potential to the ground pin of the mains, and your body. In other words:
GND is HOT!
I believe that these jumpers are only intended for use with a current limited power supply of about 52 V.
I've finally sussed out the use of the three main jumpers on the Elcon/TC charger daughter board (the one with most of the chips on it, including the processor and small red LED).
J8: short to disable the PFC stage. This is a good thing when debugging with a 52 V current limited power supply, because the MOSFETs will switch at 50 V (~2 V drop across the diodes) rather than 385 V.
J7: Short to force 240 V mains detection. Without this jumper, the mains sensing circuitry will decide that your 52 V power supply is too low, and will disable both the PFC stage, and the PWM stage. So the MOSFETs won't switch at all.
J3: Without a battery detected, the microcontroller won't enable any switching. By inserting a jumper with a 1.8 kΩ resistor (see below), you will get a moderate duty cycle. This is ideal for testing. [ Edit: I used to recommend a 3.3 kΩ resistor for a very low duty cycle, but on some chargers, there isn't enough voltage to get the UC3846 to start generating pulses. ]
So the sensible combination of jumpers is as follows, in order:
1)
All jumpers out. 50 V across the main MOSFETs, but they are not switching. Good for finding shorted MOSFETs. Leave the power supply current limit at 0.5 A or less. You may need 52 V or more at the AC input to get the Viper chip to start so you can verify that power is reaching the control board.
2a) Optional.
Only J3 in. For the truly cautious, this will give the MOSFETs a short burst of switching, then immediately stop switching them (as it realizes that the mains is not present). You should be able to measure part of the voltage from the next step at the output, slowly decaying. It might be only a half or even a quarter, so a peak of 4-12% of maximum voltage, or 5-16% of nominal voltage.
2b)
All jumpers in. 50 V across the MOSFETs, which are now switching. You should see some 15% of maximum rated voltage (about 20% of nominal voltage) at the charger output (negative output terminal and PCB pad, see below). Power supply limit can stay at 0.5 A or less.
At this point, you should be confident that the MOSFETs are switching properly, because the energy in the bus capacitors is about to increase about 8² = 64 times. Use a DSO if there is any doubt. One thing to check is that there is some dead time for both half cycles; test point T34 is for this. Dead time is when this point is low. T34 is awkward to get to under U12, so just use pin 11 of U12 (middle pin). There should always be some time during each half cycles when this test point is low. I like to use both my J3 jumpers (1.8 kΩ and 3.3 kΩ), as well as no J3 jumper at all.
3) J7 in,
J8 out, J3 out. Now there should be ~ 385 V on the DC bus (the MOSFET power supply), but the MOSFETs are not switching yet. The power supply current limit needs to be at least 2 A, preferably 2.5 A, to get started. It may take ~10 seconds to get close to maximum bus voltage, at which point the current should fall to around half an amp (it jumps around a lot on my power supply, which is two 26 V supplies in series).
4) J7 in, J8 out,
J3 in. Now there should be ~385 V on the DC bus, and the MOSFETs should be switching. The power supply limit needs to be at least 2 A, possibly 3 A. Now you should read about 110% of the maximum rated voltage (about 150% of nominal voltage) at the charger output. This could exceed the voltage rating of the output capacitors; if so, don't leave it running like this for very long. I prefer to start with a 2 A current limit, even though it takes over 10 seconds to reach maximum bus voltage. That gives me a chance to switch off if things don't seem right, and there is a slightly better chance I won't damage too much.
If it passes all this, it's time to reassemble the charger and test with a real battery and mains power.
Here is my collection of jumpers:
The jumpers appear to be 2.5 mm spacing, but I used the more commonly available 0.1" header pins (2.54 mm spacing). The slight mismatch makes them stay put without falling out. Note: there is black junk over all of the jumpers, in fact over 95% of the PCB, so you need to clean the area around the jumpers. Also, the holes fill up, which is a royal pain. I use a paper clip to push through the holes. You may need to clean the back of the board where the jumpers come through as well. A wooden chopstick, flat at one end and sharpened at the other, is useful for this. I sharpen the pointy end with a pencil sharpener, and the flat end with a small file. (Thanks for the idea, KennyBobby.)
The two jumpers at the left are shorted; the heatshrink is to keep them together and to make them easier and safer to handle. These are for J7 and/or J8. The jumper with the 1k8 resistor is for J3 only.
Here are the locations of the jumpers, and some close-ups:
The power connector on the left of the control board is useful for connecting to ground with a multimeter negative lead or DSO ground lead (though you get tons of glitches when the MOSFETs are firing). I had a plug already made up, but only plugged it into the top pin, so there was no danger of shorting the 15 V power supply. For temporary multimeter negative leads I often use the ground via circled in orange. For +15 V, the top of L1 (in the top left hand corner of the PCB) is handy.
At the output of the driver chips (U15 and U16), you should see ~ 12 V p-p on the low outputs (pin 1), and around 60 V p-p on the high outputs (pin 8). The latter is because you are adding the ~ 48 V from the MOSFETs switching (50 V from the power supply less some diode drops), plus the ~ 12 V from the boost power supply (pin 7, this should be a square wave with the low end around 12 V to around 60 V at the top end, about 12 V higher than the MOSFET outputs).
When all is fixed, you should see some DC output, but not necessarily at the actual positive output terminal. This is because the micro doesn't see a battery, and so won't connect the output relay. There is a resistor across the relay, so you should see something at the positive output terminal. But there is a spare relay position (only populated for very low voltage chargers whose output exceeds 20 A), where a multimeter positive lead can be conveniently placed:
The negative lead can be placed on the negative output terminal (not interrupted by the relay being open), or the negative output lead if it's still connected.
In my case, I was working on a 288 V nominal unit with a 13:7:8 transformer ratio (many of the transformers seem to have their ratios written on them, particularly the 2 kW units). The :7 and :8 parts add; only the higher two voltage units have this arrangement. Treat it as a 13:15 ratio transformer. Lower voltage chargers will have rations like 13:9 or 25:8. In my case, I expect roughly 15/13 x 50 V = 58 V; I was seeing a little over 60 V.
[ Edit: added sentences re test point T34 and dead time. ]
[ Edit: 2 A is a good value for the final jumper test. ]
