# Debugging a battery charger circuit

I have built a circuit of a battery charger here. Here is the circuit in the original project:

However, it doesn't seem to work. The output I measure at the battery charging clips is only 1.2 V or so while the LED2 is on. Now I don't want to bother people here with all my questions, but it often helps me to think out loud explaining what I do in order to find the problem.

I will also try to use the circuit sketch tool here to see if it can predict the voltages that I should measure at the various points in the circuit. But I see it can't simulate SCRs anyway, so that didn't work.

I do not have really good equipment. No oscilloscope. I have a cheap analog and a digital multimeter. The trouble starts with the transformer. It is rated 15 V and yet I only get 14.3 V to start with. But I don't remember exactly if multimeter readings from AC are always low? Plugged into the power outlet directly I get a reading of only 210 V. I remember in Germany as a young hobbyist I always measured higher, like 230 V AC and I even knew the reason for this. Like the difference between V ss (as they call it there, from apex to apex of the sine wave) vs. the power integral, something like that. Anyway, I don't remember measuring less.

So, let me suppose both my meters are bad. Both measure only 210 V. On a 5 V TTL circuit powered by a USB phone charger I measure 5.1 V, but that's DC. Maybe only the AC measurements are off? If I apply a correction factor of 4.8%, then I get 220 V and 15 V on the transformer output. Now after the rectifier bridge I get only 12.3 V. That is super strange, as I remember that after rectifying the readings were usually higher than the AC reading. But I'm not sure. Clearly if the output was just 12.3 V there is no way I can drive a battery charger from this.

But be that as it may, next my potential difference over the LEDs are 1.9 V LED1, 1.6 V LED2. So I should get 12.3-1.6 = 10.7 V over all of R5 and SCR2 to ground. That is what I get. OK.

But now, the voltage at the SCR2 over A to K is only 1.27 V. That is as if it was open. So what's the potential at it's gate? Only 0.62 V! And this is independent of the potentiometer.

The output at SCR1's K is only 1.2 V. Given the resistors R2 1.5 kOhm and VR1 10 kOhm the vast majority should fall over the VR1, and it definitely should vary depending on the setting of VR1. But it doesn't vary much. It is between 1.09 V and 1.16 V. Given that R2 is 1.5 kOhm, what is the resistance of VR1 for a voltage over VR1 U1 between 1.09 and 1.16 V? It is VR1 = U1 / (U - U1) * R2 = 1.09V / (1.2 V - 1.09 V) * 1.5 kOhm = 14.8 kOhm, and in the other extreme over 400 kOhm. That is definitely fishy. But still, why would the Zener diode let anything through?

I suppose I need an oscilloscope to figure this out. I have an Arduino here, but another project to fix my first project?

I suppose there is something wrong either with the potentiometer or with the Zener diode. Perhaps everything is working fine and only because the battery is not connected does the system shut down entirely? If current was passing through the battery (I don't have a battery to test, I need this thing to work because I get back to my battery at a remote place where I can not buy any electronic parts), then would the Zener diode not trigger the shut off?

To summarize the findings to this point:

• it looks like SCR2 is open, even though I measure only 0.6 V on its gate there is only 1.12 V on the gate of SCR1.
• I checked all resistors, including the VR1, they are according to specifications.
• so this means indeed that the ZD1 triggers because the maximum voltage is on the cathode of SCR1.
• but if that was the case, I should be able to down regulate VR1 setting it almost to ground, so then the ZD1 should definitely not trigger

I cannot see why there should remain a 0.6 V potential measurable on the SCR2's gate if VR1 is all the way shorted to ground? So what could be wrong?

The 560 Ohm resistor feeding gate current to the thyristor via the 1n4002 diode, is almost certainly too high a value leading to the device not being turned on properly. This in turn will increase the dissipation and cause over heating. You need to reduce the value at least by 10 to about 56 Ohms. One simple way to arrive at the value is to monitor the battery voltage and reduce the value gradually until the voltage stops increasing. You may destroy the thyristor if you reduce it too far so no less than say 33 Ohms. The best way is to monitor the gate current and set it at the manufacturers recommended value from the data sheet.

Aha, I noticed now looking closer that on my perf board I left exactly that lead of VR1 that should short to ground without any connection to ground! This would explain why the potentiometer has no effect, and why the value I calculated from the potential differences is so much higher. It is because it's all pushed through the ZD1 and R4!

OK and after I fixed that it is just about working. Now I measure 11.26 V on these battery cables. If that's true, that would still be way less than what is needed to actually charge a battery.

Why this transformer's output is so bad? It actually is a bi-volt transformer. It has 4 wires on the primary coil and 2 on the secondary. I have:

1. red
2. green
3. yellow
4. black

There is continuity between red and green and between yellow and black. I put in a 220 / 110 V switch in such a way that in the 220 V case green and yellow are connected to each other, while in the 110 V case I would connect green to black and yellow to red. This seemed to make sense when I experimented with that before I soldered it. But can I explain it? In the 110 V case the primary coil should be half the windings than in the 220 V case. And that seems what I am doing. So, why would I get such a bad result from a transformer rated at 15 V and why do my power outlets give me only a 210 V reading? Despite this being Brazil, I am in quite a modern city, it shouldn't be off by so much.

Anyway, I am much further along now, this is why it has helped me to write this all up here. Without you guys potentially reading it, I would not have taken the time and discipline to explain it and measure it once and again to be sure. Thanks for existing, even if you haven't read until here.

Success: now I am with my battery and it's working well. Much better than the commercial charger I had bought before, which actually destroyed the battery not doing the trickle charging properly. This one does. And it is capable to force a charge into a dead battery. Nice.

PS: I was alarmed at first to find the full 12 V potential on the heat dissipator of SCR1, but I checked the data sheet that this is considered lead 4 and that is connected to the anode A. I find this very strange and if I had put this into a metal case I could have shortened it to GND. Very odd.