I recently bought the S18V20AHV regulator, visible here. I used it to generate a 24V output given a 12V 2.1 A regulated input. The 24V would then go to a H bridge that would drive an old bell mechanism composed of two 500 ohms coils commutating at about 30 hz during operations.
First, and as I had only that at my disposal, I used a 9V cell. I was able to generate the 24V. I then removed the cell, plugged the 12V regulated source and rotated the screw a bit to obtain the same 24V. The chip then got really hot (but it seems that is normal) and started to produce a little noise, very high pitch. I assumed it was coming from the the coil inside the circuit, nothing serious. After that it seems that no matter how high I would rotate the screw I would still get 12V in the output, but somehow the system was still working fine for a 9V input. After many swaps and different configuration, it seems the system got even weirder as when the 12V was on the measured voltage of the source would drop to 3V only …
What is seems is that somehow I fried the component by touching it too much (there were a lot of wires, and possibly a few arcs when I was measuring voltages), and that for some reason it kept on working for 9V input, but was in short for 12V, causing it to draw maximum current and leading my regulated source to go under and generate only 3V.
Can someone please confirm that behaviour, or at least what I thought possibly happened ?
Also, has anyone used this component and identified case scenario that are to be avoided, or very sensitive points on that chip? Any help would be much appreciated, I will order one more and start over again in two weeks and I just need to confirm that my burnt chip is only my fault and the the thing is robust enough to survive, but with careful hands
The potentiometer should not need to be readjusted when switching to a different input source. What kind of an H-bridge are you using? Did you have it connected when you first tested the 24V output with the 9V battery? Does the regulator output 24V now with no load attached and your 9V battery as an input? If not, can you post pictures that show your board and connections including any soldered connections you made?
I don’t have the “thing” right now with me, but yes I soldered a circuit that includes the regulator and the H bridge which is a basic L293D found in the arduino starter kit (can take as much as 36V in). It was connected all the time and the whole system was working fine when ringing the mechanism controlled by the H. The regulator is now dead but yes it was outputting the 24V using the cell, and it worked for like 15min with the 12V regulated from sector.
If you have not checked the regulator output with no load and a known good battery, you might try that to see if the regulator gets hot or fails to output a consistent voltage to verify it is damaged.
If the regulator is damaged, you might want to check your circuit for large transient voltages using something like an oscilloscope. It sounds like something in the regulator might have failed in a short and damage like that can be caused by short voltage spikes over the voltage limits. Generally loads like electromagnetic coils have high inductances, which can create spikes like that. I would generally expect a system powered by a battery to better protect against transient voltages than one powered by a regulated power supply. Also, power supplies can generate voltage spikes when first powered. The L293D has integrated flyback diodes in parallel with the transistor switches that form the H-bridge, but the L293 does not, so you might also double check which version of the IC you are using.
Adding electrolytic capacitors and over voltage protection like Zener diodes to the voltage rails of your circuit can help suppress transient voltage spikes. Also, if you get a replacement regulator, you might test the output voltage of the regulator with both supplies before connecting the driver circuit.
Yeah I think at some point a short must have happened … I remember seeing one arc when measuring using a multimeter. Also the fact that the regulated 12V output only 3V in the end prove that a high current must have been pulled and brought it to its knees I think.
Also I checked and I the H bridge I use is the L293D and there are flyback diodes, so the spikes inside the H bridge are definitely not a problem I think.
If I understand you right, the quality of the regulated 12V may be in cause. I don’t have any Zener but I have capacitor that I can use to prevent possible spikes. I’ll put this together in a few weeks.
In the meantime, here is a procedure to ensure I am doing things right :
- put a 10kOhms resistor in the output the of the regulator
- plug the 12V regulated input on the regulator
- rotate the screw accordingly as to obtain desired output, 24V
- don’t touch the screw anymore and assemble the whole thing, possibly with decoupling capacitors on the input side
Do you agree with all that?
Yes, those all seem like good steps.
Hi there, in three days I will be returning and be able to do more testing.
One last minute thought though, should I care about the fact that I have a 12V regulated source powering the S18V20AHV & an Arduino as well? I usually don’t pay attention to that, but maybe input impedance should be similar ? That could be the source of the weird 3V output on the 12V regulated source ?
Do you have a datasheet or any more details for your 12V source? Power supplies commonly provide a consistent output voltage and it is generally acceptable to connect different loads in parallel to the supply, so long as the maximum output current is not exceeded. It sounds like the regulator failed in a short, which was trying to consume more power than your 12V source could supply and dropping its output to 3V.
Also, if you post pictures or schematics of your connections, I can let you know if they look OK.
You mean the regulator in the 12V source? Yes I’ll post schematics as soon as I have them.
I do not have the datasheet for the 12V source but it’s a regular “big black box” one, coming from china …