Need advice on overvoltage circuit

I had a situation where a voltage spike occurred when the lithium batteries in my solar system finished charging, and the voltage spiked until the charge controller shut it off. This spike causes the inverter to shut down, necessitating a reset. I know that is what is going on because I caught it on my oscope. Currently, I have a TVS diode combined with a 22K uF elctrolytic cap across the lines from the charge controller, and that appears to work ok, but I am mystified as why it works. Either the TVS or the cap by themselves don't reliably prevent it. I saw the below circuit somewhere, and for the life of me, can't find the link now.

Anyway, the circuit is shown (hopefully) below. As I understand it, the R1 is there to limit current flow through the zener, and R2 is a 1 ohm inrush limiter to limit the current through the transistor to 30 amps max.

The nominal voltage is ~27-28 volts, and the zener voltage is 30 volts. My questions are, in no particular order:
1: Will this work?
2. What type of transistor would be best suited?
3: Is R3 necessary?


Do you want to suppress over voltage transients/spikes?
Do you want to prevent the power supply malfunctioning or wrong supply being connected from damaging your circuit?

If the latter, google crowbar protection

Thanks… Tom… :slight_smile:

The circuit has certainly been drawn incorrectly since the base of the transistor will always be at 0 volts irrespective of the supply voltage.


Out of curiosity, I had a look at the crowbar type protection circuits as recommended by @TomGeorge.
Most appear to rely on a thyristor, the main differences being only in how it is triggered.

They use Thyristors/Triacs so they become a locked short across the supply rails and ensure the series fuse blows.
I had one on my 13.8V 30A linear power supply, thankfully it never had to trip.

Tom… :grinning: :+1: :coffee: :australia:

Thank you all for the replies. I couldn’t reply earlier because the forum was down. The circuit is indeed drawn wrong, as I forgot to delete the line from the zener to ground. I’ll repost what I think is the correct drawing later.

I did consider a crowbar circuit, but I don’t want the system to shut down completely, just absorb these small spikes.

Here is the updated drawing…

Most people would simply use a TVS diode, cheap, very fast and available in a wide range of clamp voltages. See this discussion on their use in automotive circuits.

That is what I am using ATM, coupled with a 22K uF capacitor. The TVS by itself isn’t sensitive enough. I thought a setup like this would be more exact.

I’ve been researching BJT transistors, and it seems to be a matter of balancing the transistor gain against the current rating of the zener.

What do you mean by that?

The TVS did fine protecting the arduino’s power supply, rated at 35 volts max,
but did not reliably prevent the inverter’s overvoltage circuit being tripped at 30 volts. Adding the cap to the circuit eliminated that last problem.

I’m having a hard time understanding the transistor’s datasheet, specifically the gain values. I think this transistor will work, but I can’t figure out if the amount of current (~300 milliamps) I can safely pass through the zener diode will switch it satisfactorily.

Sounds like you have a broken charge controller - it should be limiting the output voltage automatically - its not very clear but it seems you charge controller overvolted the batteries and you are trying to mitigate this by adding a shunt regulator.
Is the charge controller actually matched to the particular batteries you are using? What is it?

Can you post link to specs/data of your solar charge controller please?
Some controllers can be programmed or have a parameters menu to set things like battery characteristics, for example different battery types.

How much PV have you got?
How much battery capacity do you have?

Thanks… Tom… :grinning: :+1: :coffee: :australia:

Thank you guys for your replies. I really do appreciate you taking the time. The charge controller is not the issue. I saw this problem last year, and went 'round and round with the controller manufacturer, inverter tech support, and the engineer at the battery factory. I tried 4 different controllers from 4 different manufacturers, and the original controller manufacturer sent me a new one. All had the same result. It is just one of those maddeningly intermittent issues that seem to just be a combination of factors. The guy at the inverter tech support admits that the inverters are just too sensitive, but there wasn’t much he could do. It was him that told me about the TVS/capacitor and shunt regulators. I went with the TVS and cap because it was the simplest, and seems to work fine. Now that I have more time, I’d like to explore the shunt, both as another fix for this problem, as well as gaining some experience with transistor circuit design.

Can you show in a diagram how everything is wired, and particularly how long the various cables are? Normally a big battery will absorb pulses very solidly, so I’m suspecting there’s
an issue with layout and EMI generating transient spikes.

1 Like

What voltage do you want to clamp at ?


This topic was automatically closed 120 days after the last reply. New replies are no longer allowed.