I’ve helped a few other owners that saw their battery light flicker when the AC compressor clutch would turn off. The fix was to have them install a 18 volt, 5 watt zener diode to clamp the inductive spike you get from the compressor clutch turning off quickly.
Well I found that spike problem getting into my battery relay controller. It would put the microprocessor into a reset. I fixed the engine kill problem with a software change but I still got the battery relay doing an extra power on cycle. The battery relay stays on and when the micro is reset it defaults to turn on the battery relay after a five second delay.
So I wonder what other things that unclamped voltage spike could affect. Probably misfires the ignition ECU. Probably hits the lambda ECU. Probably hits the idle ECU. We just don’t notice those problems since they only would occur when the AC is turning off.
So I decided to install that same zener diode fix even though no real problem is noticed without it.
Installing that zener did stop my battery relay circuit from resetting. The zener part number I used is 1N5355BRLG if you want to do this change. The striped end goes to the compressor wire and other end to ground. It’s best to put the zener as close to the noise source but I would think you could also mount it anywhere on that circuit.
I've been working on a vehicle-powered project recently and have researched a lot on the kind of electrical noise found in cars. What's happening when the compressor turns off is probably a small load dump - the alternator's output voltage rises really high (>100V) for a couple hundred milliseconds before it reacts and settles back down again. I'm using a 1kW TVS (transient voltage suppression) diode in my design, and it can still be destroyed by large load dumps, so I've included a resettable PTC fuse in series. I've yet to get the PCBs back, but they should be here in a few days, so I'll test the design then.
TVS diodes are basically big zener diodes that clamp voltage spikes to a safe level. All automotive equipment has these or similar parts that keep their electronics safe, so I don't think it should affect anything else in the car.
I'm not too sure your zener is going to last very long - 5W isn't a whole lot when you're talking about trying to clamp long, high-voltage spikes. Let me know how it does, though!
Yes the AC clutch is probably the worst inductive spike and will need a high power device to clamp and dissipate all that energy. I could have driven the AC clutch with one of my solid state relays to fix the spike. My relays have slow turn on and turn off times designed into them to reduce the inductive spikes.
It's complicated to measure the power requirement of inductive clamps so I just used an educated guess at using an 18 volt 5 watt zener. I will have to try to test the component from time to time.
So far the owners that installed the same component to fix their flashing battery light have not got back to me with any problems.
For those people not knowing what we are talking about, you get high and low oscillation voltage when the current in an inductor (wire wound on metal) is removed quickly. It's generated by the collapsing magnetic field. Typical 100 to 200 volts swinging above and below ground.
Here is a screen shot of the noise measured right at the compressor clutch in stock (no suppressor) configuration.
I was expecting a positive ringing also but it looks like just a negative voltage spike over 200 volts. So a regular diode would have also clamped this spike if yours shows the same. The Zener should also do the job since it conducts reversed voltage like a diode.
Edit:
I did check it with the zener attached and it clamps all that spike.
Last edited by Bitsyncmaster; 07-20-2015 at 09:49 AM.
Thanks for the advice in this thread! I installed the zener diode with the stripe toward the compressor, however it is hot to the touch. The compressor turns on and off as expected, and I checked resistance after running to confirm that it still blocked the flow of electricity in the other direction. My question is, is it normal for the diode to be too hot to hold onto when the compressor is operating? Want to be sure I'm not going to create an electrical fire back there.
These diodes were labeled as "1N5355BRLG DO-15 5 Watt 18V Zener Voltage Regulator Diode." If I perhaps have a faulty diode, where would you recommend I buy from?
Thanks for the advice in this thread! I installed the zener diode with the stripe toward the compressor, however it is hot to the touch. The compressor turns on and off as expected, and I checked resistance after running to confirm that it still blocked the flow of electricity in the other direction. My question is, is it normal for the diode to be too hot to hold onto when the compressor is operating? Want to be sure I'm not going to create an electrical fire back there.
These diodes were labeled as "1N5355BRLG DO-15 5 Watt 18V Zener Voltage Regulator Diode." If I perhaps have a faulty diode, where would you recommend I buy from?
A zener diode only has a few micro-amps of current flowing back biased to nearly the rated voltage. These higher voltage zeners even have a sharper curve then low voltage zeners. So unless your voltage is approaching 18 volts or your zener is defective it should not produce any heat. If your battery connections are weak then your alternator may be producing some peak voltages. The battery is what filters the alternators rectified AC.
You can really use just a diode since I only saw negative voltage spikes when I tested my car.
Dave M vin 03572
Reply With Quote
Originally Posted by DavidProehl
