New to the Group, Need recommendation on suppressing arcing at the relay and fuse for radiator fan

Thread Starter

MB107

Joined Jul 24, 2016
345
I have an electric fan on my car it draws about 18 amps. When I first put it in I decided I did not want this circuit to fail so I used a 75 AMP heavy duty relay. Like this one.




Amazon.com: TE tyco BOSCH 75 Amp High Current 12 Volt Automotive Relay SPST: Automotive


After driving the car for about 6 month the relay exploded. We are talking shrapnel and the EMI effected the computers and shut down 6 cylinders. I had to limp it home 30 miles on 6 cylinders (This is a V12) and no fan. That is why I didn't want it to fail. Since that didn't work I replaced the relay with a standard 30 amp relay that you buy from pep boys for about $6.00. Its been working flawlessly ever since until now. Today I found the fuse blown. And I don't mean just blow as in pull out the fuse and replace it. I mean shrapnel and nothing in the pile of melted goo that remotely resembles a fuse. I think these events are related. Arcing across relay contact or arcing across a blown fuse. In both cases the possibility of fire was very evident.

In any case I did a little research and found that this is a problem in a lot of places. I found wire diagrams of circuits used to suppress arcing, consisting of a capacitor and resistor or zener diodes.

What I haven't found is the table that says for 12V at 20 amps use size X capacitor and size Y resistor and place between pins? and ??

Should such a device do on the relay contractors or across the motor.

It has only happened twice in about 8 years but it is a safety concern.

Does anyone have any suggestions.
 

Thread Starter

MB107

Joined Jul 24, 2016
345
So from the Electorcube table I found the present a formula to calculate capacitance = I^2/10 which for me using the published 18.5 amps comes out to 18.5^2/10 = 34.2 mfd. about the size of a standard automotive distributor capacitor from when cars had points. But then they go on and say that .47 to .5 has been found to be the optimum.

Then calculating the required resistance based on 34.2 and .5 I get < 1 and way < than 1. But the lowest resistance unit they have is 10 Ohm

For wattage they are recommending 1/2 watt for < 3 cycles per minute. In actual use the fan may cycle at most 6 times per hour. The best I can do is get 2 watts.

My calculations are shown below along with the text format of the equations used to calculate resistance.

My best fit unit is the RG1676-55

So the question I have is that it seams pretty far off but the event only happened 2 times in 8 years. If I didn't consider it to be a safety issue I would just replace the fuses and relays as required. Would this be enough to diminish the arcing to a non damaging level. Would I just put this across the two motor leads.


Suppression.jpg
 

Thread Starter

MB107

Joined Jul 24, 2016
345
Across the motor and also maybe place one across the relay contacts to protect them also.
Max.
Thanks again
If I replaced them across the relay contacts would I have an issue with battery drain when the car is not running. In my case power to the relay is hot at all times. Across the motor the circuit would never see power with the relay not active. A work around would be to take power from a key on source but I prefer not to do that.
 

crutschow

Joined Mar 14, 2008
34,281
I think an R/C suppression circuit is more complicated than necessary for this application.
Just connect a 5A diode (such as this) from the relay contact to ground (assuming a high-side relay connection) to suppress the negative spike from the motor and wiring inductance when the relay opens.
Connect diode anode to ground and cathode to fan motor relay output.
You don't need anything additional across the motor.
R/C suppression circuits are typically used in AC applications where a simple diode suppressor won't work.
 
Last edited:

Thread Starter

MB107

Joined Jul 24, 2016
345
I think an R/C suppression circuit is more complicated than necessary for this application.
Just connect a 5A diode (such as this) from the relay contact to ground (assuming a high-side relay connection) to suppress the negative spike from the motor and wiring inductance when the relay opens.
Connect diode anode to ground and cathode to fan motor relay output.
You don't need anything additional across the motor.
R/C suppression circuits are typically used in AC applications where a simple diode suppressor won't work.
Thanks for the reply

Not sure I understand what your describing. If I connect one side of the diode to the relay output and the other side to ground that would be electrically the same as putting the diode across the motor since the motor is connected to the relay output and to ground. Or is proximity to the relay important.
 

crutschow

Joined Mar 14, 2008
34,281
Thanks for the reply

Not sure I understand what your describing. If I connect one side of the diode to the relay output and the other side to ground that would be electrically the same as putting the diode across the motor since the motor is connected to the relay output and to ground. Or is proximity to the relay important.
Proximity is a consideration.
It's basically the same connection, but placing it close to the relay suppresses any additional transient from the (small) wiring inductance.
If placing it across the motor is easier, then I wouldn't worry too much about it.
 
A diode that's revered biased at the motor basically absorbs the energy in the motor coil. Close proximity is a good thing because the closer you "catch it", the less likely that length will act as antenna.

A diode that's reverse biased across the coil does the same thing to the coil energy.

An RC type snubber across the relay contacts itself are generally not critical in terms of value, but with DC a transorb (TVS diode) or ZNR rated at 18 VDC might be a better choice. The snubbers are best used for AC applications.
 

Thread Starter

MB107

Joined Jul 24, 2016
345
Not sure if this makes a difference but the 87 and 30 terminals are connected backwards from standard convention. I will be changing that shortly.
Incorrectly wired relay.jpg
 

Thread Starter

MB107

Joined Jul 24, 2016
345
My corrected solution will be as shown. If that doesn't work I will try replacing the diode with the RG1676-55 resistor capacitor.
Corrected Solution.jpg
 

tcmtech

Joined Nov 4, 2013
2,867
After driving the car for about 6 month the relay exploded. We are talking shrapnel and the EMI effected the computers and shut down 6 cylinders. I had to limp it home 30 miles on 6 cylinders (This is a V12) and no fan. That is why I didn't want it to fail. Since that didn't work I replaced the relay with a standard 30 amp relay that you buy from pep boys for about $6.00. Its been working flawlessly ever since until now. Today I found the fuse blown. And I don't mean just blow as in pull out the fuse and replace it. I mean shrapnel and nothing in the pile of melted goo that remotely resembles a fuse.
This seems strange to me being any standard automotive type of DC rated relay or fuse won't flat out explode in normal working conditions unless it's hit with an extremely high DC current as in dead shorting across the battery and even then actual 'pieces a flying' blowing up is extremely rare.

Even at a dead stall a fan that draws 18 amps running seems like it would be highly unlikely to have a low enough stall resistance and inductance to be able to wipe out a normal 30 amp relay let alone explode a 75 amp rated one without popping the fuse first which again shouldn't be able to be physically blown up by that type of over current event either.
 

Thread Starter

MB107

Joined Jul 24, 2016
345
Well I was thinking about this and it was about 6 years ago the 75 amp relay blew up and to the best of my knowledge it did not blow the 30 amp fuse. It may be possible that the 30 amp fuse and fuse holder were compromised during the event. In any case attached are pictures of the blown fuse holder. I replaced the relay and fixed the wiring to the industry standard B+ to #30.

IMG_20160723_141823812_HDR[1].jpg IMG_20160723_141823812_HDR[1].jpg IMG_20160723_141915480[1].jpg IMG_20160723_175327727[1].jpg
 

crutschow

Joined Mar 14, 2008
34,281
A diode that's revered biased at the motor basically absorbs the energy in the motor coil. Close proximity is a good thing because the closer you "catch it", the less likely that length will act as antenna.
,,,,,,,,,,,,,,,,,,.
I see no significant difference where you put the diode as far as the wire acting as an antenna from the motor inductance.
In either case the voltage in the wire will suddenly drop from 12V to -0.7V when the relay opens and the diode starts to carry the inductive motor current.
I prefer putting it at the relay since it then also absorbs any transient spike from the wiring.
Even though the wire inductance is fairly small (about 0.4μH/ft), there still can be sufficient energy stored, with 18A through a long wire, to create a significant voltage spike and high frequency oscillations when the current is suddenly interrupted.
That could certainly create some EMI radiated from the wire.

Note: As tcmtech noted, The two catastrophic failures are of concern. There would be no normal path for such high current needed for such a failure.
I suspect a momentary short somewhere in the circuit, either the wiring or the fan.
I would double-check the wiring to the fan.
Perhaps replacing or bypassing it with a new wire would be prudent.
 

Thread Starter

MB107

Joined Jul 24, 2016
345
I see no significant difference where you put the diode as far as the wire acting as an antenna from the motor inductance.
In either case the voltage in the wire will suddenly drop from 12V to -0.7V when the relay opens and the diode starts to carry the inductive motor current.
I prefer putting it at the relay since it then also absorbs any transient spike from the wiring.
Even though the wire inductance is fairly small (about 0.4μH/ft), there still can be sufficient energy stored, with 18A through a long wire, to create a significant voltage spike and high frequency oscillations when the current is suddenly interrupted.
That could certainly create some EMI radiated from the wire.

Note: As tcmtech noted, The two catastrophic failures are of concern. There would be no normal path for such high current needed for such a failure.
I suspect a momentary short somewhere in the circuit, either the wiring or the fan.
I would double-check the wiring to the fan.
Perhaps replacing or bypassing it with a new wire would be prudent.
Makes sens but shouldn't the fuse just blow normally.
 

tcmtech

Joined Nov 4, 2013
2,867
Well I was thinking about this and it was about 6 years ago the 75 amp relay blew up and to the best of my knowledge it did not blow the 30 amp fuse. It may be possible that the 30 amp fuse and fuse holder were compromised during the event. In any case attached are pictures of the blown fuse holder. I replaced the relay and fixed the wiring to the industry standard B+ to #30.
Wow! Never seen those styles of relays or fuses ever burn up like that before and I have worked around high powered industrial equipment for years that uses the same types of relays and fuses where many have been destroyed by hard dead short circuit events. I've seen them with their contacts welded together or internal metal conductors melted right through the plastic casings but never blown to bits.

Same with fuses. At worst I have seen them with big black blisters in the center where the sacrificial link is or the fuse and holders melted together from loose connections but not blown apart like that.

My first thought was loose connections getting way to hot but still that tends to just melt things together or have the high current mental components just pull out of the plastic but not cause explosive destructive failures like what you are showing in the pictures.

Very unique failures indeed. You have me stumped and I consider myself an expert at destroying electronic components in nonconventional ways. :p
 
Top