I'm working on this project that includes a relay connected to one of the Sparkfun Car Horns. However, when I try to mount the car horn to the base (metal), it seems to ground out. If I leave it off of the base, it works fine.
I've tried nylon bolts, electrical tape, and rubber to try to isolate the metal bracket (holding the car horn) and the base.
Here's a quick video of me explaining everything: - YouTube
Maybe it's a wire that's coming loose? My wiring is VERY messy, and that could very well be the case. But, I've gone over the main board several times making sure all of the connections are solid. I'm fairly confident on that.
Need a schematic.
At 8 amps and this being a car horn there will be a lot of inductive noise generated.
Need very good decoupling on things.
Need snubbing on the horn.
You have interference from the high current circuit of the horn directly influencing
your logic circuitry I think - not specifically a grounding issue, you have big loops
of wire carry high currents (car horns can take transients of 50A or more I believe.
they are very unsophisticated.).
Firstly twisted-pair - this is how all high current supply and load wiring should be
done, to get rid of all those loops (which radiate e-m interference and induce
voltage in nearby circuits). Every high current wire needs to be twisted with its
return wire so the net current is zero when not directly touching the cable. The
twists cancel out the magnetic field component that is responsible for the induced
signals.
Secondly keep high current wiring short and neat, and very important, well away from
low-current sensitive circuitry. This means never run them in the same cable harness.
You can shield the sensitive signal cables to reduce pick-up too.
And lastly a car-horn is an inductive load, a substantial free-wheel diode may be
needed to stop it pushing large voltage spikes onto the supply (100's of volts). In
a car the large battery is left to do the best of a bad job with this. Horns, I think,
have some sort of built-in snubber capacitor to reduce the spikes, but its worth
investigating.
LarryD:
Need a schematic.
At 8 amps and this being a car horn there will be a lot of inductive noise generated.
Need very good decoupling on things.
Need snubbing on the horn.
I don't have a schematic on hand, but here's the jist:
The 9.6V battery (+) is connected to the Arduino Mega's VIn port, then routed to the COM port of the relay
The (-) side of the battery is connected to the Arduino Mega's GRD port, which is also connected to the Horn's (-) side
The (+) side of the horn is connected to the NO (Normally Open) port of the relay
Need more details?
Can you explain more on the "inductive noise" and "snubbing"? Is it the physical vibrations that might be causing the issue?
LarryD:
Here are some starting points.
Keep wires short.
Keep wires to horn away from all others wires.
Wire up similar to the attached image.
Thanks for the fast replies! I tried shortening the wires, twisting them together, and did my best of keeping them away from other wires (very hard to do with such a small space). Still seemed to short out.
I'm still somewhat new to electronics. Where should the diodes be connected? I don't have any connected in my project right now.
I also didn't put a resister for the signal wire of the relay. Maybe it's getting some noisy interference?
The diodes must be connected right next to the device they are assigned to.
Make sure you have the wires connected in a star configuration as in the schematic.
LarryD:
The diodes must be connected right next to the device they are assigned to.
Make sure you have the wires connected in a star configuration as in the schematic.
Star configuration? I'm sorry, I'm still very new to electronics.
LarryD:
Here are some starting points.
Keep wires short.
Keep wires to horn away from all others wires.
Wire up similar to the attached image.
Thanks for the fast replies! I tried shortening the wires, twisting them together, and did my best of keeping them away from other wires (very hard to do with such a small space). Still seemed to short out.
Short out?? Nothing is shorting out surely??? I thought you had spurious resets.
You've wound the right wires together? Someone mentioned decoupling, which for the
horn needs to be a big electrolytic.
I'm still somewhat new to electronics. Where should the diodes be connected? I don't have any connected in my project right now.
google "freewheel diode" "inductive load", terms like this. You need one very high
current diode for the horn if that's the problem. However it might affect the
note - a snubber network would be better for this, but snubbers are trickier to
design.
I also didn't put a resister for the signal wire of the relay. Maybe it's getting some noisy interference?
Can you show how you've wired the relay, a transistor or MOSFET is needed, what
resistor are you thinking of?
Anyway the really important thing is don't power the Arduino from the same supply as
the horn, that's bound to lead to many issues and could damage the Arduino.
PS A photo of the new wiring and a circuit diagram would both be very handy to see.
LarryD:
The diodes must be connected right next to the device they are assigned to.
Actually, it turns out that is wrong!
You need to understand the function of the diode.
The diode conducts in order to keep the current flowing. While the inductive load is powered, current flows from the supply, through the load, through the switching element and back to the supply. On the moment of switch-off, the current is instead carried through the diode and the inductive load, so while the current through the load does not change rapidly, the current through the supply and the switching element does, so if there is to be any radiation caused by the sudden change in current, it will be due to the change in current in the switching element and power supply. If you attach the diode at the inductive load, then the wires carrying the current from the power supply and switching element will be subject to that transient in current. If however you put the diode at the switching element and power supply, the connecting wires to the load will not be subject to that transient in current.
Now on the moment of switch-off, the voltage across the switching element instantly (well, limited by switching speed and circuit capacitances) rises to the supply voltage plus the diode forward drop. This represents a voltage transient on the wire which connects the switching element to the load however since the transient is the same at both ends of that connection, it is immaterial in respect of the voltage transient whether the diode is at the load or the switching end.
And an important consideration is that the sudden reduction in current is experienced by the switching element and the power supply, buffered by the power supply and in particular the output capacitor or bypass on the switching circuit. This means that to avoid this transient being conveyed to sensitive areas, any other circuits connected to the same power supply should connect at the supply terminals or at least, the main bypass capacitor (the "star" arrangement referred to above).
With large inductive loads you might want to place the diode on the inductor to
prevent the safety hazard of electrocution (someone disconnects the cable whilst
the inductor is powered and a big spark is created).
