P_ower supply question

Hi all,

I am building a "digital dash" for my motorcycle, a 1999 Royal Enfield Bullet. It uses a GPS module for speed, distance and acceleration (0 - 40, 0 - 60). I also monitor battery voltage vis a potential divider and an analogue input. My problem is that it keeps locking up or resetting at random. It only does this when I power it via the bikes loom. If I use a battery pack and do not connect it in any way to the loom then it works fine (apart from not monitoring battery voltage of course!).

I am using a 7805 voltage regulator with 100 uf electrolytic capacitors between the input and output lines of the regulator and ground. There is also a 4 mH toroidal suppression choke on the input.

Any thoughts or suggestions?


I'd get rid of the choke. The whole point of the 100uF capacitors is to make up for the impedance between the power source and the 7805 input. By adding a 4mH choke you are just adding more inductance (a LOT more inductance) for the capacitor to overcome.

-- The Ruggeduino: compatible with Arduino UNO, 24V operation, all I/O's fused and protected

@Rugged: Originally it did not have the choke in. I put it in to try and cure the problems but it had no effect at all :~

@KE7GKP: I forgot to mention there is a diode after the choke and before the voltage regulator. When you say reservoir capacitor is that in series or after the diode to ground. Forgive my ignorance.



I think it would be useful if you could sketch out your schematic. I have a feeling there are a few more "gotchas" hiding in there somewhere ;)

Also, make sure your capacitors are as close to the 7805 as possible.

-- The Gadget Shield: accelerometer, RGB LED, IR transmit/receive, speaker, microphone, light sensor, potentiometer, pushbuttons

C1, C2, C4, C5 = 100uf Electrolytic C3, C6 = 100nf Ceramic

The capacitors are physically close to the 7805.



Get rid of the choke, move C4/C5 to the input rather than the output (100nF on the output should be enough given the capacitances already on the Arduino).

-- The Rugged Motor Driver: two H-bridges, more power than an L298, fully protected

Thanks again guys I have made the changes but before I could try it out the throttle stuck wide open on the bike so I need to fix that first :cold_sweat:


You might want to ensure that ALL relevant connections are made as close to, and preferably onto the actual motor cycle battery terminals. The electrical noise introduced from the ignition coil will be buzzing merrily throughout the cable loom so you need to get as close to the battery as possible, where the effect is minimised. I'd also suggest mounting everything within a diecast metal enclosure which is itself "earthed" to the bike frame. This will act as a shield to prevent RF affecting the circuit. Run all power and signal leads within screened cable, with the screening connected to the bike frame. The use of ridiculously sized capacitors is totally unnecessary providing you take sensible screening precautions.

Try an RF choke, not a low-frequency one - your circuit has varying supply current requirements, a large choke will fight to keep the current very constant and might actually cause over/under voltages. Is the choke rated for the DC current it is taking BTW? If not it won’t act like a choke at all.

10uH/1A would probably fare much better (assuming you don’t use more than 1A).

Anyway you first need to identify how the interference is reaching your circuit, it need not be the power supply at all, it could be capacitively induced onto a signal like, or inductively induced onto any of the wires… Or even directly onto the circuit board. How is everything connected? It could even be as simple as the supply dropping below 5V occasionally.