Help with voltage spikes

[cggorman]

So I am having some issues with my voltage spiking, originally I thought it could be a bug with the megasquirt which could be fixed with a capacitor, but I am realizing it is an actual problem. The voltage at idle and below about 2700 RPM is stable at between 13.2 and 13.9 volts; however, when the RPMs rise to around 3000 and higher, the voltage spikes can, at times, be as high as 20 volts. This is verified with lights and a dmm. Had my high beams burned out because of these spikes. Tried adding more grounds to the voltage regulator ie. Voltage regulator to alternator, voltage regulator to body and even on directly to the battery and still the spikes occur, what am I missing? Any help would be great.

[Matt Cramer]

Are you using low or high impedance injectors?

Where is the MS grounded?

[cggorman]

High impedance
Ms is grounded through the relay board and then grounded to the same place as the factory ECU

[LAV1000]

When lights burn out because of voltage spikes, then there is something seriously wrong.
Check alernator and regulator, start with regulator.

[Matt Cramer]

High impedance
Ms is grounded through the relay board and then grounded to the same place as the factory ECU

Not all factory ECU ground locations were really ideal. Where exactly is the same place at the factory ECU?

[cggorman]

I'm sorry, but how would the ms ground in anyway affect an otherwise steady voltage below 2700 RPM? The ms cannot magically produce 20 volts can it? Regardless it is grounded to the chassis next to the stock engine fuse box

[Matt Cramer]

Actually, it can; it's generated from the voltage spikes created by the injectors (and ignition coils, if using direct coil control). And the more times per second the injectors and coils fire, the more voltage this produces. If the ground is marginal, this can be a serious issue.

Try grounding the MS to the battery (best) or the engine block or cylinder head (second best).

[billr]

Burning out headlamps takes quite a bit of power, I (and others) aren't convinced there is enough energy from the injector/coil "fly-back" to do that; seems like it has to be a problem with the charging system.

Edit: clarify what you mean by "spikes", are these very short-duration pulses or periods when the system voltage rises above normal. Post a run log that includes "battery voltage".

[DaveEFI]

So I am having some issues with my voltage spiking, originally I thought it could be a bug with the megasquirt which could be fixed with a capacitor, but I am realizing it is an actual problem. The voltage at idle and below about 2700 RPM is stable at between 13.2 and 13.9 volts; however, when the RPMs rise to around 3000 and higher, the voltage spikes can, at times, be as high as 20 volts. This is verified with lights and a dmm. Had my high beams burned out because of these spikes. Tried adding more grounds to the voltage regulator ie. Voltage regulator to alternator, voltage regulator to body and even on directly to the battery and still the spikes occur, what am I missing? Any help would be great.

Generally, this is likely to be a fault inside the alternator regulator. They are usually sealed units and can't be repaired. Depending on make and type of alternator, they can be quite simple to change, if you can get a spare. If not you'll need a new or recon alternator.

[Matt Cramer]

Sorry, I missed the burned out headlights and thought that was just voltage spikes inside the MS. The flyback spikes aren't likely to burn out headlights.

[cggorman]

IIRC the current alternator is not supposed to have a voltage regulator inside of it. Have the external regulator wired into the circuit correctly as best as I can tell from the installation instructions

[cggorman]

No, the alt doesn't have an internal regulator. This is a Neon/Talon/Eclipse 420A standalone installation. The factory ECU housed the voltage regulator. We installed an external regulator on the firewall. Its grounded to both the chassis and the alternatpr housing. It the reg from a 87 Dakota, IIRC. brand new. The Alt is also brand new (well, reman from Autozone). We'll replace the reg tomorrow.

The external reg is wired per Symtech's diagram. One wire directly to the alternator field wire and the other goes to the other field wire on the alt and to the battery thru a switched relay.

[LAV1000]

No, the alt doesn't have an internal regulator. This is a Neon/Talon/Eclipse 420A standalone installation. The factory ECU housed the voltage regulator. We installed an external regulator on the firewall. Its grounded to both the chassis and the alternatpr housing. It the reg from a 87 Dakota, IIRC. brand new. The Alt is also brand new (well, reman from Autozone). We'll replace the reg tomorrow.

The external reg is wired per Symtech's diagram. One wire directly to the alternator field wire and the other goes to the other field wire on the alt and to the battery thru a switched relay.

Relay and connections in good condition ?
Maybe a loose contact due to motor vibration ?

Otherwise see what happens with disconnected field wires

[cggorman]

Relay is brand new, connected to brand new socket and the wires are all soldered. I've been soldering and brazing for 20 years at work. The joints are rock solid.

[cggorman]

Plugged a new voltage regulator in and voltage is now always between 13.8 and 14.9 volts. Problem solved.