Voltage Spikes in Marine Electrical Systems

Feb 6, 1998
11,665
Canadian Sailcraft 36T Casco Bay, ME
A Look at Voltage Transients In a Marine Battery System

I had been meaning to put this together for a while and assembled a system for a training class I was doing. Once that was assembled I realized I could simply add a starter motor, an oscilloscope and a Fluke 289 & Fluke 376 and then would would was able to create a video on voltage transients.

In the video I’m walking through the very typical, though not necessarily ideal, marine battery system which utilizes the venerable 1, BOTH, 2, OFF battery switch. The typical 1, BOTH, 2, OFF switch is configured for both bank charge selection and bank use selection.

With this switch configuration the battery bank selected, via the switch, serves as both a house and starting bank as well as the bank which gets charged.

Here's the video (Warning: It's 16 minutes):

“You Must Have a Dedicated Starting Battery”

This gets repeated over and over on docks in magazines and on the web, and there is a bit of good intention in it but also some myth. While a dedicated starting battery is never a bad idea it is also not a 100% solution to eliminating transients in a marine system. So what is a dedicated start battery?

Dedicated Starting Battery: A battery used and wired as such that its only purpose is for starting the motor and powering engine electronics. It will have its own isolation switch though may often have the ability for a paralleled connection to the house bank, if needed in an emergency. A dedicated starting battery isolates starting from house loads.

Unfortunately starter motors are just one example of devices that can create a voltage transient. As can be seen in the video where you connect your DC panel loads matters and can greatly minimize the potential for transients.

Potential Damaging Sources – Other Than Starter Motor:
  • Windlass
  • Bow Thruster
  • Electric Winches
  • Electric Windshield Wipers (power boats mostly)
  • Engine Driven Refrigeration Clutch
  • DC Motors (these transients can occur on the negative side too)
  • DC Refrigeration
  • Water Makers
  • A Load Dump of an Alternator (passing a switch through OFF)
  • Inverters
  • Water Makers
  • Electric heads
  • Water & Washdown Pumps
  • SSB Radio
  • Bilge Blower
  • Battery Charger
  • Battery Switch
  • Relays, Solenoids & Combiners
  • 12V Appliances – Blenders etc.
  • Electric Oil Pumps
  • Bilge Pumps
  • Etc. etc..
By connecting the loads or equipment closer to the engine or tapping into a negative near the engine or positive lead near the engine you increase the potential for equipment seeing a transient from the starter motor.

Despite all the above potential sources of transients in well wired boats I rarely measure transients above 14V – 15V but when boats have been wired haphazardly I can measure transients in excess of 25V+.

It's important to recognize that the act of using a dedicated starting battery does not absolve your house bank of the risk of voltage transients. Starter motors are not the only potential source of transients.

Spending the money on a dedicated starting battery, despite all the marketing efforts by battery switch makers etc, will not remove all the potential for transients from your marine house bank but it can remove one potential.

Bottom Line:

Keep your DC load positive & negative take off points as close to the battery bank as possible and this includes keeping your battery switch as close to the batteries as possible. Best practice is to use only use one take off point for DC electronic loads, as close to the bank as possible, and you can minimize the potential for transients. If you are really concerned about this it can be measured and tested for using the proper equipment..

There are also tricks such as TVS diodes and devices to help eliminate and or minimize the risk further but generally, in a well wired system, this is not something you need to be hugely concerned with.
 
Oct 1, 2011
172
Canadian Sailcraft 36T PCYC Toronto
More great timely info, appreciate it as I'm doing a complete AC/DC re wire this spring.
 
Oct 24, 2010
2,405
Hunter 30 Everett, WA
Question
Have you considered the starter under load? I'm thinking here of the make/break of the starter brushes with high current loads. I'm thinking because the starter is an inductive load it has potential to provide higher spikes as the magnetic field collapses when the brushes break.
The reason I'm thinking of this is because in a household AC power failure, that frequently happens as the transformer fields and fields of any connected inductors collapse. This is why power failure frequently plays havoc with semiconductors. It may also be interesting to know not just the amplitude, but the duration also. (Sorry, I do know it requires a different test than the bench)

Ken
 
Jan 30, 2012
1,123
Nor'Sea 27 "Kiwanda" Portland/ Anacortes
I love this stuff and I know for a fact that we all do. But ---

Is it the closer the battery to the motor, or the closer the battery to the victim device, or is it the battery attenuates the motor(s) transients if the battery is between the motor and the victim device? I read "the bottom line" but Is there some operative general rule(s) - which are more like commandments?
You know like 'thou shalt' or 'shalt not'?

Charles
 
Last edited:
Jan 25, 2011
2,391
S2 11.0A Anacortes, WA
I love this stuff and I know for a fact that we all do. But ---

Is it the closer the battery to the motor, or the closer the battery to the victim device, or is it the battery attenuates the motor(s) transients if the battery is between the motor and the victim device? I read "the bottom line" but Is there some operative general rule(s) - which are more like commandments?
You know like 'thou shalt' or 'shalt not'?

Charles
There are "numerous" commandments. It depends how far into the noise you want to go. We're not doing Dreamliners here so we can stay with some basic ones. Maine says to take your + -takeoffs as close to the battery as possible. Batteries make excellent capacitive filters. If takeoffs are further away from the batt, there is the inductance between the takeoff and the batt to take into consideration and the battery is not as effective as a filter. Inductance is basically resistance but in the frequency domain. So think of it as ac/rf resistance which noise would be characterized as. Do not tie or parallel your takeoffs to the large battery wiring as they will inductively pickup starter transients and noise. Brush noise is basically white noise or random noise with very high rise and fall times that can be observed with an oscilloscope. The energy content would be pretty low and the starter wire would attenuate it because of the inductance of the wire which would be pretty high due to the high frequency content of the noise.
 
Jun 6, 2006
6,990
currently boatless wishing Harrington Harbor North, MD
For the record the spikes come from the sudden turning off of a device that has a coil like the windings in a motor. The current has developed a magnetic field and when you turn the coil off suddenly that magnetic field collapses and induces (getting technical here) a back EMF (electromotive force aka voltage) in the wiring. Since it by definition can't go through the now open switch it goes through the ground wire. Which is why you want the switches on the posivite side of the circuit (among other reasons which are more important BTW). Turning on a coil does not induce the spikes BTW as there is no magnetic field when the switch is closed.
I'm not thinking that having spikes on a ground is going to cause much trouble if the spike has to travel down the ground wire to the buss and back up another wire to the effected device.
If the buss is close to the batteries they will help absorb it and the rest of the ground circuit will also help (though wire has a pretty low capacatance so don't plan on this as a strategy)
I'd be interested in hearing from anybody who has had a verified voltage spike that caused a problem. I've never expireenced one but my system is pretty standard
 
Feb 6, 1998
11,665
Canadian Sailcraft 36T Casco Bay, ME
Question
Have you considered the starter under load? I'm thinking here of the make/break of the starter brushes with high current loads. I'm thinking because the starter is an inductive load it has potential to provide higher spikes as the magnetic field collapses when the brushes break.
The reason I'm thinking of this is because in a household AC power failure, that frequently happens as the transformer fields and fields of any connected inductors collapse. This is why power failure frequently plays havoc with semiconductors. It may also be interesting to know not just the amplitude, but the duration also. (Sorry, I do know it requires a different test than the bench)

Ken
All I have ever measured is starters, windlasses, winches, fridge clutches etc. under load. This was the first time I bolted a starter to the bench unloaded. It was interesting to see the in-rush not change at all (in-rush happens before the engine has begun to turn) and the unloaded average current draw of 100A on a .8kW starter.

I've not noted transients any differently on boats while actually starting an engine and releasing the start button than I have on the test bench.. The brushes/motor do create noise on the scope, heck even a stable unloaded battery has noise, but I've not noted potentially damaging transients that approach that of the field collapse of a solenoid.

The transient duration can be captured on my o-scope but after talking with the folks at Fluke, Tektronix (my old scope was a Tek), Littelfuse (they make TVS diodes), Garmin, Raymarine etc.etc. they were not concerned with any duration shorter than what the Fluke 289 can capture transients at, which is 0.00025 seconds. I have successfully used my Fluke 289 to fight on behalf of customers for warranty denial too. I was never asked to test it with my scope once they knew I had been using the 289. Any time I come across a random death of electronics the first thing I do is test the boat for transients. It is rare that I seen in excess of 14-16V when the system is well wired but not extremely rare to poke into the low 20's when folks have haphazardly tapped a plotter or sounder or VHF etc. into the engine block for DC negative or engine circuit for DC positive.

I have only had a few boats where I had to install TVS diodes and these were very large solenoids on large windlasses causing the issue. Even then it was about 26V on a 12V system. Not a deal killer but worth throwing a TVS diode on..
 
Apr 24, 2006
868
Aloha 32 Toronto, Lake Ontario
A friend of ours had an interesting solution on one of his boats. Instead of multiple battery banks and "specific wiring" he had a single bank. High amperage loads with "inductive kick" were run from this bank but it also powered a dc to dc inverter (12 v in 12 v out) for everything else (electronics, lighting, etc ). I believe the inverter was a 50 amp unit and "high 90%" efficient. I don't know where he obtained the unit but it was pretty much immune to anything you could throw at it - electrical and physical.

Chris
 

walt

.
Jun 1, 2007
3,511
Macgregor 26S Hobie TI Ridgway Colorado
FYI, while you have that set up, and if you are interested in accurate measurements, you should try one more test.

Your scope comes with a probe with a SHORT ground lead and using this is the only accurate way to measure fast transients. For using the scope probe, you must keep that ground wire short - dont extend it and also put the tip of the scope probe on the node you want to measure. Ie, also dont use a wire to extend the scope probe tip. The scope probe provides a very high impedance at its tip and has a controlled transmission line back the the instrument measurement input. With your high end DVM, those long probe wires work just fine for lower frequency signals say in the Khz range but when you get into the Mhz range, you will get transmission line distortions and effects from the inductance of the measurement wires so the signal the instrument sees can be different than the signal at the measurement node.

What would be interesting is to do the measurement like you did with the DVM instrument (in the attached picture) and get the peak spike amplitude.

Then use the scope on the same measurement node and only use the ground wire that comes with the probe on the ground side and the probe measurement node on the positive side - also no extra wire extensions. You may want to try the scope on both AC and DC settings and set the scope up to capture a single sweep with the trigger set to go off on the transient. Adjust the time scale to match the time of the transient.

If the transients you are measuring are "slow", you might get similar peak measurements. However, if the transients are fast, you likely will get a significant difference and the scope (assuming the probe has been calibrated) will be the accurate measurement.

edit.
Garmin, Raymarine etc.etc. they were not concerned with any duration shorter than what the Fluke 289 can capture transients at, which is 0.00025 seconds. I have successfully used my Fluke 289 to fight on behalf of customers for warranty denial too. I was never asked to test it with my scope once they knew I had been using the 289
I also wanted to point out that this exercise is mostly just to see.. Damage is caused by energey deposited which is power * time. That DVM is likely band limited on its input (.00025 is 250 usec which is not a fast signal) and since the damaging transient last a longer time (since energy is power*time), the DVM instrument likely is adequate. In an engineering lab.. the scope would be the preferred way to capture the amplitude and it would also capture the actual time frame - also important for damage.

scope.jpg
 
Last edited:
Oct 24, 2010
2,405
Hunter 30 Everett, WA
Thanks Main Sail. That answers the question. I hadn't considered dragging my scope to the boat. Now I don't think I'll bother.
Ken