How to deal with transient voltage drops?

Thread Starter

kuksul08

Joined Jan 17, 2010
4
Hi everyone,

On my vehicle the electrical system experiences transient voltage drops every time the brakes are applied due to inrush current in the ABS pump module. The effect is a dimming/flicker in the headlights and ambient lighting.

I was able to measure this drop at the battery terminals both with the engine off and with the engine running. The voltage drop is up to 1.5V or around 10% of nominal 14V.




I am trying to think of ways to alleviate this. I believe the ABS module does not have sufficient capacitance, so it "shocks" the rest of the electrical system by momentarily dropping the voltage. Would it be possible to place a capacitor of some sort across the battery terminals to help provide energy in these cases? Or maybe some other approach? I know that typically you want to place caps/inductors as close as possible to the device that needs it, but putting something at the battery would be far easier practically.

Thanks :)
 

Alec_t

Joined Sep 17, 2013
12,812
The battery behaves somewhat as a big capacitor anyway, so I think you'd have to add a whopping external cap to keep the volts up when the battery has a major load such as headlights.
 

Thread Starter

kuksul08

Joined Jan 17, 2010
4
The battery is brand new, less than one month old. I also tested another battery that is relatively new, with the exact same behavior.

Does a battery act as a capacitor? I was under the impression that internal resistance will prevent it from delivering energy quickly enough. It can deliver high currents to the starter motor, but perhaps not transients in the higher frequency range? I could be totally wrong. I just know that filtering caps are typically used in an array ranging with varying magnitudes to cover a wider frequency response range.
 

crutschow

Joined Mar 14, 2008
29,800
Such a large voltage drop indicates a high transient current as the typical battery impedance around 50 milliohms.
That translates to a transient peak current of 1.5 / .05 = 30A
Since the pulse is sort of triangular, the charge transferred would be about 30A/2 * 100ms = 1.5 coulombs.
So that would take a several Farad capacitor (huge) to reduce that transient voltage significantly below that observed.

How old is this vehicle and has it always acted that way?
 

nsaspook

Joined Aug 27, 2009
9,733
A small car battery is equal to about 15000F.

Lots of equivalent F there to handle loads usually but it's not instantaneous. IF the car has always dimmed lights from ABS then it's a minor design issue that might be better cured with a high quality AGM battery. (lower internal resistance for the same size battery) If it's something that that you've noticed recently getting worse then it might be electrical (charging system including wiring and battery grounds) or mechanical like a failing ABS pump problem that a load decoupling capacitor will only band-aid if it works at all.
 
Last edited:

gerty

Joined Aug 30, 2007
1,302
Since the battery checks ok I would look at the alternator rating, making sure a smaller(amps) unit wasn't put in place of the correct one. Next I would check for an open diode in the diode trio. What is the battery voltage at idle and at approx 1,500 rpm ?
 

Thread Starter

kuksul08

Joined Jan 17, 2010
4
If I'm reading your 'scope trace correctly, your big transient lasts about 100mS. That's hardly in the higher frequency range :).
True. That's a pretty long time all things considered. Long enough to see it with the naked eye.

Such a large voltage drop indicates a high transient current as the typical battery impedance around 50 milliohms.
That translates to a transient peak current of 1.5 / .05 = 30A
Since the pulse is sort of triangular, the charge transferred would be about 30A/2 * 100ms = 1.5 coulombs.
So that would take a several Farad capacitor (huge) to reduce that transient voltage significantly below that observed.

How old is this vehicle and has it always acted that way?
It is a brand new BMW motorcycle, and always done this. Based on feedback from other owners, theirs do this as well. I am pretty sure it's a design issue.

A small car battery is equal to about 15000F.

Lots of equivalent F there to handle loads usually but it's not instantaneous. IF the car has always dimmed lights from ABS then it's a minor design issue that might be better cured with a high quality AGM battery. (lower internal resistance for the same size battery) If it's something that that you've noticed recently getting worse then it might be electrical (charging system including wiring and battery grounds) or mechanical like a failing ABS pump problem that a load decoupling capacitor will only band-aid if it works at all.
The battery it currently has is a fairly large AGM. It definitely hasn't gotten worse, but been the same since brand new.

Since the battery checks ok I would look at the alternator rating, making sure a smaller(amps) unit wasn't put in place of the correct one. Next I would check for an open diode in the diode trio. What is the battery voltage at idle and at approx 1,500 rpm ?
I do find it odd that the peak voltage drop is worse when the engine is running vs the engine off. The battery voltage with the engine off is ~13, at idle is ~14.4, and revved up is ~14.6V. I will need to double check these numbers, but when I last checked it all looked to be in working order.


Thanks all for the input. As suggested, I'm starting to think this isn't a defect, but a design flaw. Suppose I was the engineer trying to fix this on future models. How would I go about preventing the ABS pump current sink from affecting the other systems of the electronics?
 

nsaspook

Joined Aug 27, 2009
9,733
I do find it odd that the peak voltage drop is worse when the engine is running vs the engine off. The battery voltage with the engine off is ~13, at idle is ~14.4, and revved up is ~14.6V. I will need to double check these numbers, but when I last checked it all looked to be in working order.


Thanks all for the input. As suggested, I'm starting to think this isn't a defect, but a design flaw. Suppose I was the engineer trying to fix this on future models. How would I go about preventing the ABS pump current sink from affecting the other systems of the electronics?
They might have used a small alt for the battery charging system. At full ABS load the drop from full charging voltage to the sustained battery voltage would be normal if the alt can't supply the needed current at the charging voltage.
 

crutschow

Joined Mar 14, 2008
29,800
I do find it odd that the peak voltage drop is worse when the engine is running vs the engine off.
Not so odd.
When the engine is running the power is being supplied by the alternator and the battery voltage is above its normal discharge voltage.
When the large pulse of load current is generated the voltage has to drop until the battery can provide the current.
Suppose I was the engineer trying to fix this on future models. How would I go about preventing the ABS pump current sink from affecting the other systems of the electronics?
You would need some method of reducing the peak current draw and that likely means a redesign of the pump.
A small value resistor (less than an ohm) in series with the pump may work, if it doesn't reduce the voltage too much.
 

Thread Starter

kuksul08

Joined Jan 17, 2010
4
Huh... I never thought about the alternator-battery voltage differential! Always something new to learn :)

I am definitely wary of reducing the power available to the ABS pump, because functioning brakes are pretty darn important.

Perhaps the best solution here is more of a band-aid rather than systematic approach? If I put a capacitor or some similar device directly on the headlight bulbs and ambient (dashboard) lighting, so their their immediate power sources were not interrupted, could this be easier and require more manageable component sizes?
 

crutschow

Joined Mar 14, 2008
29,800
Perhaps the best solution here is more of a band-aid rather than systematic approach? If I put a capacitor or some similar device directly on the headlight bulbs and ambient (dashboard) lighting, so their their immediate power sources were not interrupted
You'd need to put some impedance in the line to the lighting otherwise the current surge will just deplete the capacitors and still cause dimming of the lights.

A Schottky rectifier in series with the wire to the lights with a large capacitor to ground will prevent this back draw.
But you will need a very large capacitor for the headlights.
 
Top