Voltage fluctuations - Megasquirt and Lipo battery

[JohnJL]

I am finishing up a megasquirt installation on a 3l and am seeing weird voltage fluctuations in my logs. I am working with a tuner remotely and he noticed this before me. I am not sure if it is effecting the MS corrections but in any case it seems something that should be corrected?

I am using a batterytender LiPo battery, and the megasquirt is grounded directly to the battery. It was grounded through the chassis previously but this change did not fix the issue.



Here is a log of a cold warmup. I have circled the voltage fluctuations in a white oval. Voltage swings from 13.8 to 12.4 and back again continuously. The alternator is a newish Valeo with internal regulator. It has maybe 5,000 miles on it without issue.

I wonder if the lightweight battery doesnt act like a voltage buffer, I thought I heard normally batteries also serve to level out voltage fluctuations. Or a bad altertator regulator? Something else?

Appreciate any help,

John

[Bob Kontak]

V-reg at first blush.

[3rd_gear_Ted]

Optimal battery max float voltage is 13.8 with max charge voltage being 14.2

The cycling between 12.4 and 13.8 is about the same range as a voltage regulator diode duty cycle.

[Jim Smolka]

Start with the basics...

Check grounds., clean and tighten
If still having issues, have alternator voltage check with another O'Scope to see if the issue really is with the alternator - regulator.

Also, if its the battery, swap with another battery.

Just my $0.015 worth...

[JohnJL]

Thanks guys. Let me drill down on these....firstly, is that fluctuation abnormal and a problem?

Originally Posted by Bob Kontak

V-reg at first blush.

Its an internally regulated alternator, so a pain to switch out. Hopefully something else but you could be right.I have a spare car I can switch it in but hoping not to have to do that.

Originally Posted by 3rd_gear_Ted

Optimal battery max float voltage is 13.8 with max charge voltage being 14.2.


The cycling between 12.4 and 13.8 is about the same range as a voltage regulator diode duty cycle.

WHen you say in the same range as a voltage regulator diode duty cycle, do you mean under normal conditions? THis is where I was going with my original question about whether the charging system is working as normal, its just that the little lipo battery doesnt smooth out the voltage fluctuations as well as a traditional big lead-acid battery?

~13.5v is what I saw in my logs in the car before rebuild. My Adaptronic log showed a solid 13.4-13.5 when running.

Originally Posted by Jim Smolka

Start with the basics...

Check grounds., clean and tighten
If still having issues, have alternator voltage check with another O'Scope to see if the issue really is with the alternator - regulator.

Also, if its the battery, swap with another battery.

Just my $0.015 worth...

Unfortunately I dont have an oscilliscope. Swapping a battery is going to be a pain. I really buried the battery down deep in the smuggler's hatch with the master cut-off on a short lead next to it. I may have to try and jumper the battery leads to a big battery sitting in the front.

[Jim Smolka]

What type of wire did you use?
A lot of the Auto wire now is aluminum with copper
Clad... junk IMHO for 911s

I like marine grade copper wire that has been tinned

[JohnJL]

Originally Posted by Jim Smolka

What type of wire did you use?
A lot of the Auto wire now is aluminum with copper
Clad... junk IMHO for 911s

I like marine grade copper wire that has been tinned

This is all mil-spec wiring from a friend who owns Interconnect Systems in Sydney, a ballistics supplier.

Like this

[Nux]

Interesting!

Where do you measure your voltage? Sure its not interference from the engine bay (CDI ect.)? I would probably try the easy way out and simply install a capacitor for 2 bucks......

[dad911]

Originally Posted by JohnJL

I am finishing up a megasquirt installation on a 3l and am seeing weird voltage fluctuations in my logs. I am working with a tuner remotely and he noticed this before me.
......
I wonder if the lightweight battery doesnt act like a voltage buffer, I thought I heard normally batteries also serve to level out voltage fluctuations. Or a bad altertator regulator? Something else?

Appreciate any help,

John

Batteries do smooth the voltage. They are really a very large capacitor/filter.

Megasquirt in back (by engine) battery in front. You may be measuring noise and voltage drop in the cables.

Is the tuner you are working with used to 911 or is his data based on front engine cars, where the megasquirt is likely connected with a very short cable to the battery?

[72911]

Here is a quote from the Megamanual on grounding your sensors. They should all go to the same place (engine block) not the battery.

"The main grounds from pins #8, 9, 10, 11,& 18 go to one spot on the engine block. Do not ground them at physically separated locations, and do not use a single fat wire for this. Instead run separate wires from the pins all the way to the ground spot.
Pin #19 is the sensor ground. If you have two wire CLT and IAT sensors, their grounds (and the TPS ground) should run back to the DB37's pin #19 to reduce the potential for noise in the sensor signals."

[JohnJL]

Originally Posted by Nux

Interesting!

Where do you measure your voltage? Sure its not interference from the engine bay (CDI ect.)? I would probably try the easy way out and simply install a capacitor for 2 bucks......

THe voltage measurement is as observed by Megasquirt. It internally logs voltage which is what is fluctuating in the log I posted at the top.

[JohnJL]

Originally Posted by dad911

Batteries do smooth the voltage. They are really a very large capacitor/filter.

Megasquirt in back (by engine) battery in front. You may be measuring noise and voltage drop in the cables.

Is the tuner you are working with used to 911 or is his data based on front engine cars, where the megasquirt is likely connected with a very short cable to the battery?

Could be, though all the cables are new and I get a solid 13.5v when measured at the fuse block behind the seat where the megasquirt and MSD boxes are mounted. The tuner I am working with is Jeff (I dont know if its OK to use his full name) but he did some cleanup/fininshing on the megasquirt box for me and mentions some other 911s including one in GA this weekend also with a JB Racing 12 plug distributor like mine...thats pretty specific!

[JohnJL]

Originally Posted by 72911

Here is a quote from the Megamanual on grounding your sensors. They should all go to the same place (engine block) not the battery.

"The main grounds from pins #8, 9, 10, 11,& 18 go to one spot on the engine block. Do not ground them at physically separated locations, and do not use a single fat wire for this. Instead run separate wires from the pins all the way to the ground spot.
Pin #19 is the sensor ground. If you have two wire CLT and IAT sensors, their grounds (and the TPS ground) should run back to the DB37's pin #19 to reduce the potential for noise in the sensor signals."

Thanks for this, I originally ran the engine block ground cable to the firewall-side of a groundbolt. On the cabin-side I ran each of those ground wired separately as instructed to the same groundbolt. I get 0ohm resistance from the engine to the groundbolt. My tuner then suggested also running a ground from the battery to that same groundbolt, which I did. None of these changes made a difference in the voltage fluctuations.


Then I switched in a standard lead-acid battery...no change in the voltage fluctuations.
Then I switched in another fan/alternator assembly...no change in the voltage fluctuations...

Anyone see strange voltage fluctuations on megasquirt? I seem to have eliminated the battery and the alternator as sources of the voltage swings. BTW this whole time my fluke multimeter shows a solid 13.6v while running. Its only megasquirt logs that show that weirf voltage swinging.

[JohnJL]

Originally Posted by Nux

Interesting!

Where do you measure your voltage? Sure its not interference from the engine bay (CDI ect.)? I would probably try the easy way out and simply install a capacitor for 2 bucks......

Got a part number?

[cmonref]

JohnJL: I use a 4700 uOhm capacitor from a 944 electronic ignition system. Don't know the PN tho'. It is wired from the MS's power line to engine ground.

Oops -- 4700 uF (micro Farads), not uOhms.

[JohnJL]

Hi all,
I switched in another alternator/fan and another battery.

Can I ask again from someone who has a high-speed logger or o'scope if that fluctuation from 12.5 to 13.9 is normal?

cmonref, did you wind up with that solution tracking down a problem? Why did you put that in?


THanks

[cmonref]

did you wind up with that solution tracking down a problem? Why did you put that in?

JohnJL: It was part of the original planning after reading the MS documentation. There was a recommendation that a capacitor put in parallel with the power supply would dampen, or eliminate, stray voltages in the power supply. It acts on the stray voltages but is inert if there are no stray voltages. And Porsche put it into their 944 EFI systems as a regular component, so I just copied their idea. It has screws for electrical connectors and is easy to install. Mine is mounted on the left side of the engine support bar, near the power supply that I incorporated into the design. I have no idea if it works or not -- I have not had any power problem -- but the theory is solid and it just might aid you in solving your problem.

Oops -- 4700 uFarads, not uOhms. Sorry.

[timmy2]

Bitz kit EFI instructions suggest using a 4700uf 35 volt D.C. Electrolytic capacitor on all switched power going to the ECU and Wideband 02 if using one.

[targa44]

Would anybody be willing to throw in a picture or diagram of the capacitor installs? ... just connect in-line?

thanks,

[Nux]

Originally Posted by JohnJL

Hi all,

cmonref, did you wind up with that solution tracking down a problem? Why did you put that in?

THanks

I had a slight oscillating idle and some uneven running below 3000rpm after instaling the Bitz EFI kit. This was almost completely eliminated after installing the 4700UF capacitor to the wideband sensor as recommended.