TheSamba.com :: View topic - DIAGRAM: Starter & alternator circuit “The Voltage Drop Test"

Sat Nov 29, 2025 10:15 pm

Tryna keep vans off the flatbeds......

If you have a little understanding of electricity, or friends who do, here's a diagram for your Samba Vanagon files.

Here's a how-to to qualify the high-power legs of your Vanagon circuit. These are responsible for starter cranking speed, battery charging speed, headlight brightness, radiator fan speed.

The diagram shows six locations to measure voltages.
The tools required are a volt meter, and a 100 amp load tester.
The load tester is sold as a battery tester, and they cost only about $30.

1) Load-test the battery. 100Amps is not a definitive qualifier. You really want to test it with about 325 amps. But if it struggles with 100A, perhaps your battery is the problem.

Negative legs of the circuit are much easier and safer to measure voltages.
For tests 2 thru 6 you may need a helper to push the 100A button (for 5 seconds) while you read the "voltage drops".

2) Measure the voltage drop between the battery post and the steel van chassis. It's common to have losses here. 0.25v is acceptable. Maybe even towards 0.5v. If you find 1.0v your starter is slowing. My van has 0.07v drop here.

3) Measure the voltage drop between the chassis (somewhere shiny) and a starter mounting bolt. It's common to have losses here. 0.25v is acceptable. Maybe even towards 0.5v. If you find 1.0v, your starter is slowing. If you find 2 volts it's really slow, and the electricity is ruining your internal drivetrain parts. Delete that VW ground and replace it with a 4g cable from the starter mounting bolt directly to the chassis.

4) You have to be very careful measuring positive posts. Measure the voltage drop between the battery POSITIVE post and the starter positive post. This is not particularly easy. You can use jumper-cables to extend your multimeter leads. 0.25v is acceptable.

5) Measure the voltage drop between the starter positive post and the alternator positive post. This is not particularly easy. You can use jumper-cables to extend your multimeter leads. 0.25v is acceptable.

6) Measure the voltage drop between the alternater case and the van chassis. Your van may have an engine ground, and some of the grounding may come from the faraway transmission ground. 0.25v is acceptable.

The connections degrade over time, especially the five corrodable connections between the starter and the transaxle forward ground..
As they corrode, the starter slows, headlights dim,
it won't re-start at the gas station....
.....flatbed rides loom large in your future.
These are difficult to maintain or renew. Best to bypass all of them with a “$20 mod”

You may be inclined to purchase a new starter that's perhaps not required (yet!).
TEST the circuit FIRST.
The new starter may behave the same as the old one if your degraded circuit is the actual problem.
Sometimes just unbolting and rebolting your starter is all that's needed to renew the starter's electrical connection and restore starter operation. But there are other connections that may need to be restored as well.

The connections across the engine to the alternator is another daisy-chain of corrodable materials.
For that reason it's best to add a direct ground from the starter case to the chassis,
AND
add a direct ground from the starter mounting bolt to the chassis.
See this Diagram regarding grounds modification: https://www.thesamba.com/vw/gallery/pix/2534409.jpg
A couple $20 ground cables may be the cheapest and most effective mod you can do.

Sun Nov 30, 2025 10:14 am

I see some issues with your Rev A diagram and the location of the Load Tester negative lead position.

If you don't have a DC clamp meter you should get one, making sure it does DC amps with the clamp.
Then use it to verify that the assumed amount of amps are flowing where you think they are.

Mark

Sun Nov 30, 2025 10:35 am

Deleted.

Sun Nov 30, 2025 10:55 am

Losing 0.07V on your positive cable between the battery and the starter is very acceptable, while losing 0.4 volts on your ground circuit at only 100 amps seems quite high to me.

Sun Nov 30, 2025 5:45 pm

crazyvwvanman wrote: I see some issues with your Rev A diagram and the location of the Load Tester negative lead position.
Please explain explicitly. Do you mean that theres no actual chassis frame there?
Yeah I agree that could add confusion.
Can't just "assume" every person knows how to choose a good chassis ground.
I would like this diagram to be as correct and easy to use as possible.

Probably need to tell the user exactly where to clamp the load tester's neg clamp.
Where would you suggest?
The 10mm ground bolt on the left side of the engine compartment seems like a good place but hard to show in a diagram.

crazyvwvanman wrote: If you don't have a DC clamp meter you should get one, making sure it does DC amps with the clamp.
Then use it to verify that the assumed amount of amps are flowing where you think they are.

Yep i have that tool. I didn’t think to verify actually amps for this diagram but I will.
It would be useful to know if amperage is dividing off somewhere strange.
Such as leaking out across the CV axles, pitting wheel bearings & exiting via parking brake cables.

Sun Nov 30, 2025 6:31 pm

Thanks for doing this. I appreciate all these sorts of tutorials and save them on paper in my van.

Sun Nov 30, 2025 6:33 pm

Wildthings wrote: Losing 0.07V on your positive cable between the battery and the starter is very acceptable, while losing 0.4 volts on your ground circuit at only 100 amps seems quite high to me.

The voltage drops shown (2% and less) are all acceptable in this diagram.
No starting problems, no charging problems.

Going around the starter circuit:
battery output starts at 11.5v
drops 0.070 across the long positive cable
drops 0.234 across the corroded tranny
drops .150 across the battery ground near the passenger seat

leaving 11.04volts to crank the starter.

Note that the drops are additive in a circuit thus the drop from the starter (.234v) to the battery (.150) is .234+.150=~0.4v

Agreed this groundpath approaches the 0.5 volt "fix the danged thing" threshold.

Thanks Bob, glad you find it useful.
I hope it keeps some vans off flatbeds.

Sun Nov 30, 2025 6:39 pm

Sodo wrote:
[]
drops 0.070 across the long positive cable
[]
drops .150 across the battery ground near the passenger seat
That battery ground needs looking at! It should be *less* than the long positive battery cable to the starter...

Sun Nov 30, 2025 6:45 pm

fxr wrote: Sodo wrote:
[]
drops 0.070 across the long positive cable
[]
drops .150 across the battery ground near the passenger seat
That battery ground needs looking at! It should be *less* than the long positive battery cable to the starter...

Agreed.
I re-crimped the terminal ends and now it's 0.050v.
But that's battery post to chassis. I did not check ring-to-ring.

And I don't remember if the 0.070v was ring to ring or "battery post" to "starter post".

But in any case these tiny vDrops are "piddly" :roll:
My van cranks hard.
If I had a van to test with an actual "starting problem" we would see bigger vDrops.

Heres a video of a voltage drop test on the ground path of a van that has a direct starter ground.
And voltage drop test shows improvement on a 're-vitalized' battery ground cable.

I would have preferred to do a video test revealing a "bad ground" but I don't have a "slow-starter van" handy.
Video shows finding 0.150v drop on the battery-to-chassis cable. 0.150v is acceptable at 1.3% but iy IS a very short cable, thus not much excuse.
So the video goes on to show rectifying the 'non-problem' (to 0.050v......) :wink:
But you get the idea.

Keep in mind that resistances always get worse over time,
so rectifying any degrading connections when you've located one, is wise.

Sun Nov 30, 2025 6:47 pm

Sodo wrote:
But in any case these vDrops are "piddly" :roll: .
True, and are just "illustrative" anyway. ;)

Mon Dec 01, 2025 12:58 am

Sodo wrote: Wildthings wrote: Losing 0.07V on your positive cable between the battery and the starter is very acceptable, while losing 0.4 volts on your ground circuit at only 100 amps seems quite high to me.

The voltage drops shown (2% and less) are all acceptable in this diagram.
No starting problems, no charging problems.

Going around the starter circuit:
battery output starts at 11.5v
drops 0.070 across the long positive cable
drops 0.234 across the corroded tranny
drops .150 across the battery ground near the passenger seat

leaving 11.04volts to crank the starter.

Note that the drops are additive in a circuit thus the drop from the starter (.234v) to the battery (.150) is .234+.150=~0.4v

Agreed this groundpath approaches the 0.5 volt "fix the danged thing" threshold.

Thanks Bob, glad you find it useful.
I hope it keeps some vans off flatbeds.

When the amperage hits 250 amps on a cold morning you are going to be seeing a much higher voltage lost than 0.40V. As fxr mentions you shouldn't be seeing more on the ground side than the positive side.

Mon Dec 01, 2025 8:17 am

fxr wrote: Sodo wrote:
But in any case these vDrops are "piddly" :roll: .
True, and are just "illustrative" anyway. ;)

"These numbers" in this diagram, are actually a composite of two different vans. So.... real numbers from "somewhere" but NOT one "real van".
Discussion using these numbers is worthwhile to show owners how to get their own vDrops.
Then post up the real numbers and we'll all analyze them.

Wildthings wrote: Sodo wrote: Wildthings wrote: Losing 0.07V on your positive cable between the battery and the starter is very acceptable, while losing 0.4 volts on your ground circuit at only 100 amps seems quite high to me.

The voltage drops shown (2% and less) are all acceptable in this diagram.
No starting problems, no charging problems.

Going around the starter circuit:
battery output starts at 11.5v
drops 0.070 across the long positive cable
drops 0.234 across the corroded tranny
drops .150 across the battery ground near the passenger seat

leaving 11.04volts to crank the starter.

Note that the drops are additive in a circuit thus the drop from the starter (.234v) to the battery (.150) is .234+.150=~0.4v

When the amperage hits 250 amps on a cold morning you are going to be seeing a much higher voltage lost than 0.40V.

Certainly if the amperage hits 250A the voltage drop will increase accordingly.
That's justification for the anal mofo to minimize every voltage drop. :wink:

I'm curious what the a reasonable (cold) max starter amperage is.
Anal mofos never-U-mind, if someone suggested 250A.... U gotta use 250A !

We need a Minnesotan, Alaskan, perhaps a Kootenaian(?) to clamp on a lobsterclaw a few times this winter...
and get us some starter amperages at a few low temperatures:
-20°F
0°F
20°F
I'm gonna start the speculation.... that -20°F is only 125A....
( I have NO basis whatsoever for this wild guess ) :wink:

Wildthings wrote: As fxr mentions you shouldn't be seeing more on the ground side than the positive side.

If you watch the YouTube vid, I show reducing the .150v to 0.050v by re-crimping the terminal ends of the (very short) battery cable.
The .234v from the starter to the chassis is very likely the sum of multiple connection resistances (corrosion can change day-to-day).

The .070 is pretty small for 11 feet but it's a real number (for 90 amps load at the end of the Alt B+ wire).
Sorry I don't recall if the 0.070v was strand to strand, terminal to terminal, or battery post to starter post.

Everyone who embarks down this path of learning "the voltage drop test" needs to (eventually) understand the ABOVE sentence.
If you skip over any "connections" you may miss the location of your vDrop.

Thu Dec 04, 2025 10:44 pm

crazyvwvanman wrote: If you don't have a DC clamp meter you should get one, making sure it does DC amps with the clamp.
Then use it to verify that the assumed amount of amps are flowing where you think they are.

100% agreed this is another useful tool to have and enjoy
once you have started learning the Vdrop test.

And they don't break the bank :wink: .

=================

( :roll: deleted my own comment :roll: that risked blowing this technical post OFF-TOPIC
keep in mind
....there may be vans on flatbeds rightnow due to starter trouble......
....gotta keep eyes on the prize.....)

Thu Dec 04, 2025 11:52 pm

All of this loaded testing is good stuff, but while you're at it and have all the tools out, check other high-power consumers. Defroster, HVAC fan, headlights, radiator fan, ECT..
The more testing you do, the better and faster you will become.

BK

Fri Dec 05, 2025 1:03 pm

1988M5 wrote: while you're at it and have all the tools out, check other high-power consumers. Defroster, HVAC fan, headlights, radiator fan
Yes, but don't test those other items with this 100A load.
Those other circuits you need to find the appropriate 30A 40A, 50A load test devices.

This 100A load is appropriate for testing Starter and Alternator circuits because they are both 90A components and are wired for 90A.

I need better qualified assistance to design the test.
Or maybe someone else should design the test.
Here's a question for those who are good with electricity.

It seems a valid direction, considering....
~ 100amps load tester tool is appx equal to the starter-motor load on a Vanagon.
~ It's more consistent than cranking the starter (and easier too)
~ At $120/hour shop time, the $30 tool costs 15 minutes, and you OWN it.
~ Testing on the B+ alternator post is EASY.
~ And you already own a voltmeter, right?

====================

We DO need to help our Vanagon brethren/sistren stay off the flatbed.

And many would appreciate a simple test
to avoid buying new batteries, starters.

And the "current" (heh) conundrum....why is a diesel starter for 22:1 needed to crank an older 9.5:1 motor (perhaps with declining compression)?

As crazyvwvanman noted, the composite ground-path now includes the 100A load-tester,
confusing any measurements observed on the ground side.

But measurements on the positive side should be valid, right?

OK then, Crazyvwvanman, 1988M5, fxr, DanHoug, E1, and others... ?

1) can GOOD information come from this test? How would you get the useful info??

2) Is there anything "faulty" about this test? Can the error be ignored or does the fault sink the whole test?

3) Should the diagram be changed? How? The intended viewer is the "normal Van owner." Not the "normal electrical engineer van owner".

Fri Dec 05, 2025 2:32 pm

I don't really have time for this today but I'll steal a few minutes.
Yes, the voltage drop test is valuable, with some fine print reserved.

The DC Clamp Meter seems an almost essential addition.
I bought one recommended here by someone, Amazon.

I have 2 of the HF Load Testers, a 2 year old one and a 2 day old one, exact same model number.
The older one read 89 amps connected at the battery posts of my 87.
The new one read 95 amps there, back and forth multiple tries with each.

Maybe more time tomorrow.

Mark

Sodo wrote: ....

OK then, Crazyvwvanman, 1988M5, fxr, E1, and others... ?

1) can GOOD information come from this test? How would you get the useful info??

2) Is there anything "faulty" about this test? Can the error be ignored or does the fault sink the whole test?

3) Should the diagram be changed? How? The intended viewer is the "normal Van owner." Not the "normal electrical engineer van owner".

Sat Dec 06, 2025 6:17 pm

After I compared the different readings from my 2 HF Battery Load Testers, I wanted more data.
I took them inside where I had a new spare 100ah LiFePo4 battery waiting for a project.
First I tested it with my 2 year old one, reading the amps from my DC Clamp Meter.
This time it read 94 amps, not 89 like with my starter battery in my Westy.

Next tested with the new Load Tester just purchased, which gave 95 amps previously.
The DC Clamp Meter display jumped around before suddenly showing amps dropping to 0, while sounds came from the new Load Tester along with some flashes of sparks inside it near the load-on switch I was pushing on.
So I re-seated the cables and tried again, same results.
I was sure I had fried something but no.
The old one still worked with the LiFePo4 and the new one still worked back outside on my van starter battery.

What happened? I'm not sure.
One guess is that the BMS in the LiFePo4 objected to the new unit's 100+ Load Tester amp draw, even as brief as it must have been.

Clearly the HF so called 100amp Battery Load Tester isn't a precise and calibrated device.
Also it isn't meant to test LiFePo4.

I'm going to have to think about this more.

Mark

Sat Dec 06, 2025 8:01 pm

crazyvwvanman wrote:
Clearly the HF so called 100amp Battery Load Tester isn't a precise and calibrated device.

For $20 (I'm guessing at the HF load tester that you bought) wouldn't a buyer be happy with +/- 10% accuracy? For these purposes it seems the testjust needs to measure voltage drop with decent load, something greater than the high beams and less than a snowplow.

Sun Dec 07, 2025 9:03 am

If that 2nd tester is not consistent with a 12.6v start-battery either.... I'd look for a bad connection (?). It could be repairable in a fraction of the time it takes to drive to Harbor Freight. I can't imagine what could go wrong in such a simple machine.
The good/bad scale is of course depicts a 12.6v battery not 13.3v.

Incidentally if you had a significantly deteriorated ground path with 0.8 volts of "drops",
a 13.3v start battery should crank the engine as fast as a "perfect" system.
Those little Lithium "pocket jump starters" may confuse diagnosis if you don't consider their higher beginning voltage.

anyway back to the Voltage drop test....
the Load Tester's contribution to the test ....
It's easier, quieter and more repeatable than using the vehicle starter load.
And you can load specific legs of the circuit ( :oops: if they're discrete legs :oops: ).

Sun Dec 07, 2025 9:33 am

I agree, for what they are built to do for only $20 I'm satisfied.
Now though we are using them in a different way so variations in their performance may be more significant.
Does it make sense to compare voltage drop measurements down to the 1/1000 volt when the test setups can vary so much just in the Load Tester?

I'm planning to get another one today for this project, since I'm going to HF anyway.
I want to build a test box using 2 of them with matched performance, in part so I can do more tests in a shorter time without burning them out.

Mark

jimf909 wrote: ....
For $20 (I'm guessing at the HF load tester that you bought) wouldn't a buyer be happy with +/- 10% accuracy? For these purposes it seems the test just needs to measure voltage drop with decent load, .....