What causes the voltage drop between the 12V and the jump post?

Why are the voltages at the jump post and the battery not the same?

On our 2007 there is a roughly 0.2V drop between the voltage measured at the battery and the voltage measured under the hood at the jump post. This is with the car off, all the doors closed, no lights on, the car not having been run recently. In "ACC" mode (press Start only once, foot off the brake) the MFD voltage is the same as the jump post voltage (except with less accuracy, ie 12.1V on the MFD vs. 12.13V on a multimeter).

In this test there is only the background current (around 20mA) coming out of the battery. That isn't enough to drop the battery voltage. Even if we were to assume a full 1 Ohm of resistance along that path (which would be pretty surprising if true), it would still be only a 20mV drop, not 200 mV. I don't believe there are any electronics in that circuit (like a rectifier), and the voltage drop across the big fuse should be negligible.

Since the MFD and jump post measure the same, whatever causes this is between both of those locations and the positive battery post. Hmm, or maybe there is an anomalously high resistance in the path to ground? I recall long ago having tried to measure the voltage drop from the negative post to a bolt on the body near the battery, not the grounding point, and it was too small to measure. I have not tried to measure the voltage drop from the bolt head in the engine compartment I use for the jump post measurement to the ground point in the trunk. Again, there is only 20 mA current in play, so resistance seems unlikely to be the problem.

 

Exactly. The only way that much of a voltage drop makes sense with so little current is if there is some sort of junction between dissimilar materials. A Schottky diode can have a forward voltage drop in that range, but AFAIK only for signal level currents. There are power Schottky diodes which could survive in that path, which I have never worked with, which again AFAIK, have much larger forward voltage drops. I was wondering if there might be a triangle of diodes in play, so that current could only flow like: battery to post to inverter to battery, and not in the other direction. That might also explain why the Battery Tender Junior worked so poorly when attached at the post (two small voltage drops), but better when it was attached directly to the battery terminals (no voltage drops).

The exact method was to use the probes from the UT210 meters in the following positions:

1. At the battery: positive probe poked under the red cap to the outside of the clamp on the positive post, negative probe to the negative post or to the bolt at the end of the grounding strap. (I could not measure a difference between those two.)

2. At the jump post: positive lead to the post, negative lead to the bolt immediately posterior to the top of the fuse box, on the curved sheet metal.

The battery is currently out of the car. When it goes back I can get a piece of long wire and measure directly the voltage drop from the positive terminal to the jump post.

 

This car must have gremlins or evil spirits.

With the battery out I measured the resistance from the bolt near the fuse box I was using for a ground to the battery grounding strap, completing the path with a length of wire. It came out to 0.9 Ohms (after subtracting the 0.2 Ohms for the wire). Not that I believe that measurement very much - to be accurate in that range the multimeter leads should have been more firmly attached, like screwed on. In any case, not a lot of resistance.

Yesterday the battery was charged with a Battery Tender Junior, it started at 12.44V and ended at 13.12V (which was falling slowly once off the charger.) From solid red LED to solid green LED took 8 hours. In previous days I had tried to charge it up in the car, from the jumper post, and it always said it was done after an hour. It was charged oddly, like it was all surface charge, any load and it would drop dramatically and stay low. This time it seemed to actually be holding a charge. Let it sit overnight, and around noon attached it to a Harbor Freight 58759 "Viking" digital battery tester. Do I trust this unit? Not much, but figured it couldn't hurt outside the car, and it was inexpensive and one was available close by. Punched in the CCA of 320 and ran the "out of the car" battery test. Results:

State of Health: 100%
State of Charge: 100%
measured CCA: 355
Battery Volt: 12.75
Internal R: 8.44 mOhm

That internal resistance looks slightly high to me, but it isn't enough to account for the 0.2V drop when the measured current out of the battery is only around 20 mA. UT210 measured 12.76V right before the test, so those values agree. It's possible the CCA is really better than the rating in a 2 year old battery, or the measurement might be off. See "trust", above. The test is pretty quick, and this small device does not dissipate heat, so not a traditional load test.

Put the battery back in and looked at the voltage at the post and at the positive battery post with the car off. They were the same. They were the same at the post when measured to the bolt I have been using or with a wire all the way back to the negative terminal. They were the same as each other in ACC and also in IG-ON (but of course each of these was less than the car off battery voltage).

So taking the battery out, charging it, and putting it back in eliminated the 0.2V voltage difference which had been measured consistently before that. Why was it there in the first place, why did it go away after doing, well, pretty much nothing? Beats me. Gremlins or evil spirits.

Because somebody might ask - I didn't see anything resembling corrosion or burns on the clamps to the posts or the posts themselves, and the clamps were both attached firmly, so that they would not rotate, before and after taking the battery out to charge it. The other connectors (the white ones which plug into the positive clamp and to the brake box next to it) were also clean. Probably irrelevant, but it was at least 20F warmer in the car today than the last several days.

For future reference, current draw in ACC mode was around 2.3A (but oddly, because the car wasn't doing much, varied by .5A or so around that value) and in IG-ON mode, with the cabin fan, radio, and all lights off, it was 9.01A (and very stable at that value). The ACC voltage measurement on the MFD with the fully charged battery was 12.4V, corresponding to around 12.72V open circuit at the battery. IG-ON is not a good mode to check the battery voltage, as the current corresponds to ~.2C, which is enough to drop the battery voltage substantially, into a range where the battery looks iffy when it is not.

 

Under normal circumstances the battery would only be putting out (more than) 9A for a fraction of a second when the car turns on. After that it is all from the inverter. So IG-ON to load test the battery, albeit with a not entirely controllable load, but not really indicative of a normal operating state. Darn car turns things on and off at random times for a long time after the doors close!

Interesting that a Gen 3 can sometimes show the 0.2V drop too. I wonder if CR94 were to unhook the battery for a few minutes if that drop would go away on that model too. Probably only to mysteriously return at some later time.

The ACC voltage after 3 minutes was 12.4V before turning it off last night, but only 12.1V in ACC today (at entering the mode and for 3 minutes after once it settled). Perhaps the 0.2V drop has returned, or the 460mA parasitic load which I saw for a while, and which then disappeared, has returned. I'm not really feeling motivated to take the back apart again to find out today. After driving the car for a total of about 40 minutes afterwards, ACC was at 12.6V right after the journey ended, and was at 12.5V (with occasional 12.5V) at 3 minutes. So it certainly charges up OK. Will check after letting it sit for a long time and see if it falls again. Presumably an AGM battery in a normal Prius that is 12.3V when the car is turned off should still be 12.3V 12 hours later. .02A * 12 = .24Ah, out of 46Ah, shouldn't really be enough to move the needle. Conversely, .46A * 12 = 5.52Ah, which could easily drop it that much.

 

Update on that Harbor Freight tester - some care is needed to make solid electrical connections. We just tried it on my son's truck, which has slightly corroded battery connectors and a 7 year old battery. It was clamped onto those connectors/cables rather than the posts, and it read it as 15% SOC. Changed it to the "front teeth" on the clamps instead of the "side teeth", and it went up to 53% SOC. I think this is probably because the alligator clips are not as big and hefty as one normally finds on a load tester, so the side ones don't "bite" through the surface contamination as well as the front ones. (Because more force on the teeth at the front than the ones at the side.) To detect this check the voltage at the posts with a voltmeter separately. If the voltages agree then it is a good connection, if not, then it isn't.

This would not have affected the Prius battery test above as that was directly on the battery posts.