lights/power/voltage drops

[Previologist]

My analog low range is 10v, then 50. My digital low range is 20v.

I am clear on the fuse block and the 2 fuses but I am not clear what you mean by the connection between the ALT and AM1

Do you mean the metal band (conductor?) on the side that looks sort of like this? (this pic is from an earlier model and is not identical to mine) (EDIT:I will assume that's not what you mean, because on mine there is a lead directly from the + terminal to that metal band)



But this is probably a good time to mention that (courtesy of previous owner) the fuse block is not mounted normally, it was just sort of tied with wire to the battery hold down and another point on body. Not sure that would make a difference but thought I better mention it

Also I realize now I should really disconnect the plugs and inspect all the lower terminals closely, which I did not do before when I inspected the fuse block.

[John Kaufmann]

The 10V range should do fine.

With respect to your picture, I'm at something of a disadvantage, because I no longer have my Previa. But I do have the EWD, and this picture 040042-(11) Power-source.pdf shows that fuse block schematically on the first page. You see how, at the top of p.40, the battery directly feeds four of the five links -- all except the 50A AM1, the link that feeds the loads (most important, IGN and ST) that are affected by the higher loads fed by the 100A ALT link?

As Dan was saying, you have classic symptoms of a weak supply, but tests have indicated that battery and alternator are OK. That just leaves the links ALT and AM1. You mentioned some corrosion in that fuse block, so we want to test the voltage between that ALT-AM1 connection (on the meter's low side) and the battery (on the meter's high side) when you apply one of those high loads: stoplight (brake pedal) or blower on high, or headlights (or all three). Whatever voltage reading you see on your meter is subtracted from the battery (or alternator) supply.

Then the next step: put your meter's (-) lead on the AM1 output to the IGN switch. If there is a difference between those two tests, your major drop is through link AM1. If there is an reading of several volts, but no significant difference between those two stops, then the major drop is through the ALT link.

If you don't get a reading of at least 4 volts with one or both of those tests, then switch to the next higher meter range, put your meter across the battery terminals and do the same test - but this time you are looking for a voltage drop, not increase. Please let us know the results.

[Previologist]

Sadly, I don't quite understand much of your very helpful post but maybe a light bulb will go in my brain on as I clean up other things I've discovered.

I took the fuse block out, so I'll post some pictures of it that might help narrow down some of the details in your post.

I discovered that a previous knucklehead has cut and spliced all of the wires to the connector plugs. Fortunately they seem well connected and I don't see any corrosion on the ones where I can actually see into the butt connectors. I presume they did that because the connectors are a PIA to remove, so they just cut them off. But why? NVM, I probably don't want to know.

I pulled the alt fuse and there is some corrosion on it's "filament." It doesn't look terrible but I will replace it anyway. There is a dusty coating on the wire ends at one of the ring terminals that bolt the alt fuse into the block, but it doesn't look terrible either and the ring was fine. I'd like to avoid messing with those wires unless it proves necessary.

The block is missing it's lower cover, but all of the spades on the bottom look pretty good and free of major corrosion.

[Previologist]

That dusty connector on the alt1 fuse ring connector (would that be "C" -->>> EA1 in the top schematic?) runs to ground somewhere, I presume, because the other alt1 fuse ring connector goes straight to the + battery terminal. Does it run through anything that should cause increased resistance?

I can follow the wire down under the battery and it then enters a flex conduit and that's where I lose it. Something comes out of that conduit and connects at the alternator with a plastic connector plug, but that is a much smaller wire (might be 2 small wires).

But anyway I am getting variable (and sometimes puzzlingly increasing) resistance between that alt1 fuse connector and ground. I get a minimum of 2000 ohms when i ground it to the negative battery cable (which in turn shows no measurable resistance to ground). And sometimes that ohm reading increases slowly but not always. But its never less than 2000 ohms (2 on the x1k scale).

[Previologist]

I traced that conduit a bit more. Another wire the same size as the alt1 wire emerges and bolts to the top of the alternator (worth noting the alt1 end is white wire and the alternator end is black) . Thinking this may be the same wire (the alternator end also has a dusty coating), I disconnected it from the alternator and found about 2000 ohms between its end and the alt1 end. Granted, I do not know if these are the same wire but they must be connected somehow. But when I test between the alternator end of the wire and ground, I get 0 resistance.

I'm not entirely sure how to interpret this, except that if it is the same wire perhaps the alternator end has another path to ground whose connection is hidden somewhere in the conduit, "beyond" the point where the 2000 ohms of resistance is. But I don't know if this resistance is enough to cause my problem or not, so I'm not sure how much effort I should spend tracking it down.

[man_btc]

To see what's happening, don't measure resistance (ohms), measure DC volts. Measuring ohms on anything other than an unpowered wire to ground or across a single unconnected component won't be meaningful. Measuring ohms on a circuit with power on it will definitely give you meaningless, wacky numbers.

Besides, significant vehicle voltage variations occur with tiny resistance variations, so even when measuring resistances to ground with no power on, it's in a low, inaccurate range where the meter can see a low-resistance path to ground, but not whether it's low enough. If under power, that point shows zero(ish) volts, then you know it's a good ground.

My issue and continued focus on the power source is because you're not seeing dimming with only one cause (like brake lights and nothing else). I would also think high fusible link resistance wouldn't let your lights be bright enough in the first place, even before anything else started drawing power.

If I were you, I'd want to be confident that the alternator is producing stable DC volts to the battery and vehicle with different loads. If you measure AC volts across the battery (van running*) what do you see? Should not be much, less than 0.5 volts and probably a lot less. What is the DC value you see? Can you look at that DC value with and without the lights and lights+brake lights on? I guess I don't trust the tests the car parts folks did.

*...Van running at a fast idle with the lights and brakes on (&/or fan on high maybe - something to apply a significant load) that is...

[Burntboot]

Lots of great advice here.

I would only add that the only correct place to have a crimp style butt connector, is in your emergency kit.
They are notorious for causing issues.
Regardless of appearance, they should be cut out and wiring repaired properly (solder and shrink tube).

Inspect the wiring while you're in there. When you strip the end, make sure it's shiny and bright.
If its dull, it is degraded and needs to be replaced.

Clean that "dusty" end and make sure the insulation is the same colour as further away.
Overheated wiring can take on a dusty look as the insulation starts to break down from excessive heat.

[man_btc]

I'm going to disagree about soldered wires being better than crimped in the sense that either can be bad if not done properly. You'll find crimped connections all over mission-critical things that move and vibrate, but ones made with proper (read expensive) tooling used by trained technicians on approved connectors and wiring - in other words, not some low-quality connectors mangled by some unskilled rando who bought them and the crimping tool on the cheap, then didn't protect it from the elements properly. That same mook could just as well have soldered together a poorly-fused, brittle connection that wouldn't automatically be an improvement because it wasn't crimped.

Again, voltage measurement is going to tell you if enough resistance has appeared between two points in a circuit to produce a significant voltage drop. That voltage-drop measurement will show up when you put your multimeter's probes* to either side of your suspected item (assuming you can find a place to touch the wiring at both sides of it). If you see almost zero volts, it's not "stealing" enough voltage from the other items in the circuit to be a problem.

* Meter set to measure DC volts and the range that just includes 12V (20 volts?) so you get the most resolution on the display - in the same sense that you wouldn't want a speedometer that goes to 200 MPH because you could barely see the difference between speeds you normally drive at.

[Previologist]

To see what's happening, don't measure resistance (ohms), measure DC volts. Measuring ohms on anything other than an unpowered wire to ground or across a single unconnected component won't be meaningful. Measuring ohms on a circuit with power on it will definitely give you meaningless, wacky numbers.

Besides, significant vehicle voltage variations occur with tiny resistance variations, so even when measuring resistances to ground with no power on, it's in a low, inaccurate range where the meter can see a low-resistance path to ground, but not whether it's low enough. If under power, that point shows zero(ish) volts, then you know it's a good ground.

My issue and continued focus on the power source is because you're not seeing dimming with only one cause (like brake lights and nothing else). I would also think high fusible link resistance wouldn't let your lights be bright enough in the first place, even before anything else started drawing power.

If I were you, I'd want to be confident that the alternator is producing stable DC volts to the battery and vehicle with different loads. If you measure AC volts across the battery (van running*) what do you see? Should not be much, less than 0.5 volts and probably a lot less. What is the DC value you see? Can you look at that DC value with and without the lights and lights+brake lights on? I guess I don't trust the tests the car parts folks did.

*...Van running at a fast idle with the lights and brakes on (&/or fan on high maybe - something to apply a significant load) that is...

I tried these tests except for using fast idle, although I could probably do that if I cut a piece of wood the right length.

For some reason my analog meter wouldn't read ac or dc volts, so I had to use my digital which I'm less familiar with. I got nothing measurable on any AC setting, testing across the battery terminals when running. DC came in at about 14.48 with nothing electric running or with lights on, blower on and brake on (with the brake fuse back in it's place). When adding the hazard lights on top of the other loads it caused pulsing between roughly 14.3x to 14.48. Turn signal alone caused weaker pulsing, only about plus/minus 0.02-0.05 v

Since I applied the brakes using a stick jammed against the seat, any temporary voltage reduction was unknown. I did the test again, setting up a camera under the hood and shooting video of the meter while I press the brakes on and off a few times. Video showed it only dropped to about 14.1 or 14.2 when I hit the brakes.

That doesn't seem like a big drop to me.

[Previologist]

Going back to John's post, here are pictures of my fuse block that may help me understand what he was suggesting I do.


IMG_3928.jpg
This is the inboard side, the cable runs from the alt fuse to the flex conduit and (presumably) to the alternator

IMG_3927.jpg

This is the outboard side, a cable runs to the alt fuse from the + terminal.



IMG_3929.jpg



Bottom of the block, no cover but all looks good.


The abs fuse is intentionally absent.

[man_btc]

Looks like you are getting a solid voltage from the battery & alternator. No need for a fast-idle IMO, that was more if the voltage at idle was marginal. i.e. close to 12v and not 14-ish where is (and should be).

[Previologist]

Well that all makes sense: larger loads (like stoplights and blower high) causing larger voltage drops.


Good - let's use the analog meter. (I hope it's not auto-ranging.) Do you see the connection in the battery fuse block between the 100A ALT link and 50A AM1 link? Can you get the common (low) side of the meter on that point, and the high side of the meter on the battery + (the terminal that the fuse block is mounted on)? What is your meter's lowest DC volts scale?

Still trying to understand what you are asking here. Language like "common" and "low side" are Greek to me, but I'm guessing you mean the black probe, and the "high side" would be the red? Once we get that cleared up I just need to figure out what you mean by the "connection"

[man_btc]

I got curious and disassembled my donor's fusible link box, as in totally extracted all the metal brackets and connectors. I'm putting together an exploded-view diagram, but now that I understand its construction, I saw something odd in your photos. That blue 100A link sits on tabs that extend into the body of the connector. One goes to the Battery +, the other to the alternator (so neither can blow the other up). That link's tabs have holes that line up with the threaded holes on both sides of the housing - and there should be screws tightened into them! (You can see one in your "earlier model" photo.) As in you can't budge or pull that link out without removing those screws. Without that, it's not a tight connection at all, and that's THE main power connection, so just smallish fuse-ey-type spring-loaded connections aren't good enough (the connectors underneath are beefier). Unless you removed the screws for the photo, then never mind and ignore my rant.

As far as the probing goes, dinna worry about which probe goes to which side - if they're "backwards" you'll see a negative voltage, but it's just the absolute value you care about. And you want the smallest meter range that will still cover up to 12 volts, say 20. On larger ranges it'd be like trying to gauge your speed through a school zone with a speedometer from a formula one car.

Also. not sure where you probe stuff if that housing is all buttoned up. At any rate, if those screws were never there, put 'em in and retest. If they were really installed when the block was in use, you could measure the voltage from one to the other (if you can get to them) as that's the drop across the 100A link.

[Previologist]

Yeah the screws are there, they just weren't shown in the picture. But I can easily measure across them because the block isn't bolted down. I just tried it quickly and got extremely low voltage (<1.0 if I was reading it right), but being cold and night I don't want to open the garage door. So I only ran it about a minute because I also don't want to die. I will check it again in the morning.

[Previologist]

Same result today. I am only picking up 0.03v or so probing across that Alt fuse. I will follow through on my threat to replace it today, if the snow lets me get to a parts store.

Kind of surprising because that one looked better than the other fuses, but there is a large part of it that is hidden within the plastic and can't be seen.

[man_btc]

So that's not the problem I guess.
When you were measuring, were the lights and brakes on? You want a decent amount of current flowing while you measure the voltage because of the possibility that -

• poor connections
• narrowing of the fusible link by corrosion or internal degradation from previous high-current episodes
• whatever

- might reduce its current-handling capacity and thus increasing it's resistance to higher currents.


As an FYI, if you were the type to radically disassemble (i.e. bust up the housing) a long-lived (in Arizona) 100A fusible link to see the "innards," it'd look like this. No charge for the ambience.

20240110_153804.jpg

20240110_153857.jpg

[Previologist]

This mornings readings were without anything running. I tested again tonight with everything running-brakes on, lights on, blowers on, radio on flashers and both wipers on, but with wipers or flashers on it just jumped wildly from 10 to 18v. When I turned the wipers and flashers off the most stable reading I got was about 15v.

I already bought a new fuse so I'm going to perform surgery on this one anyway. I'll run the test again with the new one but I expect similar results.

With that white stuff on the ground, I can tell yours isn't in AZ anymore!

[man_btc]

Don't use anything that doesn't draw a steady current, like those flashers and wipers - they will make things jumpy and unreadable as you saw. The lights brakes and high blower is fine.
So with that current being drawn, measure the voltage across that 100A link with a small range (couple o' volts) if you have it, 20V if that's the smallest available. Shouldn't be more than say 1/4 - 1/2 volt tops, and I would hope a lot less. Then, to double-check - and use a different meter range - set the meter to 20V, then with the black on ground, touch the red to one, then the other side of the 100A link (14ish volts presumably). Then, take the difference between those two values to get the voltage across the link. Should be around the same as across the link, but the meter may not be all that accurate at very low values. If you can have the idle up a bit, that's more realistic, but not entirely necessary.

And yeah, that donor Previa has never seen the white stuff (well, snow at least) and the rust-free carcass sits cozily in my garage aggravating my wife no end. I did the outside link shot to get accurate lighting and show how even a link in a dry environment gets corroded somewhat. I understand the normal wire issue - you can only get so much current through a fixed cross-section until it starts melting, making that cross-section even more inadequate until it ultimately breaks. Now, fusible links are meant to endure short-term bursts of current and stay intact where a regular fuse would just blow, so maybe there's some internal degradation that increases resistance over time? I don't know enough metallurgy to speculate beyond that.

Now for the latest visual aid, as I know you're not a fan of schematics. This is an attempt at showing the fuse block "blown apart" so you can see how all those metal bits and bobs inside route the electrical paths (I tried my old X-Ray Specs, but all I could see were the bones of my hand wherever I looked). The graphic wound up being a little more "dense" than I hoped - my attempt to make it a little easier to follow was to tint the positive battery connections red and the positive alternator connections blue.

fusiblelinks_colorized_annotated 1440px.jpg

[Previologist]

I guess I'm about to put my electrical ignorance on full display again, but I'm over it.

Why should I expect to get such tiny voltage (1/4-1/2v) across that fuse even with things running? I would expect closer to the 15 that I am getting. That 15 is DC volts. My lowest range is 20 on the digital meter.

That's a cool diagram, useful but slightly different than mine. You must have bought your x-ray glasses from the Johnson Smith catalog, they were famous for bogus items like that to prey on 12 year old boys. I had their secret radio pen, which actually sort of worked.

[man_btc]

Hah! I've still got my radio pen! It's in the garage in the misc-stuff-in-drawers section.

Anyway, if it were easy to do, you could leave your black lead on ground and measure the voltage from where it's highest (the battery/alternator connection) all the way through the lights to ground, measuring lower and lower voltages until you got to (near) "0" at chassis ground. The difference between any two points is how much of the total voltage is "used up" by what's between. Let's say the van is running. The power is coming from the alternator putting out 14 volts. Current flows across the 100A fusible link to charge the battery and separately to power whatever's turned on. Some small amount of the total voltage will be lost along the way through each wire, connector and component to ground, with most of it appearing across the Thing You Want Powered, like the lights. Each non-major component's "absorption" of the total voltage could ideally be measured across its terminals, but that'd be a fraction of a volt that might not be measured accurately on a meter that can't go down to a small scale, so the alternative is to measure the voltage from each side of a thing to ground and subtract them (as they'll be more in the "middle range" of what the meter can display).

So things work fine with the lights on, but press the brakes or turn on the fan and they dim. That additional load is pulling current from two places shared with the light circuit - the alternator output and then at least one fusible link before the wiring splits off from the fusible-link box to the things being powered. The alternator is supposed to sense the need for additional current and then put out more, which is why the first thing done was to check that, while running, the voltage at the battery stayed steady with increased electricity use. One of the other potential problems is that maybe (?) one or more of the fusible links can't carry enough of that increased current. That might be because a fusible link isn't as "thick" as it used to be (the fatter the conductor, the more current it can carry) or something about its internal structure prevents it from passing through the increased current, especially when it gets hotter.

If there's a problem carrying more current through an intact fusible link, its going to manifest itself by taking up a lot more of the circuit's voltage across it's terminals. Let's say when just the lights are on, one side of the main 100A link shows 14v and the other side shows 13.9v, but with the brake lights and fan on, the alternator side still shows 14v, but the other (load) side now shows 12.9. After further drops along the way from wiring. etc. (call it 1 volt) the lights might only see 11.9v, vs the 12.9 they'd normally get.

On the other hand, you didn't see a drop in the battery voltage during the charging tests, so my example above isn't the greatest. If it was me I'd see if I could get a probe into a connector at the bottom of the block to measure voltage - pin 1 on the shorter white connector - white/blue wire - as that would tell you whether something in the block was causing the voltage coming out to the headlights* to be significantly lower with other loads turned on.

*Well, actually the headlight relay, then to the headlight fuses, the lights themselves, then the light switches, then ground

If something in that fusible link block isn't found to be "stealing" more voltage out of the path from the alternator to ground under an increased load, it's back to head-scratching.