How can I safely energize a suspect alternator

[Grinkle]

TL;DR Summary

Odd looking 12V waveform on my alternator - I suspect the regulator, and I want to energize the alternator bypassing the regulator to see if that cleans things up

I have a "house" alternator on my van. The 12V rail waveform is attached. The 5V drop-outs are not right, imo.

The alternator is controlled by a Balmar 614 regulator.

http://www.balmar.net/PDF/regulator info s/MC-614.pdfThe spacing of the drop outs leads me to suspect they are synchronous with the state control switching of the voltage regulator - I doubt the belt speed would produce spikes that are 2ms apart like that, it seems more like the micro-controller frequency is driving the spacing of these drop outs. Moreover, changing the van engine idle speed does not change the spacing of these drop outs. I suspect the output of the regulator that is driving the alternator field is not right. If my understanding is correct, its a pwm current output, not a voltage output, so maybe hard to monitor directly.

What I want to do is run the alternator without the regulator attached at less than full field and see if the 12V rail is clean. I am concerned about doing a full field test - I don't want to damage the alternator. The field strength is regulated by a PWM input coming from the regulator pin 4. The regulator manual says one can do a full field test by removing that connector and shorting pin 4 to the 12V rail - I guess that will saturate the coil in the alternator. I don't want to do that - conceptually I'd like to connect the field input to a bench supply and ramp it up slowly to see the output current from the alternator come up and see if its clean / stays clean as I ramp up the field current.

Any thoughts on the wisdom or folly of thei below approach much appreciated.

"connector" in my description below means pins 1-4 shown on page 3 of the regulator manual, there is a picture of "connector" on page 2 of the manual.

Step 1 is to disconnect the connector from the regulator, so the regulator is unpowered and the alternator field is unpowered. I have confirmed this just leaves a clean 12V battery output on the rail with the engine idling - the regulator is dark, and the alternator is spinning freely, field not energized.

I am thinking to take a bench supply, set it to 12V, set the current compliance very low, and use the bench supply to energize the field by conntecting it to the alternator field input with the the plug disconnected from the regulator (so I would just connect the bench supply to pin 4 of that connector, the ground of the bench supply to pin 1 of that connector, and leave pins 2 and 3 floating). I think the alternator will collapse the bench supply if I set the compliance to a few hundred uV and the bench supply will be acting as a current source until I increase the compliance enough to saturate the alternator. I'm not sure how much current will saturate the alternator coils.

Thanks in advance for any advice. I am hoping someone here may have prior experience with alternators and can tell me if my thinking makes any sense.

 

[Grinkle]

I found the source of the drop outs much more easily than the procedure I asked about above. I tried scoping the PWM output and it does capture a voltage wave quite easily. Its pretty clear the pwm field control edge (red trace) is synched up with the transient on the 12V alternator output (blue trace). Whether its coupling in the regulator or coupling in the alternator I am not sure. I don't see the transient at the battery so the cable parasitics are apparently filtering it out, or else it is somehow fully common mode when measured at the battery terminals. I doubt that because it isn't common mode at the regulator 12V / Gnd connections.

Why do I care (perhaps I should have opened the thread with this) -

There is a 10A fuse on pin 2 of the Balmar connector and I seeing this blow from time to time, and I am trying to debug that.

Anyway, my original question is now moot - if anyone finds this interesting enough to brainstorm whether these transients are normal or not, and whether they might be associated with the fuse blowing or not, I'm happy to hear any theories.

 

[Tom.G]

There is a 10A fuse on pin 2 of the Balmar connector and I seeing this blow from time to time, and I am trying to debug that.

Anyway, my original question is now moot - if anyone finds this interesting enough to brainstorm whether these transients are normal or not, and whether they might be associated with the fuse blowing or not, I'm happy to hear any theories.

The link in the first post returns an error. This one worked for me:
http://www.witzgallphotography.com/balmar/wp-content/uploads/2020/07/mc-614-manual.pdf

The occassional fuse blowing could be a corroded fuse holder overheating the fuse during heavy loading. It could also be a vibration failure of the fuse if it is mounted on a surface that shakes, even intermittently, or if it flops around and hits something. Vibration fuse failures are common if the old style glass tube fuses are used. If that is the case, replace with the newer automotive fuses tha plug in with two blades and have a plastic body.

The spikes on the Blue trace coincident with the Field pulses remind me of a Grounding problem, probably of the 'scope but could also be that the Ground (common) point of the regulator, alternator, battery is not quite perfect; as in a too-small wire somewhere or again a corroded connection. It could also be that some cable is too long and it was coiled up so it didn't flop around, not terribly likely though.

Cheers,
Tom

 

[Grinkle]

@Tom.G Thanks very much for posting a working link - I should have tested what I posted.

I do have a blade fuse. I agree grounding might be involved - a ground can never be too good. I will see if I can introduce better grounding on that particular ground wire (the wire into the regulator ground connection).

One thing I was wondering is whether I am seeing a mechanical transient as there might be some compliance in the belt and pulley's, and when the field voltage comes up the load on the belt increases and this causes a drop in the rotational velocity of the rotor as the compliance in the system is 'taken up'
as it were and absorbs a little energy. But then it strikes me as too high a bandwidth ringing to be mechanical and I'd expect some phase difference between the PWM edge and any mechanical play showing up.

 

[Baluncore]

There is a 10A fuse on pin 2 of the Balmar connector and I seeing this blow from time to time, and I am trying to debug that.

PWM voltage to the field winding passes through the brushes and slip rings inside the alternator. If the field fuse is blowing I would check for an intermittent short inside the alternator.
Try replacing the fuse with a 10 amp breaker, which will be less sensitive to momentary shorts than a fuse. If the breaker opens, reset it and tap the alternator to test for a vibration sensitive short. Also consider a short across the slip rings or in the field winding.

I keep a selection of low voltage breakers connected through long leads to used open fuses. That way I can substitute a breaker quickly for a fuse when needed. That reminds me, there are two breakers in a neighbours Chinese tractor ignition and start circuit, it has not failed since they were installed. If they did open he could reset them from the drivers seat without needing to access the fuses.

 

[Baluncore]

Anyway, my original question is now moot - if anyone finds this interesting enough to brainstorm whether these transients are normal or not, and whether they might be associated with the fuse blowing or not, I'm happy to hear any theories.

Switching transients are normal and occur when the path of the current in the inductive field winding changes from the regulator switch to the freewheeling diode.

For a 12V vehicle, the regulator voltage should never exceed 14.5V.
14.5 volt / 10 amp = 1.45 ohms. Measure the resistance of the field winding. If it is greater than 1.45 ohms then the fuse should not blow unless there is a short circuit somewhere in the field circuit.

To blow the fuse requires more than 10A current. Find out where the 10A is going. Does it only blow when a flat battery is connected, the headlights are on, or when the starter is energised? Look for a pattern that could make the alternator struggle.

 

[Grinkle]

Find out where the 10A is going. Does it only blow when a flat battery is connected

Yes, I have seen that when the battery is charged and the alternator is not heavily loaded the fuse does not blow.

For the past few days the battery has been well charged, and when I start the vehicle the alternator only stays in bulk charge mode for about a minute, then switches to acceptance mode. The current in bulk is about 170A and the current in acceptance in between 40A and 60A. The system can run like this without blowing a fuse, at least for the 20 minutes or so that I tried it.

One point to make is that the fuse is on the 12V input to the regulator. The field output to the alternator is not fused. Of course, the field output is ultimately powered by the 12V input I am pretty sure.

The alternator is billed as a 280A alternator and I've never seen the load be higher than 170A, so I haven't suspected its overloaded. That said, the fuse blowing is associated with the heaviest loadings on the alternator that I see.

Measure the resistance of the field winding.

Thanks! I will do this - great advice.

 

[Baluncore]

One point to make is that the fuse is on the 12V input to the regulator. The field output to the alternator is not fused. Of course, the field output is ultimately powered by the 12V input I am pretty sure.

Sorry about getting the fuse position wrong. You referred to a pin 2, while the data did not identify pins with sensible names.

There is another weird possibility. Maybe the alternator pulley is slightly too big for the application. Alternator output voltage is the product of field current and RPM. A low RPM motor or alternator will require greater field current before it exceeds battery voltage, and so can begin to charge the battery. There may be a small range of RPM where the voltage is sufficient for the regulator to function, but is insufficient to overcome the battery voltage. That could occur where the regulator is supplied from the battery, and not from the internal (internal secondary) alternator output.

Check that the minimum specified alternator RPM is satisfied by the pulley ratio and engine idle RPM.

Check the obvious, that the model of alternator and voltage is correct for the battery voltage.

 

[Grinkle]

There may be a small range of RPM where the voltage is sufficient for the regulator to function, but is insufficient to overcome the battery voltage.

Interesting thought!

But as long as its the case that the regulator energized the field, wouldn't the voltage have to fly up and pretty quickly become higher than the battery voltage? How could it not? Where else can the current go to prevent the voltage from increasing?

Check the obvious, that the model of alternator and voltage is correct for the battery voltage.

Yes, agreed, and they are. The system has been working more or less properly for a couple years prior to this as well. So either I have not really stressed it before, which is possible, or something changed.

 

[Baluncore]

... , I'm happy to hear any theories.

But as long as its the case that the regulator energized the field, wouldn't the voltage have to fly up and pretty quickly become higher than the battery voltage? How could it not? Where else can the current go to prevent the voltage from increasing?

Alternator output voltage is the product of RPM and field excitation. At zero RPM, during a delay while starting for example, a battery powered regulator will have a field current limited only by the battery voltage divided by the field resistance. Only once the engine starts, and the RPM rises, will the regulator start to reduce the field current. That is also why pully size matters.

 

[Grinkle]

Only once the engine starts, and the RPM rises, will the regulator start to reduce the field current.

Makes sense, thanks. Thinking your post over, I am now gaining confidence that I have a good theory on what is going on, at least conceptually.

Most of my testing is done at normal idle, so that will require the most field current to meet the regulator set point voltage. More field current means that all other things being equal, I am close to 10A on the regulator power input that I would be with lower field current. The regulator shuts the field current down when the alternator gets to 120C. This shutdown may introduce some transients that blow the fuse.

Just now, I turned on the AC to load the alternator and ran the system with the engine at high idle, 2k RPM's, not quite twice normal idle. I could see the alternator temp rising, getting to 120, the current going to zero, the alternator cooling to about 110C, the current coming back, rinse and repeat. I couldn't get the fuse to blow. I decided not to repeat things with the idle at normal because I don't really like seeing the fuse blow.

I am concluding that the basic weakness in my system is the alternator can't keep cool when its loaded and sometimes when the regulator shuts the alternator off this makes it blow its fuse. I can reduce the likeliehood of blowing a fuse by running the system at higher idles to reduce field current. If I want to make things fundamentally more robust, I need to find a way to keep the alternator cooler. It seems non-optimal to me to have the system continually limit cycling to the high temp cutoff. I was observing perhaps a 50% on-time for the alternator as it cycled to 120C, was shut off, cycled up to 126C or so before cooling down to 114C or so to get turned back on.

I may be able to track down what exactly is causing the fuse to blow under high field current conditions when the alternator shuts down, but I think I need to find a way to keep the system from getting that hot under what should be normal operation conditions.

I can limit the field on the regulator if I want to go that route - equivalent to hobbling the alternator. Since I'm already seeing a limited on time due to over temp that seems a better approach that wouldn't result in less net charging anyway. I'll give it some more thought.

I really appreciate the brainstorming / sounding board, thanks!

 

[Baluncore]

If it is a momentary transition that blows the fuse, replace the fuse with a thermal breaker.
Also consider a self-resetting thermal breaker as used in trucks.

Is there a fan on the pulley? Did you check the pulley size?

What is the make and model of the actual alternator installed? or a replacement part?
Do you have a link to a picture in a spare parts catalogue?

 

[Grinkle]

Great suggestion on the breaker.

I have this kit installed.

https://www.nationsstarteralternator.com/Sprinter-Van-Dual-Alternator-Kit-p/sprinter-dak-280xp-er-l.htm

The alternator is installed in a factory provided 2nd alternator bracket, its location is at the bottom of the engine compartment, behind and below the grill, roughly behind and lower than the passenger side headlight. I am attaching a picture I took lying on my back on the ground looking up at the alternator. The tire you see is the front passenger side tire. You can see the pulley to the in the upper left of the image. No fan on the pulley. I did not check the size, but it came from the alternator mfr themselves, I am assuming they have it sized correctly.

I also attached a picture of the regulator installation at the top of the engine compartment on the back wall, adjacent to the passenger compartment, passenger side, perhaps 3 feet away from the alternator.

 

[Tom.G]

Could be an airflow problem.

You mentioned testing at a "high idle speed" which I took to mean while parked. Your photo shows the alternator mounted rather low in the chassis and something (front bumper?) directly in front of it. You might try a road test on the open highway to see if it still overheats. If it's OK on the highway you will need to find a way for more airflow when parked. (add a dedicated fan? Or possibly a duct from the radiator fan, but that may be too hot.)

Another thing to check: Is it mounted close to the exhaust system? If so, a piece of sheetmetal mounted between the two would help.

I hope the regulator has at least a cover that was removed for the photo. It WILL get damp or wet. At least a splash shield could be added to extend its lifetime. Those connectors do not appear to be splashproof either, and as such, they will likely corrode over time.

Cheers,
Tom

 

[Grinkle]

while parked.

(front bumper?) directly in front of it.

Both correct.

You might try a road test on the open highway to see if it still overheats.

I will do that. My standard use of the alternator is never to think of it - it runs when the van runs, and I almost never idle the van while parked. On this latest trip, during which I first noticed the issue, I had at one point idled the engine for 30 minutes or so while at a campground and running the AC. It was a couple days after that when I first noticed the battery was not ending a day of driving at full charge. It was a hot part of a hot day when I had idled the engine prior, its possible the fuse blew during that idle and it took me a while to notice it. I can't tell if the alternator is charging or not unless I specifically go look at the battery current with the engine running, or I pop the hood and look at the regulator display. There is no alarm, its me visually polling the system when I am inclined to do so. There are a couple solar panels on the roof and while they can't fully keep up with the usage it will take days for the batteries to go flat even without the alternator. To tell if the regulator is overheating, I will have to have a co-pilot watch the current (or be driving while I watch the current) and see if it drops out periodically. I can't see the alternator temperature with the hood down, its displayed on the regulator under the hood.

Is it mounted close to the exhaust system?

I don't think so, its right behind the front bumper, in front of the engine, so probably well away from any exhaust plumbing.

has at least a cover

You can't tell from the photo, but the regulator is completely sealed with a clear potting compound, the connectors extrude from the compound.

 

[Tom.G]

If you can't find a reasonable explanation after the above checks, I suggest finding a service shop that specializes in automotive electrical. Don't bother with the corner gas station or neighborhood repair shop, for electrical they either hand-wave to sell parts or send it to an electrical shop (usually the former in my experience).

Or if you know someone with the test equipment and electrical knowledge, a first measurement would be of the field current under worst-case conditions; i.e. low battery and low engine speed.

Clamp-on DC ammeters are available but not many folks have one. Amazon shows some from $50 and up. A Google search (https://www.google.com/search?&q=clamp+on+dc+ammeter) shows over 1 000 000, but many of them can't measure DC current.

Back in the Old Days they were even available as a moving-needle analog meter that had a groove in the back of the case. You routed the wire thru the groove and it read the Amps. I haven't seen one in decades though.

If the field current is above the rated 10Amps then the prime culprit would be the regulator, or maybe an intermittent short in the field or the wiring to it.

Cheers,
Tom

p.s. Reading over earlier posts, I noticed this:

The system has been working more or less properly for a couple years prior to this as well. So either I have not really stressed it before, which is possible, or something changed.

Maybe the drive belt just needs tightening. At least that's an easy first check!

 

[Baluncore]

Maybe the drive belt just needs tightening. At least that's an easy first check!

A slack drive belt will usually not slip until the alternator begins to generate current and places the belt under tension. It then squeals while slip self-limits the charging current, until the belt warms up and grips, or the battery has charged sufficiently to reduce the load.

I see from the pictures that the alternator pulley cannot be reduced in diameter.

10A, Auto Reset, Standard Blade Circuit Breakers - will plug straight into a blade fuse holder.
They cost less than an autoelectrician; $7.00 from Jaycar in Australia, part# SF2323.
https://www.jaycar.com.au/standard-blade-circuit-breaker-10a-auto-reset/p/SF2323