diy solar

diy solar

Destroyed $4K of alternator & regulator with a good Fuse

48V system. 56V 160A APC alternator. WS500 Regulator.
200AH Winston cells.
Running at ~160A during commissioning, when, we think, the 200A Class T Blueseas fuse opened. Took out the alternator rectifier diodes (internal) and the WS500 regulator. Also opened the 15A ATC fuse to the regulator, destroying the sealed fuse holder.

Lessons.
More headroom for main line fuses. At least 50% more than expected load.
Consider slow blow fuses
Most spade fuses / holders don't cope with 48V systems(more than once)
There is no alternator protection for 48V systems. Only 24v / 12v

One of the reasons I recommend people on 12V system side to leave their Lead batteries in parallel with the LFP system.
The lead battery is always in the system - manufacturers actually in many cars/trucks/RVs don't even fuse from the alternator to the battery.

Yes no fuse from the alternator to the battery in most OEMs and that's for lead
 
Maybe a TVS device could protect alternator from disconnect. Maybe a capacitor would too.

160A is 80% of 200A, shouldn't have blown. Do you know why?
"200AH Winston cells" - that's lithium. anything to do with a BMS disconnecting?
Does system regulate current? Lithium could take too much.
Wouldn't be the first time to have a variance in a fuse. Fuses are designed with curves and manufacturing tolerances. Sometimes when the current ramps up to quickly they blow below their rating. Shouldn't happen but it does. I guess the manufacturer rather sells you another fuse instead of getting sued that a fuse is not engaging early enough.
 
  • Like
Reactions: me3
Further investigation showed alternator tested fine. regulator was toast.
New regulator arrives next week. Testing with adjustable DC supply to rotor shows good output.
I'll increase the fuse to 250A and install the new regulator.

I think there will be some FAQ learnings for the APS alternator

I use the Wakespeed regulator with the canbus integration which helps somewhat, but likely still not enough if the high current line were to open like you had due to fuse blowing. I believe Sterling sells a device to absorb this surge but no idea how well they work.
 
I use the Wakespeed regulator with the canbus integration which helps somewhat, but likely still not enough if the high current line were to open like you had due to fuse blowing. I believe Sterling sells a device to absorb this surge but no idea how well they work.
No 48v option :(
 
One of the reasons I recommend people on 12V system side to leave their Lead batteries in parallel with the LFP system.
The lead battery is always in the system - manufacturers actually in many cars/trucks/RVs don't even fuse from the alternator to the battery.

Yes no fuse from the alternator to the battery in most OEMs and that's for lead
Generally there is a fusible link. Wire that performs same with less flammable insulation.
 
Generally there is a fusible link. Wire that performs same with less flammable insulation.
fuseable links have a far higher overload profile then a "fuse"

They are usually sized far higher and designed for that case that the Alternator shorts against engine ground - and such the battery pumps hundreds or thousands of amps through the wiring.

So they practically only prevent fire in the case a dead short of the alternator. They don't do anything if you starter draws 300-400A.
 
I've seen fusible link for all battery circuits not including starter. It was between battery and fuse box, while starter had a heavier cable direct to battery. The fusible link was maybe 10 awg? I think it would blow in a fairly short time carrying higher starter current (like stalled), maybe not easy spinning starter current.

But fuse won't do anything for inductive kick blowing electronics such as diodes or regulator if battery disconnects.
A surge arrestor of some sort could. There are TVS/MOV devices for any clamping voltage you want, also gas discharge tubes starting around 60V. These things can carry 1000A but only briefly - any of these can only absorb a brief pulse. What you need is enough to eat the energy stored in inductance, and whatever time it takes for regulator lower output (that might not be fast). I would consider a transistor circuit clamping to maybe 20V, possibly into a resistor sized for max current.

A lead-acid battery always connected seems like a low-tech solution. Maybe a couple power-diode voltage drops from the alternator, if it needs to be floated at lower voltage. A DC to DC converter could keep it charged, and any spikes would turn on the diodes and dump current into battery. Diodes are slow (about 1 millisecond) so a MOV in addition would be good.
 
Diodes are slow (about 1 millisecond) so a MOV in addition would be good.

Use schottky ones (should be plenty enough, only a few µs max) or ultra fast recovery ones (even faster, in the dozens of ns range) ;)
 
Use with care. If you don't know what you're doing, or if the manufacturer deceives you with the data sheet, the Schottky diode burns up.

It may be that turn-on of a silicon junction diode is relatively fast, and turn-off is slow due to carries hanging around in the junction region. They will continue to conduct for a while when reverse-biased, which is a problem for rectification at high frequency, but not for this spike clamping purpose.

There are also silicon diodes fast enough for switching supplies, reverse recovery in nanoseconds. For the alternator/battery application, need something that can carry a few hundred amps. Actually even small diodes can take such current briefly, but best to handle lead-acid charging current in case it is ever run down.
 
A lead-acid battery always connected seems like a low-tech solution.
Some solutions don't need to be high tech. The Romans put Tile on roofs, those houses are still live-able after 2000 years. While all modern hightech materials fail after 20-50 years. :) just a random comment.

When you take a alternator developed for charging lead batteries and you are trying to adapt it to charge a much more touchy battery technology - LFP. You might need to start earlier in the chain. We might have to come up with a different alternator design all together.
 
Further investigation showed alternator tested fine. regulator was toast.
New regulator arrives next week. Testing with adjustable DC supply to rotor shows good output.
I'll increase the fuse to 250A and install the new regulator.

I think there will be some FAQ learnings for the APS alternator
Have you followed Sail Life's trials and tribulations with his high output alternator and BMS woes?
 
A 48v alternator can develop very high no load voltage at high rpm. It can be as high at 600 vdc unloaded.

You can use a lower voltage bridge rectifier but will be relying on regulator to keep it under control. This is for a coil energized rotor field winding alternator. A permanent magnet alternator has to use high breakdown diodes. 48v wind turbines use 1000 PIV diodes.

I believe the spec for this alternator has diodes with 300 PIV rating so if there is a momentary field winding overdrive, usually the result of intermittant output load connection, at high rpm, the diodes are at risk of popping.

The usual situation that can cause this in regular use is inverter connected to battery has sudden drop off of high load current leaving the alternator at full output for a moment. Batteries must take the momentary current surge until alternator regulator turns down the field drive. The batteries are keeping the peak voltage under control, preventing alternator rectifier damage.

Problem with LFP batteries with their BMS is if the reverse surge current causes the BMS to disconnect for any reason, excessive charge current or cell overvoltage, the voltage limiting action from the batteries disappears.. This can not only damage the alternator diodes but also exceed the breakdown voltage on BMS series MOSFET's when BMS shuts down.

Assuming battery/BMS can take the maximum alternator charging current, make sure the BMS overvoltage trip has some delay time before reacting to cell overvoltage. Make sure you don't have any intermittent battery cable connections and never disconnect battery from alternator while it is running.

I would recommend putting a surge limiter on alternator output of no greater than 80v limiting to protect BMS.
 
Last edited:
Problem with LFP batteries with their BMS is if the reverse surge current causes the BMS to disconnect for any reason, excessive charge current or cell overvoltage, the voltage limiting action from the batteries disappears.. This can not only damage the alternator diodes but also exceed the breakdown voltage on BMS series MOSFET's when BMS shuts down.
almost sounds you need a dump load with a voltage activated solid state thing which activates a few volt above normal operating conditions.

Or put a lead battery in parallel which doesn't have the issue. It can be tiny 8ah AGM batteries in series. Just something to take the surge. Lead is very good in buffering surges - it just gets warm for a moment.
 
almost sounds you need a dump load with a voltage activated solid state thing which activates a few volt above normal operating conditions.
Since the surge time is short (<0.5 sec), the surge suppressor needs to take the alternator peak current but doesn't have to dissipate much heating power.

Weakest link is the BMS MOSFET's breakdown voltage, likely 80-100v for 48v BMS. If BMS opens and inverter is still connected to alternator output it also can put the inverter at risk of overvoltage damage.
 
100A x 12V x 0.5 second = 600 joules

That's around the energy of a handgun bullet. But not far out of line with the specs of a MOV (the following is higher voltage and 480J)


Down under 24V, best I'm seeing is 120J, maybe use 5 in parallel (more or less depending on maximum current to clamp)

 
Update
Alternator was ok
Controller dead. WS support was outstanding
Back running again with uprated class T fuse.

Interesting note. The dead fuse would test ok with a multimeter, but be open under any minimal load.
 
Update
Alternator was ok
Controller dead. WS support was outstanding
Back running again with uprated class T fuse.

Interesting note. The dead fuse would test ok with a multimeter, but be open under any minimal load.
Oh man, was chasing down gremlins with an FJ cruiser -- alternator fuse was bad but would test good. In order to change the fuse, the whole electrical fuse box had to come out. Did everything I could to avoid trying to do that, but that was it. Even looked okay as well.
 
I am also working with a similar setup. In our case I made a wiring error on the shunt and the current went high enough to trip the dual breakers. (it is a redundant 48 volt, LiFe battery pack).

Same alternator / same WS setup.

We didn't damage the alternator or WS but we did knock out some of the other electronics from the surge.

So - also looking for ways to shut down in an easy way.

Looking into various diodes and protection methods, but open to ideas.
 
I am also working with a similar setup. In our case I made a wiring error on the shunt and the current went high enough to trip the dual breakers. (it is a redundant 48 volt, LiFe battery pack).

Same alternator / same WS setup.

We didn't damage the alternator or WS but we did knock out some of the other electronics from the surge.

So - also looking for ways to shut down in an easy way.

Looking into various diodes and protection methods, but open to ideas.
All feedback says it's difficult to get any device that clamps / dumps very rapidly at over 56v and less than damaging (say 80V or so).
Doesn't sound hard, but seems to be so.
I'm thinking now about putting Ultracapacitor array inline, that would soak up a lot.
I just need to find one that has a balancer etc and is rated at say 100V to handle the peak.
Another option is 4 x motorcycle AGM's as a load dump inline, but that's heavy / ugly.
 
All feedback says it's difficult to get any device that clamps / dumps very rapidly at over 56v and less than damaging (say 80V or so).
Doesn't sound hard, but seems to be so.
I'm thinking now about putting Ultracapacitor array inline, that would soak up a lot.
I just need to find one that has a balancer etc and is rated at say 100V to handle the peak.
Another option is 4 x motorcycle AGM's as a load dump inline, but that's heavy / ugly.

Based on the devices that failed, I am pretty sure that we experienced > 100 volts peak on the bus bar, so in my mind, there is plenty of opportunity to at least have less of an event.

One of the things that we have done is to limit the current to substantially less than that alternator is capable of doing, and maxing the charge voltage to be low enough to reduce the probability of the batter BMS tripping for over voltage.
 
Back
Top