Why won't fan start ? Perplexing

So, 2 new Mermaid A/C systems, one 12K, one 16.5K. both 220V. Discovered that the blowers on both units won't run when on the genset. No problem at all on shore power. Genset is 12.5 kw, it's not heavily loaded, voltage is 232 at power connection on the A/C control box. It's an FX-1 control board. Fan motors are fractional HP, shaded pole with a 2uf capacitor. Capacitors test within rating. Compressors and pump relay function normally.

It appears that this problem is related to the Magnum 2812 charger being energized and in bulk, e.g. >100A charge rate. When the charge rate is closer to or floating, the blowers run OK; if the charger is switched off/standby, also run OK. I also have an older 12K Mermaid, it runs fine, never has been a problem, so it's perplexing why the 2 new units are misbehaving. I haven't had the opportunity to duplicate the problem on shore power by discharging the bank to replicate the conditions when the charger comes on when the genset is 1st started. No problems with any other electrical gear on board.

A conversation with Mermaid was inconclusive, they will do some research. My suspicion is that there may some power-related noise from the charger that's getting into the power supply and causing a problem with the PC board and/or the phase shift in the blower motor. I'm not sure how to conclusively isolate the problem, or how to resolve it. Welcome suggestions.

I'm assuming at that the compressor is running, but the blower isn't blowing?

Since the compressor is a 240V load, it is wired hot1-to-hot2. Since, the charger is a 120V load (check me on that) and the blower is a 120V load (check me on that), they are wired hot-to-neutral.

...this makes me wonder about the neutral path when on the generator.

Questions:
-- Do you have an ELCI breaker installed for shore power?
-- At the times when the fan isn't starting, but is expected to be starting, is the compressor coming on?
-- Are the two AC units behaving the same, and are they using the same phase for the blower?
-- Is the charger using the same phase (hot wire) as the blower(s)? Or a different phase?
-- If you measure the voltage at the HVAC unit with the charger running in bulk vs with it off vs with it in float, how does it change, if at all? In other words, what voltages are you measuring across each hot wire and neutral and between the pair of hot wires in each case? (3 measurements in bulk, 3 measurements float/off).
-- Under any circumstances, is any voltage measured neutral-to-ground?
-- Do other 120V loads work on each phase? (Not just one or the other or neither)? Is 120V measured at other outlets between hot-and-neutral and hot-and-ground?
-- With all power sources off, and especially the generator off, and the source switch set for generator, what is the resistance between neutral and ground as measured at the generator? At the AC unit?

Now questions focusing in on the neutral...
-- Where is neutral tied to ground and when?

Neutral should be tied to ground...
-- On shore, when operating on shore power without an isolation transformer
-- At the output of an isolation transformer, if in use
-- At the output of the inverter, when operating on the inverter (often relay switched)
-- At the output of the generator, when the generator is in use (often switched by a multi-pole generator/shore-power switch)

Without knowing the details of the voltage measurements, it is really hard to hazard a guess. But, I'd be surprised if it is a noise issue, e.g. voltage spikes, so much as something else. But, if you have, or know anyone with, an oscilloscope, you can look. The voltage should look like a nice, pretty, clean sine wave @ 60Hz from any hot to neural or ground in any case.

Instead, my initial attention is on the fact that the neutral is wired different between the shore power and generator power and the neutral is in use for the blowers, but not for the compressors.

Let's imagine that, "somehow" when the charger is drawing a lot of power, it is floating the voltage up on the neutral. Maybe some type of resistance in the neutral path could cause this. This could lower the apparent voltage as seen by the blower.

Maybe neutral and ground are tied together when on shore power and this hides the problem because current is returning via ground and there isn't an ELCI/GFCI on-board or shore to notice?

Totally wild guessing. It'll take a lot more information to hazard a better guess. It might also help if there is a wiring diagram on the unit that can be photographed and uploaded.

I've run into several similar issues and they are truly maddening.


Can you tell if power is being switched on to the fans by the control board? That would at least tell you if it's a control problem or a motor/start cap problem. I'm guessing it's a control problem, and the fans are never being commanded on.


I hate to say it, but my experience trying to deal with similar issues has been an exercise in talking to brick walls. The problem has always been an electronic device design that is overly dependent on a pure and unlimited power source. Most vendors have no interest in debugging, let alone fixing the problem.


But there might be hope. My most recent issue was a Buderus/Bosch furnace that would error out when on generator power. The flame detect circuit wouldn't work and it never though there was a flame so the whole thing shut down and required a manual restart. Buderus was useless of course, so I wired up a lock-out to shut off the furnace while the generator was running. It was the only way to have unattended operations without losing heat.


Then, to my great surprise about 7 years later, I read that Buderus had released a new control module to fix this problem. I got one, and presto, it worked. For a variety of reasons I would never even consider a Buderus/Bosch heating system product again, but at least they fixed this problem.


In your case, since it's a Marine AC system, and boats with that product will close to 100% have generators and expect the AC to work on the generator, I think you are in a strong position to force them to fix it. The units should be under warranty, right? Hopefully being polite and persistent will get it fixed.

So to reiterate the ac units are the only things you’ve changed out and everything else worked before? There are a host of settings on the magnum inverters that can cause your symptoms if set incorrectly. If you haven’t tweaked them to add say a solar controller or some other charging or power source then I think you should follow the troubleshooting guide in post 2 for starters. One question just came to mind, If you discharge your batteries while on shore power to the point that the magnum charger goes into the bulk phase of charge while connected to shore power at the ac inlet do the fans also fail to start?

Thanks for all the suggestions & feedback. A/C units, pump, pump relays, blowers, all of the A/C system is ALL 220V. No neutral involved on any of it. Fans are being energized, they hum, but just don't turn, when this problem occurs.

Turn the charger off, presto! All works. Charger is 110V. As the charger tapers off, the problem disappears. I've not 100% confirmed this, but initial DX seems to bear this out. Neutrals have been sorted out previously, there's no co-mingling of neutrals served by inverter and the shore power neutrals. The A/C's have no neutral connection, so it's really moot. GFI shore power doesn't trip.

TT makes a point I've experienced before. As a retired HVAC guy, I've had experiences where a control board is flakey, no rhyme or reason, but a different revision suddenly works perfectly. Not at all uncommon. The fact that the older M12 unit works fine, and both the new ones are unhappy is pretty compelling. Mermaid has already offered to warranty swap boards, but I'd like to get a better handle on the root of the problem before exercising that option.

So the battery charger is acting like a resistive load but in bulk phase delivering constant current which may not leave enough to start the fans. I would experiment with lowering the ac in current setting for the charger or lowering the max charge current threshold. You might also just change the percentage of power the inverter allocates to charging or even the max charge rate for the batteries in the battery setup. These settings are all buried in the menu tree under setup for the inverter charger and battery type.

I think cafesport hit the nail on the head.

cafesport said:

So the battery charger is acting like a resistive load but in bulk phase delivering constant current which may not leave enough to start the fans.

Click to expand...

Not likely... this is a 12.5 kw genset. Charger may use 18-20 A on one leg, the blowers are 220V, they're drawing in single digit amps, if that. If it were a voltage sag issue, I'd see consequences in other areas. Voltage while A/C is starting up doesn't dip or change a bit from about 232 V. I really suspect a noise issue that the shaded pole motors are unhappy with. I can call for the blowers to run independent of the compressor relay, they exhibit the same symptoms with or without the compressor.



Magnum is on my call list, I'll post if they offer any enlightenment.
Appreciate all the suggestions.

Since the blower is humming but not turning, it sounds like you have a power factor (voltage vs current phase issue) caused by the charger in bulk mode.

Those AC's could probably be spec'd or wired for either 120 or 240v. Double check the wiring. Manual probably has directions for setting it up for either voltage.

Are you sure there is no neutral connection?? And it could be an open N on your genset.

My theory: The Magnum charger is a highly non-linear load (i.e. a switching power supply) creating a large power factor distortion when in bulk mode. It is enough of distortion to prevent the shaded pole motor from starting. Ask Magnum what power factor correction they have at large loads-my guess is none. Shore power is more resilient to the power factor distortion compared to the small generator.

Ski in NC said:

Those AC's could probably be spec'd or wired for either 120 or 240v. Double check the wiring. Manual probably has directions for setting it up for either voltage.

Are you sure there is no neutral connection?? And it could be an open N on your genset.

Click to expand...

The AC's are 220V only. They are not dual voltage. No neutral connection as there is no need for it since none of the components are 120V.

I think we'll be able to be much more helpful with those voltage measurements. It is just too hard to guess, otherwise. If you suspect noise, try to find someone with an oscilloscope who can take a look. I wouldn't bet on that, but that's just me. Looking is certainly better.

The only documentation I was able to find online for 220V mermaid AC units refers to a neutral wire. But, it is probably the wrong doc.

-- https://www.mmair.com/wp-content/uploads/2016/08/Install-Guide-2016-PDF.pdf

If you can upload the wiring diagram for yours, that would really help. They normally put them on the units, themselves.

Well once you eliminate the ac unit itself, wiring or the inverter’s charger load, all you’ve got left is the genny. I’ve seen a few with fuel and air leaks that cause power loss. I would also check out the frequencies or rpm for the unit loaded vs unloaded? Could be as simple as a fuel or air filter. Can you pull full load without an induction load?

So when the fan is energized but not turning if you manually turn the fan (use a pencil not your fingers) will it run? I assume you used a volt meter on the fan leads with the different power scenarios is there a significant difference?

gsholz said:

My theory: The Magnum charger is a highly non-linear load (i.e. a switching power supply) creating a large power factor distortion when in bulk mode. It is enough of distortion to prevent the shaded pole motor from starting. Ask Magnum what power factor correction they have at large loads-my guess is none. Shore power is more resilient to the power factor distortion compared to the small generator.

Click to expand...

Yep. Related to harmonic currents causing those shaded pole motors to lose starting torque.

Magnum 2812 specs say PFC > 95%. If it is working to spec, I'd guess that should be plenty good enough.
-- http://www.magnum-dimensions.com/si...e-MagnumEnergy_64-5024_revA-singlepgs-web.pdf

diver dave said:

Yep. Related to harmonic currents causing those shaded pole motors to lose starting torque.

Click to expand...

And would explain the transient nature of the problem. I'm waiting for Magnum to return my call.


If I give the motors a push, they stay put, just hum.



To reiterate, no voltage drops, if this is a loading problem, I'll eat my hat. The blower motors are such a miniscule load compared to the 12.5 kw capacity, and there's no voltage drop on the supply when the compressor runs. There's no voltage drop when the charger loads up either.

I don't think it's a power factor issue, but rather clipping of the AC power wave form. A lot of chargers draw power only at the top of the wave form, and flatten out the voltage by doing so. The harder you charge, the flatter it gets. Back off on charging, and eventually the shape recovers enough for the motors to start.


Your charger is 120V, so all the flattening is happening on one leg, which is probably making matters even worse.


It's not a cheap or easy experiment unless you happen to have a transformer kicking around....... but I have seen where a transformer has been used to spread the 120 charger load evenly across the 240V power sources, which is your generator in this case. You can use a 240V to 120V transformer, or you can use an autotransformer. Either will present the generator with a balanced 240V load rather than a unbalanced 120V load. Now with all that said, it's only a guess whether that would sufficiently improve the wave form to make things work.....

If we assume that the HVAC units are of the model and configuration described, wired correctly, and the voltage is not dropping, the problem is with the charger, not with the AC units. It may just be that the new AC units are more sensitive to it than the old ones.

The charger is advertised to have a power factor correction (pfc) of more than 95%. This should be plenty good to ensure that anything can run. A blower of all things should run with a 5% loss. Something has to be way off on the input power for it not to move even after kicked. Way off. This just isn't possible with a 95% PFC and no measured voltage drop. By definition, power factor is the ratio of the real power to the load divided by the apparent power. 95+% only allows for less than 5% loss due to waveform disturbance of any kind: Voltage, frequency, harmonics, etc. And, even absent the published spec, I'd be /shocked/ if a device like that had a PFC less than 90%. It is just unheard of. PFCs that far off mess with utility power generation and distribution, too.

Could it be, as twisted tree suggests, a clipping of the voltage vs harmonics? Sure. But, this wouldn't satisfy the criteria of 95+% PFC. The charger would be bad. And, more importantly, if the waveform were clipped -- it would be measurable on a volt meter. And, in this case, a voltage drop has been denied.

I'd really like to know the /measured/ voltage numbers and if they are measured as RMS or peak. I'm not sure how it is being determined that voltage isn't dropping. "Most things work", if that is the test being used, may not be the right test, because "most things" may be insensitive or have power supplies that can do their own correction that a fan expecting sinusoidal power does not have, etc. I really think measured voltage has to be changing here, even if only as a side effect of other problems.

It would be interesting to put an oscilloscope onto the power and see what the waveform looks like with and without the charger.

Regardless, if the charger affects the waveform enough to cause this problem, it isn't meeting its spec of PFC > 95%. It is defective and in need of repair or replacement.

One thought that comes to my mind. What if the HVAC units are European 240VAC/50 Hz units, instead of US 230VAC? I need to think more about this to understand if /anything/ would be expected to work. I haven't thought it through and can't right now. But, /maybe/ things would work, but only very marginally so -- but be brittle enough to break with a small change from the charger.

...I haven't thought this one through.

In any case, I'd love to see the voltage numbers and a waveform if possible. But, I think the problem is with the charger.

For the record. Power factor is different than harmonic currents. PF describes the angular difference between a load current and the supplied voltage. Harmonic current is an unwanted conducted emission from a load. These are not related. Load2 wont care about load1 PF, as one example.
Shaded pole motors already have a low starting torque. Although i have not worked with a SP motor with a capacitor. Maybe a “run” cap??

Right, a device that draws all its current at the voltage peak wasn’t necessarily reduce it’s power factor. At least that how I remember it from many years ago.

A rectifier, which is the first thing in the circuit for a charger, behaves exactly this way. Until the AC wave form voltage reaches the voltage of the capacitor bank, there is zero current draw. Then as soon as it goes over, it’s balls to the wall. Non linear is the technical
term, not balls to the wall. I’d stake a case of beer or some other beverage that this is what’s happening. And with the flattening on one leg and not the other, the motor gets a push one way, but never the corresponding pull in the other direction.

To see it you would need a scope, or some thing that can see the flattening. Maybe something that records peak voltage, but in this case only one direction will be flattened, so it might not show up. RMS will almost certainly not show it, and that’s what 90% of meters show.

twistedtree said:

I don't think it's a power factor issue, but rather clipping of the AC power wave form. A lot of chargers draw power only at the top of the wave form, and flatten out the voltage by doing so. The harder you charge, the flatter it gets. Back off on charging, and eventually the shape recovers enough for the motors to start.


Your charger is 120V, so all the flattening is happening on one leg, which is probably making matters even worse.

Click to expand...

That makes a lot of sense. I'm wondering if the capacitors on the motors might be able to be tweaked (alternate uf rating) as a workaround. And then there's the issue of the older 12K unit that's not affected. I may have a closer look at the capacitor on that one's blower or swap it to the affected 12K to see if the problem follows. I've also dropped a tech support note to Micro Aire, manufacturer of the FX-1 board.

Although I like the versatility and features of the FX-1 controller, I acknowledge that there's always potential for problems that might not exist on a simple relay package.

I think we are getting closer. The switching power supply is a highly non-linear load when at max charge current. This introduces large higher order harmonics in both voltage and current wave forms (and the higher harmonics impact power factor). This is what prevents your new fan motor from starting. The old fan motor obviously was more tolerant, maybe it had dual/quadruple shaded poles etc.

A work-around could be to dial back the max charging current when using the A/C.

It would be interesting to put a scope on the voltage/current as seen by the A/C fan motor.

It would be an interesting experiment to strip off the harmonics using a 1:1 240v isolation transformer feeding the fan motors. 60hz transformers will be lossy at 120 and 180 Hz. Rather a long shot.

I don't want to distract from this thread. It is really interesting with a lot of smart people.

But, my understanding is that harmonics will generate distortion causing non-reactive power, lowering power factor. So, whereas power factor and harmonics are not the same thing, harmonics from non-linear loads reduce power factor.

Specifically, I think that total power factor [tpf] is the combination of displacement power factor (caused by old school things things like inductive loads) and harmonic power factor [hpf] (caused by things like switching power supplies).

So, if the charger is generating harmonics of a type and level that would matter, as a side-effect it should have a lower power factor than spec'ed 95+%.

Of course, if the charger is broken, it doesn't matter what the spec says. And, /if/ everything else is right, but a simple blower motor won't start -- it is pretty clear the charger is bad. That isn't a particularly sensitive load and I can't imagine a marketable product being designed to cause such problems when operating correctly.

The experiments suggested, especially lowering the output charging current (which is free to try), seem like really interesting ideas to me.

Without knowing the delta of the voltage (even RMS) at the fan between the various power sources this is just speculation. The generator may have a low voltage output, the frequency may be off. this is always going to result in a lower voltage at the fan when the charger is running. A 12 k generator if it is functioning correctly has basically the same current supply as shore power unless this is a hundred amp supply which I doubt.

Just grab a hairdryer and discharge the batteries at the dock with the charger turned off. When the batteries are down 10% fire up the charger and the air conditioner. If the fan starts you have a generator problem but you might be able to create a workaround by reprogramming the charger.

On every boat I have had, the air conditioning was on a totally different circuit, not connected to the battery charger or inverter in anyway. On my current boat it is 220 V. I can shut off the battery charger entirely as well as the inverter and still run the air conditioning.

This sounds to me like a problem between a ground and neutral. I had this problem on our current boat shortly after buying. The previous owner had crossed the neutral and ground on the icemaker. When motoring, and when the icemaker was being powered by the inverter, I would get a polarity error, and no working icemaker. However, if I flicked the selector switch to generator, from shore power, The icemaker would power up when underway. When I switched the selector switch back to shore, I would again get a polarity air and the icemaker would stop working.

The electrician who found the issue said that when the selector switch was set to generator, the missing Ground was provided by the generator.

Good luck,
Gordon

Knucklehead mechanical engr here. Cool to have some EE's here, indeed. Instead of debating harmonics and power factor (yes I know some of that stuff), I'd put a DVM on motor leads and see what the volts are when it is in stalled condition.