Yep, I agree. I probably haven't seen the last of this. I did study the traces that I could see and they all seem really well done. They appear gold plated, but that could be the amber encapsulation on top. It's hard to see everything because of the encapsulation compound and all the parts in the way, but it looked really solid. I used a 45x scope on some of it, but that didn't reveal anything.

I didn't tell you that the heater is a pool heater. I sure as heck am not paying $1400 for a pool heater blower!!! SO I will be on top of this one!

By the way, I think the variable speed is related to basic function and not so much savings. The gas needs to have the right draft to burn efficiently and different lengths of ducting with different bends, etc can make a huge difference. There is a pressure switch that must have air pressure within a range to work or the motor is turned off. There is a procedure for adjusting a pot on the board to have it go faster or slower. I don't think the speed ever changes otherwise. Once it's set, it's set. But if a bird build a nest in the stack, you might have to turn it up!

 

Ok, thought this was a furnace blower, for the duct work.

 

It is just like the exhaust blower used on a high efficiency furnace. Same basic idea.

 

I am sure you are all less than surprised, but it still doesn't work. In a way, I'm happy about that, because I haven't forgotten everything yet, like I would have if it happened in six months.

When I was at this stage before, I disconnected the 120AC supply lines and plugged them into another line that I also tested - didn't work. I removed the assembly and after a little shock and vibe and plugging the leads into an AC socket (and getting some nice arcing), I got it to start up. This is where I was when I posted the first message.

I am certain there is something wrong with the motor's start circuitry. Yes, it could be a bad solder joint, but I have studied that pretty closely and that is not bearing fruit. SInce the board is all gooped up with an encapsulate, it's hard to actually get a meter or scope leads to make contact. I could file them down. Any suggestions if testing circuit elements is the next step and if so, how?

I ams also wondering if maybe the basement location is cold and thus causing the anti surge to be at high resistance, perhaps too much for the motor to overcome? Maybe I could take a hair dryer down there and blast it to see if it behaves better when warm? I'll go try that next, but I'm not optimistic.

I am ready willing and able to go deep into diagnosing this. It's a learning experiment for me. I'd also like to use my new little USB scope if that could be of service. If I come out of this with a working motor, I'll be happy, but if I come out with solid experience in diagnosing a fairly complex circuit, I'll be thrilled (assuming the motor works too).

 

Okay, I tried to heat up the board a bit with a hair dryer then applied current. It started immediately, probably before it could have warmed up much. I let it cool and tried to restart. No luck. I warmed it up. Nothing. I warmed it a little more and bumped it and taped it. Nada. Then I took the motor cover off and turned the motor by hand and powered it on. It worked. I tried without my help. No luck. I tried again with a little turn. It worked right away. I think I can nearly remove the intermittent behavior that is so troublesome when troubleshooting. I can now make it work by hand every time. It mostly won't work on it's own. There are the few cases where it will work on its own, but I cannot get that to consistently happen despite trying different current sources, different temperatures and even a bit of shock and vibe. The motor is fine. The blower assembly is fine. The controller is fine, except for one thing. The starter circuitry isn't working. That's the place to dig in on. I could really use advice on what steps to take next. Ideas:

Learn more about how this controller board and IC actually works so I can better understand the thermal stuff and the start surge protection stuff. Try to figure out what components or parts of the circuit would fail in the way. Maybe draw a schematic of the start circuitry? I don't think the manufacturer (Ametek) will help given that it has a big label that says, "do not open, do not service". They prefer me to drop $1400 for a replacement despite of the whole assembly despite the blower and motor working fine and just an isolated part of the start circuit of the controller being bad.

Perhaps I could use my USB oscilloscope to test what's going on in the circuit? i don't even know how to use it yet, so could be a good idea for learning or a dumb idea to destroy my new toy.

Maybe figure out some other way to test the components in the starter in circuit until I find a failed component? Ideas?

It looks like most of the actual board is used as the power supply with the bridge rectifier and bank of big caps. Getting those caps charged up could cause a surge and there is the thermistor in place for that. Maybe it really is bad? How can I do the math to know what resistance would be needed when charging the caps to protect the rest of the circuit?

I never did figure out the fuse, but I think that's not it since the start does work sometimes. I'm assuming the fuse is really just a resister somewhere and might even be this SG421 thermistor. It does function be reducing resistance with an increase in temp, but I don't know if it is doing so to spec or to the ultimate need. Also, even when I heat it up, the motor won't start, so I'm thinking it probably isn't that one.

The AC input seems to go first to a diode bridge then on to the three big caps in parallel. I assume that's some kind of charge pump needed to help reduce the ripple in the bridge, but I'm over my head. I'm guessing that the little IC takes data from the motor position hall sensors and at heat sink thermistor and if all is well, it plays round robin with precisely timed firing of the P and N channel mosfets and letting the big caps fire the coils on the motor. Does that sound right? There are six low voltage transistors in addition to the 3 hall sensors and a bunch of little diodes and resisters, plus the pot for speed setting.

If I'm on track, here are the potential culprits:

- that SG421 doohickey is bad and causing resistance so high the motor can't get going.
It is right in the middle of the action and I did see that burned area, but my gut tells me that is not the problem
- Maybe one of the mosfets is bad? Since the motor has three phases maybe two of the tree won't let it start and only if it ended up on the other one would it start automatically? Maybe the magnets are cracked or not positioned just so or a winding is not just right? With enough momentum, it doesn't matter? The motor is very fast, smooth and quiet, so that all doesn't seem right.
- perhaps the low voltage IC circuit might be signaling somehow not to start. The thermistor on the heat sink would be tell the IC to stop if it was hot, so that's probably not it...
- what about a bad cap? There are three big electrolytic caps in parallel. If they didn't have the needed capacitance, that might explain the no start issue. However, once up and running, the caps are doing more of a filtering duty, I'd assume. Maybe they are good enough for that job, but not the start? Could that explain the intermittency, though? Maybe so if they are just off a bit too much and other variables in the circuit like power supply fluctuation, motor position, heat, etc.

Any of these investigations will take a lot of time. Any suggestions greatly appreciated. Thank you for your support in my little man versus PCB challenge. the PCB is up one point, by the way...

 

I took the motor cover off and turned the motor by hand and powered it on. It worked. I tried without my help. No luck. I tried again with a little turn. It worked right away. I think I can nearly remove the intermittent behavior that is so troublesome when troubleshooting.

...

- Maybe one of the mosfets is bad? Since the motor has three phases maybe two of the tree won't let it start and only if it ended up on the other one would it start automatically? ...

Now that we know it's 3phase;
I have seen behaviour like this in a 3 phase motor before when only 2 phases were connected. It would just sit there and humm (you didn't describe any humming, but I witnessed this on a 10HP motor so that might make a difference) until I spun the shaft by hand, at which point it would spin on it's own (with drastically reduced torque; I could stop it by hand). Maybe check to see if all 3 phases are working. Your meter may or may not measure the output of the inverter circuit; I suspect you would have a better chance of getting a measurement if the motor were connected.

post#400

 

Here is a link to Amtek technical documents; http://www.ametektip.com/index.php?option=com_rokdownloads&view=folder&Itemid=120&lang=en

May be a schematic there.

Here is a link to a troubleshooting guide for Amtek; http://www.enviroequipment.com/remediation_new/documents/101-404.pdf Page four is the guide.


Here is another link that describes the way the motor and drive circuit works that might be of some help. May not be your model (you didn't give model) but the basics should be the same; http://www.farnell.com/datasheets/311553.pdf

 

Now that we know it's 3phase;
I have seen behaviour like this in a 3 phase motor before when only 2 phases were connected. It would just sit there and humm (you didn't describe any humming, but I witnessed this on a 10HP motor so that might make a difference) until I spun the shaft by hand, at which point it would spin on it's own (with drastically reduced torque; I could stop it by hand). Maybe check to see if all 3 phases are working. Your meter may or may not measure the output of the inverter circuit; I suspect you would have a better chance of getting a measurement if the motor were connected.

post#400

The second link, troubleshooting, says that in it You type faster than me

 

That is a great help! I had scoured the Ametec site earlier and didn't find what you found. Unfortunately, you were searching blind without the model and those aren't the right files. The troubleshooting one doesn't mention the no start problem. Sorry about leaving off the model number. In any event, you sent me back looking and I did find a couple of sales sheets with google that helped me determine the watt rating. There are two sheets I found with only slightly different specs. I attached a copy of these in one pdf file here. Your tip also sent me back digging at the Ametek site. Armed with new info, I think I found a users guide for the specific group of blowers including this one. It even has a block diagram on the last page. I'm studying now. Thank you!

 

Now that we know it's 3phase;
I have seen behaviour like this in a 3 phase motor before when only 2 phases were connected. It would just sit there and humm (you didn't describe any humming, but I witnessed this on a 10HP motor so that might make a difference) until I spun the shaft by hand, at which point it would spin on it's own (with drastically reduced torque; I could stop it by hand). Maybe check to see if all 3 phases are working. Your meter may or may not measure the output of the inverter circuit; I suspect you would have a better chance of getting a measurement if the motor were connected.

It doesn't seem to hum at all when it won't start.

I wonder if I might be able to meter somewhere if turning the wheel by hand and not plugged in? I suppose the windings would do something if I could figure out where to test. I could also do it on the bench under power, but I'm not sure what to test. Should I use the logic analyser on the mosfets bases (or whatever they are called on a mosfet)? Never tried that, so not sure what I'd learn. I just fear frying something. Would I be testing the voltage off the high voltage side of the mosfets with the DMM? How can I use the oscilloscope to get a picture of the wave form generated by the motor? If one phase were off, perhaps I'd see a lower amplitude? Any suggestions?

 

It doesn't seem to hum at all when it won't start.

Postulating why it would or wouldn't hum is where my knowledge ends. Could be that the 120V fractional horsepower motor doesn't humm because there's not as much juice flowing through it as there was flowing through the 480V 10hp motor I dealt with; or could be a totally different issue. I could only throw guesses in the dark from here out. were you able to measure voltage across all 3 phases?

 

My review of the docs provided a few clues. Note that the users guide is for 5.7" blowers, and this one is 8" so it may be somewhat different. Irt seems consistent, however.

Findings:
- The thermistor/ NTK resistor on the heatsink is the thermal protection and the SG421 is for inrush current. Still a potential fault there.
- Can't find the fuse in series with the 120 VAC source unless it's the big axial resistor I see in that position...
- biggest finding is mention of three modes of operation and the one I think I'm using is called analog. It says,

In Analog Mode, combination of internal potentiometer setting and input voltage prevents unit from starting.

My next step is to see if reducing the resistance with that pot helps.

 

Postulating ... is where my knowledge ends.

Wait, don't go away yet! We're getting close.

 

Wait, don't go away yet! We're getting close.

don't worry I never make it very far

 

This might seem like a dumb question, but, are you sure the problem isn't in one of the off motor controls? The controls that signal the motor to run could be the problem. Is there a wiring schematic in the housing/case or the door of the heater?

I went and looked at my high efficacy furnace and it just uses a shaded pole motor for the exhaust blower, or if worse comes to worse maybe you can retro-fit some thing like that.

 

This might seem like a dumb question, but, are you sure the problem isn't in one of the off motor controls? The controls that signal the motor to run could be the problem. Is there a wiring schematic in the housing/case or the door of the heater?

I went and looked at my high efficacy furnace and it just uses a shaded pole motor for the exhaust blower, or if worse comes to worse maybe you can retro-fit some thing like that.

Anything is possible, however I find it strange that the motor will run if you spin it by hand. I would expect a controls issue to be more like speed is "just plain wrong". He still hasn't answered whether or not he has voltage output on all 3 phases; that would be the fisrt thing I would check; each phase to ground and phase to phase with the motor connected and disconnected

You do have a multimeter, right?

 

Anything is possible, however I find it strange that the motor will run if you spin it by hand. I would expect a controls issue to be more like speed is "just plain wrong". He still hasn't answered whether or not he has voltage output on all 3 phases; that would be the fisrt thing I would check; each phase to ground and phase to phase with the motor connected and disconnected

You do have a multimeter, right?

It's a bit of work to remove and very hard to test in position. I do plan to test the pot and at the same time will test the voltage of the motor phases as you suggest. I'll have to work on my yoga to fit in there, but the results should be enlightening.

Questions. I think you are suggesting that I make 12 motor phase tests. Each of three phase to ground and phase to phase while both connected and disconnected from source supply?

I can find ground but where do I test phase? There are the 3 pairs of mosfets that I could reach the legs of, but otherwise, I'm not sure how to probe the phases. I may need to pull it out for that anyway, but nearly as good on the bench and I could use the oscilloscope that way too.

By the way, the motor needs a minimum voltage to start as part of it's control circuitry, not just to overcome the inertia. A low voltage combination of high pot resistance and low supply voltage results in no start. This makes sense and the user's guide for the similar motor says so. The issue I have with this is that the supply voltage is tested good ~124VAC and the motor speed is correct for the exhaust pressures, so if the pot resistance must be lessened to enable start up, that means motor will spin faster. A little bit is fine, but it's not actually needed for the exhaust system, so seems dumb to have to speed up the motor just to enable it to start. For such a premium cost device, they should have isolated the start control from the speed control. I see no need to have a pot controlling start voltage. The IC knows when the motor is spinning, so it could easily signal a transistor to enable a speed control pot once it was spinning.

 

It's a bit of work to remove and very hard to test in position. I do plan to test the pot and at the same time will test the voltage of the motor phases as you suggest. I'll have to work on my yoga to fit in there, but the results should be enlightening.

Questions. I think you are suggesting that I make 12 motor phase tests. Each of three phase to ground and phase to phase while both connected and disconnected from source supply?

I can find ground but where do I test phase? There are the 3 pairs of mosfets that I could reach the legs of, but otherwise, I'm not sure how to probe the phases. I may need to pull it out for that anyway, but nearly as good on the bench and I could use the oscilloscope that way too.

By the way, the motor needs a minimum voltage to start as part of it's control circuitry, not just to overcome the inertia. A low voltage combination of high pot resistance and low supply voltage results in no start. This makes sense and the user's guide for the similar motor says so. The issue I have with this is that the supply voltage is tested good ~124VAC and the motor speed is correct for the exhaust pressures, so if the pot resistance must be lessened to enable start up, that means motor will spin faster. A little bit is fine, but it's not actually needed for the exhaust system, so seems dumb to have to speed up the motor just to enable it to start. For such a premium cost device, they should have isolated the start control from the speed control. I see no need to have a pot controlling start voltage. The IC knows when the motor is spinning, so it could easily signal a transistor to enable a speed control pot once it was spinning.

Yes, 12 would be the total. it may seem excessive, but as you say, "It's a bit of work to remove and very hard to test in position." and it would be advantageous to test and record everything you can all at once to avoid having to take it apart again and test for some theoretical problem that might be suspected later. The reason I suggested testing both with the motor connected and disconnected is because, not knowing what type of meter you have, it may not measure the PWM output of the board correctly with the motor disconnected; the motor may serve as a filter to give you a better reading. Conversely, reading with the motor connected could give you false reading; voltage from 2 working phases could give you a voltage reading on the 3rd phase, even if the board is not outputting to the 3rd phase.

As far as where to check the phases; I just re-reviewed your pictures and noticed that the stator is part of the board, so that complicates this a little. Normally all this testing would happen at the motor T-leads, but you don't have any. Those magnet wires have to come off the stator and terminate to the board somewhere, so that would be your motor leads. if you wanted to test with the "motor disconnected" you would have to cut those (this would be a good time to also check the amp draw of each phase); that's up to you whether you want to do that or not. if you have a current clamp meter that is small enough to fit in there, that would be even better.

 

I took the motor off for the pot and voltage tests. Turning the variable speed pot to minimum resistance did not enable the the motor to start on its own. I was hoping on that being it, but no such luck.

Testing the T-lead voltages was a bit more complicated. I tried probing the mosfets, but didn't get far - to hard to reachespecially with the motor running. I followed the T-leads and soldered on test leads to the back and checked for equal resistance to assure my soldering was okay.

I then ran a ton of tests. I won't bother reporting all the results, unless you want any, because I found something interesting. After testing every other thing I could think of, it dawned on me what what I needed was the voltage at precisely the time the motor is failing, i.e. powered up, but not started/turning. It took a bit of time to get it to fail, but once it did, I listened for a hum and heard nothing but silence. I then tested and got only 0.9V from each phase! perhaps that low voltage is why there is no hum and no turn? It seems that the windings are not given enough voltage to overcome the inertia of the motor. With a little helping hand, the motor turns and the voltage jumps up.

Now what? Any ideas? I can provide any data you might request now that I'm set up to test. Also I do have the oscilloscope too.

We have to be getting closer...

 

Sounds like one of the phases is dead, which is why it only fails to start occasionally. If the rotor happens to stop on the dead phase, it won't have any torque to start it moving.

You might start by checking each MOSFET, and trying to see if you get output from each Hall-effect sensor.