Clippers, Resistors & Burned Slip Rings

Started by devo, December 21, 2013, 09:02:35 PM

Previous topic - Next topic

boB

Quote from: RossW on December 22, 2013, 08:36:58 PM
Quote from: boB on December 22, 2013, 07:51:41 PM
OK, so if this is NOT the problem (Clipper fail-safe) then it must be that the
clipper resistance is WAY WAY too high and needs to be raised.

Guessing that was meant to be WAY WAY too low and needs to be raised.


OOOoooPS !!! Yes, you are right !  Thanks Ross !

Way too high of a LOAD maybe...

boB
K7IQ 🌛  He/She/Me

dgd

#16
Quote from: devo on December 22, 2013, 07:06:34 PM
We don't normally see 2500 watts, but in normal operations we have seen 2500 watts. I just looked at the offline data from the controller and it shows over 2.5Kw of peak power on two occasions, with 1.5Kw on two other days and the rest were well under 1Kw. The highest Vin the controller saw was 142.5v when the turbine was producing 1.5Kw. The last day it produced power the turbine's peak was 136w.

Sorry I worded that badly, I knew what you meant. Even so I would think that slip ring arcing was probably started at one of those high output times. Once damaged the situation would deteriorate even in normal use until the ring's brushes/contacts failed.

Quote
I agree that this turbine isn't the right fit for this area. We need a three blade turbine instead of a five. I'm not convinced a mechanical furling mechanism is good though - I'm not sure it would last.

I always found tail furling systems very reliable, mechanical stressing is probably a tiny fraction of the overall stress the rotating turbine endures..
Electronic/Magnetic and blade twisting schemes for braking as the MAIN braking methods just dont do it for me. Give a furling tail every time.

Quote
Like you, I think the braking is causing the issue. It's just a pure guess on my part, but it seems like there is just too much amperage being put through the slip ring when the clipper tries to slow it down. This is why we are looking at the resistance value and trying to determine if it is too low. We've checked the wiring and settings a bunch of times. We even replaced the clipper with another one to see if that was the issue.

Not sure that this could be the reason for failure but could be. The slip rings are the weakest, highest resistance point in the wiring from the turbine to the Clipper so if there is to be a failure thats where I expect it would happen, apart from a dodgy connection somewhere..
Any evidence of lightening strikes on the turbine/mounting/pole?
Quote
Is it possible that the 4Kw load is too large, causing the turbine to basically short instead of slow? A higher resistance would reduce the amount of amperage coming through. You're using 1.3 Ohm's per phase or you rewired it higher? How high should we go for testing? I was thinking of going to 4 Ohms per phase to see how the turbine reacts.

I ended up with two 1.3 ohm resistors in series on each phase, giving 2.6 ohms per phase. Even now I think this may be just getting a little high and I may look about for 1.1 or 1.2 ohm resistors.
In you case you have 1.6 ohms per phase now and its bringing the turbine to a dead halt. If it were me I would double to 3.2 (or thereabouts) and see what happens.

Quote
The only other option we've looked at is removing the clipper and going to a full diversion load - i.e. keep the turbine completely loaded at all times. All of our other sites run this way and we have not had a single issue in the 18 months or so they have been running. Granted, this is the first time we have tried a turbine this large. The other sites use 600w units.

You mean just connect the output from the 3 phase rectifier to the battery then use a pwm controller (eg C40) to introduce resistive load as diversion?
I have done a similar setup with a remote comms site. A turbine through 3 phase rectifier to battery.
Except I use the wastenot AUX2 of the Classic to connect to a 3phase SSR that has three WW resistors one per phase of the turbines AC legs.

Anyway it seems boB has the problem being sorted, I hope you get it resolved

dgd
Classic 250, 150,  20 140w, 6 250w PVs, 2Kw turbine, MN ac Clipper, Epanel/MNdc, Trace SW3024E (1997), Century 1050Ah 24V FLA (1999). Arduino power monitoring and web server.  Off grid since 4/2000
West Auckland, New Zealand

johnnymac

Our clippers were shipped with 6 (3ph X 2 in series) .8 Ohm resistors for a total resistance of 1.6 Ohms per phase. I've been speaking to Devin and we're going to double the resistance to 3.2 Ohms per phase to see if we can cut down on the current a bit and perhaps create a smoother landing. I'd still like to know why the clipper was causing the turbine to "clamp the brakes on" though, as I expect the resultant back EMF must be horrendous for a microsecond or 2.

I won't be replacing the load today though for 2 reasons:

1. It's Tips eve
2. The voltage regulator blew on our skidozer and I'm not walking 19KM through the snow. :)



devo

QuoteI always found tail furling systems very reliable, mechanical stressing is probably a tiny fraction of the overall stress the rotating turbine endures. Electronic/Magnetic and blade twisting schemes for braking as the MAIN braking methods just dont do it for me. Give a furling tail every time.

We're always looking at other options. We may just have to try a different turbine with a mechanical furling mechanism. If you had to choose a turbine in the 1Kw range, which would you go with? Bergey?

QuoteAny evidence of lightening strikes on the turbine/mounting/pole?

We haven't seen anything that would be indicative of a lightning strike. I would expect some of our radios would also have been damaged if this had happened.

QuoteI ended up with two 1.3 ohm resistors in series on each phase, giving 2.6 ohms per phase. Even now I think this may be just getting a little high and I may look about for 1.1 or 1.2 ohm resistors. In you case you have 1.6 ohms per phase now and its bringing the turbine to a dead halt. If it were me I would double to 3.2 (or thereabouts) and see what happens.

I'm looking at perhaps testing a single 3.9 ohm 1000w resistor per phase. This would tell us if the resistance/load is causing the appearance of "jamming" the brakes on. In order to test it i'm thinking we could manually charge the battery bank with a generator and charger and see what the turbine/classic/clipper does with the batteries in float. If the batteries aren't in float we won't see the clipper engage. I'd like to actually be on-site when it happens, in case the resistance and/or load is too small. Don't want a fire!

QuoteYou mean just connect the output from the 3 phase rectifier to the battery then use a pwm controller (eg C40) to introduce resistive load as diversion?

Yeah we've done this in the past and it has worked quite well. Although now that i think about it those loads use 3 ohm resistors!

We're open to testing out different things to try to find a solution. I'm not convinced that its a problem with the clipper/classic though. If the 3.9 ohm resistors don't show any change we may just have to look at other turbine options.

I have sent the data and pictures to the manufacturer and they are looking at it as well. I must admit, they have been helpful when things have gone wrong. I don't think they would be a good choice for someone looking for a home turbine, but they are responsive at least.

Midnite has some of the best product support I've encountered in a long time. boB and Ryan have been great answering questions, as well as everyone on this forum. It's nice to see a company that takes pride in its products and really tries to help its end customers. So many businesses these days just want to sell you a product and forget about it!

Thanks again everyone!

boB

Quote from: johnnymac on December 23, 2013, 09:19:29 AM
I'd still like to know why the clipper was causing the turbine to "clamp the brakes on" though, as I expect the resultant back EMF must be horrendous for a microsecond or 2.

This is exactly the reason why I much prefer AC clippers (when possible) over DC clippers.
Turning off at 0 current pretty much ensures there is not going to be an inductive
voltage spike.   I think that if 1.6 Ohm per phase is too small that double that is also
going to be too small but I'm just not sure.  10 or 12 Ohms might be just fine.

In fact, if you were getting 1500 watts at 143 volts, that would mean that the load the
Classic was giving the turbine is around  1500/143 = 10.4 Ohms.

I seem to remember you saying you were getting either 1500 or 2500 watts at that input
voltage but I can't seem to find that info right now...

Are you using the stock HY-2000 curve that is in the Classic ?  Since your turbine is a 1500 watt
rather than a 2000 watt rated turbine, the current for each of those voltages probably needs
to be lower BUT since (I think) you really want it to stall more than it does, I would leave the
currents in the curve alone and bring down the voltages.   Here is the default array of voltages
and currents for the HY-2000 that is built into the MNGP for the Classic.

Cut-in is 62 volts at 0 amps, 66 volts for 17 amps, etc.  These values were empirically found
and could be optimized I'm sure.  I am sort of suspect at the 17 amps for the first non-zero
current step in the curve but I don't have one of these turbines or the wind to test it if I did.

WndPwrTblVdfltwindmax_2000A[] = {62,66,70,73,77,81,85,89,93,96,100,104,108,112,116,120};//V in
WndPwrTblIdfltwindmax_2000A[] = {0,  17, 20,23,27,31,35,39,44,49,54,60,66,72,77,78};  //I out

Also, maximum current is 78 amps at 120 volts and your current limit is set to 55 amps so that current
limit is going to immediately make the input of the Classic go high and unload the turbine at around
step 11 (54 amps out, 100 volts in) so it will not make it go higher than that step.

I think that if you simply lower the input currents so that they mainly stay below your Classic
current limit, it would help to stall the turbine more and keep the voltage from going so high
which would also limit the amount of clipping.

Of course, to keep from clipping the input, if you could add battery side diversion in addition
to the lower power you will probably get from the de-tuned power curve, that should
keep the clipping way down.

Higher resistance is also something you will hopefully try.

Quote from: johnnymac on December 23, 2013, 09:19:29 AM
I won't be replacing the load today though for 2 reasons:

1. It's Tips eve
2. The voltage regulator blew on our skidozer and I'm not walking 19KM through the snow. :)

Sounds like a couple of very good reasons !

boB

K7IQ 🌛  He/She/Me

johnnymac

Hi boB,

Devin has ordered some 3.9 Ohm resistors from where ever he found them for our diversion loads on the other systems. When they show up, I'll change over the clipper to the new ones and lug it, via the skidozer hopefully if I have it working,  to the top of the hill. We'll then change out the turbine with another spare (thank God for spares) and see how this works.

I'm thinking we should cut out part of the turbine body and put a small "sight" window in there so we can keep an eye on slip ring without taking the darned (I didn't swear yay!!!) thing apart.

He also ordered a 48V charger so we'll bring the battery bank up to float, then we can sit there and watch it to see what happens. I will have to remember to take my kettle this time, as the last time I had my coleman stove, tea bags, sugar, milk, water, but no kettle  :'(

Tea time is much more important to me than Devin when we're on site.

Outside of that, I think we're pretty much done until Santa comes.

That said, we want to wish everyone a very Merry Christmas and a Happy New Year!!!

Please be careful, and especially DON'T DRINK AND DRIVE!!! It really is stoopid (2 O's that's really important) you know!!! The life you kill might be your own ha ha ha


Oh.... one other thing.... DGD, with regards to aerodynamic braking, it does seem to work. At one of our sites that has a Turbine with the aerobrake on it, when the wind gusts really hard (120 Kph +++) you can hear the blades trim out similar to a turbo prop. I've been up there in a few storms now and you can watch the current from the turbine as it decreases with the sound change, and the sound is unmistakable. I'm no expert on the intricacies of the blade design, but I do know it sounds like a Dash-8 adjusting their prop pitch as the plane comes in for a landing. The Dash-8 is our most common airplane around here and most everyone who flies is aware of the prop noises. I'm sure with a bit of smoke, you could see the turbulence on the edge of the blade. Tail furling, as nice an option as it is, seems better suited to lower wind areas than we have. As Devin had stated earlier, Aliant (that's our version of Bell Canada) had tried it a couple of years ago with furling tails and did nothing but snap off a bunch of tails. It got to the point that they considered wind power unusable for their purposes. That mindset has not changed for them to this day.



John

dgd

Quote from: boB on December 23, 2013, 03:51:52 PM

This is exactly the reason why I much prefer AC clippers (when possible) over DC clippers.
Turning off at 0 current pretty much ensures there is not going to be an inductive
voltage spike.   I think that if 1.6 Ohm per phase is too small that double that is also
going to be too small but I'm just not sure.  10 or 12 Ohms might be just fine.

In fact, if you were getting 1500 watts at 143 volts, that would mean that the load the
Classic was giving the turbine is around  1500/143 = 10.4 Ohms.


An interesting calculation  :)
When I did this some time ago for my turbine I did it differently
So for 1500watts at 143 volts that gives amps as 10.4
So classic R is 143/10.4 or about 13.5 ohms
V on each of 3 phase inputs is 143 *.74. = 102v
So to load each phase to same as Classic needs about 10 ohm resistors
So I would go slightly under about 8.5 to 9 ohms
Did I get this wrong?

My turbine is 103v at 1600w as it has 3.2 metre swept dia and spins much slower at similar power from high speed 2metre hy1500
(103*.74)/(1600/104) or 75/15 = 5 ohms per phase or 3.5 to catch over 1600w events..

It seems to work for me but then my calculation may be too simple or even just not the way to go..

Happy Christmas

Dgd
Classic 250, 150,  20 140w, 6 250w PVs, 2Kw turbine, MN ac Clipper, Epanel/MNdc, Trace SW3024E (1997), Century 1050Ah 24V FLA (1999). Arduino power monitoring and web server.  Off grid since 4/2000
West Auckland, New Zealand

boB

Quote from: dgd on December 25, 2013, 04:54:15 PM
Quote from: boB on December 23, 2013, 03:51:52 PM

This is exactly the reason why I much prefer AC clippers (when possible) over DC clippers.
Turning off at 0 current pretty much ensures there is not going to be an inductive
voltage spike.   I think that if 1.6 Ohm per phase is too small that double that is also
going to be too small but I'm just not sure.  10 or 12 Ohms might be just fine.

In fact, if you were getting 1500 watts at 143 volts, that would mean that the load the
Classic was giving the turbine is around  1500/143 = 10.4 Ohms.


An interesting calculation  :)
When I did this some time ago for my turbine I did it differently
So for 1500watts at 143 volts that gives amps as 10.4
So classic R is 143/10.4 or about 13.5 ohms
V on each of 3 phase inputs is 143 *.74. = 102v
So to load each phase to same as Classic needs about 10 ohm resistors
So I would go slightly under about 8.5 to 9 ohms
Did I get this wrong?

My turbine is 103v at 1600w as it has 3.2 metre swept dia and spins much slower at similar power from high speed 2metre hy1500
(103*.74)/(1600/104) or 75/15 = 5 ohms per phase or 3.5 to catch over 1600w events..

It seems to work for me but then my calculation may be too simple or even just not the way to go..

Happy Christmas

Dgd


Yes, Happy Christmas DGD !

No, I think your calculation is OK.  Remember though if it is 10 Ohms phase to center point that this will
be two times 10 Ohms in series from a phase to phase standpoint.  BUT, there are 3 phases so it's
actually lower than this, on the average.

In addition though, the 10.4 Ohms I came up with was just an equivalent resistance presented by
the Classic for 1500 watts at 143 volts.  If the wind gets higher, then the resistance from a clipper
will want to be lower because the voltage will rise from this starting point.

I like to reduce my minimum amount of resistance necessary for safe clipping by approximately two
to allow for higher winds.  If you have any kind of PWMing, then this half resistance can easily
be made an equivalently higher value with that PWM.

It sounds like in the case of our Canadian friends here on their very windy mountain-top site,
that the resistance of their Clipper is WAY WAY too low, IF it is what is causing the slip-ring
burn up problem.  It may of course be something completely different and not because of
too low of resistance load but I just don't know yet.

Ideally, a good PWM'd Clipper can take a much lower resistance than necessary and make it
almost any resistance higher than that  for variable Clipping.  Of course that ideal-ness does
not exist in reality.

boB

K7IQ 🌛  He/She/Me

Robin

I am not a wind expert, but I do have a couple of observations to make on this.
Take a look at the picture here. It is of the slip ring assembly on our turbine under development for the US Army.
You will note that there are two sets of brushes for each slip ring. You will also notice the plastic dividing discs between each slip ring. They have a long overhang to help eliminate the carbon dust from shorting one slip ring to the other.
Now take a look at the slip ring assembly on the broken turbine. It looks like a single set of brushes. I do not see much separation between brushes either. The carbon dust that accumulates around the slip ring area is conductive. Given enough dust, voltage and current, bad things can and will happen. That is what drove our design. I suspect the slip ring assembly on this turbine is just not up to the job being asked of it. One thing that might help would be to detune to power curve such that you keep the voltage lower and thus the speed lower and thus the current lower. See where this is going? Reduce the load on the slip rings. I suspect the same thing would happen to these slip rings with or without the Clipper and classic, but with the Classic and Clipper you have a chance of limiting the output power to save the brushes.
Good luck, this stuff isn't easy.
Robin Gudgel

devo



A little update for everyone. We replaced the 1.6 Ohms per phase 4Kw load with a 4 Ohms per phase 3 Kw load.

We kinda have it working TOO well! Notice the temperatures on that picture. Things are soon going to start catching on fire. Winds up on that mountain are probably reaching 170 - 190 Km/hr today. Wind warnings are in effect everywhere from the weather office. Here at home I'm seeing gusts past 150Km/hr at sea level. This site is at an elevation of over 500m!

One question: If we turn the mode to OFF on the wind classic will that short the turbine and stop it, or will the clipper just continue to dump?

Right now it may come down to a decision between blowing up the turbine or allowing the telecom shack to catch fire!

devo

On another note, boB, we're certainly stress testing the clipper and classic!

I'd love to see just what kind of power this setup is actually generating. I expect its 3Kw+ being dumped into a building that's 8' x 8' and insulated like crazy.

At what temp will the PCB melt? 100C? 110C? At some point the thing will just explode from the heat eh?

We're not even sure we can get someone up there to open the darn door on the building and let some cool air in.

Westbranch

devo, I know it may be too late this time, but is there any possibility to automate the venting, like they use in greenhouses?  I believe they use a small opening device 'powered ' by a cylinder.
http://www.groworganic.com/automatic-vent-opener.html
KID FW1811 560W >C&D 24V 900Ah AGM
CL150 29032 FW V.2126-NW2097-GP2133 175A E-Panel WBjr, 3Px4s 140W > 24V 900Ah AGM,
2 Cisco WRT54GL i/c DD-WRT Rtr, NetGr DS104Hub
Cotek ST1500 Inv  want a 24V  ROSIE Inverter
OmniCharge3024  Eu1/2/3000iGens
West Chilcotin 1680+W to come

Robin

Ryan or Bob can tell you how to reduce the speed of the turbine although since the resistance was changed, it may not slow the turbine down. The battery temp is disturbing. The rest of it is ok. Circuit boards are rated for 130C.
I will see if they can chime in here.
This is a condition that we would like to see automatically turn down the speed of the turbine. Sustained high wind conditions should be able to be identified by the Classic and dealt with, but it is not automatic yet.
After looking at the meters again, I do not think the turbine can be slowed down. The Classic is in float. That means it would have sent a signal to the Clipper to bring the speed down so the voltage is just above float. I see the voltage is 194 volts. It may be that the 4 ohm resistors are too much resistance to control the stator. Just a guess, but that is what it looks like to me.
Robin Gudgel

Halfcrazy

What I would do at this point is go to Aux2 and turn it manually ON and see if the Input voltage drops? If it does drop a lot life should be good. If it drops very little to none than the clipper is already doing all it can to slow the turbine. I am not worried about the Classic and heat so I would not turn the mode off.

I assume there is solar on site to? And probably no way to turn that off? I would say that in the future you should have the wind set to a higher voltage than the solar.

So what this shows me is on a remote site a Exterior mounted diversion load would be an asset.

Other than turning the clipper full on or turning some load on to actually load the turbine down I am sort of at a loss. You could flatten the wind curve way down to try and stall the turbine that may help but only as long as there is a load needed.
Changing the way wind turbines operate one smoke filled box at a time

devo

Thanks for the replies guys! I think we've dodged the bullet this time ... the wind has subsided and everything is back to normal.

We really need to get the environmental monitoring system installed at these sites so we know what's happening outside.

@Westbranch - We thought about using some industrial louvers and fans about 6 months back, but we haven't been able to generate any substantial heat in the buildings until now. The only concern we have with louvers, or any other mechanical opening, is that we would be letting the wind inside the building. By allowing the wind in, we may not have a problem with heat, but the building may end up at the bottom of the mountain!

@Robin - Our big concern was the classics themselves. We weren't sure how much heat they could take before meltdown. If they can handle 130C we should be okay. The temp peaked at about 90 - 95C on the PCB and FET. The batteries are rated to 50C, but I don't know if they can take much above that. The rest of our electronic equipment is rated to 60C, so it should be fine. The 194V was a peak that I managed to catch while watching the local app. Once I did see the voltage hit 200v, but only for a second. That's what we have the clipper set to as a maximum for testing.

@Halfcrazy - I didn't get a chance to actually try to manually engage the clipper. I think there MAY have been a few volts drop if it was engaged, since the input was held fairly steady around 170 - 180V with peaks to 194 - 200v a couple times. If there was any capacity left in the clipper it wouldn't have been a lot. We've already discussed adding an external dump load to get rid of the heat outside. That might be one option. The other option we are looking at is adding a 1Kw split mount air conditioner / heater in the building. Our concern is during the summer months when it is 25 - 30C outside and the clipper starts dumping 2 - 3Kw of power into the resistors. We could overheat everything, or start a nice little fire! The AC unit could take 1Kw of energy from the system, reducing the amount of clipping, while also cooling the building without opening the building up to the wind.

There is roughly a Kw of solar at each of these sites as well. Today was too overcast to generate much power, and with the batteries in float i imagine the solar classic would just open the panels and not let them produce. Is this correct?

I think we can safely say that 4Ohm, 3Kw load on a HY-Energy Hy-1500 turbine will generate some serious heat!

Here's the site we're talking about: