Once the SUN 22-60 goes into overvolt you really don't know how high the voltage will go. It depends on the turbine. Car PMAs can get to 100s of volts and thus can easily fry the dump FETs (don't do it!). The goal of the GTI dump port is NOT THE SAME AS IT IS FOR A BATTERY SYSTEM. For a GTI we want to slow the RPM ASAP so that the voltage drops and stays under 60V where the GTI can continue to feed the grid. We do not want the dump port to be active very long (just seconds). If it is on for longer, this indicates the load is not heavy enough. At 1000w and 60V the current is 16.6A. This equates to 3.6 ohms, which is clearly not low enough to slow down the turbine as it already went to overvolt. So what lower value is best? What we don't want is for the voltage to get anywhere near 100V, the limit on the IRF3710 dump FETs. I'd target 80V. 80V x 16.6A = 1333 watts. 1333w at 60V is 22.2A and equates to 2.7 ohms. That would be the highest ohm level I'd consider, but a lower value would even be better. The (2) IRF3710 FETs are rated at 57A and together could provide 114A of capability if properly cooled. The problem is, they are poorly cooled. When not activated, they share a heatsink with the normal MOSFETs so the heatsink surface is already hot when the dump FETs are activated. Also, it's not clear if the fan continues to operate while in dump mode (perhaps someone can address this). With poor cooling, it might be wise to not drive more that 15A through each FET, which would be about 25% of their rating, for a total of 30A for the pair. 30A at 60V suggests a 2.4 ohms of resistance and a dump power dissipation capability of at least 1500 watts. Using the WN green resistor line, we can get there with 8 resistors and have enough power dissipation capability. Placing 2 of the 24V and 2 of the 48V resistors all in parallel yeilds 1.13 ohms. Add another set like this in series then yields 2.26 ohms, very close to our 2.4 ohm goal. Each parallel set will dissipate 796 watts for a total of 1592 watts at 60V for the system. Adding 8 of these resistors can get pricey - perhaps a qty discount could be considerd, or a set matched and priced for a particular SUN GTI. As for the quartz crystal resistor version, perhaps 2 of the 5 ohm ones in parallel (2.5 ohm equivalent and 1800w capability) would be a good alternative. I'm a bit concerned about the wire gauge they use on these however.
Minnesota, Do the sun gti's try to maintain any particular voltage inside the 22-60v? Will it attempt to hold a 24 v turbine at 24-30v or will it let it go all the way up near 60 before applying serious load?
434 ... generally speaking they increment up the operating voltage until they see the amps decrease, then they increment down the voltage until the amps increase. It happens very fast. Unlike a solar panel that has a pretty stable MPP voltage point, a wind turbine is all over the place, so it's been a challenge to develop an equivalent method for wind. SUN simply modified their PV GTI software to get it to crudely work with wind, but I'm sure there will be much improved ones coming, such as a single GTI that operates from 6-100Vdc or so. They know the wind GTIs are good sellers.
Thanks. This is as I thought but I wasn't sure. I was planning to wind my coils for 48v (axial flux) but when choosing a gti instead of battery charging I felt 24v would better suit the sun g. I would love to see a 2kw 100+v unit.
I noticed Mouser has 750w and 1000w resistors. Two of the 750w @ 4.7 ohms in parallel (2.35 ohm load) would be a lower cost solution than the set of 8 of the 300w WN ones as mentioned above. Perhaps WN should stock these 750w resistors to support the 22-60V GTI 1kW users. In qty they are about $30 each. There is one guy on youtube who is using this GTI with a Windtura. http://www.mouser.com/ProductDetail/TE-Connectivity/TE750B4R7J/?qs=sGAEpiMZZMtlubZbdhIBIIenWIH7rhEv5eJQQmVbCd0=
I wish I had thought about this earlier .... Do we even know on the WN 750 stock PMA how soon it overvolts the SUN 1000G? I'm thinking if the voltage can get to 60V (~650 RPM) then the amps could be in the 30s, perhaps calling for 2 SUN 1000g and associated divert loads rather than 1. I do think it is possible to hit >1.5kW with the Windtura using these GTIs as there are less losses at higher voltages. We already know when the 750 is used to charge a 24V battery bank that it can hit 1000 watts. Imagine if it could run to 60V! One might have to enable the 2nd GTI at a higher voltage (~45V?) so it doesn't hold back the 750 at lower levels. Could be a HUGE market, and is all the more reason to develop a variable voltage dual GTI switch.
Minnesota, What do you think a 48v turbine would do attached to 2 sun g 22-60 1000's? I know it would over volt using just 1 but would 2 work? Or would it hold it down in the 20-30 volt range and stall the blades? I am currently building a 10.6' axial flux mill and am stuck in a decision of 24v or 48v.
My fear going 24v and using 1 is exceeding the 1kw as in high winds I can reach 1.5kw. But I know using 2 gti's at 24volt will probably not Work.
It really depends on the RPM/V. The 24V 750 has a 10.7 RPM/V and appears to hit 60V at about 650 RPM; about the same RPM that it furls (good!). Some car PMAs that are "so-called" 24V type hit 60V at as low as 200 RPM - and that can come in light winds, so clearly not all 24V setups are comparible. If the WN750 takes possibly 2 of the Sun-1000g GTIs to keep it tame then yours could require many more since your sweep area (and potential power) is 2.6x larger. Hasn't someone implemented a GTI setup with a turbine like yours? What is your RPM/V and furling RPM point?
Rpm/v varies. @24v: 60rpm=7.4v and 150rpm =24v @48v: 60rpm=14.9v and 150rpm = 48v Cut in is obviously 150 and furl should be no more than 500rpm, 450 is preferred and around 28mph. I don't really see a good gti option for voltages under 100 without also battery charging like the outback does.
So the RPM per volt is 5.42 for the 24V and 2.72 for the 48V configuration. At 500 RPM the 24V would be at 92.2V and the 48V unit wuld be at 183.82. Appears that both are well above 60V so all of that top end power would be diverted away using the SUN-1000g. Ideally you should design to 8.33 RPM per volt which is a "15.6V" turbine if you want to use a bank of these GTIs.
So are you saying the gti will not Be able to hold the voltage below 60? I thought they imposed a load to maintain voltage and create amperage?? I thought they would be able to hold voltage until you approached rated wattage of the gti.
How did you come up with those #'s? If it reaches 24v @150rpm, that equals 6.25 rpm/volt. If its lenear, which axial flux motors are not, it doesn't pass 60volts till about 380 rpm. That would be about 22-23 MPH. Plus with my motor, I can increase the air gap to increase the rpm/volt.
They are current limited to rougly 15 amps internally I believe, so they can invert about 900 watts (60A x 15A).
I assumed basically linear; on the 24V 60rpm=7.4v and 150rpm =24v, so (24-7.4) = 16.6V over a 90rpm range so that is 5.42 revs/volt.
Was getting ready for bed and shut the light off and that new dump load was glowing orange. Checked it out and guess what???? I do not know when it happened today but the 22-60 volt GTI fried the mosfets on the dump load again. Good thing I ordered more irf3710 mosfets. I am going to make a fet board for this thing. Just ordered 10 of them. Will use 6 for a fet board. Like to see them burn out. LOL
Tom, What turbine are you running on that gti? Do you know what it was last outputting before the fets fried?
It is running the little turbine. It is a 24 volt. I do not know when it quit. I was not home. I just happen to shut off the lights and seen the dump load glowing. Winds are at 25 mph. I just took two more SSF7508 mosfets out of a junk GTI. Now it is working again. One of the other mosfets was a direct short. Hard to believe it did that. The two mosfets that were in it were 100 amp each. Of course I have seen where people all over the net have had problems blowing mosfets without even getting close to the amp max. Just going to build one of the zener boards for 55 volts and hook it to a Heavy Duty relay. Or build a fet board for it.
If you're running it with just 5 ohms then it likely exceeded the MOSFET voltage rating. 100 amps x 60 v seems impossible don't you think?
