Greentime controllers

[Jeremy Harris]

Still, it's $45 and you can get 4110s if you need. this controller handles 24s lipo no problem

Not wise. Sorry, but the absolute maximum voltage rating (i.e the NEVER exceed voltage) for the IRFB4410 is 100V. Bear in mind that you'll have ripple on the supply, plus inductive spikes from the motor, so running them at 100.8 V max is asking for early failure. You must always allow some headroom between the max, hot off charge voltage of the pack and the absolute maximum rating of the components in the controller. Ideally I'd allow between 10 and 20% derating to the component rating and use that as a maximum, so 80 to 90V is a realistic safe maximum voltage for a controller with 100 V FETs. The same would be true if it had IRFB4110 FETs, BTW. If you want to safely go to 24S then you really need to be looking at FETs with a higher voltage rating, not forgetting that other components in the controller (the 12 V power supply, commutation caps etc) will need looking at to make sure they are safe at the voltage you want to use.

 

[modder]

i just bought one of these. Just waiting for it to arrive.

if i don't use 3 speed will i just get max speed?

what if i wanted to run 140volts LiMn? is there a fet that can handle that?

 

[ian.mich]

Still, it's $45 and you can get 4110s if you need. this controller handles 24s lipo no problem

Not wise. Sorry, but the absolute maximum voltage rating (i.e the NEVER exceed voltage) for the IRFB4410 is 100V. Bear in mind that you'll have ripple on the supply, plus inductive spikes from the motor, so running them at 100.8 V max is asking for early failure. You must always allow some headroom between the max, hot off charge voltage of the pack and the absolute maximum rating of the components in the controller. Ideally I'd allow between 10 and 20% derating to the component rating and use that as a maximum, so 80 to 90V os a realistic safe maximum voltage for a controller with 100 V FETs. The same would be true if it had IRFB4110 FETs, BTW. If you want to safely go to 24S then you really need to be looking at FETs with a higher voltage rating, not forgetting that other components in the controller (the 12 V power supply, commutation caps etc) will need looking at to make sure they are safe at the voltage you want to use.

The capacitors in this controller at 100V, and i charge to 4.15/cell. Greentime sells 100v-12v 150w dc-dc converters for ~$10. this controller still handled 100.8 no problem, and has never gotten warm when using it (when i was testing at 4.2v and regular use at 4.15)

 

[ian.mich]

i just bought one of these. Just waiting for it to arrive.

if i don't use 3 speed will i just get max speed?

what if i wanted to run 140volts LiMn? is there a fet that can handle that?

Yes, but i think i read if you connect the blue and black wires (3 speed wires) then you get a slight speed upgrade, havent tried it yet. For 150v you're looking at 4115s which they have. you made the right choice, this is the best controller supplier in china.

 

[modder]

currently i used the hua tong but definitely can't wait to try this as the power and phase wiring looks to be 12ga which indicates to me they offer much less resistance and handle higher discharge than the huatong but i will have to do a comparison once they arrive. thanks for taking the time to answer and i will be back to discuss further.

 

[ian.mich]

currently i used the hua tong but definitely can't wait to try this as the power and phase wiring looks to be 12ga which indicates to me they offer much less resistance and handle higher discharge than the huatong but i will have to do a comparison once they arrive. thanks for taking the time to answer and i will be back to discuss further.

Don't think it's quite 12g, or may jus be lower strand count than i'm used to, but it can certainly take 100v 45a

 

[Jeremy Harris]

The capacitors in this controller at 100V, and i charge to 4.15/cell. Greentime sells 100v-12v 150w dc-dc converters for ~$10. this controller still handled 100.8 no problem, and has never gotten warm when using it (when i was testing at 4.2v and regular use at 4.15)

OK, so you got lucky, but I can absolutely guarantee that it's not a sensible or wise thing to do if you want the controller to be reliable. Exceeding component ratings is dumb, and asking for failure. There's no margin on an "absolute maximum rating", it is what it says, an absolute maximum. You may get lucky for a day, a week or a month, but the component will fail sometime, earlier than it would if operated inside its spec.

Remember that you'll have several volts of ripple superimposed on the supply, plus you'll have inductive spikes induced by the FETs switching and parasitic inductance in the controller. It's quite probable that you'll have several volts over the supply adjacent to the FETs, hence the need to keep the battery voltage 10 to 20% below the absolute maximum component rating.

 

[ian.mich]

The capacitors in this controller at 100V, and i charge to 4.15/cell. Greentime sells 100v-12v 150w dc-dc converters for ~$10. this controller still handled 100.8 no problem, and has never gotten warm when using it (when i was testing at 4.2v and regular use at 4.15)

OK, so you got lucky, but I can absolutely guarantee that it's not a sensible or wise thing to do if you want the controller to be reliable. Exceeding component ratings is dumb, and asking for failure. There's no margin on an "absolute maximum rating", it is what it says, an absolute maximum. You may get lucky for a day, a week or a month, but the component will fail sometime, earlier than it would if operated inside its spec.

Remember that you'll have several volts of ripple superimposed on the supply, plus you'll have inductive spikes induced by the FETs switching and parasitic inductance in the controller. It's quite probable that you'll have several volts over the supply adjacent to the FETs, hence the need to keep the battery voltage 10 to 20% below the absolute maximum component rating.

eh, maybe it is pushing it, but i've gone up some major hills at full throttle with a warm motor, but no warm controller.

 

[modder]

my hua tong is running at 96v and it didn't feel warm either. anyone here compare the 2?

 

[ian.mich]

my hua tong is running at 96v and it didn't feel warm either. anyone here compare the 2?

I believe they are pretty much the same (same mosfets and microprocessor) but the greentimes have a huge number of features.

 

[Jeremy Harris]

eh, maybe it is pushing it, but i've gone up some major hills at full throttle with a warm motor, but no warm controller.

The controller case may well not get hot, but you can be sure the FET silicon will be. If you read some of the posts where we've looked closely at the thermal resistance of multi TO220 FET controller dissipation, you'll see that the major roadblock to getting heat from the silicon to the controller case is the combined thermal resistance of the silicon to the FET case, plus the thermal resistance of the FET case to the insulator, plus the thermal resistance of the insulator to the heat spreader bar and finally the thermal resistance of the spreader bar to the case. The net result is that the FETs can cook before the case gets hot - especially with multi-paralleled FET controllers, where heating will almost certainly be uneven between the paralleled FETs. Using FETs with such a poor Rdson exacerbates, this by a significant degree.

It's just not good advice to others in this forum who may not understand the risks as well as you seem to, to recommend using a controller with such poor FETs at this sort of power level and marginal voltage, as it's going to fail sooner, rather than later.

Finally, it isn't voltage that heats FETs, it's current. The heat in the FET silicon is from the power dissipated by the square of the phase current x the FET Rdson. It matters not what the controller voltage is at all, as it's only current and resistance that determines FET dissipation and losses.

 

[Gordo]

Hi Jeremy,
At first I was wondering why you were taking the time to lay out the error of running components at maximum ratings. I have to keep reminding myself the dangers of leaving bad advice floating around. You are most generous with your time and wisdom, correcting the record for the newbee's who may be looking for information. Especially those who are penny wise and pound foolish. I^2R will bite them with 4410's in TO220'S.
Although we never met, I think of you each time we pass the water, looking across at CFMETR.
Gordo

eh, maybe it is pushing it, but i've gone up some major hills at full throttle with a warm motor, but no warm controller.

The controller case may well not get hot, but you can be sure the FET silicon will be. If you read some of the posts where we've looked closely at the thermal resistance of multi TO220 FET controller dissipation, you'll see that the major roadblock to getting heat from the silicon to the controller case is the combined thermal resistance of the silicon to the FET case, plus the thermal resistance of the FET case to the insulator, plus the thermal resistance of the insulator to the heat spreader bar and finally the thermal resistance of the spreader bar to the case. The net result is that the FETs can cook before the case gets hot - especially with multi-paralleled FET controllers, where heating will almost certainly be uneven between the paralleled FETs. Using FETs with such a poor Rdson exacerbates, this by a significant degree.

It's just not good advice to others in this forum who may not understand the risks as well as you seem to, to recommend using a controller with such poor FETs at this sort of power level and marginal voltage, as it's going to fail sooner, rather than later.

Finally, it isn't voltage that heats FETs, it's current. The heat in the FET silicon is from the power dissipated by the square of the phase current x the FET Rdson. It matters not what the controller voltage is at all, as it's only current and resistance that determines FET dissipation and losses.

 

[ian.mich]

eh, maybe it is pushing it, but i've gone up some major hills at full throttle with a warm motor, but no warm controller.

The controller case may well not get hot, but you can be sure the FET silicon will be. If you read some of the posts where we've looked closely at the thermal resistance of multi TO220 FET controller dissipation, you'll see that the major roadblock to getting heat from the silicon to the controller case is the combined thermal resistance of the silicon to the FET case, plus the thermal resistance of the FET case to the insulator, plus the thermal resistance of the insulator to the heat spreader bar and finally the thermal resistance of the spreader bar to the case. The net result is that the FETs can cook before the case gets hot - especially with multi-paralleled FET controllers, where heating will almost certainly be uneven between the paralleled FETs. Using FETs with such a poor Rdson exacerbates, this by a significant degree.

It's just not good advice to others in this forum who may not understand the risks as well as you seem to, to recommend using a controller with such poor FETs at this sort of power level and marginal voltage, as it's going to fail sooner, rather than later.

Finally, it isn't voltage that heats FETs, it's current. The heat in the FET silicon is from the power dissipated by the square of the phase current x the FET Rdson. It matters not what the controller voltage is at all, as it's only current and resistance that determines FET dissipation and losses.

If it isn't voltage that causes heat, then this controller is fine for 24s, there is no chance that 45A is going to harm these fets, they may be lower quality but there is a lot of them. Ok i guess it is bad advice so i'll say this
use this controller for 20S and you're safe
use this controller at 24s at your own risk.
but its 43 freakin dollars, you could run the controller into the ground and buy a 4115 from them when you have the money, or buy two and carry an xtra around with you lol, or just go 4115 off the bat for less than the price of a 6 fet 4110 lyen

 

[Jeremy Harris]

Hi Jeremy,
At first I was wondering why you were taking the time to lay out the error of running components at maximum ratings. I have to keep reminding myself the dangers of leaving bad advice floating around. You are most generous with your time and wisdom, correcting the record for the newbee's who may be looking for information. Especially those who are penny wise and pound foolish. I^2R will bite them with 4410's in TO220'S.
Although we never met, I think of you each time we pass the water, looking across at CFMETR.
Gordo

Glad CFMETRs still there, Gordo, more than can be said for the similar facility I ran in Scotland back then. My time there was only marred by being sat in a hotel in Nanaimo when Princess Diana was killed - somehow I always seem to relate the two. Still, I have happy memories of flying out to the island on the little Hughes 400D chopper, seeing a bear wandering along the beach from the air and flying the Beaver floatplane out of the harbour.

 

[neptronix]

Many people on this forum have blown their 4110 infineons running 24S 3.6-3.8v nominal cells, as if exceeding component ratings and never leaving any headroom in the spec isn't a problem :lol:

4410's also have very poor current handling ability. So when you order up a controller with those, expect say, a 24FET to perform like a 15FET. I'm not sure why anyone would want those FETs. It just makes for a larger and more expensive controller that performs like a cheaper and smaller controller. I fail to see why they even offer them as an option.

 

[Arlo1]

If it isn't voltage that causes heat, then this controller is fine for 24s, there is no chance that 45A is going to harm these fets, they may be lower quality but there is a lot of them. Ok i guess it is bad advice so i'll say this
use this controller for 20S and you're safe
use this controller at 24s at your own risk.
but its 43 freakin dollars, you could run the controller into the ground and buy a 4115 from them when you have the money, or buy two and carry an xtra around with you lol, or just go 4115 off the bat for less than the price of a 6 fet 4110 lyen

Jeremy is one of the smartest on the forum (Blue Name). Don't argue with him he knows his shit! Ok so as you up the voltage the controller has to turn the fets off for longer times or PWM more often to keep the amps below 45. Voltage = push if you up the push you need to up the resistance to keep the amps down! More resistance within the controller = more heat. Thats lamens terms!
When you increase PWM or turn the fets on and off more to keep the amps down the controller builds more heat from switching losses, which meens every time the fet is in a transition it has hi resistance and becomes a weak link untill it is fully turned on. As well when the fet first shuts off the inductance of the motor forces current to keep flowing and so because of this fets have a Diode inside them that will alow that current to keep flowing and this Diode also gets hot.

So in short you have more on and off time causing more heat and the more times on and off the more times you have a chance of a voltage spike going over 100v and blowing your fets.

 

[Jeremy Harris]

If it isn't voltage that causes heat, then this controller is fine for 24s, there is no chance that 45A is going to harm these fets, they may be lower quality but there is a lot of them. Ok i guess it is bad advice so i'll say this
use this controller for 20S and you're safe
use this controller at 24s at your own risk.
but its 43 freakin dollars, you could run the controller into the ground and buy a 4115 from them when you have the money, or buy two and carry an xtra around with you lol, or just go 4115 off the bat for less than the price of a 6 fet 4110 lyen

Others have said the same as me, it doesn't make sense to exceed the voltage rating, and it doesn't make sense to use poor FETs in a controller, either.

Lyen builds capable controllers using Xiechang boards, populated with high quality components. You can buy the same boards Lyen uses direct from Keywin Ge for a lot less money. For example, a 6 FET board, less FETs, but with case and all the other components is around $21. Add a bunch of decent FETs, like IRFB4110s (if you want 80 + volt capability) or IRFB3077s (if you are OK with 65 V + capability, but with greater current handling capability) and you'll have a controller that is significantly more reliable than one of these Greentime units, costs about the same and is programmable using a serial port and the free programming utility available here.

As an example, I bought five 6 FET bare boards from Keywin a year or two ago. They cost me around $21 each inc shipping. I run one with 6 IRFB4110 FETs, at a total cost of around $50, another on 6 IRFB3077 FETs at around $45 total cost, and a third runs 6 IRLB3034 FETs at a total cost of around $45. All are capable of 50A +, with the IRFB4110 version being OK for 80 + V.

The price of the controller is really not that relevant if it blows up when you're ten miles from home. If you're happy to carry two controllers around, because you know you're running something that's liable to fail at any time, then that's fine, but many people wouldn't want that level of hassle or expense.

 

[HypnoToad]

I would rather have a 6-fet made with 3077's for that voltage range of controller than a 15-fet made from 4410's, which you need twice as many of to equal a 4110.

Pretty much an unbeatable combination for a small controller at this spec point, in my view. Bare board 6 FETs from Keywin are ~ $20 each, add in 6 IRFB3077s at around $25 and you have a 65V + capable controller at 50 A + with very low losses.

This sounds ideal, I'm not planning to go above 60V anytime soon and I quite like the idea of more energy going into my motor as opposed to the heatsink in my controller. I'm waiting for this controller to arrive from the Hong Kong, I ordered it a few days ago:

http://www.ebay.co.uk/itm/290621145469#ht_3490wt_909

Is this controller similar to the Keywin unit? If not do you know who sells the bare board 6 FET units as I'm pretty usless at unsoldering FETs, especially on cheaper PCB's where tracks can be lifted. For the sake of $20 I may as well order a bare board.

Also, I'm not really sure how to calcate heat output of FET's compared to resistors, as I wanted to see how much difference swapping the FET's in my current (currently non functioning) controller would make. If I do fix this controller and replace the 11mOhm P75NF75's with IRFB3077's, I'm guessing I would have higher current output with less heat. My setup is pretty low power, and 40A is more than my motor would output.

Assuming 40A load, 6 FET's, 0.0011 Ohm resistance per FET
Power loss in a single FET: P = I^2 x R = 40 x 40 x 0.0011 = 1.76W per FET

Now for the IRFB3077, 3.3mOhm:
40 x 40 x 0.0003 = 0.48W per FET

Now I'm guessing that a 6 FET controller uses 2 FET's per phase in parallel? So maybe I should half the resistance in all my calculations? Also, I'm assuming the FET's are simple resistors here, now FET's essentially switch on and off vary quickly, so I may have got my calculations wrong here. Still, It's not hard to see that with higher power setups, going for a low loss FET gives more gains.

 

[Jeremy Harris]

Contact Keywin Ge (e-crazyman on ebay) and ask him to give you a price for controllers with no FETs fitted. His email address is ecrazyman(at)gmail.com (make the obvious anti-spambot changes) but be prepared for a slow response from him - he seems to run this business part time and often takes a while to reply. I've bought a few from him like this, some 6 FET ones and some 18 FET ones, and I suspect this is where Lyen gets his parts from (BTW, Lyen's prices are OK when you factor in the cost of good FETs and the labour to build these things up). The boards come from Keywin with all the components fitted except the FETs and the wires, usually, and he supplies all the mounting hardware, case, screws etc if you ask. You can then fit whatever FETs you want. He can let you have 6, 12, 18 or whatever size units you want - often the bigger ones aren't much more expensive as it's the FETs that are far and away the most expensive components in these things (hence the reason so many cheap controllers have rather nasty cheap FETs).

 

[HypnoToad]

Contact Keywin Ge (e-crazyman on ebay) and ask him to give you a price for controllers with no FETs fitted. His email address is ecrazyman(at)gmail.com (make the obvious anti-spambot changes) but be prepared for a slow response from him - he seems to run this business part time and often takes a while to reply. I've bought a few from him like this, some 6 FET ones and some 18 FET ones, and I suspect this is where Lyen gets his parts from (BTW, Lyen's prices are OK when you factor in the cost of good FETs and the labour to build these things up). The boards come from Keywin with all the components fitted except the FETs and the wires, usually, and he supplies all the mounting hardware, case, screws etc if you ask. You can then fit whatever FETs you want. He can let you have 6, 12, 18 or whatever size units you want - often the bigger ones aren't much more expensive as it's the FETs that are far and away the most expensive components in these things (hence the reason so many cheap controllers have rather nasty cheap FETs).

Thanks, I've sent him an e-mail, I'm in no hurry as it's mainly something to experiment with. Both Farnell and RS sell the IRFB3077 they usually deliver pretty quickly so I'll order these once the controller arrives.

 

[HypnoToad]

Contact Keywin Ge (e-crazyman on ebay) and ask him to give you a price for controllers with no FETs fitted. His email address is ecrazyman(at)gmail.com (make the obvious anti-spambot changes) but be prepared for a slow response from him - he seems to run this business part time and often takes a while to reply. I've bought a few from him like this, some 6 FET ones and some 18 FET ones, and I suspect this is where Lyen gets his parts from (BTW, Lyen's prices are OK when you factor in the cost of good FETs and the labour to build these things up). The boards come from Keywin with all the components fitted except the FETs and the wires, usually, and he supplies all the mounting hardware, case, screws etc if you ask. You can then fit whatever FETs you want. He can let you have 6, 12, 18 or whatever size units you want - often the bigger ones aren't much more expensive as it's the FETs that are far and away the most expensive components in these things (hence the reason so many cheap controllers have rather nasty cheap FETs).

Thanks, I've sent him an e-mail, I'm in no hurry as it's mainly something to experiment with. Both Farnell and RS sell the IRFB3077 they usually deliver pretty quickly so I'll order these once the controller arrives.

 

[ian.mich]

Thanks. So have you tried it in sensorless mode?

works flawlessly with or without sensors.

 

[SamTexas]

Great. Thanks.

 

[ian.mich]

Great. Thanks.

any time. eventually greentime will have programmable controllers. oh well until then we have our solder irons. anyone know how to modify LVC?

 

[ian.mich]

these controllers works flawlessly with 24s