Teaser Speed Controller

[sn0wchyld]

@Planky Maybe it's this controller then as it did this with the 9C I had and Beavinator had the same problem with his HS3540. Anything above 35amps and it loses sync really bad so I have to feather the throttle. Once the volts go down by a volt or two then it gets a lot better. Not sure if mine is a mark 1 or 2, have had it over a year now.

@gwhy! Sorry I meant 16S my fully charged voltage is 66.2v. I charge to about 4.12v and discharge to 3.8v on my 17.4Ah of Lipo and still apparently have 40% left in the tank. 12 mile WOT ride with a none modded HS3540. Think I will drill some holes in it like the 9C.

66.2 volts is still boarder line for 75v fets I know on paper they should be fine and yes you could probably get away with it with a hub motor , OK I will keep it in mind when the time comes

Yea I'd be less interested if I could only run 16s. Just ordered one of those huge 5404's, and was hoping to run it at ~100v....

 

[gwhy!1]

@Planky Maybe it's this controller then as it did this with the 9C I had and Beavinator had the same problem with his HS3540. Anything above 35amps and it loses sync really bad so I have to feather the throttle. Once the volts go down by a volt or two then it gets a lot better. Not sure if mine is a mark 1 or 2, have had it over a year now.

@gwhy! Sorry I meant 16S my fully charged voltage is 66.2v. I charge to about 4.12v and discharge to 3.8v on my 17.4Ah of Lipo and still apparently have 40% left in the tank. 12 mile WOT ride with a none modded HS3540. Think I will drill some holes in it like the 9C.

66.2 volts is still boarder line for 75v fets I know on paper they should be fine and yes you could probably get away with it with a hub motor , OK I will keep it in mind when the time comes

Yea I'd be less interested if I could only run 16s. Just ordered one of those huge 5404's, and was hoping to run it at ~100v....

The controller has 100v rated caps so the fets will need to be rated at least 100v they do exists but they are also rather pricey ( plenty of peeps do run controllers @100v with only 100v rated fets and caps, but I wouldn't ). The whole Idea behind this controller was to make it easily configurable dependant on budget/power/voltage, higher voltage caps can be used and also higher voltage fets.

 

[gwhy!1]

Your controller chucks out insane amps though? I would probably utilise 50% of that so should help with headroom

How come you can get so much power without the overheating and over sized usual Chinese type offerings....without giving away any secrets

There is no secret, The little standard (modified ) 6fet controllers that I have been using are good upto 80A and maybe more (I never have really pushed the limit ) The heat issue goes away if the controller is mounted into a more suitable box with a better setup for sinking the heat, I have a modded ( same fets that is installed into my 6fet controllers ) 12fet controller still in its original case but it gets very hot very quickly at 80A but I can, and have pushed 140A through it without a problem.The fets in this monster controller has a lower Rds than the fets im using in my 6fet controllers it also has a bigger case package and it is capable of handling nearly twice the current also the operating temp is a tad higher, so all this should add upto a Very high current controller.

 

[gwhy!1]

Update:
The throttle interface is working well, ( unsure now if it really needs a momentory button from 50% speed to 100% speed as the ramped throttle makes it much more user friendly ) I just need to finalise the lipo monitoring part.

I have also made the heatsink for the controller and it lends itself very nicley for water cooling in needed 8)

hopefully there will be more news tomorrow :wink:

 

[gwhy!1]

UPDATE:
The throttle interface has been finalised and now I can get back to finishing the speed controller .

I need to make myself another interface so I will cost exactly how much it will cost, as the interface stands its optimised for 12s lipo monitoring but I can set this up for any voltage. The battery monitoring do require a small resistor to be soldered into the controller on the B+. I bought myself a motorcycle kill/light handlebar switch this takes care of all the switches needed for 50% speed override, disable controller/throttle output and throttle ramped output.




I will look into making it a Current based throttle interface but this is not a priority for me so may be a long time away. The chip Im using only has 1 analog output so to have switched current limits will involve some addition hardware, again this is not a priority for my needs ( but I think maybe in the future I will incorporate some form of current limiting switch ).

 

[barryfzr]

Looks good! but whats CA stand for?

Pm on its way

 

[gwhy!1]

Looks good! but whats CA stand for?

Cycle Analyst, very similar but with more functions than a watt meter.

 

[gwhy!1]

Update:
I have had a small setback on the controller front , whilst doing some final testing on the bench I managed to blow a fet which caused a fair bit of damage to the fet driver circuit, On investigation the failure was down to a short on the heatsink . The insulators that I am using are not very good and alignment is crucial (as I have just found out ). Anyway The good news is its back up and running and the insulators have now been changed something better. I am going to remake the heatsink so this can not happen again. I am going to mount it a slightly bigger box ( just for when Im putting on my bike for testing ) just to give me more room so I can easily do mods ( if I need to ).

 

[richerson]

It's great your developing this In development there are many problems. But hopefully small ones.
Keep up the great work.

 

[nieles]

hi gwhy!,

what about using a INA138 OP AMP for a programmable current limit. this way we keep the bandwith for cycle by cycle current limiting
and use a multiturn pot for setting the gain. eg between 1x and 10x.

let me know what you think.

 

[gwhy!1]

hi gwhy!,

what about using a INA138 OP AMP for a programmable current limit. this way we keep the bandwith for cycle by cycle current limiting
and use a multiturn pot for setting the gain. eg between 1x and 10x.

let me know what you think.

Yes the theory will work but I'm not sure this chip will do the job, the shunt is in the low side of the circuit. I haven't checked the current limiting out yet on this controller but when I do I will have a bit more insight as what can and cant be done. The next controller chip up in the range ( the one with the motor brake ) has a fault indicator output and goes high with various fault conditions, now this could be very useful for a number of things like letting the current overshoot the shunt setting for a preset time. I have one of these chips and will be making a board for if so I can have a play.

 

[gwhy!1]

Well times been getting on!! I was taken off this controller project for a little while :? but now im back on it.. it is ready to go on the bike and WILL be road/offroad tested this week, I have re-made the power boards so they fit together better, I also bought a controller chip that has a few more functions that I will be having a play with , after the power boards and driver board has been real life tested.

 

[gwhy!1]

Update:
Tested controller with a limit of only 20A on a bike ( I did not get out as often as I would of liked this week for testing but I had other things on the go ) . Verdict is very smooth throttle control ( unlike some other controllers I have used ) and very strong acceleration even with only a 20A limit ! , I can stall the motor at 20A without issues or strange noises from the motor ( like you get from other controllers ), With the limited real life testing done so far, it is looking very good indeed. The next step is to increase the current limit to 80A and give it another thrashing for a week

 

[nieles]

good to hear this project is going so well!

i have been working on the the Current controller project for rc ESC's

maybe parts of the Current controller project can be used in the future for phase limiting on this controller.

it works with 3 Hall sensors for current sensing.

Niels

 

[gwhy!1]

update:
Ok some bad news I adjusted the shunt for approx 80A but it blew a fet streight away without any load on the motor :? so the fet needs replacing now this is the same fet that blew when i very first put this controller together ( which i blamed the heatsink as the cause ) but now im wondering if it maybe some wrong in the driver circuit for this phase to blow the same fet again or it may even be that I got a solder splash on the fet legs ( fingers crossed ) from adjusting the shunt, either way it will have to be investigated and a new fet installed so now its the waiting game again

 

[gwhy!1]

i now have the controller up and running again at 20A limit ( with some much cheaper fets <$1.50 each I would rather blow these up the the expensive ones), There is something wrong with the gate drive on the phase that blew the bootstrap isnt working as it should , the motor never starts up first time when power is connected its very strange as the other 2 phase are fine and all three are identical circuits. I think I will just replace the driver chip and dc-dc converter but they do seem to work as they should when tested. :?

 

[gwhy!1]

More testing done ( but not on the bike, for 1 simple reason it keeps blowing fets ) I have upped the current to 100A limit and this appears to be fine but I am unable to really test under a static/dynamic load which needs to be done, if I start the motor up with any sort of load at this battery current limit the phase current goes through the roof and pops fets, I have tried some cheap and dirty ways of getting the motor running then loading it down ( once the motor is over about 4k rpm ) I have loaded it down to around 50A and as long as the rpm do not drop below this 4k rpm this 50A load is a walk in the park for the controller. I have also stacked 2 power boards (200A battery limit but still the phase current is still killing the fets! with the big outrunner motors) and carried out the same dirty tests as before ( once upto around 4k rpm ) 50A load the signals on the gates are perfect and no heating problems with the fets.I have a few more tests to try and get done but my next step is to try and run the motor up with the maximum load I can put on it ( whilst on the bench which will only be around 60A )on a much lower voltage which should reduce the phase current. I have ordered a 12fet controller to butcher to try and put the powers boards through there paces on my bike as these controllers have some sort of pseudo phase limiting that may just get around the startup problems, but in the long term some realtime phase current monitoring needs to be incorporated which brings me to this thread http://endless-sphere.com/forums/viewtopic.php?f=2&t=28935 and I like the way that the phase current have been scaled down to a manageable level .

..... nieles.... is the current sensor board a stand alone board i.e the 1 board out of the 3 or do the sensor board need to have the other boards to read the current ? I have a plan and require something just to read the massive phase currents..

 

[nieles]

IIRC you only need 1 board for the 0-5v signal from the Phase current limiter, i will make sure when i get home tonight

my idea for this was simply use a 50:1 "precision" divider resisor, to make the 0-5v signal a 0-0.1v signal. And use it as a input for the current sense input instead of a shunt.
not sure this will work, but is worth trying

 

[gwhy!1]

IIRC you only need 1 board for the 0-5v signal from the Phase current limiter, i will make sure when i get home tonight

my idea for this was simply use a 50:1 "precision" divider resisor, to make the 0-5v signal a 0-0.1v signal. And use it as a input for the current sense input instead of a shunt.
not sure this will work, but is worth trying

That would be cool thanks, Now that I have slept on this, the problem with my controller. Im beginning to wonder if I am over looking something a little more simple to solve ( It may not even be to to with the phase current popping the fets ) phase current limiting will be a must for reliable operation but I need to have a look at a few things in the driver board that just may be causing this problem.

 

[nieles]

dont know if it helps, but when i first tested my controller with 20A, one of the fets also failed, and my shunt acted as a fuse and burned in two. this also happend before i moved barely 20cm

this was with a similar layout as jerremy's controller but with an 2-layer pcb.

after the first prototype burned i stoped this project until i get the current controller fully working


i really hope it is not the Phase amps, but i fear it is.

i will see if i can make some schematics for you with all the needed parts for the Phase limiter parts if it needs more than one pcb

i did not think of this before, but we need a 5v line for powering the Phase amp sensing circuit
if we can use the 5v line from the brushless controller that would be helpful. the 5v circuit of the current controller right now is on a seperate pcb

Niels

 

[bigmoose]

I scanned this thread quickly, but did not see a schematic. Your problem seems to me to be a leg shoot thru event. Perhaps a driver timing issue. ... but I don't have enough data to be sure.

 

[nieles]

I scanned this thread quickly, but did not see a schematic. Your problem seems to me to be a leg shoot thru event.

so that could be solved with Low ESR caps, eg. ceramic or film caps right?

Perhaps a driver timing issue. ... but I don't have enough data to be sure.

what is the procedure to solve this? adding more resistance to the Gate resistor?

Niels

 

[Lebowski]

what kind of algorithm is running on the brains of the controller ? Does it change PWM frequency with rpm ?

 

[gwhy!1]

dont know if it helps, but when i first tested my controller with 20A, one of the fets also failed, and my shunt acted as a fuse and burned in two. this also happend before i moved barely 20cm

this was with a similar layout as jerremy's controller but with an 2-layer pcb.

after the first prototype burned i stoped this project until i get the current controller fully working


i really hope it is not the Phase amps, but i fear it is.

i will see if i can make some schematics for you with all the needed parts for the Phase limiter parts if it needs more than one pcb

i did not think of this before, but we need a 5v line for powering the Phase amp sensing circuit
if we can use the 5v line from the brushless controller that would be helpful. the 5v circuit of the current controller right now is on a seperate pcb

Niels

Thanks nieles, At a 20A battery limit on the bike it was fine, I have changed the gate resistors values and the bootstrap cap I have also lowered the pwm frequency a tad this have improved things a little , I really dont think its phase amps killing the fets at the moment the 4 fets that I have blown are all low side fets and all have failed in the manner.

I scanned this thread quickly, but did not see a schematic. Your problem seems to me to be a leg shoot thru event. Perhaps a driver timing issue. ... but I don't have enough data to be sure.

Thanks BM,
This is the conclusion that I am coming to , but to have it confirmed by you helps. the schematic is the JH circuit ( I will dig it out and post it in this thread )

 

[nieles]

Thanks nieles, At a 20A battery limit on the bike it was fine, I have changed the gate resistors values and the bootstrap cap I have also lowered the pwm frequency a tad this have improved things a little , I really dont think its phase amps killing the fets at the moment the 4 fets that I have blown are all low side fets and all have failed in the manner.

i was only using 4 220uF caps at the time, and no ceramic/film caps so that could have very well been the problem! the rest was stock, as JH schematics
in the next iteration i will also be adding 10k at the gate-source, and a 15v zener diode for extra protection