This is greek to me, but here are the fets for those two versions
75100 uses 6x TO-220 MDP10N027
ESC continuous current = 100A
ESC max current = 120A
HV200 uses 18x DIRECT IRF7769L2
ESC continuous current = 200A
ESC max current = 300A
So are the mosfet ratings multiplied by 2 on the 6 fet version and 6 on the 18 fet version? Which rating do we use? How far off are the ESC “ratings”?
Those ratings are the ones most often quoted but are the least useful or completely useless. The important ones are the drain-source on-state resistance (which determines how much heat each creates), the thermal resistance specs, and the SOA graphs.
But there is much we can’t determine since we don’t know how well the ESC’s manage the temperature of the FETs. There are also things like how long the FETs take to turn on/off (which affects how hot they get via “switching losses”). More FETs in parallel can help to reduce the heating when the FETs are fully on but it takes longer to turn on/off paralleled FETs and that creates more heat.
There are so many inter-related things at work. Looking at datasheets unfortunately only provides a small fraction of the info we need to compare ESC’s.
Controlled testing is the best way to sort things out. There’s just no good way otherwise to know which are exaggerating their ratings or not.
Mind as well add some more pics. Lol. The thermal compound actually looked pretty good. Artic silver-type stuff. Better than the typical generic white silicone “star” brand stuff.
BTW, I thoroughly cleaned the existing compound with 99% IPA and reapplied with my own Arctic Silver 5 because removing and reinstalling the PCB into the enclosure caused the existing stuff a little thin.
You have me curious on the disabling of the traditional square/trapezoidal PWM method with BLDC ESCs. Not the gate driver EG3112 datasheet They look to be designed just for the purpose of DC to DC controllers and battery chargers.
I was looking back over my pics and something stood out that I should have corrected when I had it apart. Actually, Im going to take both of mine apart and apply a fillet of solder where I highlighted. Beefing up the supply rail traces with a “buss bar” is a good Idea for multiple reasons. Current capacity and to an extent more thermal mass but damn if your going to do it do it all the way. IMO, this is like a kink in a garden hose.
Is pA short for phase amps? If it helps I set my phase amps to 120A at ±40V but it exhibited a time out like period when I attempted to gun the throttle from a stalled rotor position which required a lot of torque to get going. I can only assume this is caused by some current limit being tripped.
This is my first foray into VESC land plus I dont have bluetooth connected yet for logging.
Overall, Im extremely impressed with the FS 75100 FOC. The torque produced insane. Launch mode in a Tesla or like being in a roller coaster that uses linear motors to accelerate from a standstill. Its a smooth, powerful, rapid linear acceleration that I’ve experienced only in those two examples and now w the upgrade to this controller.
I installed another into an identical NB Max today. Seems this ones battery is in much better condition, again no logging but seems to hold voltage under load better. More low end grunt and more speed.
Its addicting. Cant decide if Ill go 12s or 14s next. I love this shit!
Update, I decided to build a 3s6p using Panasonic IMR18650BD cells. I have a spot welder but couldn’t find my freaking nickel strips anywhere. Decided to solder the cells, I have a crap load of soldering equipment but for this I decided to try a 70 yr old 250W American Beauty iron that used to be my grandfather’s, it’s made for soldering sheet metal/copper roofs. Turned out to be perfect for the task. Add a tiny bit of rosin flux, pre tin both wire and cell, add a tiny bit more flux to the pre-tinned cell and voila, in and out in less than a second with the cell staying remarkably cool. The irons tip is a interchangeable plated copper slug 3/4” in diameter and 4” long, I chose one that tapers down to a 1/4” chisel point, more than enough thermal capacity, highly recommend if you need to solder cells. Knock off versions are sold on Amazon/ EBay for cheap, should work just as well. I have no reservations about this method.
No BMS, added balance taps to charge using a RC balance charger. Have voltage set to end charge at 4.02V, same as the existing 10s pack does, kinda sucks but I guess it’s set low to increase cell life? To fit the extra cells without being exposed I removed the Ninebots max has a built it AC charger located under the deck, I just charged with an external 10s charger. Tight fit to get everything in there, took a little grinding with a Dremel but I got it to work, wouldn’t be able to do it if it wasn’t for the FS75/100s compact size.
Anyway, I used VESC tool to rerun the FOC wizard. For some reason I needed to renter voltage cut off limits as they didn’t seem to carry over from the FOC wizard.
Took it out for a spin and there is a definite increase in speed. Low end torque isn’t what it used to be but it’s fun, really fun. Probably will keep tweaking but it works well as is.
I’m amazed all that power is being pushed through those fets. Crazy if you think about it.
Next thing is to get nRF Bluetooth up and running I have nRF51 and a NRF52 module’s. both should work. Any thoughts on which would be better?
My diy scoot already fried two xt90-s loopkey switches. By fried I mean there’s no fireworks. It’s just that the antispark capability of the xt90-s got disabled permanently. Seems like the resistor inside got busted. Was it because of the 75100 or the 20s?
There’s a whole thread on loopkeys dying at 20s voltage. Flipsky has a 20s antispark which may be helpful.
No, your 75/100 didn’t cause your loopkey to fry. Well, technically it did but its normal. Sounds like your antispark resistor took a shit.
Im powered by14s now using a regular xt60 as a loopkey. Using a 10W 10ohm resistor to charge caps immediately before plugging the loopkey in, Not elegant but it works. Antispark xt90s are out of my budget ATM. Yep, Im a unemployed, previously successful, recently divorced, broke 48yr old father of two living back at mom and dads fighting an unknown illness that causes extreme fatigue that is bad enough where I couldn’t be reliable enough to hold down a job. . No disability for me as no doc can figure it out. Life just great!! Dont know what that has to do with antispark connectors other than Im broke. Lol!
I use the 75100 now in a 20S build with a scooter motor (detailed information are following) and tried to max out the stock FW specs. So I set it to 120pA, 120bA, 160A absolut max and 150k ermp. In this setting I’ve gotten over current faults, so I set it to 100pA and left the rest like it was. The interesting part is that, the max battery AMPs pulled on 120pA were 36A and at 100pA 34A. I don’t get it, since the battery is good for well over 100A. I also got a temperature of 85C on a short 100m pull. Everything seems quite odd.
The motor is rated for 48V 450W though, but this shouldn’t affect the low bA to pA ratio, or am I missing something?
Here are some logs with 40pA, 50pA, 60pA and 70pA.
So far I would rated this ESC as a 20A controller.
So, I fried my first two 75/100s. Ordered two more, just received them yesterday. Looks like a V2 version. To me the most obvious is changes in the power supply and additional current sensing resistors??? stacked in the bottom side of the board. Oh, the PCB is blue.
Wonder if this is why they disappeared of FSs site for a short time?
Hi
I’m new in forum 
read from few hours about 75100 across forum. It’s good to read your posts. (I also hope your private affairs will go well soon).
Again thanks for all info and foto. If FLIPSKY use only kapton as FET insulator… hmmm not pro solution.
I ordered one and wait for delivery, mavbe will be V2
I bougt this VESC becouse need really quiet work of motor, need replace BLDC Castle ESC 8S powered in RC roller with sensored Castle motor. I hope will be quiet in FOC mode. After reading several pages I think i leave original firmware.
Total weight of platform maybe will be 30 kilogram.
… tbc…
Regards
Jerry
sp3swj
Thanks… You will definitely notice a HUGE reduction in noise and possibly better performance when using FOC. Its even possible to make it quieter by increasing the switching frequency to 30khz from 25khz which is in the FOC advanced menu.
Yeah, the Kapton tape insulator isn’t great but, it is what it is I suppose. Probably a cost issue for the FETs. Not sure.
I agree with others here that the "floating electrolytics’ weren’t such a great design either but I added tiny bits of foam to support them. Added the foam in such a way that its not covering any surface mount components which could then overheat.
Old Castle Creations… Yep, they definitely make motors whine. IMO Castle were great at one time but they were too expensive. They also fumbled and missed the drone market big time. By the time they realized it they were too late. They stalled with the same 8-bit processor for many years, incremental innovations, nothing huge. I had many of their ESCs and BECs but once I got into multirotors and found castle ESCs couldnt handel the high refresh rate needed for good control, not to mention the expense of spending $200 for four of their 30A ESCs I bailed. I found guys were flashing dirt cheap ESCs with SimonK or BLHeli firmware whose performance was MUCH better. SimonK seems to have disappeared but BLHeli moved on to 32bit and is still being developed. Its the defacto FW for “drones”.
Castle still sells a good amount of air ESCs and also seemed to pivot to the RC car market but still, assembled in the US or not they are still too expensive IMO. I could be wrong but I think Castle is still using the same 8-bit processor for all of their lineup, air, and land that they used back in 2003! But hey, they have great marketing!
Hi
In my car I use XL2 for speedrun. My not fo fast car pass is 116mph 8S powered 350A current from battery and not big motor 4082 size.
From some time i’m interested VESC for speedrun, maybe i realise some test. Castle XLX2 now rules in this type of hobby, up to 600A peek 8S cars.
I need really quiet besose camera platforms need quiet drive.
I think i firtst check out og box if work 75100 … then i fix favrory fuckup like add MICA and ptotect capacitors.
… waiting for delivery…
Wow! 600A 8s! That’s about 20kw peaks!!! I thought I was badass pulling pulses of 150A from a 4s 2200 with four 2206 motors on a freestyle/racing quadcopter.
I bet you added lots of expo to that transmitter huh! Not trying to be negative but I hear almost all gear noise, not noise from square/trapezoidal switching from motor commutation. Maybe a lot of the noise travels through the motor shaft and is amplified by the drive train. I dont know, will be interesting to find out.
Mica is a good idea. Some good idea here. Insulators-Kapton, mica, silpads. - Page 1
While I had my ESC out of its case I removed the AL heatsink and applied a slight amount of artic silver 5 thermal compound between each FET and the Kapton. Not sure if it helps but as long as it isn’t dripping I suppose it cant really hurt. Sounds like straight up anodized AL is best under 400V.
About 600A pull - you can find logs of end of movie 180MPH Limitless carbon FC100 body test run - YouTube this one of the fastest men now.
Now i start testing new motor SSS size 5684, first test on 4S only 42pinion 29spur and diff standard 13/43 - test 5686 1100kv 4S 42-29 13000rpm 94A 15V - YouTube just quick test. On Dyno I have voltage and temp and preparing current. Now I’m on holiday and some test waiting when back home…
Sound in my movie, camera was mounted to plastic body and every sound is equalized.
Vesc - definitely wil be tested to on my cars. I little disapointed short time overload of VESC, in Tekin i have 220A and 880 burst shirt time. ZTW esc on 2S i short time overload to 360A (metered by FLUKE clamp) for 120A current and 760A declared as burst.
Maybe for my needs I must something change in vesc firmware
Havent watched your videos yet. I will though.
Definitely use the stock 5.2. 5.2 works well for me despite having no field weakening. Maybe you have the skills to tinker and get 5.3 working… I dont!
EDIT: 5.3 works just fine. Just remember to disable phase filters in VESC tool!!! You will let out the magic smoke if you don’t!
Plus, the default FOC switching speed is 40khz. Motor is even quieter now, 10khz higher than the default on the 75/100FOC V1.
Im kinda cross posting between this and Jaykups hacked FW thread. More info there.
Keep us posted please!
