Is anybody tested new FSESC 75300 from Flipsky ??
From few days on my desktop - plan to use with 2028 CASTLE 800kv 12S sensored motors.
Any remarks to this VESC - welcome.
Becouse of XMASS still lack of time. Just first test done - working.
tbc…
4 Likes
some test done, connected 4kW load to motor
Testing in progress 
Battery CNHL 16s Lipo
MOTOR 2028 800kv 4 pole inrunner max power 15kW
now loaded about 4kW - foto below
Time to create 16 kW for my motor…
to be continue …
1 Like
How are you finding it without water cooling? I thought that ESC was designed for water. Also that high kv motor might be pushing it a bit, at 12S fully charged that’s 161k ERPM. 75300 is normally limited to 150k and flipsky probably isn’t the company I’d look to for advanced engineering of reliability
For me 70000 ERPM is ok, I plan spin 4 pole motor to 35000 RPM.
Cooling 1,3 kg aluminium concrete block. Water cooling this is 4 hole 8mm though enclosure.
During my tests ESC is very cool, even Inload 4kw warm up slow. Till now not exceed 30⁰C for short 15…20 second load.
One “FET key” is build by three transistor, in total 18 transistor
Finnaly ESC will be mounted on aluminium plate, I expect no problem with heat flow
now problem is how to generate 16 kW load
1 Like
Maybe I’ve done the maths wrong here, I think I calculated ERPM with 4 pole pairs instead of 4 poles
1 Like
yes, this is not clear described in a lot of materials
4POLE motors => we need
ERPM / 2 == >> RPM
on begin i do this same mistake
I Searchin good descriprion how to configure BLCD mode properly. I observe that FOC on HIGH RPM looks not so good like BLDC.
For me most important is good strong power on high RPM. Smooth start is less important.
ESC will be used for RC car - something like this
1 Like
This is really cool testing.
Check out this thread, the 75_200 no limit firmware should also work for that unit (same layout just 2 mosfets instead of three) and might perform a bit better
A propeller would probably be the easiest way to get a 16kw load without messing with two motors and resistive loads. Cool calculator here. I have the non-demo version so if you give me the parameters I could run it for you.
For BLDC mode, run the auto detection stuff on the BLDC tab, once you switch to that mode and write. The motor wizard sets up FOC only.
1 Like
For comparision
600A 8S rc car ESC
300A 20S VESC
XLX2 I personally tested up to 500A, but I saw log from 600A run.
NO LIMIT firmware - is available in VESC tool and Flipsku confirmed - should work.
Testing with motor is good because no crazy wind in garage 
I will have second 2028 800kv exact same motor to connect as LOAD nad will check.
TP5660 1450KV 6pole motor have very strong magnets, I’m very interested how CASTLE motor will work as LOAD.
LOW reststance Y load is made from ordinary stell wire 5mm. I have thre “resistors” 2,3miliohm, that give Y 4,6 miliohm load.
Looks maybe crazy but this is 10Euro hi power LOAD
BLCD - currently use with no SENSOR
FOC - use with sensor, automatically deteced properly by Vizard
HALL SENSOR in BLDC create strange big current start motor… need to investigate.
ERPM limit now I stet for 62000 31kRPM
Tested 90kERPM spin with soft load.
16S Powerin is intentional for 12S motor
I Try use VIZARD as JAYKUP advice… with no succes, try again.
Next test planned, compare 8S for both ESC
Test 6pole motor from VESC - 35kRPM will be 105kERPM still save for motor.
Work in progress
Thanks for any advice.
2 Likes
Short video from one test
I know this is a thing but their newer generation of controllers with aluminium boards are much higher quality than previous efforts. Need more data for sure but I think they will turn out to be much better than what came before!
5 Likes
Is this some Gen Z thing I don’t know about?
Just remove one S… and live will be easier
Yeah I saw @jaykup 's very interesting testing of the
newer 75100s, the thermal performance is impressive. Maybe I’m missing something but I thought tolerance of high switching frequencies was less to do with thermals and more about FETs and drivers with good rise and fall times, good trace layout, voltage tolerance of inductive spikes, basically things that are a bit more involved and hard to solve than thermals.
I get the (possibly unfounded) impression that flipsky doesn’t do a huge amount of novel design. For a long time they’ve worked from the standard 4.12 design, the allegedly stolen 75100 that they put out, the antispark that was borked and reworked by Gamer then reworked again but largely just stuck on an aluminium PCB. Maybe I’m underplaying their thermal design work or misattributing problems, but bleeding edge new innovation doesn’t seem like their thing
1 Like
This controller is totally sealed by some rosin or other coating. When we look under cover we can see only shunt resistor and silicon protecting cable … and metal bars on pcb for big current.
And some description that this is V2 pcb.
Totally sealed.
I hope will work strong.
i need pull up to 400A for maybe 20 second … then slow down…
Like on this video… but much faster
1 Like
I know prolems with spike, in past witx XLX2 8S powered I add extra vapacitors LOW ESC
16 X 1000uF 50V 0,017ohm in parallel
This give me additional reduction of sipke 2 times
Using CASTLE ESC on log we have special plot for
RIPPLE and we can check how looks
One interesting observation with pass on motor with stron magnets - with stronger battery (very strong) rppple is bigger than with weak battery.
Adding 16000uF capacitor Ibreduced ripple twice, tested with oscilloscopes and LOG from CASTLE
I will meter ripple with VESC too… work in progress
1 Like
The original TO-220 75100 was designed by a college kid. Flipsky bought one from him anonymously and reverse engineered it. The original designer posted a bunch of detail on endless-sphere. Performance wise it got hot very quickly.
The Aluminum PCB version is totally different. It’s not a simple “aluminum PCB” version of the original. Flipsky’s nomenclature is pretty terrible. I don’t know if it’s an in house design, or if MakerBase designed it, or if they commissioned it or what. It’s an impressive ESCs though. This 75300 should be a beast.
The current problem is that they put 75_300_R2 firmware (from a Trampa 75/300) on them, which is specified to use phase shunts (these units have low side shunts), is reading from 3 temp sensors even though only 1 exists which gives bad “average” temp data, and at least on the models I had, the input voltage reading was off by a volt or two. The hardware config I came up with, first on the original, then adapted for the aluminum PCB designs, addresses all those things. Now that firmware in the VESC project… but it has to be manually loaded to get it off the 75_300_R2 firmware.
Long story short, @sp3swj, if you are currently on firmware 75_300_R2, then try loading the FSESC_75_200_ALU no limit firmware from here
That should help the 75300 to perform better. It’s only been tested on the 75100 and 75200, but since all of those are shipped with 75_300_R2, and the hardware layout is the same except for the number of mosfets, then I see no reason why it wouldn’t work… and you’d be a great tester for a true Flipsky 75_300 firmware.
Does the VESC tool, with 75_300_R2 loaded, show three temperature sensors but only gives a correct reading for one of them?
2 Likes
