I know this is old , but I find the figure quiet interesting , the source pin of the mosfets met at 2 point , point 1 switch node where main current is passing and point 2 is at the gate driver leg . doesnot that make a loop / antenna and radiate noise ? doesnot that also welcome the switching node current at least a small part of them to come to your gate driver board ?
HESC V1 150v 600Amp VESC based motor controller
- Thread starter Hackey
- Start date
I wrote this on my build thread but posting it here too maybe one of you can help me.
Also anybody got firmware for cl700? mine is on firmware 5.03
Today I played with the controller a lot.
I tried sensorless mode and this way I am able to spin the motor to 24000 erpm with no load, but with the bike on the ground it cogs and is not able to move the bike. If I put it in encoder low rpm are fine but as soon as it gets over 5000 to 6000 erpm it cogs. There is a négative current spike that stop the motor then it gets going again. So the encoder is good because it works flawlessly in lower rpm. When the cogs happen the motor is already switched to sensorless( I tried différent settings between 1500 to 4000rpm). I even tried to make it switch a 15000rpm and the cog still happens at around 5000 to 6000rpm
Also anybody got firmware for cl700? mine is on firmware 5.03
Today I played with the controller a lot.
I tried sensorless mode and this way I am able to spin the motor to 24000 erpm with no load, but with the bike on the ground it cogs and is not able to move the bike. If I put it in encoder low rpm are fine but as soon as it gets over 5000 to 6000 erpm it cogs. There is a négative current spike that stop the motor then it gets going again. So the encoder is good because it works flawlessly in lower rpm. When the cogs happen the motor is already switched to sensorless( I tried différent settings between 1500 to 4000rpm). I even tried to make it switch a 15000rpm and the cog still happens at around 5000 to 6000rpm
TorgueRPM
100 W
It should work with the newest vesc firmware off the vesc website, but if the controller is one of the earlier versions, I'd doublecheck with Hackey. New firmware might fix the issue on its own
j bjork
1 MW
Dont you need a controller specific file too? I dont think Hackey has shared those, so they can be included in the standard vesc package.
TorgueRPM
100 W
For firmware the manual just says "latest (firmware update supported)"
Probably best to check with Hackey, I've reached him through Facebook messenger
Probably best to check with Hackey, I've reached him through Facebook messenger
Hummina Shadeeba
10 MW
Can I use the 26-pin ignition wire as an on-off switch and leave the main leads connected?
did u program the motor with wizard? It didn’t work well for me and was told the wizard is no good for big motors.
Last edited:
I programmed the motor with the wizard and also in the motor categorie but both do the same result. Yes pin 26 in on off swith
Update in 6.02 firm
And do the automatic motor setup and after play with the parameters manually to search the good one, 3Shul usually are very bad at the auto dectect, you always need to tune the parameters manually and searching the good one to run properly
Also Ortega Observer is more facile to use and have better tolerance at wrong motor parameters so try to run with Ortega observer first, but MXLEMMING observer with very good settings are more efficient and have better torque per amper than Ortega if you manage to working with it
Typically you need to manually thune in Foc > General :
-Resistor (generally to reduce)
-Inductance (generally to reduce)
-Const Time who change the KP/KI value (in Detect and calculate parameters) (Also to reduce, usually from 1000 to 300~500
That's my expérience with E-trott motor hub and also Surron motors, but i never trying your QS motor
JRP3
1 kW
thomasates
1 mW
What would you recommend as a robust limit for Ciss/Crss? I am using IRFB/S/SL3307ZPbF for a project currently, and it is really bad. It has Ciss 4750nF, Crss 190nF (ratio of 25), and it reliably produces the parasitic turn on, so I am in the process of selecting an alternative. I am looking for components with a ratio of around or larger than 200, is this overkill?
Thanks, Thomas.
Hummina Shadeeba
10 MW
How are you able to assess if it’s running properly or efficiently?
mxlemming
100 kW
- Joined
- Jul 17, 2020
- Messages
- 1,209
Typically, heuristically, the ratio should be approximately equal to the voltage. It can be better to look at the gate charge at threshold compared to the miller charge. But MOS datasheets do not have a standardised way of quoting these numbers... it can be very hard to calculate exactly whether it will self turn on.
Your MOS is not great though for parasitic turn on, that is sure. Try IPP026N10NF2S which is probably more expensive but I am pretty confident does not self turn on.
thomasates
1 mW
Thank you for your advice - will try and report out! I already also realized that gate charge plays an important role, but flipping through different FETs in LTspice did also show that the numbers in the datasheet is one story, how the components behave can be a different one...
Greetings, Thomas.
Greetings, Thomas.
thomasates
1 mW
Ok so I went off and ordered a few of these FETs. After exchanging them, I did the same pulse tests that I used to characterize my halfbridge with the IRF3307 before to have a direct comparison. And I can truly say that it is a large improvement. The gate voltage (channel 2 on the attached pictures) stays much cleaner during transitions and all other signals are quieter as well. Thank you again for this helpful recommendation.
Thomas
Attachments
Last edited:
