FLIPSKY new 20s 100A tiny controller (vesc based)

bhartshorn said:
Been out of this game a little while so I can't tell - is there any indication whether this controller is based on VESC4 or VESC6?
See the email I posted earlier from flipsky. Its based neither on 4 nor on 6, it's based on 75_300 :)
https://vesc-project.com/node/311 -> all open source VESC schematics (afaik)

district9prawn said:
The controller losses will not be different when running at say 25% and 50% duty, assuming phase current is the same. Switching frequency is always the same and if the phase current is the same mosfet conduction losses will also be similar.

Yes, as you correctly pointed out - if phase current is the same.
But if you also increase the number of turns in the motor coils (so you have the same top speed) you need less phase amps for the same power output. (Later Edit - I removed some totally incorrect stuff from this phrase)

So, in the end, if you also redesign the whole system around the higher voltage, (altering just the number of turns in the motor windings, so the top speed is the same) you need less phase amps for the same power / torque output, so conduction losses in controller mosfets and phase wires (assuming the mosfets and phase wires stay the same) are lower.

But then again, if you also have a controller designed for your voltage, higher voltage FETs have higher rdson, so the fet losses difference also disappears. And all you're doing is slightly changing the losses in the phase wires :lol:

No free lunch here. Bummer.

Br,
 
badgineer said:
john61ct said:
OK I was making assumptions about that 84V, 75V÷20S is 3.75V figured close enough to li-ion nominal.

Well, good news!
Flipsky replied to my inquiry and it seems it really is 20s capable. So you were quite a bit closer to the truth than me!

As this esc is based on vesc (Benjamin Vedder esc opening source) 75, we call it 75**. The max voltage can be 90V during our test, safe continuous 84V.

We use 6pcs mosfets for this esc, MDP10N027,MagnaChip. As the chips, drivers and mosfets keep up rising, we are trying to find some replacements with similar specification performance. Thanks.

MDP10N027 is 100V 2.8mOhm rdson TO220 FET, quite nice. Would have probably picked something similar myself :)

so it looks really promising!

Br,
The mos isn't really the limiting factor in 75300 designs, it's the ina240 opamp.

I guess they found an alternative, or are running it right up to the wire of spec (INA240 absolute max common mode input voltage 90V, differential 80V...) This doesn't provide much safety barrier over 84V20s and with through holes FETs there will not be negligible ringing and overshoot.

They may of course have found other ways around it like clamping the opamp inputs, or they may have just put the shunts on the low side like the vesc 100/250.

Please someone open this up for a critique!
 
mxlemming said:
The mos isn't really the limiting factor in 75300 designs, it's the ina240 opamp.
....
(INA240 absolute max common mode input voltage 90V, differential 80V...)

I always wondered where the 75v limit on the 75300 was from. I couldn't tell from casually looking at the schematic and components, I kinda thought that slapping a 100V FET would make it a 20s controller.... (would not have been able to tell even if I invested time in it, tbh).

Thanks for this info.

Br
 
So guys do you think it will allow to set max motor AMP to Somewhere around 160A?

In the end that is what gives your startup acceleration, this controller will replace my Kelly kbs-x 72121 wich is 50amp cont and 120 max motor current.
 
atkforever said:
So guys do you think it will allow to set max motor AMP to Somewhere around 160A?

Just a suggestion: Ask them :). They've responded to my emails pretty fast.
Probably better than speculation, at least than MY speculation anyway :)

So ill stop speculating before I say something dumb again. :)

Br,
 
badgineer said:
atkforever said:
So guys do you think it will allow to set max motor AMP to Somewhere around 160A?

Just a suggestion: Ask them :). They've responded to my emails pretty fast.
Probably better than speculation, at least than MY speculation anyway :)

So ill stop speculating before I say something dumb again. :)

Br,

I'll allow myself a small amount of speculation... It has 6 to220 MOSFETs, which means each mos takes the full current.

I suspect 160A is around the current that would fuse the mos legs considering the duty cycle. They're rated for 120 each. I don't think it's likely to survive this.
 
So rated for drawing 100A continuous from BATTERY

does that mean 180-200A is possible to the motor when low RPM is dictating say 36V internally?
 
john61ct said:
So rated for drawing 100A continuous from BATTERY

does that mean 180-200A is possible to the motor when low RPM is dictating say 36V internally?

No.
 
Then we really need to learn the continuous vs peak phase amps it can withstand in a high torque low rpm context.
 
they replied on IG : "Hi, the motor current also not exceed 120A."

So i'm afraid max phase amp is 120 which is sad
 
atkforever said:
they replied on IG : "Hi, the motor current also not exceed 120A."

So i'm afraid max phase amp is 120 which is sad

120A is the package limited max amps of the FET. Bad things happen to FETs that exceed their package limit max current.

"For instance the “package current rating” of the FET is for the most part, independent of ambient temperature, and is a function of the internal connections between the silicon die and the plastic package. Exceeding this value does not immediately guarantee that the FET will die, so much as prolonged use above this limit will begin to reduce the lifetime of the device. Failure mechanisms above this limit include but are not limited to wire fusion, thermal degradation of the molding compound, and issues cause by electromigration stresses."
from https://e2e.ti.com/blogs_/b/powerhouse/posts/understanding-mosfet-data-sheets-part-3

BR,
 
IMG_20200901_214317.jpg

Lol at that ti note. That's a ti csd19535 with a200A package limit. The thin section of the leg and PCB just vaporized with a 40A fuse in the battery line. Photos pre burning show it was soldered well. Not masses of solder but well enough.

Yes there was a firmware issue that resulted in pretty high currents but there was a hardware comparator that trips out at 300 ish amps... You get the idea regarding that kind of mos and current limits.
 
Package limit is given for continous current, since in our case current is pretty much a sinewave are we still limited by this value ?
 
atkforever said:
Package limit is given for continous current, since in our case current is pretty much a sinewave are we still limited by this value ?

I was under the assumption, that when you set a 120A limit in the VESC (don't own one yet, so *assumption alert* ), you set a limit on the RMS current - meaning "continuous equivalent". The actual signal is a PWM of very short much higher current pulses, modulated in an sine or trapezoidal envelope.

So yes, exceeding it by a large amount will eventually cause some magic smoke and frustration (assuming you're not after the magic smoke)

But even if I'm wrong here:
1) the thing is still designed for 120A max, so you probably wont be able to be able to go beyond that without some mods
2) you still have the 12 AWG wire which is normally (industry) rated for much less, and for example hobbyking usually mate it with 60A connectors (implying 60A is max what hobbyking is expecting it to handle reliably). (https://hobbyking.com/de_de/xt60-male-w-12awg-100mm-silicone-wire-5pcs.html)

......

So long story short: If it's designed for 120, exceeding that will probably get you into danger territory in a lot of specific areas.

I for one don't like disappointments so I'm trying to keep expectations realistic.

Also, again, for the $ and size, 120A, at least to me, seems excellent.

BR,
 
mxlemming said:
They may of course have found other ways around it like clamping the opamp inputs, or they may have just put the shunts on the low side like the vesc 100/250.

They said it has 3 low side shunts.

Also said that firmware is 5.2 & cannot be updated.
 
afzal said:
mxlemming said:
They may of course have found other ways around it like clamping the opamp inputs, or they may have just put the shunts on the low side like the vesc 100/250.

They said it has 3 low side shunts.

Also said that firmware is 5.2 & cannot be updated.

3 low side shunts is a good solution. Easiest way of getting 20+s voltage capability.

I'm not even sure how they would stop firmware updates. There's a usb going to the VESC chip... It's going to be updateable. If they try to eliminate that in firmware they'll quickly arouse the wrath of Ben and Frank.
 
mxlemming said:
I'm not even sure how they would stop firmware updates.

I don't think they are actively trying to stop updates. I think it's more something along the lines "Well, we adapted the code a bit to our particular implementation and can't be bothered to release it. And what do you mean by GPL? ".

mxlemming said:
If they try to eliminate that in firmware they'll quickly arouse the wrath of Ben and Frank.
If it's a GPL violation, or some other such shenanigan, I sincerely hope they can do something about it.
Go go Ben and Frank :)
 
Are these controllers being sent from China? I went through PayPal checkout and it's only £75 to get one. I am in Scotland and there may be customs charges when it reaches England.
 
sdobbie said:
Are these controllers being sent from China? I went through PayPal checkout and it's only £75 to get one. I am in Scotland and there may be customs charges when it reaches England.

yes and you better pre-order since theres Chinese national day on the 1st of october so no shippement for 7days.

Fedex is the carrier and idk how is england law but in France fedex complies with the authorities and you pay 20% taxes every time
 
I might leave it then if they are using fedex. That 20 percent tax would make it more expensive than the ones I will be making for people in Scotland and England to buy.
 
sdobbie said:
I might leave it then if they are using fedex. That 20 percent tax would make it more expensive than the ones I will be making for people in Scotland and England to buy.

But aren't taxes more depending on the country you're in and the commercial ties it has with the country the goods are shipped from? At least I for one expect to need to pay tax, independent of shipping method, with a 95% probability when buying from China.

Ow, and hope your controller ends up as inexpensive as you plan. :)

Br,
 
Some countries see import duties as an important tax source need to pay Customs bureaucrat salaries.

Others don't bother much for consumers, once they see the cost of collecting it is greater than what they collect.
 
Back
Top