Yes i thought something like this
I think pid controller would also be possible but may be overkill
Those seem pretty cheap and fast. Iāve never dealt with isolators. However these are not photo, but opamps I believe. Not sure it makes any difference.
These seamed the easiest solution to me and 2.5 kV isolation is good enough.
Edit: i donāt really care how they work but as far as I know the signal i put into it comes out on the other side, just galvanically isolated
Just to remind you this is over-voltage case. You still have to solve the other case where you actively turn off the VS or the mosfets and freewheel through the diodes to brake downhill and pipe the current through the brake resistor. In that case the chopper will control the braking power. And if you use a battery, you channel the voltage to a regenerative circuit to charge back the battery.
I think Iām missing something again, but for breaking just slow down the rotating magnet field (lower frequency), so the magnet field lags behind the rotor.
I used this kind capacitance charge isolators before, to connect two TLL logic circuits where the one of the circuits had a floating 19v as its ground reference and 24v high side (5v TTL lol). It was done to be able to place a DAC within voltage range to accurately activate some gates in linear rangeā¦
Probably all the above might as well be written in greek for some of you 
I think I get what you mean.
This sums up my problems pretty good and is in general a pretty interesting read.
This is how it operates. Itās common for shared DC bus systems to use a single braking circuit at the AFE to regulate the dc bus. This allows other connected drives to consume the energy before itās being dumped to the resistor. It doesnāt really need a special logic or chopper. Just a switch that turns on when voltage rises above a threshold. And can be done with a mosfet, zener and few other passive components.
@Roiki do you think itās better to controll this via the MCU because i want input voltage range from 24 to 230v or via hardware?
Both have positives and negatives. Hardware isnāt prone to bugs but has a fixed threshold voltage. Mcu control is adjustable but needs to be isolated and requires isolated voltage measurement. Which is a bit of a hassle. And is subject to bugs.
Iād do it with the mcu as its more flexible. All it really needs is a few wire wound resistors and an isolated mosfet circuit if the working current isnāt too high.
Iāve been designing a high voltage AFE for servo drives so I made it with an isolated gate driver and power igbt.
@Roiki with the igbt module Iām using, the MCU is referenced to dc- on the DC bus so i can just use a voltage divider into the MCU adc to measure the voltage.
The implementation doesnāt really matter. You can do it without isolation too. Even a logic level fet works. I just prefer isolation when working with high voltage.
And you donāt need a pid loop or anything. Itās actually detrimental since fets generate heat with rapid switching. A simple on/off switch is the simplest.
Question, on a large motor spinning at high RPM opening a switch at 100% braking would create a massive mechanical stress. Wouldnāt that be too dangerous? Electrically the circuit can handle it but mechanically it seems dangerous. @Franz_Schmidt board will control a 10kW motor, if that brakes suddenly at 100% it will rip the foundations off. Iāve had large motors braking before, and had a case where the motor ended in the parking lot of the facility and smashed a car. Not a joke. Happened in Texas 20 years ago. Luckily no one got hurt.
Iām not sure what youāre referring as breaking as theres two kinds of ābrakingā. Thereās the braking done by slowing down the field control and thereās freewheeling where you shut the high side fets and the motor acts as a generator. This generator generates electricity on the dc bus which then trips up the breaking resistor and dumps the power somewhere. Like a large resistor or a battery or flywheel.
I donāt see how this could cause anything on the motor in any case. The resistor only bleeds off excess energy, it doesnāt act as a load for the motor. To me that sounds like a design flaw more than anything to do with the drive braking circuit.
āBrakingā is really a misnomer as its more of an overload protection than actual breaking. That has to be done by the drive.
Thatās exactly what happened. The motor was under heavy load (huge industrial centrifuge) and instead of slowing down gradually the circuit shorted (conductive dirt built up inside the controller for years) and the braking resistor (that thing was as big as a brick) literally lit up and melted, while the motor ripped from the foundation, broke off the shaft and bounced across the floor through a wall into the parking lot.
The breaking resistor didnāt do that though, did it?
The circuit was supposed to turn the motor off, the centrifuge turn by inertia, and slowly increase the share of the generated power through the resistor. Instead, it sparked, it shunted the whole slow-down circuit and directly created a path through the PCB going into the resistor. The resistor could have taken it however it used liquid cooling (oil) which was all gone, poor maintenance, and as it heated up the resistance increased to a point it suddenly absorbed everything coming out and melted. The whole accident was due to lack of maintenance. Anyway these are details.
Therefore my question, if you just shunt the motor with the resistor suddenly at 100% that would stop it so fast it may create a mechanical problem.
I think if the resistor isnāt a really low value, maybe 50 Ohm (Standart odrive uses 2ohm) and the MCU is fast enough,
it gets to overvoltage like 220v instead of 200v and the resistor with igbt is turned on and it shuts off fast enough when it gets back to 200v so it doesnāt pull it down lower than 200v itās fine i think.
What youāre describing is a short between two legs, which causes the motor to stall. So thatās a driver fault, has nothing to do with the breaking resistor whose only job is to maintain proper voltage range when the motor acts like a generator. Iām honestly confused what youāre after here.
Itās typical to use 50-100w 50-150 ohm wire wound resistors. And you can use multiple in parallel.
You can also set different points for turn on and turn off voltages for breaking if you want. Though you should leave a little headroom with absolute max voltage and use overvoltage shutdown as well.
You can also look at how odrive does it. Itās a simple implementation that works.