Motor control 180VDC 2HP, 9.6 AMP. Currently 1 part is being burned up. If anyone has any ideas I would be so thankful.

Thread Starter

onesys01

Joined Feb 10, 2021
5
For reference, I have included the full schematic as an attachment and the board layout for reference.

The general theory of operation: I have a Rasberry pi to act as a UI and the pi is connected to an Arduino via USB serial connector. The Ardunio sits on the board as a hat and this works well. The purpose of the system is to control the speed of a motor that is 180VDC and 9.6 AMPS steady and 30A surge. The speed of the motor is supposed to be actively controlled through a schedule, i.e. so many RPM for so long, different RPM for another period of time. The time periods are in hours and the speed is in the 20RPM to 300RPM range. The motors currently are doing this via manual control, that's what the goal is to take the man out of the middle and have the "recipe" run automatically. I will attach a photo of the device to this posting also. The device will also go into what's known as preservation mode, where the rotation will drop to 20RPM to keep the items turning. The purpose of this is to breakdown zirconium powder in the presence of deionized water, into smaller particles, so that they can be cast into a dental crown substrate.

The issue: I hope I can be clear here, the board itself when hooked into the 220 seems to be fine. However, the moment the USB cable is plugged into the Arduino board R12 lights on fire. I don't have a clue why that is happening.

Test Conditions: When the USB is plugged in only the Arduino correctly controls the lights. It also correctly controls the PWM and dims and brightens D9 as expected. Also, suspecting the ground at R13 being attached to the high voltage side, I am not stuffing R13. The idea being that the U1 buffer is only allowing high or low, never floating. The circuit actually appears to properly control the TP9 test-point and I can see the amplified PWM signal. The reason I am running the gates at 12V is so that I can get them to turn on all together within the avalanche time. Any ideas would be very welcome and there is a possibility of compensation as well.

If you have some ideas please help me with this. If you have other questions, I am very happy to answer them. I am actually a computer programmer, not really a hardware guy, although I am learning. Thank you so much.
BoardLayout..PNGRoller.jpg
 

Attachments

Alec_t

Joined Sep 17, 2013
12,080
Welcome to AAC!
With four FETs in parallel you have about 800nC of gate charge. A single CD4069 will struggle to drive that properly unless the switching frequency is pretty low. What is the frequency? What is the value of R12?
 

Papabravo

Joined Feb 24, 2006
16,165
The what seems fairly straightforward. The high voltage supply is finding a low impedance path to ground through the USB connector.
EDIT: missed the PDF with the schematic
 

Thread Starter

onesys01

Joined Feb 10, 2021
5
Welcome to AAC!
With four FETs in parallel you have about 800nC of gate charge. A single CD4069 will struggle to drive that properly unless the switching frequency is pretty low. What is the frequency? What is the value of R12?
Thank you so much for the welcome and I appreciate it. The PWM frequency is 980Hz, R12 is 1Kohm. In testing it seems to work well on the scope, but take that with a grain of salt!! :)
 

Thread Starter

onesys01

Joined Feb 10, 2021
5
The what seems fairly straightforward. The high voltage supply is finding a low impedance path to ground through the USB connector.
EDIT: missed the PDF with the schematic
Thank you for responding. When you say you missed the PDF did I miss the PDF or did you not see the attachment? Did I forget it?
 

Papabravo

Joined Feb 24, 2006
16,165
Thank you for responding. When you say you missed the PDF did I miss the PDF or did you not see the attachment? Did I forget it?
My first response which was only up for a shot time missed the included PDF file attachment. I changed my initial response when I read the response from @Alec_t and realized that you had provided it. Could you perhaps enumerate the total number of power supplies present in the entire system. In particular which side of the USB cable provides the +5V Power and USB Power Ground
 

Thread Starter

onesys01

Joined Feb 10, 2021
5
My first response which was only up for a shot time missed the included PDF file attachment. I changed my initial response when I read the response from @Alec_t and realized that you had provided it. Could you perhaps enumerate the total number of power supplies present in the entire system.
Love your quote by the way. There is 220 coming in. That drives the motors it is also used to run a 12VDC power supply which is to drive the MOSFET gates and a cooling fan, eventually. That 12 VDC is used to drive a 5-volt regulator to provide power to the Arduino, the display, and a separate Rasberry pi. The Pi provides the UI and controls the Arduino via serial connection.
 

MaxHeadRoom

Joined Jul 18, 2013
23,125
If the Main AC power is not galvanically isolated, there will be a conflict with the USB 5v being referenced to earth GND.
This point in the circuit is referenced directly to earth conductor.

1612982535550.png

IOW you do not have galvanic isolation.
 

Papabravo

Joined Feb 24, 2006
16,165
Love your quote by the way. There is 220 coming in. That drives the motors it is also used to run a 12VDC power supply which is to drive the MOSFET gates and a cooling fan, eventually. That 12 VDC is used to drive a 5-volt regulator to provide power to the Arduino, the display, and a separate Rasberry pi. The Pi provides the UI and controls the Arduino via serial connection.
I'm sorry to hear of your distress. There are two kinds of engineers, those who have burned something up, and those who are about to.
The original was about pilots, but you get the idea. I've had a few WTF moments in a long career. My favorite was working in a factory with a high ceiling, popping the pressure relief plug on a Big Blue Capacitor and watching the steam rise 25' to the metal ceiling along with the whistling noise. WHOA! -- cool.
 

Deleted member 440916

Joined Dec 31, 1969
0
And I am afraid when your done with the above problems you need to learn a whole lot more about how to use power mosfets, you cannot possibly switch even a single power mosfet with a grotty little cmos gate and expect the mosfet to survive, the switching speed will be so slow the peak power dissipation will simply destroy the device, in fact it is quite probable that one or more of your devices already has a gate to drain short compounding your other problems. Your mosfets have a gate charge of 110nC and they need a pretty hefty driver (several amps plus) and certainly not a 1K series gate resistor!
 

Thread Starter

onesys01

Joined Feb 10, 2021
5
This is the final schematic. What we did, in the end, is to have the RasberryPi function only as a user interface to an Arduino, which is running the attached sketch. The Arduino is connected to the Pi via a standard USB connection and is run at 115,200 BAUD. The Arduino loop reports the current speed, the desired speed, the current state (i.e. Slow, Normal, Fast, or off). The only input to it is an integer which is the desired RPM of the device. The pi then simply reports this. We have an automotive-grade sensor that reads pulses as gear teeth pass by. There are 24 teeth per revolution, which means we have extremely frequent updates from the unit. This results in a very low dead-band time. simple serial communication is all that is necessary between the two boards. The PWM output of pin 5 is passed through an RC filter and the PWM percentage is converted into a linear 0-5VDC. Then for the original motor controller, we purchased an optional board that accepts the 0-5VDC as an input to drive the actual speed of the motor. The feedback from the rollers comes back through the sensor into the Arduino to report actual speed. As speed falls below optimal the Arduino steps up the PWM percentage, effectively "stepping on the gas". If the speed gets too high, the opposite occurs and the speed is reduced by decreasing the control percentage. All of this occurs in the Arduino as the RasberryPi is passively showing the speed as reported from the Arduino. In even the Pi was disconnected, the Arduino would continue to actively control the speed at the last set speed. I may implement a timeout so that if the RasberryPi is disconnected for greater than 15 minutes, the Arduino drops the speed down to "preservation mode", as discussed previously. The final board design is not complete, but as soon as it is, I will post that also for your reviewing pleasure.
 

Attachments

Deleted member 440916

Joined Dec 31, 1969
0
Then for the original motor controller, we purchased an optional board that accepts the 0-5VDC as an input to drive the actual speed of the motor.
So that is the answer to your original question about the mosfet, you didn't fix it, you gave up & bought someone else's solution.
 
Top