Hello Community Members, I would First of all like to wish you all Very happy Christmas and New Year, May this Year fill your life with positivity and health and keep your family safe and happy too.
So recently I encounter a machine which needs Servo drive and Servo action to work properly. This machine works on OLD Siemens DC Servo motor with Tacho Generator, The problem is " The drive which came with the machine isn't in a good condition and I thought to build a new drive using IGBT based H-Bridge and Optical Encoder long with PIC MCU. I would be pleased if your all can support me with your Ideas and Experience to build a Servo Drive.
Specs Of Motor-
Input Voltage - 130v DC
Current- 15 Amps

Material That I selected for the Power Stage-
Power Switch - Discrete IGBT IKW30N60H3
Gate Drivers - Isolated 10A/10A IGBT Gate Driver UCC5390
CPU - dsPIC33EP256MU806

 

DC motors in servo drives no longer use the tach, these were termed velocity type drives, now the tendency is towards torque mode (trans-conductance) amplifiers for servo applications.
This is presuming this is a closed loop application?
There is quite a few application notes on the PIC website.
There is also a 18F2331/4431 series that has a QEI (Quadrature Encoder module) featured.
There is the DC motor Picdem Mechatronics board also with associated software code.
Max.

 

Max, Thanks for your valuable feedback,I agree with you that Tacho sent uused in Servo drives , So I will remove the Tacho and replace it with an quadrature optical
Encoder.
I want to use an closed loop system to for this application.
As per your notes I had selected dsPIC33EP512MC806 which do have 2 QEI and 2 seperate ADC with Fast PWM

DC motors in servo drives no longer use the tach, these were termed velocity type drives, now the tendency is towards torque mode (trans-conductance) amplifiers for servo applications.
This is presuming this is a closed loop application?
There is quite a few application notes on the PIC website.
There is also a 18F2331/4431 series that has a QEI (Quadrature Encoder module) featured.
There is the DC motor Picdem Mechatronics board also with associated software code.
Max.

.

 

If the application requires a constant speed drive then why suggest a constant torque drive? What is the logic there??? A DC drive using a tachometer generator can work very well if the generator is still working. Of course it may be fun to experiment with code and try to produce a software driven VS drive that actually works. The first question is about reversing directions. If the motor does not need to reverse then an "H" bridge is far more complex than required. If the machine seldom needs to be reversed, like a lathe, then switch reversing would be the way to go, and the drive winds up looking a lot like a programmable power supply.
So the selection of the drive depends on what the motor actually needs to do. Does it need to run in an actual servo mode? Does it need to be reversible?, does it need to be precisely controlled at a number of different speeds? Lots of questions helping to decide what sort of control it needs.

 

It is almost impossible to purchase a DC motor with a tach these days, they are obsolete.
The down side with the velocity drives was the double tuning required, first the inner loop, the drive/tach and then the outer, PID loop.
With torque mode a simple drive can be used and the tuning done in the controller.
The OP states it is a servo drive so I suspect it could be a form of CNC or close.
For servo action, (positioning) then obviously a reversing drive is required.
@ashkarmalik I use a circuit similar to Tahmid's for DC motor control . http://tahmidmc.blogspot.com/2013/01/using-high-low-side-driver-ir2110-with.html
fig. 7 bridge.
Max.

 

Finalised Semiconductors-
MOSFET - IRFP4868
MOSFET GATE Driver - IR2110
And I have Attached the Priliminary Schematic.

 

It is almost impossible to purchase a DC motor with a tach these days, they are obsolete.
The down side with the velocity drives was the double tuning required, first the inner loop, the drive/tach and then the outer, PID loop.
With torque mode a simple drive can be used and the tuning done in the controller.
The OP states it is a servo drive so I suspect it could be a form of CNC or close.
For servo action, (positioning) then obviously a reversing drive is required.
@ashkarmalik I use a circuit similar to Tahmid's for DC motor control . http://tahmidmc.blogspot.com/2013/01/using-high-low-side-driver-ir2110-with.html
fig. 7 bridge.
Max.

OOPS!! I had not considered a CNC machine, only the ones like lathes and mills and drill press types. You are correct, some of them do require full servo capabilities. I have never used a CNC machine, only the machines that are completely operator controlled.

 

It is almost impossible to purchase a DC motor with a tach these days, they are obsolete.
The down side with the velocity drives was the double tuning required, first the inner loop, the drive/tach and then the outer, PID loop.
With torque mode a simple drive can be used and the tuning done in the controller.
The OP states it is a servo drive so I suspect it could be a form of CNC or close.
For servo action, (positioning) then obviously a reversing drive is required.
@ashkarmalik I use a circuit similar to Tahmid's for DC motor control . http://tahmidmc.blogspot.com/2013/01/using-high-low-side-driver-ir2110-with.html
fig. 7 bridge.
Max.

Thanks,, @MaxHeadRoom, As Suggested by you, I have Placed IR 2110 Mosfet driver along with IRFP4868 Power Mosfets.
So I have 3 Things to Sort out-
1) Does Galvanic Isolation need to be implemented inside this Circuit as I will be Using dsPIC33EP512MC806 which is a 3.6v Device?
2) What Precautions are needed for EMI and ground Loops when Circuit will be driving a 1.5Kw DC motor and there is the lot of interference inside the Circuit?
3) As I read somewhere that Initial Motor Current is around 20-30 Times its Rated Continous Rating.
So if My motor is rated at 120v DC @ 13 Amps then I need an H-Bridge Capable of Driving 325 Amps Peak, Is that Correct?
I have Chosen Mosfets that Says 49 Amps @ 100 degrees C So I need around Six Mosfets on One Quadrant of H-bridge or As Datasheet specify Id as Pulse Current for 280 Amps So I would require 2 Mosfet Per quadrant?
Total gate Charge is 270uC max
@42A
@150v
Gate Voltage Vgs @ 10V
Is the MOSFET Gate Driver Sufficient?

I will be Using Optical encoder for the Position Feedback so No Worries for tacho Generator.

 

Galvanic isolation is optional, and can be done with H.S. opto isolators if necessary.
You need to earth ground the motor frame and any other metallic parts of the equipment etc.
Initial motor surge current is often not an issues as with a PWM, the full current at switch on can be managed by ramping the current/rpm up in a controlled fashion.
If current limit is used, the motor current can be limited to the manuf. rated continuous current.
The motor appears to be similar to those currently used in a typical treadmill application, if so I don't see a reason to double up on mosfets.
Max.

 

I have Some Updates, I have Merged the H-Bridge With my UpcommingProject with 3 Phase One, So I can Drive DC Brushed Motors, BLDC and ACIM Motors From One Power Board only.
On This Post- Link

 

Hi Ashkarmalik - I admire your wish to build your own servo drive. I've spent a long time fiddling with H-bridges etc, but I've decided that the amount of time it takes to do it properly is best left to someone who does it for a living. As such, I've just purchased this, and it's really responsive (I design slider control systems which use these now) - link: https://granitedevices.com/digital-servo-drive-argon/

Saves the headache of fingers in AC inductive solder joints! And no, I have no link to these guys, other than I use their low voltage drives a lot.