I am using 4 - 24V wormgear motors rated at 4A, 50 rpm, 50 Kg/Cm torque with levers that are 40mm center to center, a 600 watts PSU and VNH5019 drivers on a motion simulator.
I am still on testing phase. Changes of direction must be fast and frequent, sometimes 2 or 3 per second. I use a load that would be average on final product.
After some reversing, it will hesitate on reversing and sometimes stop for maybe ½ a second or less which is not acceptable at all on a simulator, maybe the over-voltage protection is taking over? I reduced the voltage to the minimum I could (21.6v) and the hesitation seems less a problem but it is still there.
Is there a way to eliminate this with large capacitors or are the VNH5019 not suited for the task near 24v ?
Can you clarify a few things about your motor specifications and how you are using it? What voltage are that speed, torque, and rated current for? What is the stall current of that motor? Is the motor going from full speed forward to full speed reverse and back again? Also, what are the DIAG pins doing when the motors behaves this way?
From what you described it sounds like the motor driver’s over-current protection might be kicking in. In our experience, the over-voltage protection cuts in immediately and disables the output while the voltage is above the threshold, which does not sound like what you are seeing.
You should be aware that fast, full power reversal of direction is the very worst case for a motor controller to handle. The motor current can be TWICE the stall current during a fast reversal, so I suspect than an overload setting is being tripped in the controller.
If you absolutely require this sort of performance from your motor and system, then the VNH5019 may not up to the task.
Those motors are GW6280 and I was not able to identify the maker yet. Unfortunately, they do not publish the data your are asking like others do. However I think they are right for my application and they also have a bigger 12V version the GW7085.
It is the first time I use 24V motors and this was not expected because it never happened with 12V motors. How am I using it ? For testing I am using my own Windows app that send data to the mc just like X-Sim does but I am pushing to find the limits. When my mc receive the data (every 25ms), it filters every values to eliminate spikes and average the last values to smooth things out, more values on lateral g and less on vertical and longitudinal so it reproduces bumps and accel/decel in as realistic as possible. With those averages it calculate a target for each motor. Next step (10ms) it recalculate set point for each motor, this time averaging values even more up to last 15 rounds that is 150ms (more make the sim look disconnected with the game), again for the same reasons. The final result is a simulator that is not sluggish but quick in changes and smooth enough that you can race at will, for hours.
I have set a maximum duty of 75% and limit the I (PID) value to eliminate winding up too high.
Am I testing to a point that will never be attained when racing ? Maybe, I will only know later on complete assembly of this version.
As for what is the status of DIAG pins then, I do not know since I use them to disable the drivers when the simulator is stopped. The PSU is also disabled with the use of a SSR. Since it is just momentary, can I have a usefull reading of some kind ? How ? Can it turn a led ON/OFF ?
Maybe I was not clear on first post because of lack of knowledge but I read many datasheet for H-Bridge lately and most if not all have a cut-off at 27-27.5V for overvoltage and I guess that is why reducing the output on my PSU also reduce the frequency on this situation. Maybe it will make it react too slow ? I must also say that it happens most of the time when I simulate repeated bumps or a very rough track surface.
What I did with the VNH5019 is that I installed them on a simulator of first generation that has 12V motors and a 15V PSU and they do a real good job running in the 30-50°C range without a heat sink. You and STM have done a great job with those, thanks. I will try the current prototype with the first prototype drivers to compare.
I think next time I will stick with 12V motors but with those 24V, do you think of anything else I can do or look at ?
It sounds like you observed that this behavior happens more during bumpy and rough terrain in your simulator, which seems like it should correspond to the motor reversing direction rapidly more often.
Do you have a multimeter you could use to measure the stall current? You can approximate the stall current of that motor by stalling the motor at a low voltage, measuring the current draw at that voltage, and extrapolating linearly to 24V.
That 24V motor looks beefy; do you know how the 12V motor you were using before compares to it? If the 12V motor drew less current, it makes sense that it would not have had the same issue.
Also, I do not think this is a case of the over-voltage protection being triggered, but you could test it by reducing the voltage to something like 18V to see if this behavior goes away completely.
You are suggesting it could be a current problem, very interesting. Could it also be the PSU not powerfull enough at 600W ? I think I can make a quick test first by testing with 3 motors only then using way you suggested or an ammeter. The project include monitoring many parameters and sending data back to the PC for graphing and fine tuning but has not been fully implemented yet. The Arduino Mega can still loop at 50K/sec with the task it has to do now without sending data through BT but then it drops below 38K. If not up to the task, the STM32F103RC can do it at well over 250K but a few things still have to be implemented before making the swap.
The key with racing simulator is fast and powerfull reversing but it must be done smoothly and must not feel like floating in the air or a chartered bus. Too brisk and you want to get off after 1 minute and too smooth seems disconnected, not in sync. I think those motors are OK for the task with the VNH5019 to drive them but if not I could use the GW7085 12V which is rated at 6A with torque of 60Kg/Cm and slower which is not bad since I always try to lower the speed by limiting the duty at 75% but then would the VNH5019 still fill the bill or would I have to use the VNH2SP30 with input at 3.25V ? Will the STM32F103RC have high enough output for this ?
I hope I can make a few tests today but I will need a little more time to modify the code in the mc and my Windows app to receive and graph the data.
Yes, it sounds like a current draw issue. I suspect that your power supply can source enough current for your application, but it is difficult to be more certain without knowing the actual stall current of those motors. In the same way, it is hard to tell if the GW7085 would work any better with the VNH5019, since we do not know its stall current, either. (Although, I suspect it might be worse, since its rated current is actually higher, which could indicate a higher stall current.)
Once you determine the stall current of the motor you plan to use, I might have a better idea if the VNH5019 is appropriate.
