blown AX2550

tmachi
Topic Author

15 years 1 month ago #25026122 by tmachi

blown AX2550 was created by tmachi

Could you guild me as to what might cause an AX2550 to be damaged? The documentation suggest that the controller protects itself from over current and over voltage, but reading through the online forum, I see that that’s not true for all cases. I have a very low impedance motor (16mohm) driving a high inertia load.
 
I have damaged two of these controllers so far. I don’t want to try another one until I can get some idea as what is causing them to fail. Detail about the system and failure follows.
 
Appreciate any help you can offer.
 
 
Problem:
Damaged AX2550 (SC) motor controller. 
 
The system has two controllers, only one of the two fails (on the same side). Two separate times. 
 
Not sure when it blows, accel? Decel? Or other?
 
Selftest fails
 
Firmware version: v1.9b 06/01/07
Hardware code: 2-0-1 (4 fet per leg, single ch)
 
7 seg displays a 'J' in one direction and an upside down 'A' in the other.
 
None of the transistors are visually damaged. (Though I can’t see all of them. I have not removed the brd from the heat-sink.)
 
Drain to source is not shorted , nor it’s open
Gate to Source is ~4.4Mohms on some of the transitors and ~500Kohms on some.
 
The motor is not shorted to it’s frame. If it has a winding short, I have no way to tell (since it’s only 16mohms normally).
 
System:
 
Motor controller:
 2 AX2550 (SC) motor controllers, 
 analog mode
 except for analog mode all other configurable parameters are factor default 
 
Battery:
 36 volt. 
 75 AHr
 
Motor: 
 Perm motor GmbH
 PMG 132 – S43
 R = 16mohms
 L = 0.019 mH
 I = 0.025kgm2
 
Two of these motors drive a ~350lb rubber track robot. One motor for each track, one AX2550 (SC) for each motor. Driven to as much as 10 ft/sec, with various (but fairly aggressive) accel and decel rates. 
 

During operation, one controller is ramped to a 50% power command via it’s analog input. The other controller’s analog input is determined by the output of an external PID that closes the loop on that tracks position. The PID controlled track follows the open loop tracks position.

Please Log in or Create an account to join the conversation.

Adamj12b

15 years 1 month ago #25031358 by Adamj12b

Replied by Adamj12b on topic Re:blown AX2550

This thread might be of some help. www.websitetoolbox.com/tool/post/roboteq...9&highlight=perm

Please Log in or Create an account to join the conversation.

cosma

15 years 1 month ago #25054798 by cosma

Replied by cosma on topic Re:blown AX2550

These motors have indeed caused us problems in the past. At the time, they would cause systematic damage during rapid decelleration because of their high inertia: we measured up to 500A of current burned off in the MOSFETs. This was also causing negative voltage spikes that killed the MOSFET drivers.

These two problems have been addressed by limiting the negative Amps with firmware change (starting with version 1.7) and adding protection device that absorbed the negative spike.

These motors are still capable of great stress to the controller, however.

If there is no visible damage, measured short, or smell of burn, there may be hope that damage is limited.

Connect the controller to the PC and check it with Roborun. Verify that it can communicate with the PC. Using the Run tab, verify that internal voltage is 11.8V. Use the sliders to exercise the motor.

Please use our support form at <A href="dev.roboteq.com/dev1/form-support.html" target=_blank>dev.roboteq.com/dev1/form-support.html for personalized assistance.

Please Log in or Create an account to join the conversation.

tmachi
Topic Author

15 years 1 month ago #25062333 by tmachi

Replied by tmachi on topic Re:blown AX2550

Thanks for the reply,
 
I have checked the controller with Roborun. Yes it can comm. The internal voltage is 11.8. It will run in a limited capacity. This was true on the robot as well. It’s as if only one channel is bad, though I can not discern a difference between the two sides with an ohm meter. 
 
There is no burn smell, and no visual damage. I wonder if that tells us that the failure mode is a high (or neg) voltage event instead of a high current (thus thermal) event?
 
When you write of testing with this motor, and of it’s high inertia do you refer to only the motor inertia? The “system” inertia for this robot is many times the motor inertia. 
 
Limited damage would be great, it should be easier to repair. But the most important question is, “What caused it?” 
 
I need to understand if the AX2550 is undersized for my application or if there is some characteristic about this motor that damages the controller.
 
Where can I find a specification for the controller that tells me what currents and power are within it’s limits (for re-generative braking decel cases)? 
 
Do you have any detail on the neg voltage spikes failure mode.
 
Can you guide me to a similar motor that would be more compatible with your controller?
 
Do you have another controller that would be more compatible with this motor? 
Appreciate any help you can offer.

Please Log in or Create an account to join the conversation.

Moderators: tonysantoni

Time to create page: 0.068 seconds

blown AX2550

World Headquarters

European Applications Center

Motor Controllers

Navigation Sensors

Power Management