Burned power part of servo KRS2552 HV
Burned power part of servo KRS2552 HV
by EngineerD
Thu Feb 23, 2012 1:57 pm
Hi, guys.
During adjusting of stand up movement of KONDO, one of the leg servos became hot, and some smell appear.
So, when I disassemble the servo, i saw that one of chips was burned. The marking on this chip is "3004 8G5 63Y8". I tried to find what is this through alldatasheet.com but had no success.
Does anybody know what is this chip, and what I could use for change it?
P.S. 1- place of disassembled burned chip
2 - same chip on board.
During adjusting of stand up movement of KONDO, one of the leg servos became hot, and some smell appear.
So, when I disassemble the servo, i saw that one of chips was burned. The marking on this chip is "3004 8G5 63Y8". I tried to find what is this through alldatasheet.com but had no success.
Does anybody know what is this chip, and what I could use for change it?
P.S. 1- place of disassembled burned chip
2 - same chip on board.
by i-Bot
Thu Feb 23, 2012 3:36 pm
I would guess it might be a HAT3004R MOSFET pair from Renesas.
A post here suggested an alternative which is more available, but you should check spec. http://robosavvy.com/forum/viewtopic.ph ... &view=next
Check the motor is OK before you spendtime finding and replacing IC. Often the MOSFET failure is secondary to motor shorting due to overheat.
Good luck.
A post here suggested an alternative which is more available, but you should check spec. http://robosavvy.com/forum/viewtopic.ph ... &view=next
Check the motor is OK before you spendtime finding and replacing IC. Often the MOSFET failure is secondary to motor shorting due to overheat.
Good luck.
by PedroR
Thu Feb 23, 2012 7:26 pm
Hi EngineerD
We're also doing a project with 2 KHR 3HVs and managed to kill a couple of servos ourselves.
For the reference of others, this is our experience with both using (and killing them) and also fixes:
- The Knee servos are very prone to being damaged if the wrong screw is used on the back horn. This happened to a customer of ours and also to us.
One of the screw sizes included is a couple of milimeters too long and although it seems to fit the horn those extra milimetres will puncture the PCB.
(exactly where the processor sits on the other side of the board; it is essentially pushed from the board).
- The other known issue is servo overheating: if you leave the robot in a static position for too long or stress the servos by pushing them to extreme positions the excess current will heat the servos and burn the MOSFETs and/or the motor.
What iBot mentioned (burning the motor) has happened to us as well. Although we were able to salvage a few servos, others had the motor burned and were beyond repair.
I'm not sure how we tested the motors here at the office but if I recall on some of the servos we actually removed the motor to see if we could move it freely with our hand (on a burnt servo the motor feels heavy when you try to turn it but this test needs t be done after removing any burnt MOSFETs as far as I know)
- Finally and most importantly VOLTAGE: because we needed to operate the Robot for a long time and the KHR does not support this scenario (it only runs on battery) we improvised a Power supply using a 12V rail from an old computer PSU.
The KHR comes with a 10.8V although the specs for the 2552 servos say it can handle up to 12V.
However we don't recommend going up to 12V. An easy way to drop the voltage is to put a Diode (that will handle high Amps) between the +12V and the Robot.
This should drop the voltage by about 1V.
That was the solution we used although I don't have the exact specs/model of the diode we used.
Needless to say 3 cell LiPo upgrades are to avoid at ALL cost. We've tried them with KHRs on a previous project and after completing the project we had burned 4 servos.
A final note to damages caused by burned MOSFETs and Motors: this type of damage is usually not covered by Warranty. Most servo manufacturers consider them to be caused by misuse or user negligence although repair options are usually available at a fee.
As for our project, after taking all these precautions (especially adding the diode) the Robots are now operating safely and happily with no other incidents to report (and let's hope it stays this way!).
We're using them on a 23 servo configuration (adding 2 DoF in the arms and 1 Dof to rotate the torso; we left the Hip Pivot out) and it looks very, very nice.
It's a mechanical design that is very unique, especially on the arms where we get one more DoF than usual.
Regards
Pedro.
We're also doing a project with 2 KHR 3HVs and managed to kill a couple of servos ourselves.
For the reference of others, this is our experience with both using (and killing them) and also fixes:
- The Knee servos are very prone to being damaged if the wrong screw is used on the back horn. This happened to a customer of ours and also to us.
One of the screw sizes included is a couple of milimeters too long and although it seems to fit the horn those extra milimetres will puncture the PCB.
(exactly where the processor sits on the other side of the board; it is essentially pushed from the board).
- The other known issue is servo overheating: if you leave the robot in a static position for too long or stress the servos by pushing them to extreme positions the excess current will heat the servos and burn the MOSFETs and/or the motor.
What iBot mentioned (burning the motor) has happened to us as well. Although we were able to salvage a few servos, others had the motor burned and were beyond repair.
I'm not sure how we tested the motors here at the office but if I recall on some of the servos we actually removed the motor to see if we could move it freely with our hand (on a burnt servo the motor feels heavy when you try to turn it but this test needs t be done after removing any burnt MOSFETs as far as I know)
- Finally and most importantly VOLTAGE: because we needed to operate the Robot for a long time and the KHR does not support this scenario (it only runs on battery) we improvised a Power supply using a 12V rail from an old computer PSU.
The KHR comes with a 10.8V although the specs for the 2552 servos say it can handle up to 12V.
However we don't recommend going up to 12V. An easy way to drop the voltage is to put a Diode (that will handle high Amps) between the +12V and the Robot.
This should drop the voltage by about 1V.
That was the solution we used although I don't have the exact specs/model of the diode we used.
Needless to say 3 cell LiPo upgrades are to avoid at ALL cost. We've tried them with KHRs on a previous project and after completing the project we had burned 4 servos.
A final note to damages caused by burned MOSFETs and Motors: this type of damage is usually not covered by Warranty. Most servo manufacturers consider them to be caused by misuse or user negligence although repair options are usually available at a fee.
As for our project, after taking all these precautions (especially adding the diode) the Robots are now operating safely and happily with no other incidents to report (and let's hope it stays this way!).
We're using them on a 23 servo configuration (adding 2 DoF in the arms and 1 Dof to rotate the torso; we left the Hip Pivot out) and it looks very, very nice.
It's a mechanical design that is very unique, especially on the arms where we get one more DoF than usual.
Regards
Pedro.
by EngineerD
Fri Feb 24, 2012 10:55 am
Hi to all!
Thanks for help to everybody.
It's really HAT3004R and we found an inexpensive replacement of the one, named FDS8958. The IFR7389 isn't a full analogue of HAT3004, so we decided do not use it.
Do anybody know, what kind of motor is used in KRS2552HV. I mean - is it brush-less motor or ordinary DC motor. The marking on the motor is "1620FE-062-11.0 2039".
I've tried to rotate motor by DC power supply, but no success, the current goes up to the current limit in 2A.
Do anybody know, how can we make sure that the motor is alive?
Thanks for help to everybody.
It's really HAT3004R and we found an inexpensive replacement of the one, named FDS8958. The IFR7389 isn't a full analogue of HAT3004, so we decided do not use it.
Do anybody know, what kind of motor is used in KRS2552HV. I mean - is it brush-less motor or ordinary DC motor. The marking on the motor is "1620FE-062-11.0 2039".
I've tried to rotate motor by DC power supply, but no success, the current goes up to the current limit in 2A.
Do anybody know, how can we make sure that the motor is alive?
by PedroR
Fri Feb 24, 2012 12:59 pm
I have re checked with Marco what happened to us:
- The Motors are normal brushed DC motors. They have 3 pins but the third pin is just for Grounding the exterior case so it doesn't matter for testing purposes.
To check the "health" of the motor, you need to unsolder it and apply power to the terminals (something in the range of 8V would be enough).
- Another point that I wanted to clarify is that when the H bridge burns and is in a "short circuit" state it is essentially behaving like a generator (a dynamo).
In these circumstances, trying to turn the motor by hand (while the Servo is still completely assembled) will feel heavy.
This doesn't necessarily mean the motor is burnt; it's just the H bridge that's behaving as a dynamo and thus makes it difficult to turn the motor.
With regards to replacement motors I don't believe we know the exact model though.
Our set of "damaged" motors was a mix of burnt motors and burnt H bridges. We diagnosed all of them and transplanted the healthy parts of each of them to rebuild as many servos as possible.
As a final note, we now have equipment in our Lab to replace/repair burnt H bridges on Kondo servos.
Therefore if you (or anyone) is interested in having their servos repaired (replace burnt MOSFETs), just email us on support@robosavvy.com
Regards
Pedro.
- The Motors are normal brushed DC motors. They have 3 pins but the third pin is just for Grounding the exterior case so it doesn't matter for testing purposes.
To check the "health" of the motor, you need to unsolder it and apply power to the terminals (something in the range of 8V would be enough).
- Another point that I wanted to clarify is that when the H bridge burns and is in a "short circuit" state it is essentially behaving like a generator (a dynamo).
In these circumstances, trying to turn the motor by hand (while the Servo is still completely assembled) will feel heavy.
This doesn't necessarily mean the motor is burnt; it's just the H bridge that's behaving as a dynamo and thus makes it difficult to turn the motor.
With regards to replacement motors I don't believe we know the exact model though.
Our set of "damaged" motors was a mix of burnt motors and burnt H bridges. We diagnosed all of them and transplanted the healthy parts of each of them to rebuild as many servos as possible.
As a final note, we now have equipment in our Lab to replace/repair burnt H bridges on Kondo servos.
Therefore if you (or anyone) is interested in having their servos repaired (replace burnt MOSFETs), just email us on support@robosavvy.com
Regards
Pedro.