Tormach Driver Failure

RJSakowski

H-M Supporter - Gold Member
H-M Supporter Gold Member
Joined
Feb 1, 2015
Messages
9,516
A couple of weeks ago, I was running a part on my PCNC770 when there was a strange loud noise during a z axis move. The noise only lasted a second and the program continued. There was a second strange noise and the head dropped about 1/4" further than programmed. A 1/4" carbide end mill plunged about .080" into my 5C collet chuck and cut a 1/2" long slot in the nose of my collet chuck, a 7/8" collet, and my workpiece before the E-stop killed the program.

I reset the controller and using the jog shuttle, I confirmed that the x and y axes were still live. When I opened the control panel, the red warning LED on the z axis driver was lit. When I pulled the motor connector there was a very obvious burn on pin 5. Leadshine Driver .JPG
I switched the motor and control connectors to the y axis controller and verified that the z axis stepper and controller signal were good.

I then pulled the driver and removed the cover. A resistance check of the output transistors showed that Q14, an IRF540 MOSFET transistor, was shorted between the source and drain leads. I removed the transistor and confirmed the it was shorted.

I placed and order with DigiKey for five replacement IRF540's and two pairs of the Phoenix connectors. Two days later, the parts arrived. I installed the new transistor and the new connectors. I installed the repaired board and powered up with the green LED on the z axis driver now lit. I reset the controller and homed the machine with the z axis now behaving normally

The next day, I talked with customer service at Tormach. They have been aware of a problem with the connectors on the Leadshine controllers for some time and made a production change in 2014. Part of the change was using a dielectric grease on the motor connections to prevent oxidation of the contacts.

A postmortem of the Leadshine 3ND883 suggested the following scenario. The bifurcated phosphor bronze contact and brass pin are tin plated. The contacts are rated for 15 amps. The pin has a square cross section so the entire current through the connector passes through two opposing lines across the pin. The integrity of the connection is also highly dependent upon the spring force of the contact. When oxidation of the tin plate occurs, the resistance of the connection increases which gives rise to heating. The increase in temperature of the contact will eventually burn the surrounding insulation. Raising the temperature of the contact can also destroy the spring temper of the phosphor bronze, weakening the contact pressure and further increasing resistance. Eventually, the connection becomes intermittent which has the effect of adding extra steps. In addition, Leadshine states that making or disconnecting the connection under power can create transients which can destroy the driver. Failure was most likely due to a chain of events which began with oxidation of the tin plate and ending with destroying the output transistor, shutting down the driver.

A few suggestions about a repair and replace.

  1. Both halves of the connector should be replaced if burning is evident. While the board connector may not be damaged, it most likely will have oxidation and surface pitting from the subsequent arcing. There is possibility that the product from the burning of the housing have coated the surface of the contact pin resulting in a thin insulating layer.
  2. There are static sensitive devices on the board so it is prudent to take proper antistatic precautions.
  3. The connector is soldered in plated through holes and it can be very difficult to remove without special equipment. If the special equipment isn't available, it would be a good strategy to cut the connector apart so each pin can be removed separately.
  4. Heat sinking of the output transistor depends upon positive contact with the thermal pad. Insert the transistor on the board and bend the leads slightly to hold it. Then mount the board to the heat sink and solder the leads and trim them. This prevents the new part from biasing the mount and possibly decreasing the thermal conductivity of the other transistors.
  5. There is a temperature sensor which is thermally connected to the heat sink with heat sink compound. Make sure that there is sufficient heat sink compound on the surface where it contacts the heat sink.
  6. Use of a dielectric grease or antioxidant gel on the connector contacts is highly recommended. Legrand sells a copper to copper antioxidant , OR-AJCC, in 1/2 oz. tubes. Harger HCAJC1/2 is another brand. Tormach recommends a straight silicone dielectric grease.
  7. The OEM connectors were made by Degson, P/N 2EDGK-5.0-02P-14-00AH and the mating connector is 2EDGRC-5.0-02P-14-00AH. The Phoenix numbers are 1754504 and 1754494 for the plug and socket.
As a preventive measure, it would be wise to pull the motor connections on a regular basis to check for any burning. If burning is evident, the best practice would be to replace the connector. Removing and reinserting the connector several times will also serve to wipe any oxidation from the contacts.

On a related note, early Tormach machines had a setting of on,on,on for the 1,2, and 3 positions of the Leadshine driver dip switch. This corresponds to a drive current of 8.3 amps. The latest 770 manual shows on,off,on for the switch setting, corresponding to a drive current of 6.6 amps. Tormach verified today that 6.6 amps is the correct drive current for all three axes on the 770.

20171026_132559.jpg
 
Great, thorough write-up Bob.

The periodic contact wipe is a good idea to extend the life. However, the contact normal force will also decrease with age....add in the heating due to the current flow and contact life is limited.

This connector system is two parts;
1) the left part solders onto the PCB and has male 90 degree pins
2) the female contact and screw terminal plug into part #1 and have the screw terminals for the wires

Is there an option to get rid of one contact interface by using a solder-on screw terminal connection?

-brino
 
Brino, I just saw your comment. I have already identified a substitute direct wire connector. I haven't ordered them yet but will on my next order to DigiKey . Phoenix P/N 1935349; DigiKey no. 277-1595-ND, or Mouser no. 651-1935349. It will require removing the individual wires for servicing but that isn't a big issue as there are only five and the wires are terminated with ferrules.
 
Back
Top