Ikegai 5t servo alarm on boot-up

[gandp]

I have not run this machine for a couple of months but last night was going to use it and after I homed it and started to move the turret around the Z axis shot back about 2 or 3 inches and not ready lamp went out. I cursor over the alarm lamp and it shows 1 under the servo. Looke at the board and the alarm Z bulb is lit. All parameters are there so what took a dump? I have a spare velocity Z drive so should I swap it out? Every time I shut down and power back up the Z shoots back in positive direction and alarms out. Strange

[Dan Fritz]

Lots of things can cause a servo to run away on a system 5. A bad servo amplifier board is one of them, but it's unlikely that the board went bad just by sitting idle. You also said that you zero-returned the machine, so it sounds like the servo ran OK for a little while, no?

Look inside the CNC cabinet for the main power supply. It will have several cables with white Molex connectors on them, and there will be two cables going to each of the circuit boards. These cables conduct power to all this boards, and the old 5Ms had a problem in this area. Look for "browning" of the white plastic on the Molex connectors. That's a sign that the connector is getting hot. Later 5Ms had gold plated pins on these connectors, and they don't seem to have this problem, but the nickel plated pins would corrode and then heat up, causing a loss of power to some of the boards. A power loss to the lower "B" board makes the pulse coders lose power, causing a servo runaway.

Another problem on the 5Ms comes from some square, black hybrid ICs on the "A" board. They're in the upper-left area of the board, and they have numbers like: A-PC03, or A-PC06. Any one of these hybrid ICs can fail, causing servo troubles. Replacing them is easy, but FINDING one is a real pain. If you can get the CNC to sit stable with the servos on for a little while, try tapping on each one lightly to see if you can get the servo to jump.

If you can't power up without the Z servo running away, try this: Pull the fuses on the back of the Z servo before powering up. If you have a larger motor on Z, there will be 3 cartridge fuses as well as 3 1.3A indicator fuses. The smaller motors just have 3 15A indicator fuses. With the fuses removed, you should be able to power up the CNC and get a "ready". If you then try to Jog the axis, it will alarm out (becuse of no power), but at least you can test the velocity command (VCMD) signal from the "A" Board to the servo. Use a voltmeter and look at the voltage on pin #1 on the velocity board. If it's about 0V, then try to step-jog Z a few thousanths back & forth and see if you see a very small voltage + or - on that Z axis VCMD signal. It should smoothly transition from a small + to a small - voltage as you step jog back and forth past zero. If the VCMD signal is anything other than about zero volts, you have a bad hybrid IC on the "A" Board. If it is about zero, and it behaves correctly with the step-jog test, then you're problem is on the Z servo somewhere.

If it's on the servo, be sure that you have all the voltages on the servo board. You should see check pins to test +5, +15 and -15v. Use check pin CH3 as a ground reference for your voltmeter. The Z servo gets it's power from the power supply on the X servo (through a ribbon cable), so be sure that's plugged in. Also, remove the board and see if any of the pins on the Molex connector has "pushed back" when installing the board. That's a common problem also.

[gandp]

Dan,

Everything was fine with the servo until the second I started using the Z pulse-wheel to touch off my tool. Had not the turret ran away in the positive direction I would off got my hand smashed against the chuck. I looked at the molex connectors and they are not brown from what I can see. I remember the guy I bought it from just replaced them and I don't think I've put 200 hours on this machine in the 3 years I've had it. I'm not by no means sharp with electronics but you have explained pretty good what to check so I'll give this a try. It's just too bad the iron in this machine is immaculate and it's come down to this when I start to think all it's worth is scrap. I've been thinking of retrofitting it with a newer control but I might as well get a different machine with those costs in mind to do so. Anyhow, thanks for taking the time to explain this.

[gandp]

Another problem on the 5Ms comes from some square, black hybrid ICs on the "A" board. They're in the upper-left area of the board, and they have numbers like: A-PC03, or A-PC06. Any one of these hybrid ICs can fail, causing servo troubles. Replacing them is easy, but FINDING one is a real pain. If you can get the CNC to sit stable with the servos on for a little while, try tapping on each one lightly to see if you can get the servo to jump.

If you can't power up without the Z servo running away, try this: Pull the fuses on the back of the Z servo before powering up. If you have a larger motor on Z, there will be 3 cartridge fuses as well as 3 1.3A indicator fuses. The smaller motors just have 3 15A indicator fuses. With the fuses removed, you should be able to power up the CNC and get a "ready". If you then try to Jog the axis, it will alarm out (becuse of no power), but at least you can test the velocity command (VCMD) signal from the "A" Board to the servo. Use a voltmeter and look at the voltage on pin #1 on the velocity board. If it's about 0V, then try to step-jog Z a few thousanths back & forth and see if you see a very small voltage + or - on that Z axis VCMD signal. It should smoothly transition from a small + to a small - voltage as you step jog back and forth past zero. If the VCMD signal is anything other than about zero volts, you have a bad hybrid IC on the "A" Board. If it is about zero, and it behaves correctly with the step-jog test, then you're problem is on the Z servo somewhere.

If it's on the servo, be sure that you have all the voltages on the servo board. You should see check pins to test +5, +15 and -15v. Use check pin CH3 as a ground reference for your voltmeter. The Z servo gets it's power from the power supply on the X servo (through a ribbon cable), so be sure that's plugged in. Also, remove the board and see if any of the pins on the Molex connector has "pushed back" when installing the board. That's a common problem also.

Dan,

I swapped out the board that had these hybrid IC's on and now the control stays on and servos don't run away. I entered all my parameters in and the Z axis moves but I can tell it needs tuning. (It seems a bit faster but not smooth) The X axis now servos out when I rapid and I never had issues with the X axis before. I think I'm at a standstill here now. Maybe a visit from my tech guy now but once he or someone gets this thing up and running I think I better unload it while I can.

[Dan Fritz]

The 5T has a bunch of little plastic jumper bars in the upper-left area of the A board (the one you replaced). Be sure the jumpers are set the same as the old board. These jumpers affect position loop gain and tach feedback for the servos, so they are very important to servo performance.

[tc429]

I dont know if the 5 had auto drift comp, but ive seen zero-analogs where a axis stuck from not being used in a long time, and auto comp tried to null out the servo- till it jumps, then throws a reverse comp in, and every time its reset the comp reverses/gets worse...like param 546 in a zero if i recall.
first time I got stuck with this issue, took me a few hours to figure out what the heck was going on...so now, any time I get a analog jumping at ready, and no apparent drive failures, I always look in the 540 area to make sure no big numbers in the parameters there...

doubt this even applies to a 5, but thought i'd toss it out there as maybe something to watch for.
Tim