isolated Tx - no service neutral

[gar]

140822-0926 EDT

GITRDUN:

Last night I stopped by the shop and looked at our 3-96 VF-3. In the manual all the solenoid controlling items are drawn as a conventional electromagnetic relays. But looking at the IO board it is clear some are ordinary relays and others are solid state switches. I could not read the part numbers.

These relays switch the solenoids from one of the 115 V secondaries on the output of the 3 phase transformer. I believe these are on the output side of the 3 pole 115 V circuit breaker. You did not mention the optical coupler failing. Thus, the circuit where the failure occurs is the SSR in series with the solenoid supplied by the 115 V source. No part of this circuit should be grounded (connected to chassis).

There is no high current path here. The maximum current is defined by the solenoid. This impedance will be low if the solenoid gets stuck or is shorted, then a higher than steady state current will flow. Otherwise not more than V/Zclosed currrent should flow thru the SSR.

If the SSR shorts, then the solenoid is continuously energized. If the SSR fails open, then the solenoid never closes. That the circuit breaker trips implies high current. How is high current generated without other failures?

I also tried a very simple experiment of monitoring the 3 phase 115 V output with a Fluke 27 in min/max mode at power disconnect, i.e., powering down with the off pushbutton. Less than a volt variation and that is well within the short normal line voltage variation. Thus, upon a sample of one I did not see a likely switching transient problem.

There is nothing so far that seems to make sense relative to the cause of your failures.

Do these machines have any external equipment connected to them? Computers, robots, etc?

.

 

[GITRDUN]

Nothing external connected to any of the machines. Im telling you this is some extremely strange stuff, but there is a root cause of these problems.

 

[gar]

140823-0030 EDT

GITRDUN:

There is something you are not tell us so far, and I do not believe you know what that something is.

Some review:

Many different CNC machines exhibt somewhat similar tripped breakers and/or damaged components. These machines consist of HAAS and Okuma. Probably no two machines were manufactured in the same year.

HAAS machines are essentially an ungrounded delta load and required a grounded wye, or a grounded delta. A delta source can be corner grounded, a grounded wild leg type, or grounded in some other way such that line to ground voltages are not greater than in a wild leg type.

All machine frames (chassis) are grounded via a valid EGC.

The computer electronics including the computer, the display logic area,, and RS232 are all referenced to the machine chassis. There have been no failures in this circuitry.

The failures in the HAAS machines are apparently in the IO board area. This circuitry is ungrounded, supplies power to 115 V input power supplies, solenoid valves, lights, servos, and I believe can be considered floating relative to ground. Some failures just trip a breaker related to the 115 V supply, and other failures are related to SSR and circuit trace failures.

HAAS has overvoltage detection just following the main disconnect/breaker. On a 240 V line-to-line machine a peak voltage of around 400 V between legs or legs to EGC will unhold (unlatch) the main contactor (CRM). There are three 1/2 A fuses associated with this circuit. At least one should blow on over voltage. But one alone might not drop CRM. In most cases at least two fuses blow, and that will drop CRM. None of these fuses have blown.

At our shop we have had this over voltage trip from lightning with no failures within the machine.

How can a large voltage and/or associated current get into this circuit failure area?

How accurate are my summary comments?

.

 

[GITRDUN]

I would say thats accurate except we also experience failures on the cuircut boards in the spindle drives on the HAAS's. But by far the majority of the failures occur in the IO/PC boards.
Makes no sense, dont it.