Vector drive slow startup

Our minimill is suffering from bus undervoltage errors unless it has been on power for a long time, was an hour but its getting worse, yesterday it took 6 hours.

Its a 2007 machine with a 20HP drive.

I have had the drive in bits and found some thermal damage to the driver board on the area that the 4 wire wound 20k resistors for the LED are located, it has been hot enough to discolour the board and solder around them.

I found in another post a helpfull circuit diagram for another similar hass vfd with very usefull info (hass closed loop vector drive .pdf) , and am begining to suspect that the isolation amplifier (HCPL7800A) located besides the droppers has been cooked by the dropper resistors over heating, this device has according to the HP data sheet a max operating temperature of 80c.
The iso amp is an opto coupled device and as such is temperature limited by the leds which if heated excessively (like about 90c) will permanently loose emmision and the gain/linearity of the device will be severely compromised. I think this is a likely candidate for the fault.

The machine has been left on overnight and is up and running so it has been temporarily put back into service.
I shall not be touching it untill the current work tasks are finished.

It is my intention to have a look at reverse engineering the high voltage detection circuitry to get a better grip of that circuit segments operation.
My plan is to then replace any components in that part that may have been cooked beyond sensible limits, and to shift the dropper resistors off the PCB to hopefully prevent a re-occurence of the failure.

Does any body have any sugestions or comments about this?

regards
malcolm

You should connect a meter to the DC bus to see that it is indeed at the correct voltage. Then you can look at the output from that opto amp. If that is the cause, the output of the amp should slowly increase until the fault clears. I think what you have described makes sense.

Good luck

Thanks Keebler,

The bus sits at around 330v, it rises to this within the time it takes to get from the front of the machine to the control box round the back.
I have also tried (as suggested in other posts) connecting a 3.6v lithium cell to J17 of the mocon board and the display indicates 370v so this (in my simple way of thinking) puts the problem in the VFD, a dvm on the plug from the vfd shows readings which concur with the front panel display.

I need to do a little more tracing of the driver boards circuit to be more confident in the first guess iso amp as being the real culprit, I think there is another amplifier between the iso amp and the remaining circuitry, there may also be a small electrolytic used in a filter in this segment, awful practice, but then so is putting big fat watty and hot resistors next to measuring circuits, buck your PCB layout ideas up Mr Hass. I also note that there are several surface mount caps which may be tantalums, a type of capacitor which I have past experience of failing catastrophically as a result of age and elevated temperature.

This seems to be a common problem and at somewhere around £3000+ to get Hass in for the repair is in my opinion a shade expensive, the HCPL7800A costs around £10.

When this problem first manifested itself a year or two ago I replaced all four big elctrolytics to little or no effect and we have lived with the 1/2 hour it took to wake up, it is now staying on 24/7 since it took 6 hours to wake on Monday.

Any more suggestions or circuit details will be most appreciated.

Hi Fungus,
I'm dealing with the same problems on three Haas SL-10 Lathes, did you get anywhere with your minimill??

Just replaced the HCPL7800A and there it is fixed all problems gone. Now to relocate the 20 risitors away from the HCPL7800A and maybe a cooling fan.
It even fixed the spindle fault when the first time i start the spindle. Thank you Malcolm

This is a SL 20 2000 with a 2007 smart drive.

We have been suffering with this problem on our 2007 VF4 for a couple of years. It had started as a 5 minute delay to start up the machine because of the present alarms 160-low voltage, 648-Dc buss short and 649-Dc buss under voltage. Over time the start up delay increased to 10, 20 and up to 45 minutes. The machine was out of warranty and it didn't interfere too much with normal business operations so I just let it be. This last fall when the weather cooled the startup delay shot up to hours to almost a whole day. It was time to seriously deal with this problem.

I started to search the forums and was surprised at how many people had this problem. I used the Haas Technical Documentation - Vector Drive Troubleshooting along with other forum member suggestions to take voltage measurements of the 20hp vector drive. I was getting the correct voltage at the 2-3 terminals (320 volts) but not at the P17 connector. Also on the diagnostics screen, the DC Buss voltage would start out at 160 volts and slowly climb to 230 volts. Just past 230 volts is when the alarms would clear and normal start up could continue.

From reading other posts, it looks like most people corrected this by replacing the vector drive, bought a refurbished board or sent out one for repair. Before I go down one of those paths I figured I would try the suggestions found on this post. Thanks to 'Fungus' and 'engine tech art' , I have successfully fixed my start up delay problems.

As 'Fungus' explained the iso amp component HCPL 7800A had been compromised by the severe heat generated by the (4) 5watt 20k resistors that are next to it. I purchased it for $11 and had a electronic repair tech solder it on for $35. The machine starts up instantly and the DC Buss voltage on the diagnostic screen now reads proper at 320 volts.

Here is a picture of the 20 hp vector drive and the board that the component is on.

https://www.cnczone.com/forums/attac...d=455378&stc=1https://www.cnczone.com/forums/attac...d=455380&stc=1

In the center of the picture is the HCPL 7800A component. The (4) 20k resistors are the tall ones to the left. This area was so hot it had melted/degraded the nylon standoff that is located in the center of the group.

https://www.cnczone.com/forums/attac...d=455402&stc=1

On the back of the board, the tech had added some lead solder to the 20k resistors to possibly aid as a heat sink. He was surprised they had not fallen out.

https://www.cnczone.com/forums/attac...d=455404&stc=1

This last picture shows the old discolored HCPL-7800A next to a new one.

In the future, I will probably install a fan to directly address the heat in this area. Only time will tell if this fix holds.

Thank you to everyone who posted useful information about this issue!


https://www.cnczone.com/forums/attac...d=455406&stc=1

Thanks for the post, We have one at work with the same issue and now it's purring again. Haas wanted over $7K for a $20 part.