What's new
What's new

Haas HL-1 Brushless Servo Amplifier Repair

Rocan

Plastic
Joined
Jan 8, 2014
Location
United States
Long time lurker, first time poster. I had been using reddit for some help on this issue but figured this would be a better bet.

This machine is a new- to me Haas HL-1 from 1997. I just powered it up for the first time using a 40hp, balanced, homemade Rotary phase converter and was immediately presented with fault 102 (servos off) and a fault for the Z axis drive. The drive showed an error, so I removed it to find these blown capacitors (shown after cleaning).

I know for a fact that this machine was working before being taken offline. I did smell an acrid smell on first start up, that I imagine was this capacitor. The machine has sat idle for over a year waiting for me to get power, so I'm wondering if it was just old/degraded caps finally letting go? I made sure to change the voltage taps to match that of my RPC output before the first start. The other drives show no faults, but I can't clear the Fault 102 or jog any axis, even after enabling the "jog axis without homing" option. Additionally the hydraulic pump doesn't appear to be running, but I don't know if that starts after homing the axis. So I can't clamp the chuck and test the spindle motor.

I did check the cable and motor for any shorts and everything looked fine there. no shorts to ground, and 0.9ohm between the motor 3 phases, measured at both the motor terminals and the cable when connected. I want to confirm this is just a spurious failed board and not some time of supply issue before I power up my far more expensive 2014 Haas VF-2SS.

Additionally, any and all help is appreciated. I am inclined to replace the caps myself, if the parts are available. I was able to track down some information on the caps and believe i have the correct ones from digikey (links below). However, I'm having a hard time determining the correct zener diode to use. It is marked Z2 and I've got the following information thanks to some reddit user, quoted below.

"
106 on the capacitors = 10μF

20K and 10K mean 20v and 10v ratings with 10% tolerance

the 643 and 609 should just be date codes.

the zener should be a 5.1v regulator diode but I do not know what power rating you'll need"



I found the following diodes that seem to fit the bill on digikey, but no clue how to select a power rating. One other user suggested reverse engineering the IC-

You could try looking at the datasheets for the HCPL-7800A and HCPL-3120 and see if you can reverse engineer the circuit a bit. I think the 3120 is more relevant, it is an optocoupler. It appears like the ground trace and everything on it connecting either pin 5 of the HP 3120 (ground) or pin 4 of the HP 7800A (also ground) burnt up. Guessing one of the caps shorted and the entire ground trace got super hot. If you can clean some more of the soot off the traces it might be easier to tell.

ISO5 should be pretty similar. Judging by ISO5 it looks like the Z2 SOT-23 is a zener diode, as pin 2 is not connected to anything as far as I can tell, and diodes typically only use pins 1 & 3 in a 3-pin SOT-23. You can verify that with a multimeter though. I couldn't tell you what specific zener diode but hopefully it's one of the ones on this list.
 

Attachments

  • signal-2024-04-01-192234_002.jpeg
    signal-2024-04-01-192234_002.jpeg
    506.6 KB · Views: 5
Last edited:
Nothing wrong with trying to repair your amp but, blown axis amplifiers are a thing with Haas. If it's a brushless-era amp, I'm 99% sure you could call Haas and order a brand new, current generation, "smart amp" and be back up and running. Or eBay for a used one.

You can also pull an amp out of your VF-2 and use it for testing. Does it have a 4th axis amp you're not currently using? That's almost everyone's spare tire in their shop.

No, you won't be able to do anything else if there's a drive fault. The hydraulics are also tied to the error circuitry. You're basically in an emergency stop state and can't clear it until that amp is fixed.

You should be checking your incoming power voltages and then setting the transformer taps at the bottom of the cabinet to match. The first time you switch on the power, I'm pretty sure the incoming power board on even a 1997 will have phase-detect lights. Those tell you if the phasing is correct for things like the hydraulic system.
 
Nothing wrong with trying to repair your amp but, blown axis amplifiers are a thing with Haas. If it's a brushless-era amp, I'm 99% sure you could call Haas and order a brand new, current generation, "smart amp" and be back up and running. Or eBay for a used one.

You can also pull an amp out of your VF-2 and use it for testing. Does it have a 4th axis amp you're not currently using? That's almost everyone's spare tire in their shop.

No, you won't be able to do anything else if there's a drive fault. The hydraulics are also tied to the error circuitry. You're basically in an emergency stop state and can't clear it until that amp is fixed.

You should be checking your incoming power voltages and then setting the transformer taps at the bottom of the cabinet to match. The first time you switch on the power, I'm pretty sure the incoming power board on even a 1997 will have phase-detect lights. Those tell you if the phasing is correct for things like the hydraulic system.
Thanks Donkey Hotey. It is brushless and I do have a spare 4th axis drive on the VF2ss- I also understand I can swap it over to test- I'm just fearful of blowing the new one. Thanks for the clarification on the hydraulic / emergency state, I figured that to be the case.

I did adjust the voltage taps based on my incoming voltage (242/260/260, as per fitch's guide), and did so prior to first startup. No phase detect lights that I could find - the manual even states "phasing does not matter", which obviously is not really the case, given the hydraulic and coolant pump need the correct phasing. I made sure to put the wild leg on B as well.

Mainly at this point I'm hoping to find the correct zener diode to attempt a repair. I may try my spare drive, assuming others have experienced drive failure without any other components being at fault.
 
Last edited:
You sound like you're electronics hip so here's some reassurance. The amps have 320V DC fed into them. That's the wires going from amp to amp on the screw terminals. The Vector Drive supplies those rails. Make sure the vector drive LED is out before messing with those wires or screws. Or check the rails with a multimeter. Caps in the Vector Drive store voltages for 10-15 minutes. The other three wires on the screw terminals go to each motor.

One of the plugs is the logic cable going to the Mocon board. I'm pretty sure that's nothing but RS422 (balanced RS232) logic connection.

The remaining harness goes back to the DC power supply. It also supplies everything else in the machine with power. If there were a problem there, you'd have other problems (or it would have blown everything else up). Check that with a multimeter if you care but, you'll find familiar 12V and 5V rails.

Not a whole lot to 'blow up' if you pull a drive out of the VF-2 and want to get the HL running.

Drive failures without other faults? All the time. For no reason. Pisses me off. Given the age, you might have a good chance at the caps or other nearby components having been the cause. I've not attempted to repair one but, my understanding is the IGBTs can fail and due to how they go, it zaps upstream control circuits that were never intended to see 320 volts. So that kind of failure is often not repairable. But caps or other failures may have blessed you with a soft, repairable problem.
 
Donkey, greatly appreciate the reassurance. I may get impatient and risk the 4th axis drive in the HL-1 to get things moving- good to know that spurious drive failures are common.

My wife reminded me that my father in law has spent his entire career designing chips and PCBs- I sent him what info I had and he said he will do some digging today and see if he can't identify the correct components. As best as I can tell, I found the correct zener diode, but I'm definitely at the limits of my EE knowledge, so I'm not 100% on the voltage or power requirements.

 
@Rocan, any photos attached to your post aren't showing up for me, so it's difficult to judge the component selection. Tantalum capacitors like the ones in your Digi-Key links are sensitive to inrush current, and tend to fail at power on. They are typically used for power supply filtering. If they fail open the circuit may continue to operate, sometimes manifesting more subtle error conditions. If they fail short, which often happens, this can prevent the power supply from starting up.

The EEVBlog forum is a good general resource for component-level diagnosis and repair.

 
@Rocan, any photos attached to your post aren't showing up for me, so it's difficult to judge the component selection. Tantalum capacitors like the ones in your Digi-Key links are sensitive to inrush current, and tend to fail at power on. They are typically used for power supply filtering. If they fail open the circuit may continue to operate, sometimes manifesting more subtle error conditions. If they fail short, which often happens, this can prevent the power supply from starting up.

The EEVBlog forum is a good general resource for component-level diagnosis and repair.

Thanks for the heads up- I uploaded a photo of the blown cap to my OP again.

I probed my X axis board and found the voltage at both sides of the diode to be -2.5V to ground. I think that means the 5.1v zener diode is fine, but I'm waiting on confirmation from my father in law. Interestingly, the X axis board has been repaired before- I see a few new tanatalum caps and some of the HP IC's have been replaced with Motorola IC's.

The fact that they failed at power on, and based on feedback from others, I'm guessing it was just time and that there are no other faults to worry about. I'll report back when I have replaced components and note if I have any success. I'll record the repair, hopefully it can save someone else some money as well, assuming I'm successful.
 
If you can, check the working voltage of the tantalum caps. There was a time when the capacitor manufacturers told engineers that they did not need to have a large safety margin on working voltages. It was not uncommon to see caps rated 16V on a 15V rail. These initially worked fine but tend to fail over time with a catastrophic short. If they don't explode they at least pull the voltage way down. Nowadays we would put a 25V or higher rated cap on a 15V rail. Tantalums are highly reliable when operated within the current guidelines.
 
Amaranth- Thanks for that tip. In hindsight I should have replaced the 10v cap with the 25v caps.

Some great news! I was able to repair the board with the components above. I accidentally ordered the wrong size for the 10v tanatalum caps, but was able to make it work anyway, given that one trace was competely missing. 10 mins of soldering (if that), and after reinstallation the board seems to be working perfectly. No faults, and the machine is back to operational condition! Thanks to all who helped.

$1.00 in parts sure as hell beats a $500 repair or $1500 replacement!
 








 
Back
Top