Howdy.

I have an older Delta VFD (VFD075A23H) that just started giving me issues. It's connected to the spindle on a CNC router. For the entire time I've had it, it's been happily working, driven from a PC running Mach3, connected via a PMDX-107 speed control board connected to a PMDX BOB. I fiddled with some settings this morning because I realized that I could reduce the minimum spindle speed by futzing with the control frequency. After this, the VFD stopped turning the spindle off as it's supposed to.

As part of my debugging effort I've now removed the router control wiring and am controlling it just with the VFD's onboard panel. Even like this it won't shut off.

I can start the spindle and increase the frequency just fine.
I hit the stop button and the Run light starts flashing and the Stop light illuminates.
But the spindle keeps spinning.

I've tried resetting the spindle to factory parameters which I think worked to reset the values - it's a bit hard to tell - but it didn't change the result. Still won't turn off.

I'm starting to think that something got fried inside the VFD when I changed the operating frequency, but that just seems odd.

Anyone have suggestions for how to get this thing working again?

-Ben

 

Check keypad switch continuity (with power off) to rule out a failed STOP switch on the panel.
Power off, open the panel, measure continuity of the STOP button while pressing it. If the STOP switch contact is open or not changing, replace/fix the panel switch membrane.

 

Yeah, the button clearly works because when pressed the STOP LED comes on and the RUN LED starts blinking. It just doesn't change the behavior of the spindle. Argh.

 

So.
What settings did you fiddle with?
Were they in the VFD or some other component?
I don't think setting to default values did you any favours.

 

Fixed! Turns out the default value for Over-voltage Stall Prevention was "on" and that was causing the thing to just keep running for some reason. Disabling that allows it to ramp to a stop. Yah!

 

Well done.

 

Fixed! Turns out the default value for Over-voltage Stall Prevention was "on" and that was causing the thing to just keep running for some reason. Disabling that allows it to ramp to a stop. Yah!

Sweet. Nice drive.

 

Heh. Would be a lot nicer drive if it would actually ALERT you to the overvoltage trigger, rather than just silently not stopping

 

overvoltage sensing sounds like a rather vital feature. what makes you think that turning this off is 'fixing' the problem and not merely masking it?

btw. most drives i have worked with have built in braking resistor of low power (only some have none), and option to add external braking resistor of high power. braking resistors are used to absorb excess energy from spinning motor while decelerating. failure to do that is risking blowing up the IGBTs. that is why warning is there - to make you aware of the issue so you can do a proper fix instead of continued blows to the IGBTs. you need to choose reasonable decel time or use external braking resistor of high power (and lower internal resistance).

 

overvoltage sensing sounds like a rather vital feature. what makes you think that turning this off is 'fixing' the problem and not merely masking it?

btw. most drives i have worked with have built in braking resistor of low power (only some have none), and option to add external braking resistor of high power. braking resistors are used to absorb excess energy from spinning motor while decelerating. failure to do that is risking blowing up the IGBTs. that is why warning is there - to make you aware of the issue so you can do a proper fix instead of continued blows to the IGBTs. you need to choose reasonable decel time or use external braking resistor of high power (and lower internal resistance).

From what I've read since I started chasing this, this internal overvolt protection parameter is not reactive, it's proactive. It looks at the deceleration time you set, and then ASSUMES that you're connected to a high-inertia load like a fan or conveyor or something, and if the deceleration time is too short for that use case it trips the protection. But I'm not driving a high inertia load. I'm driving a CNC spindle. It has very low inertia. So none of the internal assumptions about the load are correct, meaning it's being over protective.

I take this with a grain of salt since my understanding is informed in part by ChatGPT's often-flawed analysis. But in this case it's making sense.

There are other, reactive protections that are still in place to protect the drive.

 

It's reactive to the link voltage (the software normally sets the trip limit), it doesn't really care about the amount of stored energy, only the effects of what you have on the link voltage. If you want to stop quickly, the stored energy in the motor has to go somewhere, braking resistors are what we used to have very quick stop times.

https://www.galco.com/products/motors-drives-accessories/braking-resistors.html

 

also look at the linked video about 4:26. you will see staircase graph, going down. the steeper slope is what you desire but during this 250ms or so there is no correction - this is running purely on time.

in 250ms voltage can shoot WAY above limits of the drive. imagine wave traveling and then slamming at the wall. you will see upon impact that water is shooting upwards (just like voltage). the more energy the wave has, the higher the water reach.

then think about failure modes... material destruction due to high voltage is extremely fast. even if it device still appears to function, it IS scared... think Lichtenberg figure. then thing what happens over time. how many such events will take till total destruction?

i work with robots so normal D bus is 600 - 650VDC, IGBTs tend to handle 750-800V, though some may be 1200V models. without braking resistors, they do get destroyed. not every time but plenty of the time.

in comparison to overvoltage damage (nano to micro seconds), overcurrent damage is much slower... several ms to several seconds (or slower), because temperature is a much slower process than pressure (or voltage). so 250ms or so shown in video is an eternity... may as well be 2 hours.

 

I was going to suggest investigating the brake torque setting, which on some drives is linked to the bus voltage over voltage setting. I am not sure exactly why, but the seem to be related.