[alexdobert]

Hello,

I apologize if this was addressed in another post, but I am having some issues testing my G540 and motors. After eliminating a bunch of things (correct parallel cable, EPP mode and working from a desktop), I can't seem to get any response from my motors when I jog them in Mach 3, with the gecko.xml profile loaded.

I have the charge pump enabled and once started up, the controller goes to ready, which leads me to believe that it is getting a signal from the 16 pin. Mach 3 also resets if I disconnect the parallel cable or unplug, so there is definitely some communication between 'puter and controller. I am pretty sure that I soldered the wires properly (although not very pretty). Cold it be that? A Mach 3 setting? Something I may have missed?

I am at a loss and would appreciate any help.

Thanks in advance.

Alex

:drowning:

[Mariss Freimanis]

Do you have a green LED or a red LED lit?

1) Disconnect all your motors, the parallel port cable and set the Charge Pump switch to "off" if you have a red LED. Make sure you have a jumper wire from E-Stop to power supply GND (terminal 10 to terminal 12). Fire up the G540; you should get a green LED. If not, the G540 has "issues" and must come back for repairs.

2) Connect the parallel port cable and turn the Charge Pump switch "on". You should still have a green LED. If not, you have a Mach3 configuration problem.

3) Plug in only one motor axis. You should have a green LED. If not, your motor is miss-wired.

4) Add new motor axis one at a time. You should get a green LED. If you get a red LED, the last axis you added is miss-wired.

Mariss

[alexdobert]

Thank you for the response. I went through step by step and everything seems to be fine - outside of the motors actually turning. The controller faults without the parallel cable plugged, but once plugged in, goes to green. I try to jog the motors, and nothing. I can't imagine there being an issue with all three motors. Are there any other ways to isolate the possible problem? Thanks again!

[darlly01]

Does anyone can help about Geckodrive G540 and controller MN400 connections. I did everything and my motor do not run. MN400 are communicating with my pc but not with Geckodrive G540. I tested Geckodrive with parallel port and it is fine.
Any help will be great.
thanks

[Mariss Freimanis]

darlly01,

You know what an "MN400" controller is but I don't. Can you provide a url for this controller so I would know what it is and could then try to provide an answer for you?

Mariss

[Daley]

My apologies for reviving an old thread, but I seem to be having the same issue.

I've had my G540 working great for just under a year, and decided to add a slave to Y. I soldered everything up and powered it on. Now, I have the same symptoms described in the first post - power up the G540, it goes from fault to ready like normal. I hit the reset button in Mach, and everything *appears* okay (except that I'm missing the typical whine from the steppers). I try to jog, and while the DRO moves I get no action from the steppers.

In troubleshooting, I unplugged everything except Z (which hasn't been touched), and I get the same responses: Mach's DRO will move when jogging Z, but no motor action (and again, no whine from the Z stepper).

Suggestions/advice would be awesome!

[wendtmk]

Well, since everything "worked" until you soldered up the slave to the Y, I would begin troubleshooting the connections you soldered.

Are you trying to run two steppers off one drive? If so, that's not really a good idea. Each motor should have its own drive, and you should do the slaving within the control software.

Mark

[Daley]

Hi Mark,

Thanks for the reply. As you stated, driving two motors off a single drive is a bad idea, so that's why I slaved Y to A.

Also, your advice about retracing my steps back to a known working configuration is obviously spot-on, so that's why I performed the steps above to eliminate potential points of failure.

To clarify, I once had a working X, Y and Z. I decided to add a slave axis to Y, so I wired up an identical stepper as the existing Y and slaved A to Y. Without having A and Y physically connected to the moving gantry, I powered up the G540 and quickly noticed that I didn't have the familiar whine in the steppers (in fact, nothing - silence). I cycled the E-stop a couple times, hitting the reset button in Mach between each attempt. Nothing I did made the steppers power up.

Suspecting a soldering mishap, I powered down the G540 and removed all axes except Z - thinking that I should at least be able to get something from an axis that I hadn't messed with. Nothing.

Scratching my head, I began to dig a little deeper. I probed the power pins at the G540, and confirmed that I'm getting adequate power (+36V).

From there, I made sure that the G540 initially faults at power-up as it should, then goes "green" within a second or two. All good. So just for grits and shins, I tried to jog Z (as it's the only axis plugged in right now). The DRO moves, but no stepper movement.

Having a relatively high level of comfort around a circuit board, I cracked the case on the G540. I didn't see anything obvious (no blown components, no burnt traces, etc.) but probed around a bit to make sure I'm getting power. Without having the schematic, I probed what I thought should be receiving power - all seems good.

At this point, I put the case back together and posted here hoping someone could chime in.

I intentionally took the next couple weeks off work so I could play in the man-cave, only to come to the realization that I've somehow screwed myself into spending less time in the garage, and more time shopping for new CNC toys. (nuts)

[wendtmk]

When you were probing the pins on the G540, were you getting any voltages at all on the output pins?

Mark

[Daley]

Going on Mark's advice after some probing I found that the fuse had blown so I replaced it with a 7A fast-acting fuse. Powered up, all is good. Connected Z (untouched so far) and got movement, so I pressed on. Disconnected Z, and for grins I powered up with nothing connected but the terminal strip (still untouched) and the parallel cable. I heard a distinct "pop" and saw a small flash come from under the A drive.

At this point I think I'd like Gecko to take a look at the G540. I'm not sure how I'd hear/see what I did with nothing connected except what had been working (which remains untouched at this point), but to avoid further damage I'm throwing in the proverbial towel.

Geckodrives appears to be on vacation for the holidays, so I'll just have to wait patiently until they're available.

Man this sucks. I appreciate Mark chiming in and offering his advice. I'll update this thread with any updates as I receive them.

Merry Christmas all!

[wendtmk]

Ah that's a bummer to hear that. I'd double-check all your soldering and wiring too, since that may have caused the initial problems with the new setup. I've got a niggling suspicion you may have a short somewhere in the wiring that caused the initial problem. May be a solder whisker, or a wire touching somewhere it shouldn't. Might have to look under magnification. I do that on circuit board traces after I do component repairs. Sometimes a small sliver of solder goes where it shouldn't and causes all sorts of problems, or maybe a cold solder joint that doesn't have a good mechanical as well as an electrical connection. Ohm out your wires and your connections too, as well as seeing if any wires short to any other wires.

Mark

[Daley]

Again, good advice Mark. I've now triple-checked, and while there was a connection at one of the D-Subs that I didn't like, it electrically checked out (I re-did it anyway). I get 2.8 ohms across the windings on all motors, with no shorts and no opens.

My days of soldering under a microscope have long since passed, but I've still got a pretty good set of eyes and a halfway-decent magnifying glass.

I went over everything several times and haven't seen anything obvious, which is why I'm looking to have a (more experienced) second set of eyes take a peek.

I wouldn't have suspected anything on the G540 to cause this initially, as nothing had changed there; however, as you said, my changes could certainly have prompted a chain of events leading up to the current situation.

Side-note: speaking of solder whiskers, I recently read an interesting article on tin whiskers - I'd never heard of them before, but I'm glad I have now!

[Daley]

Just emailed support with the details outlined here. I know they're probably busy as heck today, so I'll stand by until they have time.

I'll keep the thread updated with any useful information.

[Daley]

Waited a couple days and called direct. Was immediately transferred to "support" (by a real human), where my call was taken directly by Marcus who (even though swamped by the post-holiday rush) took the time to speak with me for a few minutes and gave me a no-nonsense offer to "just send it in" to have GeckoDrive look at it.

Shortly thereafter, I received an email from Marcus offering the same - send it in. Additionally, he said that all repairs are offered free of charge (INCLUDING RETURN SHIPPING).

Man, if this doesn't prompt someone to choose GeckoDrives, I don't know what would - A+++ customer service!!!

Marcus/Gecko, y'all have earned a customer (and a supporter) for life!

[Daley]

Got a call from Gecko the other day, they said that I'd blown the power supply section of the motherboard along with two of the four drives. They further explained that this can be caused by one of two things: reversing the power supply wires or an overvoltage condition.

Knowing that I hadn't touched the wiring with regard to the power supply connection to the G540, I can only be left to believe that the PSU itself went haywire and blew up the G540.

They did upgrade the motherboard and drives to the latest version, for a 50% discount totalling ~$125 shipped. Still cheaper than a brand new replacement, so I'm okay with that.

I also went ahead and ordered a PSU from Ahren at CNCRP (although the order is still pending).

So of course with no chips flying, this gives me plenty of time to shop for other stuff - nicer enclosure, panel-meters, etc.

Is anyone employing any sort of over-voltage protection? After this, I'm a bit gun-shy.

[Mariss Freimanis]

This subject is driving us nuts. It seems there is a whole herd of cheap Chinese 48V switching power supplies out there that fail open-loop or don't close the loop on power-up.

What does this mean? It means your mild-mannered 48VDC power supply either fails entirely or it puts out 80VDC for a few tenths of a second on power-up respectively. In other words it becomes a G540 killer.

The G540 is rated at 50V, it's tested at 60VDC and it comes apart at 68 to 70VDC depending on the MOSFET lot batch we get from Fairchild Semiconductor. Either way, 80VDC will kill every G540 we build.

The problem has become serious enough that we bought 5 switching power supplies from 5 different vendors that appear in various ads from vendors here on here on CNCzone. We tested them rigorously to the limits of their specifications.

What we found was the cheapest one failed right out of the box. It went open-loop to 72.5VDC on power-up. Another one failed at 68.2VDC after a week at full power. The remaining 3 passed test after 10 days and remained at the rated 48VDC. They were the 3 more expensive ones.

All G540s contain a tell-tale circuit that pops if the G540 has seen more than 68VDC applied to it. It's a 68V zener diode in series with a 33 Ohm resistor. The resistor burns open if more than 70V is applied to the G540 for more than 1/10th of a second. It is a diagnostic we use to figure out what went wrong.

Two out of five is a bad record. We can't tell you who the bad guys were for obvious reasons but let's just say if you pay less than $0.33 per Watt (less than $125 for a 48V, 8A power supply), you may have some issues with power supply quality control and design integrity.

Mariss

[Daley]

Mariss, that's extremely good information, and is much appreciated. After being helped by Marcus and Lani, I immediately bought a "better" power supply to avoid just what Mariss describes above.

If I can suggest something here, I think Mariss' excellent explanation should be made available to anyone purchasing a G540. Initially, I bought my PSU, G540 and motors as a kit with the thought that "a PSU is a PSU is a PSU"... That ended up being somewhere in the neighborhood of a $200 mistake (not counting the time the machine was down as a result of this). If I'd have known this from the beginning, I'd have gone to Ahren to begin with.

Thanks to Gecko for a great drive, thanks to Gecko for great customer service and support, and thanks to Mariss for the explanation - I hope it prevents future customers from making the same mistakes!

[fretsman]

Mariss, will you then be recommending a power supply brand or company to go with on your website rather than us just guessing which one to buy?

Thanks,
Dave