[The Aspirator]

Hey guys, I'm hoping that some of you might have been through some of these newbie problems that I'm having and hopefully can give me some insight. I'm converting my Grizzly G0602 lathe using a stepperworld.com FET 3 board that a friend gave me, and 2 Keling 425oz motors. I've got them wired up as unipolar, although I'm not sure I have nailed down the proper wiring order yet. Here's the spec sheet:
http://www.kelinginc.net/KL23H286-20-08B.pdf
I'm pretty sure I have the two center taps correct, basically connecting A-C- and B-D- and putting those to power. The other 4 are just a guess, but I've tried all combinations with very little change. Right now for the test I'm giving the board +5v and the steppers +5v, though they're rated for 6. Eventually I'll step up to my 24v power supply, but right now I just want to make it work properly.

Anyways, the main problem that I'm having is that the steppers are just extremely jerky and slow, sometimes not responding at all and just jittering along the table. Here is a video of them while using the stepperworld VBStepDirXP.exe program.
[ame="http://www.youtube.com/watch?v=TUpzGW6kp70"]YouTube - CNC testing[/ame]

As you can see the stepper closest to the camera (Z) rotates fairly well, although it pulses a lot, for about 14 seconds until it just stops rotating. The other motor (X) just makes a pulsing noise and vibrates. It should be noted that the two are wired up differently, although wiring them identical seems to give the same result.

And here is with trying to use Mach3Turn to control them with the arrow keys, the settings are posted below. Now the motor that didn't work before (X) sort of rotates, although only in one direction, but the Z motor doesn't:
[ame="http://www.youtube.com/watch?v=gW8OL7ECdt4"]YouTube - CNC testing[/ame]
http://www.cravingboost.com/cars/lathe.43.jpg
http://www.cravingboost.com/cars/lathe.44.jpg

For a computer I've got a dell pc with 650mhz processor and 500mb ram. Got it from a friend, he had just installed a fresh version of XP on it and it didn't have any extra programs at all. I tried all of the hacks and mods in this article to no avail:
http://www.machsupport.com/downloads...timization.txt
Just last night I reformatted the computer, put XP pro SP2 on it and followed those mods to a T, same exact results. One thing that just occured to me that I forgot to try after reformatting it is to use that "specialdriver.bat" file. I tried this before the reformat and it made no change.

DriverTest.exe shows a consistent 25393 +-1, then every 10 seconds or so it shows 25593 for one tick, that's +200.

My question for you fine gentlemen is, is this a computer problem or a stepper driver board problem? If driver, what should I get next? I need 2 axis for cheap, I'm pretty tapped out from getting this far on the lathe. Cheap meaning if I can do it for sub $100 I'll be happy. The easy route would be to get two of those small Geckos, that'll run me $150 shipped. I'm very comfortable with a soldering iron so I'm not at all opposed to soldering my own kit, but I don't have the time or resources to mill/print my own board from scratch right now, so it would be nice to buy a board and a Digikey parts list. I'm interested in the PMinMo projects and I'd be very happy to hear some thoughts on it. Even just showing me a link to some good threads about them or similar technology would be great. I've been searching and cramming for days/weeks now. Since I have the 24v power supply ready to go, it sounds like chopper circuits are totally the way to go, what a cool technology! Makes perfect sense. I was reading something about how the A3977 chips are loud, how does that work? Is it unbearable? The price is right. Ohh and for the PMinMo boards will I have to program my own chips? Because I don't have that technology.

Thanks for all your help and advice! If you need more info I'll be happy to provide. Off to the dentist right now... yay...

John

In other news, my lathe conversion is coming along swimmingly! All the machining is done, just gotta wire it up and make it work properly.
http://www.cravingboost.com/cars/lathe.41.jpg
http://www.cravingboost.com/cars/lathe.42.jpg

[Crevice Reamer]

Hi John. Welcome to the Zone!

WOW! Your conversion looks SUPER! I hope you are going to show all of the details and pics in a post.

Well, for starters, Mach3 likes to have at LEAST 1 GHz. of computer speed.

You have an 8 wire motor there and it can be wired 4 ways. Here is a diagram exclusively for the 425:

http://kelinginc.net/KL23H286-20-08B.pdf

Blu and red/white should be connected together.
Green and Black/white should be connected together


This motor requires 83V and 2A for best performance. Is your driver SET for 2A? What size power supply are you using with it?

CR.

[Crevice Reamer]

John: If you HURRY, you could snag one of the last remaining Overstock Special G540s for just $225. That would GUARANTEE total driver reliability and best performance in the future.

Order Line: 714-832-8874

CR.

[The Aspirator]

Thanks CR!! I've been a lurker for about 6 months, learning as much as I can while preparing for this conversion. It's a huge forum with tons of great info.

As for the computer speed, good to know. I just plugged everything into my main PC, which has a 2.2ghz processor and 1gb ram, should be plenty, right? Same exact results, nothing but jittery motors that try to vibrate off the table.

Blue/redwhite and green/blackwhite are wired to positive 5v, just like the unipolar diagram on that page. I read that the stepperworld controller is only good for unipolar.

How did you come up with the 83v 2a numbers?? And how do I tell if my driver is set for 2a? I've got 2 power supplies that my buddy gave me, both old computer junk. One of them is spitting out +5v at 20a to the motors (5v 2a to the board), this is the one I used in the video. Is it important to have the proper amperage rating in a power supply? The other PS I have is a bigger one that will supply 24.5v at 5.5A. I just plugged this one in for the motors and used 5 ohm 50 watt resistors on each centertap. Same result, but WAY WAY stronger, the vibrations are crazy! I'm really starting to see how more juice makes them wickedly more powerful. If neither of these are adequate or proper, and may be causing my problems, do you have suggestions for an affordable alternative?

I wish I could afford that G540, it's a great great deal and I do want to build an ATC turret sometime soon, but neither are in the budget right now. Any other thoughts as to what the problem might be? I'm all ears!

Thanks!!

John

[Crevice Reamer]

John, I think you should concentrate on the 24V PSU. That is the one that will give the best performance in the long run. The motors can only use 2A in unipolar wiring mode. (its in the spec sheet link I gave you.) Formula for best power is 32 times the square root of the inductance. Your motors have inductance of 6.8 mH, so 32 times 2.6076 is 83V for the best motor voltage. You are running 24, so about 1/4 of potential.

Let's talk about your current limiting resistors. This is what controls the amperage to the motor. There IS no adjustment on this board.

For a 2A unipolar motor, you need a 9 Ohm CL resistor not a 5 Ohm. The resistor CANNOT be a wire wound one. Is it possible that your 5 Ohm resistor is wire wound? Were you using a CL resistor with the 5V PSU?

CR.

[The Aspirator]

Wow I hadn't read that formula! These things would be BRUTAL at 83v, I'll have to keep my eyes peeled for a bigger PSU and a board that can handle it.

I see the problem might be my lack of understanding for how volts/amps/watts all relate to each other. I've read the descriptions but they don't quite sink in.

Now to answer your question, this is where I hang my head in shame and say that these are the resistors I'm using (not this exact one, this is just a random google search, looks the same though):
http://www.microwise.com.au/drz400/resistor.jpg

I assume those are wire wound inside? I went to my local good electronics store knowing that he had some of those gold aluminum bodied ones on the shelf, turned out the were way under rated for my needs. This was all he had and I really wanted to play over the weekend, so I got 'em. And to be even more honest, they only had 10ohm 25w, so I'm using two in parallel for each center tap, giving me roughly 5ohm 50w. 8 of those added up pretty quick, $25 bucks! Now I'm having trouble finding the proper ones on digikey, we are looking for the gold aluminum bodied ones right? The only ones like that I can find are wirewound, like this one, RHRB-50-ND
http://search.digikey.com/scripts/Dk...ame=RHRB-50-ND

Unless I'm looking for the wrong kind?

I only ran these resistors with the 24v PSU, I ran no resistors with the 5v PSU because I figured they're rated for 6v, so this should be fine for testing.... Also that means that I was feeding in 5v 20a with the smaller PSU, and 24v 5.5a minus the 5ohm 50w chunky ceramic resistors.

It just occurred to my possibly why it can't be a wire wound resistor. Maybe because the driver is spitting out pulses so fast that when it goes into the coil of a resistor, the signal gets smoothed out? Or something like that, I'm no techno genius. But eager to learn! What you've said so far has been a huge help!

So if it's the resistors that are the problem, I could order the proper ones from digikey no problem, once I figure out what those are. But I'm also very interested in this chopper stuff. Is it easy to build a chopper board?

[Crevice Reamer]

The reason not to use wire wound resistors is that they add to the inductance of the motor. Your motors are already pretty High inductance and you are undervolted. Adding MORE inductance may require more voltage than you HAVE to overcome it and ultimately may cause the problem you are having.

If you are going to have to spend a lot of money on resistors, PLEASE reconsider getting the sale priced Gecko G540. There aren't many left, and after they are all gone, the price is $74 more. You won't save any money by continuing to throw it down the sewer.

The G540 would run those motors pretty well even on 24 volts--and the resistors that the G540 would need from Digikey would cost 50 cents.

A little wiring with the supplied DB-9 connectors and you would have a fine piece of RELIABLE CNC machinery.

CR.

[The Aspirator]

It's just impossible at the moment, unfortunately. Gotta schedule another dentist appointment for a root canal (fun) and it'll be $600.

Why would the geckos need different resistors than anything else? And what resistors should I be getting?

Ohh and does it matter what amps the power supply is spitting out?

[Steve-Tee]

Howdy John,

Perhaps there is a problem with the power mosfets on the circuit board.
In your videos, it sounds like the motor is running on '3 cylinders' instead of 4.
This will cause them to rotate erratically, and even change direction unexpectedly.

To check the mosfets, disconnect the motor from the outputs, then connect only one of the 4 outputs to the motor - on any motor terminal. Power on, run the driver at about 50 to 100Hz. Grab and twist the shaft to get a feel for what the motor is trying to do.
Move that same motor terminal to the next output. Power on, repeat above process.
If all is well, you should get the exact same results on each of the 4 outputs.
I am guessing that one of the mosfet drivers is not functioning, or is somehow not making a connection to the motor.
This can be done on low voltage, so a specialised PS will not be needed.


Inductance in a wire wound resistor can be seen in a stepper motor setup, but you certainly won't be feeling it's effects at the low speeds you are presently working with.
My motors can run at 50 turns per second, at 10000 Hz, and those WW resistors are just fine .


Here's hoping this helps.


Best wishes,
Steve.


P.S.

I know how you feel about dental appointments. 3 days ago, my dentist reached into my mouth and extracted $500, along with 2 molars.
Ouch!! .

[harryn]

I think you can pull this off with a linistepper driver. They are setup for unipolar designs and pretty cheap kits - 2 for well under $ 100. 12 volts DC will work.

[The Aspirator]

Well I went ahead and ordered a HobbyCNC kit, seems like they've got a level head on their shoulders and produce a good product, plus the price was perfect and they take paypal. $85 shipped for 3 axis, awesome.

Steve, thanks man! That's really helpful. I just tried out your test. I mean if this stepperworld board is still salvageable that's great news, cause technically it's not even mine. And if we can make it work again, he's got a little pcb milling machine that's perfect for it.

You forgot to mention that I need to provide power to the motor, so I gave +5v to both center taps, then grabbed one of the four motor wires and touched it to each board output for all 3 axis's, all 12. Ohh and I turned on the VBStepDirXP.exe program which pulses the motors to help testing, not sure what Hz. The motor is not locked up because it just has power, no ground yet. For the X axis touching the wire to the first mosfet did nothing. The other three fets made the motor audibly pulse, but the shaft didn't move. When I tried to rotate the shaft I could feel it pulsing. I assume this is how it's supposed to feel, and that the one dead FET is causing problems.

Y axis and Z axis, all 8 output pins, only made the motor lock up. It didn't pulse or make any noise other than lockup. I know that the program is pulsing these drivers too, but that's not getting through to to the motor. Lemme swap chips around and see what happens, hold on...

Weird. I just swapped the X and Y chips (black 14pin pictured here) and ran the test again. Now on the X axis, the first FET is still dead, the second one locks the motor up, and the third and fourth make it pulse. Y axis is completely dead, and Z still locks it up on all 4.

Now I swapped the Y and Z chips (this Y was the X in the first test). Y axis is still completely dead, even with the Z chip. On Z axis, the first FET causes it to lockup AND pulse, I can't turn it but I feel it pulsing. 2/3/4 cause it to pulse and I can turn it.

Swap everything back to the way I had it in the first test and....
It's almost the same. X, first one is dead, other 3 pulse. Y locks all 4 up. Z, first fet locks up and pulses, other 3 pulse (first test just locked 'em all up).

A few months ago I was playing with this board on my buddies pcb machine and had similar problems (with tiny steppers). So I talked to the guy at stepperworld and had him ship me some new chips, he mentioned that they are programmed and flashed. Makes me really wish I had en eprom programmer! Other than the chips, the transistors and the FET's, there's nothing that can really break on this thing, so it has to be one of those.

Ohh well, hopefully my hobbycnc board will show up soon. I'll have a blast soldering that together. Yesterday I worked on my wiring harness for the lathe. I found a tach circuit online that uses a QRB1134 optoswitch, so I made that on a breadboard and got it all wired up and working properly. Got a little relay box with 2 relays and a 2 prong household plug to activate my spindle and my coolant pump, that works great too. Can't wait to start machining again! I'm having withdrawls, plus orders are piling up like crazy and I really need my lathe to work again.

John

[budP]

DO NOT hook or unhook your stepper wires with the power on to the motors, you will blow the driver chips up.
budP

[Steve-Tee]

Hi John,

It looks like you have gained some interesting observations from the tests I suggested. At one point, you may have discovered a faulty mosfet, or level translator.

Now that I know what these drivers look like, I can see that they have a 'Wave Drive' mode. This will be really handy in locating the fault.

Could I get you to put 4*12V car lamps in place of the motor?
The 'festoon lamp' that is often used inside the cab of a car, is a good choice for this.
Each lamp would be rated for about 5 to 12 Watts, and would be set up so that one end is connected to a DC supply, anywhere from 5 to 12V will do.
The other end of each lamp is then connected to each of the mosfet outputs, as if you were wiring the motor to them.
Now, once WaveDrive is selected - and the program is pulsing the board very slowly, about 1 pulse per second, observe the lamps.
This will be a lot easier than my original stepper motor suggestion, as you should see the lamps following a simple '1,2,3,4' sequence.
These should take turns at lighting. None should stay on - and none should remain off, as this indicates a problem with either the PIC chip or the circuits that it drives.

If all is well, you should see this '1,2,3,4' sequence occurring without any form of irregularity. If it's working ok, alter the direction control to ensure that it reliably reverses the sequence to a '4,3,2,1' pattern.

This will tell you a lot about the nature of the fault.
Thereafter, a multimeter will let you find exactly what is wrong.
If you have one handy, I can talk you thru the process of fault finding.
Oh, by the way, can you tell us what type of PIC chip is in use?
The 14 pin device part number appears to start with PIC16F.......
This can be handy in identifying the signal path through the driver circuits.
While we are at it, perhaps you can read off the number on those little 'D' shaped transistors next to the mosfets. I suspect that these are NPN types, functioning as level translators.


I hope this will be helpful to you .


Best wishes,
Steve.

[Steve-Tee]

Howdy Budp,

Yes, that is very good advice. The electrical kickback from the motor coils can send the chips to 'silicon heaven', if the connections are interrupted while the power is on.
Always best to be on the cautious side here.

This driver is different, as it has mosfet outputs that handle the motor currents. The sensitive electronics of the Microcontroller are protected by the resistive coupling that exists between it and the level translators.
Accordingly, this renders it largely insensitive to incorrect handling.

I designed mine along similar lines. It can handle huge sparks when I pull the motor connections off while it is running. After a year of such intentional abuse, it still comes back for more .


Best wishes,
Steve.

DO NOT hook or unhook your stepper wires with the power on to the motors, you will blow the driver chips up.
budP

[The Aspirator]

Just got my HobbyCNC package today (5 days after ordering it, to Canada!), so I'll be busy for a while soldering this up, installing it on my lathe, then machining like crazy! So the stepperworld board might take a vacation for a while, but we are interested in getting it fixed, and your help is greatly appreciated.

For part numbers, the 14 pin chips are:
PIC16F630
-I/P (e3)
M(logo) 0813MSA. I have 5 chips here, 2 are that 08 number, 3 are 07131MY.

The transistors are:
2N
3906
-G31

The FET's are:
NL1 03 (ON)
IRF
540

For the lamp test, could I use LED's? Or do they need to have a sufficient current draw like a regular bulb would? Cause I've got tons of LED's and they'd be easy to wire up. But like I said before, it might not happen for a while cause hopefully I'm a few hours away from having a running CNC lathe :banana:.

[The Aspirator]

The HobbyCNC board works GREAT! Took me about 2 hours to solder up and test, it's a really cool little device. The chopper whine is noticeable but not annoying, and I was surprised that the noise came from the motor. From what I had read I assumed it came from the driver circuit, nope.

I'll bolt it all up to the lathe tonight and finish up the wiring.

[Steve-Tee]

Hi John,

Congrats on the new driver boards .
You'll be up and running in no time.

Thank you for the data on the parts on the Stepperworld board. Its design is very much as I had estimated. Very tough and quite versatile.

Yup, you can use leds for the tests on the board.
You'll need to add 2 resistors to each led to protect them from harm, while making them draw higher currents.

Use 470 ohms in series with the led, and 22 ohms in parallel with the led-resistor assembly. The 22 ohm resistors will need to be rated for 10 watts for prolonged testing at 12V.
At 5 volts, the power level will be much lower - - about 1.1 watts, so heating will be much less of an issue.
This will give you a good way of trouble shooting those outputs.

I rather suspect that the old board will take a back seat while you go totally nuts with the new ones.
Hehehehehe. I know how you feel .



Best wishes,
Steve.

[The Aspirator]

I've got a wiring question and I'm searching and searching but I can't find the answer. I feel dumb for asking it, but I just can't figure it out. How do I wire in the e-stop? My grizzly lathe has 120v AC going through the e-stop, it just has two wires. I can't figure out how to use this to estop Mach3. What's stumping me is the fact that it's AC, and the driver board and computer is DC, so it's not like I can just slap an input wire on the switch and call it good, right? That just wouldn't work. I don't wanna mess with and re-wire the e-stop switch cause right now it stops spindle motion perfectly. I just want to make it also send a stop signal to Mach3. Ideas?

All the schematics I'm finding just show the e-stop circuit the same as a limit switch, ground and input. Thanks for any help, I'll keep searching too.

John

*edit*
Or do I just re-wire the e-stop button? I'm using a relay to control the spindle, the relay outputs AC power to an electrical plug where I plug the lathe into, so if I just wired the e-stop as DC ground/input then Mach3 will shut off my relays and kill the lathe. I guess that'd work, just wondering if it's kosher. Or are there big capacitors in the lathe that might keep it running for a few seconds if I just "unplugged it" (which is basically what the relay is doing).

*double edit*
I think I figured it out. I found a diagram that shows that some e-stop switches have 4 terminals, this would make life easy cause you could use two for AC and two for DC, but mine only has 2. But then doing more digging online and on my machine I found the huge 8 terminal contactor (?) in the back. It's like 8 individual relays all in one, that all get triggered at the same time. So after lots of checking with the multimeter for ohms and volts, I figured out that each pair of contacts is definitely separate from the rest, and there are a few that aren't being used. So I can easily wire a DC ground to one side, and an board input wire to the other side, sweet!

Funny how I can figure it out within an hour of posting my problem...