Need Help!- 100% jog fine, but goto stalls?

[Wolfesmetalfab]

MOTHERB... look, it runs in Jog mode. I'm running at 30ipm and it seems to jog fine. Jog is set at 100% and reports as being sped up to 745mm/min (30ipm). BUT, if I move X,Y,Z far from the origin and start some code, OR use the "Goto Z" to return to origin, I get a stall all the time. This is true even if only the Z axis is the only one that needs to move and also if it's on the downstroke. On the Z axis, 5 sec of runtime and it will stall. Now let's stop and consider this. I've seen it jog. The controller, mill, and motors are doing the same damn thing at the same speed either way. I think we can rule out ALL of that for good. There's a software/computer thing here. There HAS to be. It's like the parallel port might have a buffer that fills up IF the machine's "working" on something then it's filling up and pulse timing gets out of whack when it comes out. There's just no other answer.

I have a similar problem as this guy, although he never really got an answer. Im running a two atx psus at 24v for approx a 12a supply. I have motor tuning at 150ipm accel at 10 and jog tab at 100%, using only x and y and it jogs and runs parts fine, but if i click goto zero, or input g28, it stalls my g203vs and i have to reset the drives. What gives, its the same speed, so why not when jogging or running programs?

[H.O]

Hi,
Can you jog both X & Y at the same time? Can you do long G0-moves with both axes moving?

One possibillity is that the current capacity of the suply is not enough to accelerate and run both motors at full speed at the same time. You say the powersupply is rated 12A which sounds plenty but you don't say what the current rating of the motors are. Try putting a volt-meter on the supply and see if the voltage drops when the motors accelerate.

/Henrik.

[TOTALLYRC]

Originally Posted by Wolfesmetalfab I have a similar problem as this guy, although he never really got an answer. Im running a two atx psus at 24v for approx a 12a supply. I have motor tuning at 150ipm accel at 10 and jog tab at 100%, using only x and y and it jogs and runs parts fine, but if i click goto zero, or input g28, it stalls my g203vs and i have to reset the drives. What gives, its the same speed, so why not when jogging or running programs?

Correct me if I am wrong, but it doesn't stall your 203v's, but causes them to fault out. Motors can stall, but the drive can't.

When you jog, are you getting a display of 150ipm? when you do a g00, are you getting a display of 150ipm? The g28 or goto home are doing basically the same thing as jog, but require a little more of the computers interaction.
Try a g00 move of 2 axis at the same timeand see if the drives fault out. 24v is near the bottom of the allowable input for a g203v(18V) and you may just be dropping below this for a short period of time. If it faults out, change your accel to 5 and try again.

As the other guy suggested, use a volt meter to see if there is a voltage drop. Although if it is quick enough, a digital meter may not catch it.

Mike

[Wolfesmetalfab]

Correct me if I am wrong, but it doesn't stall your 203v's, but causes them to fault out.

Correct, that is what I meant.

I do get a display of 150ipm, unless more than one axis is moving, then it is above, because of the angular movement of the 2 axis making it a faster angle travel.

Are you thinking it could be a computer problem, like what?

I will do some more playing, and see what I find.

Edit: Ok, did some more testing with Voltmeter hooked up to DC out.

X velocity = 275 accel=18
y velocity = 235 accel=18

Volts at Idle = 22.4
Volts when x & y both jogging full speed = 23.1
Volts when y full speed = 22.6
volts when x full speed = 22.9
volts for 1 sec after stop jogging = 21.9, then back to 22.4 idle volts

When click goto zero, meter shows 23.1 and then drives fault and voltage goes to and stays at 11.1 volts.

Then I unplug supply and replug it in and were back to 22.4 idle volts.

I am curious what is happening? One of the atx supplies is maybe shutting down for over current or something and causing the drives to fault? I wouldn't think it would be over current though, because the drives are limited to 6 amp each, 12 amp total, but more than likely I shouldn't be using more than 8 amps max (12 * .67), right?

When I chain 2 12v 6amp supplies together, am I getting 24v 12amp supply, or is this were I goofed, and it is a 24v 6 amp supply, and I am overcurrenting the supply when I run close to my 8 amp calculated max usage?

I am using the daisychained atx supply (2 200watt 6a 12v rail each). Could this be the problem? Also I just want to use this to test what ideal supply voltage is. This being said, would it be an accurate test to use this supply running programs and seeing what verify in mach brings up and when I find an acceptable speed that I am not loosing steps (also how do you determine between machine accuracy and lost steps when verifying), and then calculate the supply needed for wanted speed. Say 150ipm is what is good with this supply, and 300ipm is desired, so a 48volt supply is what I want. Motors are good to 56v, and drives to 80.

What do you think?

[TOTALLYRC]

I would say that the lag in the volt meter is such that the voltage has already dropped below the acceptable value and that one of you pwer supplies is droppingout and staying out.

I am not thinking it is a computer problem, just trying to get a feel for what is going on.

When you use 2 devices in series you will double the voltage, but not double the amps. In parallel you would double the amps, but not the voltage.
So you have a 24v 6 amp power supply, and you must be spiking itno an area where you shut down 1 of the supplies.

Your acceleration of 18 is very aggressive, you could try dropping it to 10 and see if the problem stops or is at least less of a problem.

As you must know by now, a real power supply will make a world of difference.

I am not sure if the speed/voltage relationship is that linear, but I am sure somebody else has the answer.

The easist way to see if you are lossing steps is to run some code, starting from a known position such as 0,0 and then see if you can go back to 0,0 and get the same reading on an indicator.
There is a better description somewhere on the zone. If it doesn't make sense to you, I will give a better descritpion of how to do it.

Mike

[Wolfesmetalfab]

Yeah, the psu is pissing me off. i was hoping to be able to use it to find appropriate finally power supply needed so i dont have to spend $50 on a small supply to tune and then when I know what I need, spend $120 on it.

I knew, but overlooked, forgot or whatever about the series/parallel volts/amps doubling, I hate it when i have a dumba$$ attack. LOL

Thanks for pointing it out.

I may do a little playing within the limits of the ps (maybe even 1 axis at a time running a part) and see if i can get an idea what I will need for psu. I am thinking it a 48v 12a supply will be ideal and let me do the speeds i want easily, and I could probably avoid some headache and just buy one, and then see what happens for running and if it needs to be more, or it is way more than what I need and motor temp is a problem, then sell it and buy the appriate supply after testing with a good 48v supply.

I did run some amp tests too with x and y, this is what I found.

Overall amps pulled from dc supply main wiring. figures did vary slightly of course.

x veloc = 275 accel = 18
y veloc = 235 accel = 18

amps @ idle = 1.8
Y running = 1.97 or about 2 amps
y decellerating = 2.95a
x running = 2.3a
x decellerating = 2.8a
Both running by jog = 2.3a
Both decellerating by jog = 4a

then when hitting goto zero and both running
it is pulling 2.8 a for about 1-2 secs and trips the drive error and motors stop

It doesn't seem like a prob, or make alot of sense, but like you said, maybe the digital cant read a spike.

Edit: i am playing with just the y (gantry). I disabled x and z in ports and pins. I tried 235 vel and 18 accel run a part, verify result was .oo44 off.
To get less error, I tried slowing down accel to 15, started to run again, got about 5-10 sec into program and drives triped?

Dont make sense to me? Shouldn't be a power shortage with just 1 motor running, should it? Why did it trip trying to accel slower, but didn't at faster?

Edit again: More playing, I can run the y at 457ipm vel (max for mach at 45khz) and accel at 30 and jog back and forth with a foot a handful of times achieving top speed and accel and decelling, and verify, and i get .00000 error? Yes the drive will trip, whenever it feels like it. I am about ready to try even faster accel values. If this is the case, maybe a true good 24v supply with 12-15 amps would be fine instead of a 48v?

[Wolfesmetalfab]

After you read my previous post, if I was to just buy the PSU that I think i will use permenantly from antekinc.com. which do you think i would be happiest with for speeds and not heating the motors more than needed?

desired rapids are about 300ipm, accel values, the higher the better, I dont know what a good accel is for plasma though, tough to find info on what anybody is using.

The supplys they have that will work with my 6.1amp motors limited on g203vs at 6amp x 3 motors = 18amp * .67 = 12.06 amps

28v 14.28amp 400watt $94
30v 13.3amp 400watt $94
35v 14.285amp 500watt $107
45v 13.3amp 600watt $120
50v 12amp 600watt $120

What do you think? I am trying to consider the crappy results from the daisychained atx and the speeds accel that I can get, but its just kind of hard to come to a solid conclusion. I do think in the back of my head, that whichever option i pick, will work great! And right now, I cant even feel any heat on the geckos, and the motors are just barely warm, so I dont think heat should be an issue.

Would the 35v be a good choice (were only talking $13 more than 30v) and it wouldn't generate too much more heat? Or does the 45v look more apealling?

[TOTALLYRC]

The motors can be allowed to get quite hot and still be OK.

If you know the mh rating of the motors, the max voltage can be calculated by the formula of sqrt of the mh times 32.
Since you are driving a plasma table, 400ipm on real thin stuff could possibly be needed. So I would consider going up to the max that the motors can handle.
If you lived close by, I could let you borrow one of my spare 50v power supplies.

Mike

[Wolfesmetalfab]

The inductance is 3.3, supply max 58v. I dont think i need quite that, at first, I was thinking maybe 48v. 50 or 56 would be okay also ( I dont like running at max whatever it may be, I guess its just my logic of it is likely to cause problems. I live near York, Nebraska 68467, google map it to see were its at. What type of supplies do you have? If your not close, would it be something that you wouldn't mind shipping, if i paid for s/h, and let me play with it for say like a week, and if It worked well, you would want to sell it? If not, I could ship it back ( i also understand if you are uncomfortable with shipping it in general)

Just a thought, I really appreciate your thoughtfullness!

Brian =)

Edit: By the way, I really like your signature, I just noticed it! It explains me very well!!! ROFL

[H.O]

Another thing you could try is to fit a fairly large capacitor on the output of your supply, this will help filter out fast spikes and surges so the overcurrent circuitry in power supplies doesn't trip. A 40-50V (or higher) capacitor and a few 1000's uF's would be fine.

Now, depending on the desing of you power supplies this MIGHT trip one or both supplies when you power it up as they need to provide quite a large current initially to charge the capacitor but it might be worth a try. There are ways around that too but then we're getting more complicated than it needs to be a "proper" powersupply is the correct approach anyway.

Another thing to concidder here is that AT/ATX-powersupplies usually have their output GND (black wires) connected to the case. This means that when you connect two supplies in series the case of the second supply is at +12V and if they happen to contact each other you get an instant short. If both supplies are connected to a grounded outlet I'm surprised it works at all....

/Henrik.

[Wolfesmetalfab]

I may try the capacitor if i dont just buy a power supply right away.

The ground is only connected to one of the cases, I read on the net the proper way to daisychain them, so we should be good there.

There are ways around that too but then we're getting more complicated than it needs to be a "proper" powersupply is the correct approach anyway.

LOL...you must be reading my mind! Thats why I want to just buy a supply if I can figure what I should get. 48v, 50v more? less?

[TOTALLYRC]

Originally Posted by Wolfesmetalfab I may try the capacitor if i dont just buy a power supply right away. The ground is only connected to one of the cases, I read on the net the proper way to daisychain them, so we should be good there. LOL...you must be reading my mind! Thats why I want to just buy a supply if I can figure what I should get. 48v, 50v more? less?

Since the g203v have idle current reduction, running it at max is not going to be too big of a problem as the motors won't cook when just sitting there. The difference between 48 and 50v is small enough to not matter.

I will look around today and see exactly which power supplies I have and what there exact power outputs are. I think I have 2 in the 50v volt range but don't remember the amps. Even a 50v 6 amp power suppply should give you a great idea on what to purchase.

Mike