I could have gone too the Xylotex boards but figured most guys here are running or have a Xylotex board on a Taig or similar machine. I have the 3-axis system with a 24 volt power supply, C11 BreakBoard board and running three Keling KL23H276-30-8B motors.

So i'm trying too set-up everything and get it tweaked in EMC.

Pretty much, my goal is too obtain 30 IPM. So in EMC I set the velocity too .5 which would be 30 IPM.

I have the potentiometers on the Xylotex board too around 2.80 volts, which means that it would put out around 2.1 amps which is the amp rating for my steppers.

I also have a stock setting on all axis of 1/8th microstepping, using a 1:1 ratio pulley set-up; 20 TPI thread on all axis.


When I jog in EMC, the view shows it going the speed, but I know for a fact that it's not true and up too speed. I know i'm not hitting 5 or 7 or 30 IPM. I'm going very slow and can't figure out why.


My power supply is on full too and everything is good. Can anyone help me please?


-Jason

So I can't figure out how too get the speed too be true. Is it something with my microstepping? Should it be more then 1/8th? I'm not sure whats going on.


-Jason

Not an EMC-guy myself , but what about your drive screws? Pitch (Threads per inch), number of starts, diameter? All "force" the ipm and resolution of your system. Are they set right -- and can they give you 30 ipm? Fewer steps should be faster -- at the penalty of resolution.
Sorry - can't shed more info for you.
:cheers: Jim

I have 20 TPI @ .5 in diameter I think....


I don't know if this is normal or not, but when my boards are at 1/8th micro stepping. But in EMC when I put .125 for the microstepping, the system is so so so so slow. When I put a higher number like 5.0 I start getting much faster!

So yea, I don't know if I'm on the right track. But I'm pretty convinced I can reach 30 IPM.


-Jason

Perhaps you need to put 8 in for 1/8 step, as in 8 microsteps per step. Sorry, not an EMC user.

I have the same setup as you just no c11 board. As ger said you need to put in 8 for the micro steps. Also you should know that if your acceleration doesnt accelerate fast enough the machine might not reach top speed before it reaches the end of the desired direction. Hope it helps you out.

Ok! So i'm on the right track. I put in 8 today earlier and was getting much higher speeds. I want my speed too be around 30 IPM. I will need too mess with the acceleration because I'm losing steps it seems. I don't think I should be losing steps though, I mean 30 IPM is not that far from being possible.

Maybe just some more tweaking. Do you agree Regnar?


-Jason

Ok so, 8 microstepping is working.

But I'm losing steps even at 19 IPM. I really thought I could be going higher, shouldn't I be? The only thing I can think of is, power supply even though at 24 volt / 6 amps it's not giving enough.

But will going too 30 help that much more?


I just feel like I should be able too obtain more then 20 without skipping. :(


-Jason

Hi Jason,

Let's see you have 8 * 200 * 20 = 32000 step signals per inch. So (32000steps/in x 19ipm) / 60sec/min = 10,133 steps per second at 19 ipm or (32000steps/in x 30ipm)/60sec/min = 16000 steps per second at 30 ipm. I don't know what the maximum steps per second that you pc can generate.

Then lets look at rpm. 19 ipm x 20tpi = 380rpm and 30ipm x 20tpi = 600rpm. So if you don't have resonance dampers on you stepper motors it is possible that you have hit a resonance problem.

The other possibility is a power supply limitation. That may just be the limit of what you can get with 24volts. I am running a 32volt supply with about 15 amps and different drivers (not necessarily better). On my lathe I have pac sci P22 steppers and on the z axis I can easily hit 75ipm with a 5 tpi ballscrew but on the x axis with a 20 tpi screw and a 4:1 reduction belt drive I get only 8ipm. With direct connection that should give 32ipm at the same motor rpm. (I don't have resonance dampers so my stepper/drive combination seems to max out about 600 rpm).

I hope this helps.

Alan

Acon- Yea it makes sense. I still feel the computer is capable. It's an older HP but everything on the comp is strictly for EMC. Nothing else is on it, no music or anything. It's possible though.


Here is what I noticed. On this new system, (New board I mean) I'm using the 24 volt @ 6 amp PS. Ok, so as of now....It seems like I can only get maybe 15 IPM (Thats a rough estimate) Because at 19 IPM I start skipping, so lets say 15 IPM.

My old system (From ebay) with a output of "ONLY" 9 VDC @ 17.0 Amps gave me a jogging speed of 12 inches a min with no Skips. I would start skipping in programs running higher then 7 on rapids.


But even if I only obtain 3 -8 more IPM almost doesn't seem worth it.


So there is something tricky going on. One PS with 9 VDC @ 17 amps jogging at 12 IPm and 24 volt @ 6 amps only getting around consistent 15 IPm.



Sound weird???


-Jason

After reading where your at Jason, I think we need to remove the microstepping. At the very leat move it to either 1/4 or 1/2 microsteps. When I had problems with the Z axis missing steps, I moved it to 1/4 and that solved the problem before the counter wieght.

That was the same setup you have +1 axis on a SX3 with a 100lb head. And I could make it move up at 15ipm.


The only difference between what I have and you are the motors and the leadscrews and EMC2. Mine were .200 pitch (1/5") I was getting 30 very easily. I was using Mach3 and the 425oz motors.

I have heard that the Xylotex boards did not compensate for resonance damping. Maybe building a damper would help you out for the axis. Are your steppers dual shaft?

Yes the motors are dual shaft. So that requires the actual board settings too be changed. I couldn't find any information on the microstepping in the manual. There is a PDF but it leads too nothing with no explanation.

I assume it's the jumpers behind each heatsink. It has too be. Hopefully this will increase speed.




-Jason

Something else I noticed. I also noticed with my set-up now...Once everything is powered up. The motors seem too buzz louder. They don't move when everything is idle. But not like before, they seem too buzz a little louder now. Don't know if this means anything.


-Jason

Here is the PDF for the Micro stepping.

http://www.xylotex.com/StepSetting.pdf

Also, in the Xylotex manual, there is a "Microstep Truth Chart" at the bottom left of the first page.

Also here is the FAQ page about the driver.

http://www.xylotex.com/FAQ.htm

I have already created a g-code file for a dampener, I have been meaning to make some just to play around. I could get a couple made and send them to you if we can't get this solved.

Oh duh! I looked at that and thought they were links. I never thought they were charts for the freaking jumpers lol!


-Jason

Something else I noticed. I also noticed with my set-up now...Once everything is powered up. The motors seem too buzz louder. They don't move when everything is idle. But not like before, they seem too buzz a little louder now. Don't know if this means anything.


-Jason


That should be because the higher voltage. Mine do that on the Gecko's as well.

Here is a tidbit from the Faq. The 269 oz motors he has are close to what you have in spec.

Here is what it says
"
17) What RPM can I get with your motors
There is no real RPM rating for the motors. Unloaded they will spin very fast. All stepper motors loose torque as they go faster, and will have more torque at lower speeds. Your actual traverse rate in IPM (inches per minute) will depend on the mass and friction of the machine. Cutting rate will also depend on cutter sharpness, width and depth of cut, spindle speed, hardness of material, etc. When using 10 TPI screws with the 269's you can get between 50 and 75 IPM. When using the 425's you will get around a max. of 50 IPM, but with more lower speed torque. If you have 5 tpi screws, then you are better off with the 425's. With the 425's and 5 tpi screws on a smooth running machine you can get over 100 IPM. Your actual rapid speed will depend on the screws being used and the motor selected (as well as the dynamic of the actual machine).
"

Yea but with 20 TPI, I should still be able too hit between 25-45 if for 10 TPI it's 50-75. I wonder if you can change the screws for the taig, but then again. More work then I think is needed.


-Jason

Exactly, that was my point. There is something else that is the problem here. Have you looked at the microstepping settings yet?

Ok. I didn't have much time too mess with it. I'm not sure by the diagrams where the jumper needs too be placed. And for instance. Since I was using 1/8 microstepping in EMC and was using the number "8", Does that means for say the 1/2 microstepping: I use 16...Make sense?


-Jason

It should be the opposite. If you use half micro stepping, it should be 2. You are just dividing the single step pulse by the number of micro steps in that case. So it would be 0 for Full, 2 for 1/2 step, 4 for quarter, and then 8 for 8.

Looking at the heatsinks, you will have 1 set of jumper blocks underneath each one. Just to the right of the blue pot.

You will see 3 rows of 2 pins. The first 2 rows (below the heatsinks) will not have any jumpers on them. That means that you will be in 8 microstep mode. That is the weakest, but most minute amount of travel the motors can have.

The rows should be labeled

MS0 on top
MS1 in the middle
EN on the bottom. (The EN jumpers will disable that particular axis if it is removed)

For 1/8 step. There should be no jumpers on either MS0 or MS1
For 1/4 Step. There should be a jumper on MS0 but NOT MS1
For 1/2 Step. There should be a jumper on MS1 but NOT MS0
For Full Step. There should be jumpers on Both MS1 and MS0

If you go to full step, you would put 0 in he EMC2 steps and then i believe your steps would be 4000 with the 20tpi screw.


I would start there and then work my way back. Your accuracy should still be good with the 20tpi screw.




Hopefully that makes sense. :) Xylotex didn't send any extra jumpers with them, so I hope you have some around there. I am in the PC biz so I had a few hundred from back in the day. lol...

Ok now that makes sense. So i'll want too leave the stock jumper where it is and then just install the other jumpers where they need too be? Sounds good. I have lots of computer parts, hopefully there will be some on some motherboards!

-Jason

With the full step, I can't put in 0. It won't allow me. But when I put 1 in (The bottom showed 4000 step rate). I did not mess too much on this setting.

I found 1/2 and 1/4 too be pretty good. But didn't have enough time too mess. With the 1/2 I was able too obtain like 25 IPM consistently.


Hopefully I can keep going higher. We'll see.


-Jason

Jason, when you're "testing" the speed, are you actually cutting something, or just air?

By the way, how come you pulled your auction for the cnc stuff on ebay, I was hoping to grab them from you- :)

I got banned from Ebay like 8 months ago. It was my brothers account and then I had too confirm stuff too put my item up. They found out and banned him.


It's such bull because I got suspended the first time for no reason. I e-mailed them.


I'm upset because I was getting offers too.


Funny how you knew that was me. lol.


I'll be putting it back up. Or Pm me. I'll see what happens.



I'm testing with air, but losing steps when I push force on it :(


-Jason

So Frets, Lets see what happens with Ebay and stuff. I bought my system with the steppers @ over 300.00 So I was just wanting too see what I could get for the that stuff and within hours I got a few questions. So we'll see.



Now onto my problems.


With what steps should I "theoretically" receive the fastest speed? 1/2 is working, but I need too do more testing. I still feel like 30 IPM consistent is possible.


-Jason

Speeds, Sorry I didnt respond but I was out of town this weekend. I can tell you that I have played around with 18, 10 and now 5 tpi leadscrews. For the best bang for your buck do something with the leadscrews. With 18 tpi screw I could never get it about 24ipm without loosing steps, 10tpi was around 60ipm but I could get it up to 70ish area. Now with the 5tpi I can easily hit 130ipm on my X2 but for me that is way to scary and keep the jogs around 40ipm and 10-20ipm for cutting. My new build that I am working on right now is going to have 2tpi leadscrew so I am looking foward to see how that works out. I dont know if changing out the screws is possible on your mill but it is something you should really consider, I think it will save you alot of heartache.

Yea...:( It just sucks because I have seen many "Speedy" Taigs. I have been thinking about the Leadscrews. It doesn't look that hard, It's a brass bolt underneath the carriage with a certain screw. Maybe I just find the right dimensions and Boom!


-Jason

Was your taig cnc ready or manual?

Also, do you have a pc there that you could load Mach3 on and try the setup with it? I wish I could help more, but my system was using Mach3 and that worked great short of the heavy mill. So I don't really know what else to tell you on the settings. I have no machines left that I can load EMC2 with.

And also, earlier in the post. When I said 30ipm, I didn't mean rapid air moves, I meant cutting.

It was manual, but I bought as the CNC ready. Meaning it came with the motor mounts etc.


-Jason

It was actually pretty sweet. I have it at 1/2 microstepping and the "stock" file for when EMC opens is a EMC2 AXIS logo.

I engraved it into aluminum going @ 20 IPM. It was finally nice too see it come up and go too another spot quickly. I engraved it in like 50 seconds. It was cool! But BUT BUT BUT, I think higher speeds are still possible and at 24 IPM...I Lost steps. So still need too work out the Kinks.


-Jason

Guys I need your Opinion on this. So the Xylotex website says. "When using 10 TPI screws with the 269's you can get between 50 and 75 IPM. " If I'm using 20 TPI, I should still be able too hit between 25-50 Consistently with my 280 Oz motors.

Now this makes me wonder. Could it really be my computer? The computer was picked up form a neighbor who was throwing it out (It was great deal, it ran my cnc stuff and works still!!!) But I wonder of that could be it. Losing steps because of the PC.


How would a 2.8 Ghz with 512 Ram running only EMC2 and nothing else perform? (Price: 200.00) The comp I had is older and probably no where close too that.


What do you think?


-Jason

Jason, I have no experience with EMC2 but from what I've read, that should be MORE than enough. Especially if you have nothing else running. As for your particular setup, I'm sorry I can't be of more help to you-

I'm stumped otherwise. I don't see how it can be anything else. I fist want too test it though. I'll use one of my nice computers too see if it's true.


-Jason

By all means, Jason, try everything you can think of.

Just curious about another thing, you mention this was a CNC ready Taig you got from ebay, does it have the actual larger anti-backlash nuts installed, or the standard nuts? If they are the standard nuts, could this be part of the problem?

No, I bought it from Nick, brand new. My CNC components were from Ebay new, but from ebay.

-Jason

Oh, nick carter? So it was a manual mill first then and not a CNC "ready" mill? (there is a difference in the nuts).

Guys I need your Opinion on this. So the Xylotex website says. "When using 10 TPI screws with the 269's you can get between 50 and 75 IPM. " If I'm using 20 TPI, I should still be able too hit between 25-50 Consistently with my 280 Oz motors.


Not necessarily. Your motors would be spinning twice as fast, and would have far less torque than when spinning half as fast with 10 tpi screws. Torque is not linear with rpm.

It's about the Torque at the speed too.

Yea Nick Carter, it was CNC "ready", just was doing manual work first.


-Jason

Even losing steps at 15 IPm on Rapids....Parts are getting all messed up because it's not going too the right area. But I don't hear the motors buzzing and stalling or anything....


-Jason

I hope I can get this fixed. I would like too Rapid much higher. Much Much Higher. I want my Taig too be a real nice CNC machine. It limits me in cutting things because I know it would take too long.


I gotta figure this out.


-Jason

Jason, I really don't know what to say that can help.

I guess it is time to start with the basics. If the motor is missing steps, and you are not hearing them bind or drop, then I would have to suspect something with the PC or maybe the BOB. Lets go through the basics. Maybe there is something we are missing.

As far as speeds, maybe you should sign up on the Yahoo group for xylotex and ask what the expected speeds should be with a 20tpi screw. I know there are guys on that newsgroup that use taigs. They could be more personal in there help I would guess. It is possible that the stepper motors do not like high rpm? These aren't the ones that are sold with the xylotex kit, so I don't know. I do know that the 425s should be used with .200 pitch leadscrews. So sending you the motors I have wouldn't help.



I am sorry you are having trouble with it. If you get to a point you just want to do something else. Box it up, and send it back to me. I will refund your money and you can find something else for it. I'll build a mini router or something with it.

I never actually connected that set to any equipment. It was just put together on the bench and tested. I guess there could be something out of whack with the 3axis and I wouldn't know since I only used it on the bench for a couple of hours playing with mach3. The 4 axis kit is the one that I used on the SX3. It was used for about 3 months.

I still have the 4 axis rebuilt unit. The buyer hasn't contacted me after we talked. If you want to trade for that one, let me know. I know it works and works well for my mill. But I couldn't imagine that it would be any different. Xylotex boards either work or they don't. :confused:

If you have time though, lets go through the basics again.

Are you sure that the gibs are not to tight?

Motors are connected well to the couplings? Both mechanical and electrical.

Have you tried connected directly to the Xylotex and not through the c11?

Are we sure that the pdf link to the motors was for the right model number?

Have you adjusted and measured the Pots on the Xylotex?

Are we sure they are wired in series?

Have we tried a different PC?

Have you any way to try with Mach3 just as a test?

Are there any places in EMC2 for step pulse length? (That does help in Mach3 if the pulses are to fast.)

Are you sure that your power circuit is not overloaded? Have you ever tripped a breaker in that room or plug?

What are your current micro-step settings on the xylotex? What are your EMC2 microstep numbers?

Are there any settings in the EMC2 that have a "kernel speed"? Maybe we could slow that down a bit and see if it helps?

I will try to think of more as I go and post those. We will either figure it out, or you will get your money back. I will take care of my customers. Even at my own expense.

I think the board is ok. I am happy I bought the kit from you. I don't think it has anything too do with the board. It has too do with other outside circumstances.

Too the Basics:

1. Gibs are not too tight.

2. Yes, they are connected ok and everything turns smoothly.

3. No, I have it connected too the C11

4. Yes, the PDF is too the right motor, I took the part # right form the motor.

5. Yes, Pots are measured and ok.

6. Wired in a series, yes.

7. I have NOT tried a new Pc. I am building one this week, 3.0 GHZ processor, 2 Gig Ram, brand new. Will try EMC with it.

8. No, I have not tried Mach3, my comps for CNC are linux based.

9. I believe EMC does in fact have these options, I need too post in the EMC forum and find out. Or check the manual.

10. I have indeed tripped the breaker numerous times, even with my old cnc system. I have multiple things running from one outlet, I must check this.


11. I have it on 1/2 mircrostepping with "2" for the microstep trying too run .33 velocity

12. I need too read the manual, but sounds very familiar.



Thank you for being so patient and trying too help me out :)


-Jason

Guys I gotta say, After reading a huge thread about resonance dampers. I have too try it before anything else. I have seen results from videos and the posts in the thread were amazing.

People were getting twice the speeds without even changing settings in Mach3 or Emc like they planned on it. It just makes perfect sense too try this out.


-Jason

What is your motors rated current? Wiring them parallel at a lower current may get you more speed, because torque falls off using series wiring much faster than parallel.

Also, Xylotex recommends always using 1/8 step mode, because it's much smoother and less likely to suffer from resonance than the lower modes.

Provided your PC is outputting steps fast enough, no mode should be any faster than another.

But go ahead and try the dampers, they may give you what you want.

If not, you might want to look into buying 3 Gecko G250s during the $29 introductory offer. You can run bipolar parallel at up to 3.5a and 50V. Without worrying about resonance. Should easily double your speed for $100 + a new power supply.

I believe there rated for 2.1 Amps when wired in a series. I didn't know I can wire them differently, I thought they would fry the board???


I'll change back too 1/8th step and mess with it.

I'm also def going too do that dampers too. I got my stock, Just need too finish the design.



Hopefully this all works. Will update tonight.



-Jason

Can you post the model number for the motors here on the board. I don't have that pm anymore since I had to clear them.

Here is the link:

http://www.kelinginc.net/KL23H276-30-8B.pdf


-Jason

I don't even know anymore. I freaking dropped the Velocity too .2 which is max 12 inches min. I dropped it down so I can cut my resonance dampers.....Whats the point. I can even cut them. I'll lose steps all over the place. I'm getting so fed up with everything. My machine in general. I can't even cut the damn resonance dampers at 4 inches minutes. Losing steps.


-Jason

Are your wires shielded? Are current and signal wires running near each other? Are the shield drains grounded? If so, did you make sure are they are grounded only at one end? Do you have the ability to test it with a known good control computer? Have you tried a different signal (parallel) cable?

If you have eliminated the mechanical, you are pretty much stuck with signal/logic issues if it isn't a base problem with the board or supply.

Yes the wires are shielded. I have the power supply a few inches away from the board the axis wires run basically right over the wires from the Power supply. Except not right on top of each other. The wires that are shielded are only the ones on the stepper motors too the axis board. Nothing else is shielded. I just tested with my brother 2.8 GHZ dell that runs a Gforce 7600 (For latency purposes) I have not tried a new cable....


Yes, nothing bind on the table. At least I don't think anything does..... Everything feels smooth.


-Jason

Ok, so you have tried a new PC. Was that also EMC2 with linux or was this Mach3?

Is the Xylotex connected directly to the PC or still through the c11?

Can you post a photo of the configuration you have right now?

Yea, I tried a new PC. Still EMC2, I don't have Mach.

I tried with and without c11. Oh and when I didn't use the C11 I used your parallel cable.


I'll post a photo tonight.



I just don't know what going on. I can't believe I couldn't even cut though at 3 IPM. I couldn't even cut a circle. It was all messed up :( :(


-Jason

Jason, I got a pc in today so I can load Linux on it. I'm downloading ubuntu right now and later this evening, I will have a better idea about it.


Also, have you read this page and checked the settings?
http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TweakingSoftwareStepGeneration

I actually have not seen that section. I'll have too go over more things again. and do what those areas say too do.


Thanks.



-Jason

I can't open the excel file because the computer my mill runs off of is not connected too the Internet. So I can't use Linux too open that file. I'm doing this on my upstairs comp. I would like too use the excel sheet too find the base_period and everything.


-Jason

I re ran my Latency test: My Max Jitter on the the Base was: 450,750 ns and according to the manual:

If the numbers are 100uS or more (100,000 nanoseconds), then the PC is not a good candidate for software stepping.


I just hope I took it from the right place: When I commanded for the LatencyL I took my number from:

Base Thread: (25. us) and the number from MAX JITTERS (ns) 450,750




-Jason

I re ran my Latency test: My Max Jitter on the the Base was: 450,750 ns and according to the manual:

If the numbers are 100uS or more (100,000 nanoseconds), then the PC is not a good candidate for software stepping.

I just hope I took it from the right place: When I commanded for the LatencyL I took my number from:

Base Thread: (25. us) and the number from MAX JITTERS (ns) 450,750

-Jason

Jason,

I have a PC with a built in video chip set that gave similar numbers. I added a graphics card to it and greatly improved my latency results. My numbers went from 4-500,000 down to around 14-15,000.

I used an ATI Rage 128PRO 32mb PCI video card. Model #VAR128P-32P.

Alan

What If I put in a really really good video card? This comp is old, but I hear of people running way older comps then me.

I can't believe you went from 400,000-500,000 too 14-15,000 That must help.

So with my 450,750 thats almost the the MAX 500,000 allowed in EMC

Acon, You have always seemed too help. I hope you stick around and we can continue working on this.


Get back too me.


-Jason

I actually dont' think the computer can handle that.


Am I better off getting a 250.00 comp: With 2,8 Ghz CPU, 1 Gig Ram and then putting in a video card?


Please let me know :(


-Jason

Jason, can you give me an idea what pc specs you have? Also what motherboard are you running? I have several cards laying around here. Let me know and then be looking for a gift. :) If you get a chance, let me know tonight and it will go out in the morning.

I'm certainly not using them, maybe you can.


Happy 4th of July everyone.

Happy 4th of July!

I don't remember the specs of the comp and can't seem too find them using Linux. Because I turned the PC all too Linux. I'm going too run EMC on my comp in my room and see what kinda Latency I get.


BTW- The computer was picked up from a neighbor who was throwing it away .


-Jason

Is there a brand name or a model number on it? How about any labels that say what the processor is or something.

It may be what is called a homebrew, (Without big name brand) but we should be able to find something out about it.

If you take the cover off, how many slots are there open for the cards? What colors are they?

I actually dont' think the computer can handle that.


Am I better off getting a 250.00 comp: With 2,8 Ghz CPU, 1 Gig Ram and then putting in a video card?


Please let me know :(


-Jason

One of my computers is a 1 gHz cpu with on board video that runs about 14,000 latency with the on board video. The other one that I mentioned is a 1.6gHz now using a pci graphics card.

It couldn't hurt to have a newer computer and more memory. Like I said above, some computers will even run quite acceptably with on board video, it depends on the on board video implementation.

There is also a page in the wiki that gives some typical latency figures for some specific models of PCs.

Alan

Right, BTW- The comp I was thinking of getting will only be used for the Mill. It will not have anything on it. I'll look at more details and post them. I'll go tomorrow.


I need too figure this out soon.


So pretty much 450,000 NS is horrible! I can't believe it was that much :(

Setting this up is so important, I mean.... I have been thinking about it and it's got too be my computer. I mean, I really couldn't expect it too be anything else. The latency tests proves that my computer this entire time has been...incompetent of delivering the steps.

I just hope that getting the new computer will not leave me screwed again.


-Jason

Setting this up is so important, I mean.... I have been thinking about it and it's got too be my computer. I mean, I really couldn't expect it too be anything else. The latency tests proves that my computer this entire time has been...incompetent of delivering the steps.

I just hope that getting the new computer will not leave me screwed again.


-Jason

I cut at 75 ipm on my router with no problem losing steps. I am running old round steppers rated at 200oz/in with 5tpi screws. I am running a 32volt supply with about 15 amps.

Observations:
1. A Taig mill probably has a lot more friction than a cnc router running linear slides.
2. You already have more powerful steppers than I am using.
3. Your voltage is lower than mine.
4. I am using fewer threads per inch on the screws than you.

I suspect you may be seeing the effects of more than one thing going on. I don't imagine you are anxious to change your lead screws. I don't think it is likely that you will go to linear slides on the Taig. So what changes are reasonable given the above.

The secret is incremental improvement by eliminating one problem at a time. I would change out the computer first. Get the latency into a reasonable zone. Next I would look at a slightly higher voltage power supply. Then, if you are still having problems, I would look at a different set of drivers (like Gecko G540 when released).

The disadvantage with this approach is that you could use a higher voltage supply with the Geckos than with the Xylotex. So, if you decide to go for more voltage, you might want to do the higher voltage power supply and Geckos at the same time.

Alan

I totally understand what you mean.


BUt lets be honest here. My computer hit 450,750 LAtency. The manual says that that not even acceptable too run a CNC mill.


I just ran EMC on my brothers computer and Mine. ANd my brothers computer hit 10,081 latency. I put it through it paces! Internet, music, games...for over 10 minutes. The manual says thats AMAZING.


I am running my computer right now with LAtency TEst- My computer is hitting 9445 latency.


I'm going too take it step by step, but starting out with the computer.



-jason

Sounds right to me.

Good luck,
Alan

Thanks Acon. I mean, it's the only thing that makes sense.


-Jason

Here are the Specs:

* AMD Sempron processor LE-1250 operating at 2.2GHz
* 512KB L2 cache
* 1600MHz frontside bus
* 1GB of PC2-5300 DDR2 SDRAM memory (1 x 1GB), expandable to 2GB
* 160GB 7200-rpm Serial ATA II (SATA II) hard drive with 8MB cache
* DVD SuperMulti drive with Double Layer support (reads/writes DVD±R/RW, DVD-RAM, DVD±R Double Layer, CD-R/RW)
* 10/100Base-T Ethernet interface (RJ-45 port)
* 56K V.92 modem
* 6 USB 2.0 ports (2 front, 4 back)
* 1 parallel port
* 1 serial port
* 2 PS/2 ports
* nVidia GeForce Go 6100 graphics with up to 128MB shared memory
* 6-channel (5.1) High Definition Audio
* 2 microphone input jacks, 2 stereo front/headphone output jacks, stereo line-in jack
* Multifunction keyboard
* 2-button PS/2 wheel mouse

Microsoft® Windows Vista® Home Basic with Service Pack 1 comes installed, providing enhanced security and productivity, and new ways to organize, find and share information. Other software included:

* Microsoft Works 9.0
* Adobe® Reader
* Vista integrated DVD playback with 6-channel audio
* CyberLink® Power2Go™ (burn DVDs)
* eMachines Connect™
* Google™ Toolbar
* Google Desktop Search
* Norton Internet Security (trial, with 60 days of free updates)
* BigFix® (identifies and resolves potential system problems)




But should I run EMC off the disc or just do full install too wipe out Vista?????


-Jason

One last thing. I tested my system with my brother computer where he was getting close too 11,000 latency. If I didn't pout that number in Step-conf and left it in the 450,000 would my mill still be weird? Cause it was, I did this a few days ago without checking his latency first and never changed the settings too 10,000


-Jason

Lets just say I'm little disappointed too say the least. I went out and bought that 300.00 computer. My lantency was 14,000. EMC2 says anything between 10,000-20,000 will give good results.

In stepconf, I program the latency and use the XYLOTEX board number (the timing) etc...


What do I get???? Skipped steps at 30 IPM, Skipped Steps at .4^2 .3^2 and .2^2 works ok but who knows if that will skip steps too!

The machine is not binding, why is this happening???? :( Now I just spent 300.00 and don't now it hasn't even solved the problem.


-Jason

I'm going too narrow it down too power supply and boards. It can't be the computer now.


I can't even test my old system because it's not intended too go more then 12 inches, so it doesn't even help.


-Jason

Does this PC still have vista on it? Are you running from the CD? If so, please load the Mach3 program on it so I can try and help you. I am at a disadvantage here because I can't seem to figure out the caveats of linux. At least not overnight.

With mach, I can give you the settings and test with you.

I can't send anything to you now, because the post office is closed. But I can send you the 4 axis kit on Monday. The 4 axis is what I was running the SX3 on. I know it works. At least well enough to snap the 1/4" shaft on my 425oz motor.

I remember reading somewhere about step timing when connected to Xylotex and other BOBs. In fact it may have been on the Mach2/3 yahoo group. someone was having trouble with that and they suggested increasing the time on the step and direction pulse. In Mach the setting was set to 2 and 2 and it solved the problem. but that was many moons ago and I don't even know if it was the same board.

I'll keep looking.

Also, with your multimeter,

Can you power up the board and post what the Dc Vref voltage is for all 3 axis. Instructions say to take the measurement with black to the GND pin on power connector and then red on the single pins that say tpz, tpy, tpx

Also, please measure the voltage across the power connector and see what the running dc voltage is.

I know we discussed this in PMs, but also post what your wire colors are connected to from the motor.... Just to double check.

color to A
Color to A#
Color to B
Color to B#

And what wires are connected together at the motor.

Lets start there.

I will post everything tonight. Thanks for being of so much help!


-Jason

I'm going too check it all tonight, But I don't know where the test points are. I don't see any labeling. I think I have been taking it from the right place (The Vrefs) but I don't know where the test points are...?

It's wired right, the Black and orange together and yellow and red together.

I'll also check the Volts form the PS.


Thanks Davis!


-Jason

I'm going too check it all tonight, But I don't know where the test points are. I don't see any labeling. I think I have been taking it from the right place (The Vrefs) but I don't know where the test points are...?

It's wired right, the Black and orange together and yellow and red together.

I'll also check the Volts form the PS.


Thanks Davis!


-Jason


This picture should clear things up for you. Let me know if it looks different.

No problem, I want this to get sorted out as well.

OMG! I was not doing this....I was doing it differently. I will measure! Thanks Davis!


-Jason

I'm looking at my stuff. Again, what should I set the Vref again too for the Amps that should be delivered too the motors. I deleted the Pm's.


I'm not sure if I'm reading it right.


-Jason

According to the manual, at bipolar series. You motors are 2.1 amp. With the Xylotex manual, it says take that and multiply by 1.44.

That means your Vref should be set to 3.024 volts DC.

What are your readings now?

I'm looking at my stuff. Again, what should I set the Vref again too for the Amps that should be delivered too the motors. I deleted the Pm's.


I'm not sure if I'm reading it right.


-Jason

Xylotex spec sheet says the number is;
Vref = Motor Current * 1.44

Vref = 2.1a * 1.44

Vref = 3.0 v

Hope that helps!

Ack! Simulpost! :p

Thats exactly what I thought. I had them probably like at 2.5 maybe less. That could have been the problem. Thats what I calculated. But I did not have them at 3.024 volts. Hopefully this was the problem :)


I will test tomorrow morning! Thanks everyone for the help!





-Jason

I just thought of something. Am I suppose too supply the board with 5VDC? From say the USB port in my computer? I just though about this....


-Jason

Nope. It's been awhile since I had one, but I remember just hooking up a 24v supply, the wires to the steppers, and the parallel cable was enough. I assume the reference signal must be coming from a voltage converter on board or something - you'd have to ask the electronic guys where it actually comes from - but you don't need to worry about it.

There is a DC to DC converter on the board. It gives the other inputs and outputs +5dc for reference.

I just thought of something. Am I suppose too supply the board with 5VDC? From say the USB port in my computer? I just though about this....


-Jason

Actually, it may depend on the version of the C11 breakout. I saw someone else recently that had an older version of the C11g that needed an external 5v supply to get things working right. I would check the manual for your breakout board and verify that your board has the on board dc to dc converter to provide the 5v dc. Better to be sure.

Alan

MY c11 has the external 5vdc one form the USB and an actual external one as well.


I meant for the xylotex. But I don't think I do.


-Jason

MY c11 has the external 5vdc one form the USB and an actual external one as well.


I meant for the xylotex. But I don't think I do.


-Jason

OK, sorry, I thought you were talking about your breakout board.

Alan

Yea, I have one on there. Going too test now!


-Jason

Vrefs have changed nothing. I set them too 3.0 Volts. Still losing steps and can't keep decent Velocity.

I changed the parallel cable and even directly hooked it up too the xylotex board. I bought a brand new computer yesterday. I checked the Vrefs. I changed multiple things around. 1/8th stepping, 1/2, 1/4 etc...


Still losing steps.

So it's not the computer, EMC is a fresh start booted from the CD. However, I can't install Mach. I'm not using Vista right now don't have internet access in the shop, so installing it may be a pain unless I put it on the CD.


So I'm stumped, it's gotta be board/power supply. I don't have other stepper motors too test, but I'm pretty sure it's not that.

I really thought the new computer would be it, but hey I did fix latency.


-Jason

Ok then, I'm stumped to then.

I'm going to send you the 4 axis box I have then. I will preset the Vref down to 3.024 and you can just swap the plugs you have to it. It has the Power supply mounted in the case with the board. I'll disable the 4th axis and leave it on 1/8th microstep.

I will get that out in the morning priority.

Davis- I hope this is not an inconvenience too you?... I will disable my system and slap my old system on. I will install the new stuff and test everything. I will send back the 3-axis once I can test everything and hope it all works out. We'll go from there and figure everything out. What I will send back, if you need payment for anything. We'll work it all out.

Is the 4 axis system any different as far as Power supply goes?





One thing that got me was that I was using a 9 vdc PS with 17 amps and getting good cutting speeds.(And when I say good, I mean I could cut) I don't know if that confuses anyone else but I thought with the extra voltage going into the steppers it would increase speed.

I don't know, I hope it works out and again, it's not trouble for you.


-Jason

Wow. If you got ANY performance out of a 9v supply it was quite an accomplishment! Even with the Xylotex's maximum of a 24v power supply and well matched steppers, cutting speed is only pretty generally adequate. The jump to 24v should see a REALLY dramatic improvement in speed from 9v! I didn't even know 9v could work...
Don't know what to tell you here. Have you actually put a meter to your supply and made sure it was actually supplying 24v? Is the parallel supplying the right signal voltage? Other than that I have to say I'm stumped, you did everything I could think of that normally solves the problems, it's very likely it is just the board now.

Yea. Well No. Listen, The whole reason I upgraded was because my cnc system was getting me "by". I mean, I cut aluminum at 3 IPm or 4IPm and my Rapids lol, were like 5 or 6 IPM. But I was cutting and it was doing it's job and was being used for my business.

I put the meter up too the power supply and noticed that It was reading more like 28 volts. Thats what it was reading at full PS! It's an analog one but it was reading the Vrefs at 3.0 but reading the Power supply at around 29 volts. Stumped here too....


How do I check the parallel signal. Meter it up or actual computer screen testing?


-Jason

Jason, I promise. It is no trouble.

My intention is to get you some rapids. I completely understand being stuck at less than 10ipm. I had that problem at first on my SX3 with the 4 axis I'm sending you. That was until I added the counterweight. Then it went from 10 to 40. (until I snapped the shaft)

I'm not worried about the money. If we can get you up and running, I'll call it even. Then once your up and going, I can get the 3 axis back, I'll then see what I can do with it. It may have been bad from the start. I wouldn't have known since it wasn't used on a machine. Just on the bench. If so, I'm sincerely sorry this has caused you so much trouble.

Yes the Power supply is the same model and voltage.

One thing I would recommend though. I wouldn't use the analog meter very often. They create resistance when measuring and can cause problems. I use to work in Car Audio back in the day and watched a guy pop an airbag in his lap while measuring leads for constant +12vdc. He was using an old analog meter because his 9 volt was dead in the fluke he used. (And I hid his test light. the guy was dangerous.)

If you can find a digital multimeter, it will make life better. If you can't, I have one that I will send as well that I got cheap from Harbor freight. I would rather give it to you that risk messing anything up when checking voltages, etc.

Ok sounds great. I'm going too buy a new meter anyway tomorrow. The analog one kinda stinks, I want too make sure it's accurate.

So thanks for sending me the kit. Once I get everything working, I'll send the 3-axis back too you. Or when I get the 4 axis kit. Whatever works best for you.


Thanks for helping out.


One last thing, is the series wiring the only way too wire this. Series wiring... I mean for the steppers??

-Jason

No problem.

As for the steppers. Xylotex shows them in series for that amperage rating. That is how I have mine wired, even on my Gecko's. I'm not sure how well that would work. I would think that in bipolar parallel, it wouldn't produce enough power to operate the motors any better.

Maybe someone can shed more light on this question.

Ok, I'll leave it at that. I put on my old system and am cutting again. I'll be using this in the mean time until the new board comes.


Cutting slowly, 4 IPM but not losing steps and is very smooth.


-Jason

Jason, man I feel for you dude!

I have been thru what your dealing with now on my maxnc. 20 tpi screws, low speeds, loosing steps everything!

A few questions......
when you raise the speed. do the motors start to growl and eventually lock up (Missed steps?) around 15 to 19 ipm they sound like total garbage?

No matter what you try you always loose steps and the motors run like crap?

if you answered yes to any of these questions.... then you got the famous resonance!

Trust me on this one! First thing you need to do is make some sort of dampener!!!!!!!!!! Dont try and chase ghosts!
YouTube - 20 inches per minute resonce in stepper

I'm in the process of making them. I read that thread religiously. But the 4th axis Xylotex is being sent anyway. My old system is on it, i'm making them!


-Jason

now remember, with the 20 tpi screws, your motors have to crank up the rpms to get the wanted rapids. I was able to cut wax at 650mm per min which is right around 25 ipm. Anything higher would miss steps. Maybe the taig will get a bit more.

But you cut wax, I cut aluminum at around 4 IPm sometimes higher. But If I can get high rapids, that would be awesome.


It's nice because I know you were in the same boat and when I saw your videos, I was so impressed.


Waiting for the new Xylotex too come in but cutting with my old system, very slow though.


-Jason

i have never run a taig, I know that it is more structurally sound then the maxnc, but I am still going to think that you cannot get much higher feeds in aluminum unless you take smaller cuts. It is still a desktop cnc machine with a fairly weak spindle. To much feed ad you run the possibility of stalling the spindle motor.
I have cut copper with the maxnc and while it did to it, I had to take into account the weaker points of the machine. shallow cuts, slow feeds ect.

back to the high rapids you want. You also have to remember that with 20 IPM screws you have alot of motion inside the backlash nuts. The faster you travel the quicker they will wear out. With the maxnc I had aluminum dove tail ways. The taig also has dovetail ways. So again, the faster you move the axis, the more wear you will get on the ways. I very quickly wore the hard anodizing off of my machine running it fast. Machine construction is very important here is what Im getting at. keep the machine dripping with lube if you are insistent on running high rapids.

Right the feeds I leave the same, the rapids I want too be faster with. I totally know what you mean, I will make sure the machine is lubed constantly.


-Jason