MechMate CNC Router Forum

MechMate CNC Router Forum (
-   MechMates already cutting (
-   -   Experimenting with options for vacuum clamping - Taibao, Taiwan (

liaoh75 Thu 25 September 2008 11:13

Irfan, the inverter they are using is made by a company called "Sunfar". I will get model numbers and specs later. When I had my lengthy conversation with Mr. Tzao, he indicated that the spindle I will get will be the same as the one that was mounted to his commercial machine but the inverter may not be the exact same one. I'll inquire more about why before I part with my 30,000NT (~$1000 U.S. dollars).

liaoh75 Thu 25 September 2008 11:19

Gerald, you really don't miss much. The Laser pointer is just experimental at this time. It is neither wired up or appropriate setting made in Mach3 yet. Still figuring that out. I have Mr. Tzao checking on the availability of bearing kits for the Chinese Spindle. Sorry it's taking so long as the day job is really getting in the way of all this fun.

BTW: The Module 1.5 rack is working out great without a hitch. Do you see anything wrong with the placement of any of my components?

Gerald D Thu 25 September 2008 12:09

David, no rush on finding out about the spindle bearings - we have 7 spindles lying around at the moment. But I think that other people should also be interested in the bearing aspects.

No, I see nothing wrong with your component placements. You have even got a tiny 0,0 written on the right corner ;)

DMS Thu 25 September 2008 14:05

David, thanks for your input. This is the problem with Alu I was thinking of. May be Onsrud Alu bits help

liaoh75 Sun 12 October 2008 04:36

HELP PLEASE!! + Picture Update
5 Attachment(s)
Last weekend, my computer wouldn't boot up. I took it in and they said nothing was wrong so I brought it back and the same problem occured. It freezes in the middle of a cut and the screen gets all blocky and patchy colors. The motherboard was an ASUS with integrated graphics. I did some research and read somewhere that integrated M/Bs are not good for running Mach3. So, I bought a new motherboard, memory, and reloaded everything thinking everything was going to be just like before. Was I in for a shock:eek:!!

Everything appeared ok until I started jogging the X and y axis. What's happening is so strange that it's hard to describe. If I hold down the left or right arrow keys to jog the gantry from one end of the table to the other, it seizes up after a few seconds and makes a strange whining noise. Initially I thought it was the pinion getting stuck so I removed the springs and let the motors hang and tried it again. Again I get the same noise. It seams as if the pulse stream is getting interupted but the strange thing is, it's not just one motor. If I hold down the arrow keys, X1 and X2 seizes up at the same time under absolutely no load. Now get this, the problem doesn't show up when I'm cutting something. It only shows when doing jogs on X and Y axis. Z axis seems fine at the moment. If I set the jog speed to 20% of full speed, it doesn't happen:confused:. Is there something going on with the pulse stream?? Does anyone think trying a PCI based parallel port would help?

My motors are tuned according to instructions, I'm fairly sure everything is grounded.

Any assistance on this would be greatly appreciated. I remember reading about a similar situation many months ago on this forum but, I can't find it. Please Help!!

J.R. Hatcher Sun 12 October 2008 07:14

David my guess is it's in the motor tuning setup. It sounds like you are over driving the motors and missing a lot of steps thus the "strange whining noise". We need more info. What type of motors, are they geared, what's the setup in motor tuning, how fast are you jogging when it happens. etc. etc. etc.

Gerald D Sun 12 October 2008 07:23

Have you tried another PC?

Roadkill_321 Sun 12 October 2008 09:18

Hi David, I think it was Doug's thread where he had a similar problem.

I also had this problem and it turned out that I was running the motors too fast.

Try this link

Doug_Ford Sun 12 October 2008 20:35

You've got a good memory John.


Same thing happened to me. I tinkered with the settings in Mach3 and if I recall correctly, once I slowed down the acceleration rate, it started acting right. Or maybe it was when I slowed down the jog speed. That was so long ago, I can't remember for sure. But I'd bet a six pack that it's just one of the speed settings in Mach3. Good luck.

Marc Shlaes Sun 12 October 2008 22:05


You are an artist. Your machine is simple, elegant and beautiful. I love your pc box. Where did you get that industrial trackball??? Very nice.

Good job,


gmessler Sun 12 October 2008 22:44


Your setup looks great. I like the way you have your computer mounted.


Here is a site I found for track balls.

Hope it helps.


Gerald D Mon 13 October 2008 00:10

That "swinging" computer mount is strong enough to carry the control box as well. Nice job!

liaoh75 Mon 13 October 2008 00:46

Thanks to all who have made suggestions and compliments. I started out with the intention of making my build self contained.

Gerald, my intention is to try another PC tomorrow.

JR, Here is the detailed info on my machine:

Motor OM PK296A2A-SG7.2 geared motor
Motor Drives: Gecko G203V X4
B.O.B.: PMDX-122
Power Supply: Antek 35VDC 500VA

Mach Settings:

Metric Mode
Velocity: 17,000 units/min ~ 283mm/sec ~11.15 Inches/sec
Acceleration: 700

When I changed the computer, I didn’t change any settings. I just copied the configuration file over to the new computer. Just a note though, this problem did occur on the old computer as well, but once in a blue moon. The whining noises I mentioned earlier only occur when the motor seizes up.

Another observation: When I had the motors disengaged from the rack as I was doing slow jogging, I noticed that the sound of the motors is not smooth. It almost seems as if the pulse stream is getting interrupted somehow. Also when the motors seize up, they seize up at the same time. It doesn’t seize up one motor at a time.

I read somewhere a while back (not on this forum) that the parallel port that comes with the motherboard is inferior to PCI based add on boards. I remember reading about the fact that dedicated PCI parallel cards pulse smoother and takes some of the processing away from the CPU. Has anyone heard of this?

Marc Shlaes Mon 13 October 2008 07:59

JR has been using one that is integrated for about 7-8 months with no such problems. His motherboard is the Intel "Little Valley" mini ITX.

Richards Mon 13 October 2008 08:24

I'm betting that the problem is being caused by Mach 3 settings. The PMDX-122 break out board should "buffer" any substandard signals being generated by your computer's parallel port.

In your Mach 3 settings, reduce the top speed by 25% and see what happens. If that cures the problem, then play with some numbers between your current settings and the lower settings until you find a number that always works.

You may find that the acceleration needs to be reduced. The ideal acceleration for a stepper motor is not linear. If you plotted an acceleration curve, it would have a steep, almost straight portion, followed by a long curve that allowed the motor to get to its highest speed without missed steps; however, that kind of curve does not work in most CNC environments because you need to move various axes at various speeds. That means that linear acceleration is required. That also means that the acceleration "ramp" needs to be longer than optimum.

There is the possibility that you have electronic noise, but if the problem is repeatable, then electrical noise can usually be ruled out. Noise tends to be random.

DMS Mon 13 October 2008 09:00

1 Attachment(s)
Nice setup David,
that the parallel port that comes with the motherboard is inferior to PCI based add on boards.
Have to disagree with that, as most of MM are running on onboard parallel card. Besides all printers are running fine on that port.

I took photo of my TFT with stone age voltage and current detectors - Red- current detector, black - voltage detector.
shows presence of voltage and strong current fields.
Do you have screen between monitor and motherboard ? I can't deduce whether this is the reason of your problem.May be other members may through some light on it.

ps - oops, while I was typing and loading image, Richard posted. Anyway this can be taken as information.

javeria Mon 13 October 2008 10:39

Hi Sharma, that might bee the cause - but usually not - manjeet on one of his plasma cutter had a problem where the problem was a parallel port cable with really thin wires inside and that caused a lot of problem for him - it was solved when thicker wired cable was use for the PP.

Sharma, where did you get the voltage and current detectors i need to get one.
I have high power cables running near my house and would want to see how much we are in the problem.


DMS Mon 13 October 2008 20:12

where did you get the voltage and current detectors
In kabadi bazar:)

javeria Wed 22 October 2008 08:39

David did you solve your problem - if so can you share what had happened.

liaoh75 Thu 23 October 2008 23:34

I finally solved the problem. It had something to do with the computer. I changed it out to another motherboard with INTEGRATED video and everything is back to the way it was before I changed out the first time. I wish I had a better technical answer for those that have been tracking this problem.

Interesting observations:

1) The problem still rears it's head once in a blue moon when jogging, but usually for the first few minutes of powering on. It almost seems as if powering on is producing some ultra strong pulse of EMI that has an effect on the control system. Then as I get going, the problem goes away. It's almost like a capacitor that charges and discharges and after the (EMI "charge") fades away, the problem goes away. I will be moving the PMDX and Geckos to the other control box on the left (see pics above) and all high voltage stuff will be in one box and the low voltage stuff will be in the other.

2) I've lowered my settings in mach to 15,000 pulses per min from 17,000 and that has helped and my acceleration setting is now at 550 vs 700.

Any other suggestions would be more than welcomed.

DMS Wed 29 October 2008 22:44


What bit are you using for Alu ?
I 've tried various bits, feed rate, router speed, oil, but still not satisfactory cutting.


J.R. Hatcher Thu 30 October 2008 06:23

I have not tried these parameters. I watched a video and it was cutting perfect so I ask for this information. Here is what I was told. Hope this helps.

Cutter: 1 fulte, Belin 33060
Lubricant: None
RPM: 18,000
Feed Rate: 48 ipm
Cutting Depth: .020 per pass

Gerald D Thu 30 October 2008 09:18

Different types of alu cut completely differently. Some alu's are easdy to cut, others are nearly impossible. I don't have some names/specs over here at this PE airport. (Hennie, I'm watching you!) The lubrication/cooling of the bit is also critical.

CNCQuest Thu 30 October 2008 10:00

Hi David,

Nice MM built.

I have encountered the same kind of seizing and whining noise when doing a continuous jog in my cnc mill. It was solved by reducing the velocity and acceleration in the motor tuning page.

Quote: "I've lowered my settings in mach to 15,000 pulses per min from 17,000 and that has helped and my acceleration setting is now at 550 vs 700."

Is that 15,000 pulses per min the velocity or should that be 15,000 mm per min? What kernel speed are you using? Try lowering your velocity by half and gradually slide it back up until it whines again. Then, back off by about 10 to 25% lower.

Can you give more technical info on the Chinese spindle?

Thanks and 謝謝,

DMS Thu 30 October 2008 12:12

1 Attachment(s)
Thanks J.R, and Gerald,

I have tried on 3mm Alu sheet with as low as 50 mm/min and 15000-20000 rpm.
With end mill, drill,vbit, ball now, etching bit, carbide 2 flute etc with gummy results.:(
J.R. I could not find Belin 33060 on google. Only this post of Gerald on shopbot -

Are you talking of these bits?
Belin 33060 search returned nil on this site.
If not then a link may help.


As WT mentioned "Seizure problem", Initially I've tried higher acceleration, and had same problem, with bit research I set to 3000 mm/min for x,y, A and 500 for z, which is comfortable speed for gear 7.2 and 36 teeth pinion. Kernel speed 25000.
I used this sheet for calculations but when I used speed given in this sheet (20,000) I had noise and stall. For this geared steppers this seems to be very high. I checked pdf of motor spec where I found permissible speed 250 rpm for pk296-sg7.2.

What I actually did - I did axis calibration in Mach with help of "set steps per unit. I found 125 against excel sheet of 127.324, I used this formula for calculation -

Speed per mm per minute = drive micro steps X motor steps per rev. X motor max.RPM / steps per mm

= 10 x 200 x 250 / 125 (actual)
= 4000
I set for lower side i.e. 3000

Now in your case, you have 20 teeth pinion so your speed comes to nearly 2200.
May be this is helpful. I have done this 3 months ago and it is working fine till now.
So guys correct me if I am wrong any where. (I am bit forgetful these days :D and my wife has long list of complaints about undone works.)


Gerald D Fri 31 October 2008 00:05

The alu cutting posts have been copied to a new thread, where some previous posts on the topic are being collected:

liaoh75 Fri 09 January 2009 18:13

I've been very busy over the last few months. I run a language school in Taiwan and from October to December is a very busy time. Just a quick update on what's going on. I've taken apart the entire machine after satisfactory testing for a re-paint.

Gerald, I seldom spend much time on the main homepage but had a quick glance to find my build listed with a picture. I am honored you would place my build on your home page. I'll supply a better picture with appropriate Logo as soon as I finish the paint job. Thank you for providing the plans and thank you to all that helped on this fantastic journey. Thanks Team Mechmate!

Gerald D Fri 09 January 2009 21:11

Hi David. Good of you to call back in and let us know what's happening. Thanks.

domino11 Mon 07 September 2009 11:40

Did you ever get your labels? You are really close to a serial number. :)

liaoh75 Tue 08 September 2009 11:34

Heath, Mechmate Logo has been on my Blue Beast for a good while. It has been very busy cutting away for quite some time now. We are finally moving to a Chinese Spindle after several bearing changes on the Porter Cable Router which held up surprisingly well. Some might find it interesting that the main reason for making the change is easier access to wider collet sizes available for an industry standard collet like the ER20. It is cheaper to buy a 3mm 0-flute solid carbide bit with a 3mm shank than a 3mm bit with a 6mm/.25 inch shank. We are going to keep the Porter Cable as a spare in case something happens to the Chinese Spindle.

I also have a two year old son and my better half that needs some of my time. I think I've been elgible for a serial number before the list broke double digits. However, I think I've just been so busy with the two growing businesses, I haven't made it a priority to get my serial number. After getting the spindle mounted, I will take pictures and post all the modifications with dust foot, replaced spider plate, and completed control console.

Thanks for the bump!

Charlie Tue 08 September 2009 12:32

Also make sure you set your steps per rotation. The 7.2s I believe are 1440 steps per rotation (0.25 step angle) not the 2000 that my Mach3 defaulted to. My motors ran so much better when this was set. If you have done this Great, but I didnt see it mentioned so I thought I better make sure.

domino11 Tue 08 September 2009 14:12

Good to hear everything is going well. Look forward to your new posts when you have the time. I have two little ones as well, so I know what you mean on the time issue. :)

liaoh75 Wed 09 September 2009 23:17

Hi Charlie, thanks for the numbers, but believe my machine has already been set correctly but I'll double check!

Gerald D Thu 10 September 2009 01:20

Charlie, a typical stepper motor and gecko combination is 2000 steps per rev. Add a 7.2 gearbox to that motor and it becomes 14 400 steps per rev.

If one makes a mistake in this area during the Mach3 setup, it is immediately visible in the scale size of the parts you cut out........add a 7.2 gearbox and watch your parts get 7.2 times smaller! :)

Charlie Fri 11 September 2009 19:23

OMG what?
Well I need to figure this out then. Im looking into it right now but if you could post more info Ill take it. Maybe mine smoothed out because it was closer with ten steps at a time? See I had a rough spot in the rotation. It would come around then and make a little magnetic rumble. Ill try adding another zero. *embarrassed**
This is all still bench top work.

Gerald D Fri 11 September 2009 22:51

See Understanding & Selecting gear pinions - setting the steps per mm[inch]

Charlie Sat 12 September 2009 00:24

I have been looking at this for hours. I now have it set at 3055.777488 steps per "unit". I had a full step/micro step lesson to learn. I will study the link Gerald thank you. Most likely I still have errors. The rough spot in my motor went away from something else because I reverted back some settings but the rough spot didnt return. Who knows?..I have been having a blast messing with them and making them sing.

Edit: Read the thread you linked and I was glad to know that I managed to get my 3055.777488 number before I read it. I thought you should know Gerald :)

liaoh75 Fri 26 March 2010 11:31

7 Attachment(s)
It's been a good while since my last posting to my build log. I've been busy working on building a business around the machine and running my Language School. I've got a little more time now so I'm going to start working on catching you guys up that have been following my build.

I'm going to start putting together a vac hold down system with a target budget of $1500 total. Over the last several months I have been experimenting with a number of pumps. I've tried:

1) shop vacs (industrial - probably more robust than the revered Feins albet a bit more on the noisy side, but not too bad) They all seem to fall short in some way or another. Vac wasn't that great (6-8" Hg at best with no leaks) and air volume was marginal. Couldn't really hold parts very well. I setup a test bed for these experiments as I didn't want to commit to setting up vac plenum and sealing everything without deciding if it was finacially feasible to do this.

2) dust collectors 2hp and 12 hp These had tremendous air volume but gave about 2" Hg on the best test run. Cheap but completely out of the question.

3) Regerative Blowers. These actually would work if they were big enough. I tested a 7hp and 15hp unit and like the shop vacs, they didn't pull very much vac but was top dog when it came to air volume. The sales guy told me flat out not to use a regen blower for my application and suggested I look into roots blowers or oil filled vein pumps. He said that many have tried using regen blowers with marginal success.

4) roots blower. Here is where things started to get interesting. I found a vendor about a 1.5 hour drive from my location that sold referbished Japanese (hand-me-downs) I spoke to the owners for quite a while about the pros and cons of all the pumps he had on hand. I asked to test out his units with my experimental setup. He agreed and seemed quite confident that his 5 hp roots blower (2" port) would probably do the job. One week later, I showed up with my test board (4'x4' with routed out plenum and one hole in the middle (single zone). The bleeder was an import (Thailand) 1/4' run of the mill MDF board that was the same size as my experimental plenum. I had one metal ball valve with a vac gauge before the valve so I could see how much vac the pump delivered. I was quite surprised. After sorting out the plumbing, we fired up the 5hp roots blower. It was a noisy thing to be honest as the guy didn't put on the muffler for the test. The gauge jumped to about 17" Hg. I opened the valve to the plenum with bleeder only and the gauge fell to about 1" Hg. I started putting flat objects on the bleeder one at a time and slowly the gauge started to climb again. When the board was almost completely covered with pieces of board, the gauge leveled out at about 9". From reading on the shopbot forums, this seemed to be the magic number: 9". All the flat pieces adhered to the bleeder quite nicely and took a good amount of force to move them; Way more force than a 12mm (1/2") cutter would exert of a piece of board. I was blown away. Price: ~$800 U.S. dollars with a 1 year warranty

I then asked if I could try the next step up from the 5hp roots (2" in and out) to a 10hp version (3" in and out). Thats when I came to the realization that a $1500 competent vac hold down was possible. When he fired up the 10HP monster of a pump, the gauge instantly jumped to 23" Hg, I opened the valve with only the bleeder and still pulled 5" with nothing on the board. I put a 2' X 2' by 1/2" piece of acrylic on the board and couldn't budge it with two other people helping me. The rest of the bleeder was completely open. The vac gauge read 15" Hg. Price: ~$1000 US dollars with 1 year warranty. (This was the pump I ended up buying.)

We then moved to a final test with roots blowers which was a monster 18 hp blower with 4" ins and outs. The vac was about the same but the air volume was frighening. Gauge read 23" with valve closed and 19 with a fully open bleeder. Did the same test with the acrylic and it litterally sucked the acrylic out of my hand onto the bleeder board as it got close. I could litterally hear air being pulled though the 1/4" piece of MDF used as a bleeder. Just for grins and giggles, I put a 3" round puck on the board and even that was difficult to move. But, it also pulled a staggering 45+ amps - 3 phase 220 during the test. Also the price was out of my target budget range: $1800.

5) Becker 7.5 hp vein pump and also a variety of Alcatel pumps yeilded excellent results as well and was much more on the quiet side but they wanted an arm and a leg for them. I was quoted $2800 for the cheapest 7.5 hp (second hand) model and upwards of $5000 for one of similar performance to the 10HP 3" roots blower I ended up with.

That's long enough of an explanation for today as it's getting late so I'll leave you with a few pics.

One last thing: I got a great deal on a 10HP inverter through my Brother-in-law. $150 referbished "Rich" brand. I got this for soft start, RPM control, and power savings. More on this later. At this point the complete system is finished but I have not had time to take pictures. Here are the earlier pics of the blower itself:

vishnu Fri 26 March 2010 12:34


That's a nice Vac pump and worth the price. Can you tell me the source of your purchase, i would love to have this setup. Waiting to see some videos of the pump working.

Gerald D Fri 26 March 2010 12:57

David, that is probably the best written research I have seen on vac holddown pumps!

(PS the correct spelling in this case is "vane" - seeing that you are a language school person :))

All times are GMT -6. The time now is 13:35.

Powered by vBulletin® Version 3.8.3
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.