Motor/driver upgrade

[gc3]

Is the 4g shopbot upgrade the best way to go to upgrade a pr machine. I have a older pr tool that was upgraded to a prt style z axis along with the motors and control box. What other options are out there?


Gene Crain
www.plantasymaderas.com (http://www.plantasymaderas.com)

[taskins]

Gene-
I am updating an early PRT 2000 vintage. Take a look at which motors you have installed. I found out that upgrading to a 4g board would do little to improve performance due to the high inductance rating of the motors-PK299-01AA on my machine.

I settled on the new shopbot RBK board and had it wired for the gecko G203v drivers and am pairing that with the PK-299-03AA motors and 4:1 belt drive transmisions. I hope to have it up and running by the end of the month.

[ken_rychlik]

Gene what do you want the machine to do? Cut faster, cleaner, run smoother, ect...

What motors are on the unit now?

My prt was helped the most by welding up all the joints of the errector set machine to make it tight. I think the PR is even more sloppy than the prt from what I've heard.

Looking at the wiggle points and finding a way to either weld or bolt bracing to take the slop out of it would probably be your best cutting improvement.

Making it faster without improving the frame may hurt more than it helps.

Kenneth

[billp]

Gene,
One of the best "upgrades/modifications"I have seen for the PR units was to rotate the wheels on the Y axis 45 degrees. I owned a PR for years and always had "issues" with the way the wheels rode on the outside edge of the X rails. And I am sure that by now you have had the fun of having to lever your entire Y carriage back on to the rails after some sort of unforeseen "collision" with a clamp, etc.
Here are two shots of one such upgrade. Yes you will have to re-rig your motor mount, but that's an easy fix. Once you've done this there is much less of a chance for the machine to bind or jump the tracks. You also won't have the issue of wearing a deep groove into your X rails anymore as well.
498


499

[rb99]

What did you have to do with the motor mounts and what i that rail the V is now tracking?

Thanks,

RIB

[billp]

Richard,
In retrospect this all went on awhile back, and my memory may be cross referencing two different scenarios. It may not be necessary to do anything with the motor mounts at all. I was probably thinking of when some of us did the original upgrade from a cable drive to the PR. Then we HAD to do something with the way the rails worked because we were moving up to the original rack and pinion setup. If someone were starting with a PR the mounts might be fine.
The V rails you see the rollers riding on can be bought from a number of sources ( maybe even Shopbot themselves?) . I know that Bishop Wisecarver was one vendor, and places like McMaster Carr, or Reid Supply might also stock them.
Guys like Bob Dodd, Ed Coleman, and Jimmy Arneson all made some sort of modification to this kind of setup, so maybe they can chime in here. I'll add another picture or two of how one upgrade was done. New rails can be bought that fit along the outside of the existing frame, or with a "lip" that bolts directly to "C" channel. If I remember correctly there was the option to buy them undrilled, or pre-drilled. Unless you really like laborious work, I'd go with the pre-drilled.


500


501

[gc3]

Looking for a little more 3d cutting speed...

Here is a cheap but very effective rail upgrade(?) I did that works very well I still have not put the hold down rollers back on and have not had any derail issues jogging or cutting as fast as the machine can go, the problem has that I am loosing strps when cutting above 4in/sec or jogging above 5in/sec


Gene Crain
www.plantasymaderas.com (http://www.plantasymaderas.com)
502

[rb99]

Gene I think your speed issue jogging over 4" per sec is from the motors being under powered. I have taken the motors off and run them at higher speeds and at a certain point under no load they choke. They also get very easy to stop with bare fingers. (with the old original controller)

Mike Richards is the "man" with the motor knowledge. He has told me lots of info about the motors, (only some of it soaks in!) but my old motors where wired straight, and when they get wired bi-polar they run on 1/2 the power, and get stronger and faster.

I bought a new G4 controller with Gecko 203's. I have not installed it yet but it wires the motors bi-polar.

RIB

[ken_rychlik]

Gene, the 4G will give you more jog speeed without loosing steps, but by it's nature 3d cutting will not reach those speeds anyway. It is either ramping up or ramping down in speed at all times and never reaches full speed anyway from the 3d stuff I have ran.

I can jog at 8 ips without loosing steps with the 4g on my prt, but as far as cutting 3d stuff it will still be under your current 4 ips place where you are now.

Do you know the voltage to the controller for your steppers? Mine has a 48 volt meanwell power supply in the box. If your voltage is below that, a 4g board and upgraded power supply would help quite a bit.

Still the 3d stuff is going to take a long time to run unless you use a large step over and then you will have lots of sanding to do. This plaque is a 10 hour run with a 1/16th bit and and a 10% step over.


503

[gc3]

I have some 36X70 signs to do for a local restaurant similar to the photo. I am hoping to find a low budget way of upgrading the speeds. Have you seen a better cut with the 4g box?

Kenneth I wish I could afford your machine, but we have been crunched this past year so the play $ is gone...

Thanks for the responses

Gene Crain
www.plantasymaderas.com (http://www.plantasymaderas.com)
504

[ken_rychlik]

The machine runs much smoother with the 4g upgrade, but I ran some files back to back, before and after the 4g board was installed and I really couldn't tell them apart. The smoothness of the cuts came by adding some braces and taking out the slop.

I understand about the play money. For me, selling the prt will help dampen the blow though.

Mainly all those growling noises you are used to hearing are GONE with the 4G.

Kenneth

[jerry_stanek]

One of the 4G benefits is there is nu stopping every 2000 lines of code also my move and jog 9speeds jump way up. I could only cut at 1.79 IPS on my PR and after the upgrade I can cut at 4.5 with greater accuracy. My PR has the belt drive gear boxes. Now I will put my PR up against my PRS standard and on 3 D work the PR seems to give be results But my unit values on that PR is 3 times more.

[rb99]

Jerry I would bet the reason the PR is better is the belt drives...less backlash.

RIB

[jerry_stanek]

I think it has to do with more steps.

[richards]

There are a lot of things that are part of a stepper 'system'.

The number of steps required to turn the spur gear one revolution is one of them. A Geckodrive stepper driver requires 2,000 steps per MOTOR shaft rotation. You would multiply that number by your gear ratio and then divide the product by the distance that the axis moved per spur gear revolution, i.e., Spur Gear = 25-tooth (1.25" pitch diameter = 1.25 * PI = 3.92699"), Gear Ratio = 3.6: 2,000 X 3.6 / 3.92699 = 1,833.4653 steps per inch. That also means that 1/1833.4653 = 0.0005454" per step.

The inductance of the stepper motor is another. A motor with low inductance will have 'snappier' performance than a motor with high inductance. A motor that is wired half-coil or bipolar-parallel with have 1/4th the inductance of the same motor wired bipolar-series. The inductance determines the maximum voltage of the power supply. 32 X SQRT(Inductance) = MAX power supply voltage. (Keep in mind that a stepper motor is rated to run at 80C or higher. 80C is too hot to touch. Using the maximum voltage allowed by the insulation of the motor is going to result in a very hot motor. A motor with 1.5mH inductance could be used with a 39VDC power supply, i.e., 32 X SQRT(1.5) = 39.1918. The same motor wired bipolar-series would require a 78VDC power supply to make it step lively. The motor that was shipped with the pre-4g machines was usually equivalent to the PK296A1A-SG3.6 motor. As ordered from the factory, that motor only had four wires, meaning that it had to be wired bipolar-series. That motor also had 30.8mH of inductance, i.e., it would have required a power supply voltage of 177VDC to make it step lively. The Geckodrive G20x stepper drivers are limited to 80VDC, so the stepper drivers and those particular stepper motors are not a perfect match. They still work, but the PK296A2A-SG3.6 motor wired half-coil would have greater speed and more torque because it is a very good match with the Geckodrive G20x stepper drivers.

The type of gearing is another factor. Some machines were sold without gearing, like my PRT-Alpha. Some machines were sold with 3.6:1 gearboxes, like many of the PRT standard machines. Some machines were sold with 7.2:1 gear boxes, like the PRS-Alpha (and the retro-fit kits for machines like mine). Some older machines were furnished with belt-drive transmissions. The 7.2:1 Alpha stepper motors and the 3:1 belt-drive transmission that I built for my PRT-Alpha give identical results - and those results are MUCH, MUCH better than the PRT-Alpha without gear reduction. The resolution between a Geckodrive driven 3.6:1 gearbox and the Alpha driven 7.2:1 is identical because the Alpha uses 1,000 steps per revolution and the Geckodrive uses 2,000 steps per revolution. Gearboxes have some backlash. The 7.2:1 gearboxes on my upgraded PRT-Alpha have no apparent backlash. The 3.6:1 gearboxes on the Standard machine MIGHT have some backlash. The four 3.6:1 motors that I have on my test bench have no apparent backlash, but some users have stated that their 3.6:1 motors have developed backlash. A belt-drive transmission, if properly tensioned, has no backlash. In addition, a belt-drive transmission can handle more torque than the 3.6:1 gearboxes on the PK296AxA-SG3.6 units.

What does this all mean? A machine that uses motors and drivers that are closely matched and motors that have a belt-drive with 3.6:1 or 4:1 gear ratio will give you all the torque and speed that a stepper motor driven machine can deliver. The software, if properly set up, will allow the mechanical components to move smoothly and efficiently. If you want a belt-drive transmission. you'll have to build your own. I've built several different models. Most cost between $100 and $200 per unit (depending on the material used). All were fairly bulky. All improved the cut quality of my PRT-Alpha (little or no 'chatter').

What about the Alpha Mode of the Alpha machines? Some like it. I have chosen to slow my cuts down so that the machine isn't pushed to work harder than it is able. The Alpha Mode caused some 'divots' in cabinet parts. Some of those parts had to be trashed. Since slowing down the cuts, the machine has never entered 'Alpha Mode' and I haven't had to trash any parts.

[jdgrahamwaldorf]

Before Christmas I upgraded my PRT with 4G board and PK299 motors. I used 7.2:1 motors and an 80VDC power supply from shopbot. My business is changing from cabinets to 3D work and I was looking for improved carvings. My observations are that the 7.2:1 motors are way smoother and cooler than the PK299s. Saw dust is no longer being vibrated off of the Y carriage and my carvings are coming off of the machine without needing sanding. The tradeoff was in speed. Move and jog speeds are lower which affects my old files but is worth it for me. I think the 3.6:1 motors might be a better choice for those people doing general work.

[richards]

John,

On the label on the back of the motors is the part number, i.e. PK296A1A-SG7.2, or something else. Would you let us know what it is?

The PK296A2A-SG7.2 has been used by several do-it-yourselfers all over the world. The A2A is a lower inductance motor and the A1A is a higher inductance motor. The A1A is rated 7.7mH when wired half-coil and the A2A is rated 6mH bipolar-series or 1.5mH half-coil. So, the A1A version, when wired half-coil would work quite well with an 80VDC power supply. The A2A, when wired bipolar-series COULD be used with 80VDC, but 70VDC would be better. If the A2A is wired half-coil, the power supply should be 39VDC or less.

The A1A motor pulls up to 1A when wired bipolar-series and 1.5A when wired half-coil, which means that the A1A motor can be used with a Geckdrive G203v without requiring a heatsink on the G203v.

The A2A motor pulls up to 2.1A when wired bipolar-series and up to 3A when wired half-coil, which means that a G203v should have a heatsink attached for proper cooling.

When comparing the SG3.6 against the SG7.2 motors, the big difference is torque. The 3.6:1 motor is rated to have a maximum of 22 lb*in of torque (352 oz*in). The 7.2:1 has 44 lb*in of torque (704 oz*in).

I think that it's safe to ignore the difference in resolution between the two motors. The 3.6:1, with a 25-tooth gear moves about 0.00054" per step. The 7.2:1 with a 25-tooth gear moves about 0.00027" per step. A piece of normal copy paper is about 0.003" thick. I doubt that most machines could mechanically repeat the per-steps resolution of either gearbox.

The limiting factor may be the ability of the Shopbot controller to generate pulses. To spin the SG7.2 motor at 5-IPS would require about 18,518 pulses. The SG3.6 motor would require about 9,259 pulses. I would guess that the controller could generate at least 20,000 steps per second (and probably more).

I'm guessing that you have the motors wired bipolar-series because you said that move and jog speeds are lower. If you wire the motors half-coil, you should see all the speed that the pulse generator can produce. (I regularly run my PK296A2A-SG3.6 motors at 1,500 RPM, which equates to about 27-IPS. The SG7.2 would be running about 13.5-IPS if it is run at 1,500 RPM. Running at that speed requires that the motor be wired half-coil and that the power supply be well matched to the motor. It also requires a pulse source that can produce 50,000 pulses per second, i.e. the G100.)

[jdgrahamwaldorf]

Mike,

My motors are labeled A6488-9412KTG which is the same motor that came with my indexer. The other info on the motor is 2 phase, .25 step, DC1A, 4.39ohm. My G4 board has g202 drivers.

I went with the shopbot motors because they are easy to exchange( 4 wires in, 4 wires out ) but if I can get more speed and keep the resolution I would like to try it if you can walk me through the process.

John

[richards]

John,

I don't have a cross reference between the A6488-9412KTG motor that have a special part number and the equivalent PK296 motors.

Let's start with what we know. You stated that the motors have four wires, that the gearbox is a 7.2:1 gearbox and that the Geckodrive stepper drivers are the G202 model.

We've got a problem. A four-wire motor can only be connected bipolar-series. You do not have the option of wiring the motors half-coil. That means that you have motors that have pretty good low-end torque but motors that cannot be run at high speed. (Bipolar-series wiring dictates that.)

So, bottom line, what you have already done is the best compromise for the motors, the power supply and the Geckodrive stepper drivers that you have.

Sorry, but you would need to have six-lead motors to convert from bipolar-series to half-coil wiring.

The G202 is an excellent stepper driver. I have four of those on one of my test benches. They work very well with motors that pull up to seven amps of current. Your motors, as presently configured, pull only 1A each. Because the current draw is less than 2A, they can be used without adding a heat sink to the stepper drivers.

My recommendation would be to replace the motors that you have with four new PK296A2A-SG7.2 motors and to replace the power supply with a 30VDC to 35VDC power supply. Then you would need to wire the motor as half-coil (Black to A, Yellow to /A, Red to B, White to /B on the G202 stepper drivers - remember, one of the X-axis motors MAY need to have the black and yellow connection reversed to reverse the direction that the motor turns - check the existing wiring to see if Shopbot has reversed two of the leads on one motor or whether they have incorporated the change on the circuit board. If all of the motors are wired identically, you would wire the new motors with the same wiring plan. If one of the existing motors is wired different, you would switch either the A, /A leads OR the B, /B leads). You will also have to add a heat sink to each motor. (Mounting all four motors onto a piece of 3/8" x 3" X 16" aluminum and then connecting the motors to the 4G board with connecting cables will work. You will also have to change the internal jumper on each G202 from positions 1,5 to positions 2,6 to allow the stepper drivers to automatically reduce current with a motor is idle.)

Replacing the motors will cost a little more than $1,000 for the motors, if you order them direct from Oriental Motor and about $125 for a new power supply if you order it from Antek or build one yourself.

You may wonder whether the change is worth the cost. I can't answer that question. If you mostly run 3-D, then the answer is 'probably not' because most of the 3-D files that I run use low speed moves. On the other hand, if most of your work has moves of 6" or longer then having motors than can spin at top speed would probably increase your throughput significantly.

Also, remember that the pulse generator that Shopbot uses on YOUR controller board has a lot to do with the top speed. The only two pulse generators that I've used are the original PRT-Alpha pulse generator and the V201 pulse generator that I now use. Both of those pulse generators can generator pulses fast enough for good efficiency with the motors, power supply and Gecko stepper drivers that we're discussing.

[cutitout]

I cut mostly foams and would like to replace my controller box (possibility the steppers) to get the fastest cut speed possible and keep a accuracy of around .03in or so. Cutting foam I also don't need the strength of others cutting harder materials.

My bot is one of the first "PRTs" from mid 2000.

any ideas let me know .. plan on moving my bot soon and would like to make the changes then.

[richards]

Eric,

With only 0.03" required, you could eliminate the gearbox entirely and use direct drive.

If you used Geckodrive G203v stepper drivers you could drive any motor up to 7A. My choice would be the Oriental Motor PK299-F4.5A wired half-coil. That would give you more than 600 oz*in of torque (at standstill, which is a little more than the PRT-Alpha motors pre-gearbox had). When wired bipolar-parallel or half-coil, that motor has 2.5mH of inductance, so you could use a power supply up to 50VDC. If you decide to wire the motor bipolar-parallel, I would suggest a 35VDC power supply. If you wire the motor half-coil, I would suggest a 35 to 48VDC power supply. (That is one of the motors that I have extensively tested on my test bench and those are the power supply values that I commonly use.)

With a 20-tooth spur gear (1" pitch diameter), you will get 0.00157" per step or 2X better resolution than you've listed. Your unit value would be about 636.6197. (You'll need to test that because of the normal variations in mechanical parts.)

With 600 oz*in of torque (half-coil) or 880 oz*in (bipolar-parallel), that motor with that Gecko stepper driver will compare very favorably to the PRT-Alpha before they started using the 7.2:1 gearbox.

Be aware that you will have chatter on multi-directional cuts, but the chatter marks will probably be less than 0.03".

That motor, which draws between 4.5A and 6.3A will require that you add a heat sink to the Geckodrive.

The PK299-F4.5A motor has eight leads, so you can wire it bipolar-series, bipolar-parallel or half-coil. It is the largest motor that I have ever used with the Geckodrive and it is one of my favorites. I prefer the PK296-F4.5A when I use a belt-drive transmission. That motor is smaller and makes a less bulky package.

I can't tell you how to interface the Gecko G203v to your controller card. You may be able to pick off the step and the direction signals directly from your controller card. You may have to purchase or build an interface card. The Gecko G203v uses a common ground signal and expects active High signals to its inputs. (The G201 and the G202 use a common 5V signal and expects active Low signals to its inputs.)

[gc3]

Mike,
My motors are PK296A1A-SG3.6, 2 phase, 1.5A DC. What can I do to upgrade the resolution, speed, ect of these motors...is the 4G option my only option?


Gene Crain
www.plantasymaderas.com (http://www.plantasymaderas.com)

[richards]

Gene,

If those motors have all six wires, you're in luck because you can wire the motors half-coil. The 4G option can be used, even if you wire the motors half-coil.

Here's the difference. Many of the motors supplied by Shopbot only had four wires, Black, Green, Red, and Blue. The Black wire and the Green wire internally are connected to the motor's 'A' coil. The Red wire and the Blue wire are internally connected to the motor's 'B' coil. Your motors have two extra wires, a Yellow wire and a White wire. The Yellow wire is the center tap for the 'A' coil and the White wire is the center tap of the 'B' coil.

If you wire the motors using the Black wire and the Yellow wire for the 'A' coil, you will have wired the 'A' coil using only half of the coil and that's why it is called half-coil. You do the same for the 'B' coil by using the Red wire and the White wire.

Many people think that if you wire a motor and only use half of a coil that you only get half of the motor's torque. That is not correct; you get 70% of the motor's torque. Sometimes that is a concern. With the PK296A1A-SGxx motors, that has no effect. The gearboxes attached to those motors cannot handle all of the torque that those motors can produce, so wiring the motors half-coil does not reduce the available torque.

What you do get is the possibility of higher speed. The Oriental Motor torque charts show that when wired bipolar-series, a PK296-01AA motor (electrically equivalent to the PK296A1A-SGxx motor) is good for about 150 RPM before torque takes a complete nose-dive. That same motor when wired half-coil has the same amount of torque at 750 RPM.

Depending on what size spur gear you use, 150 RPM is about 2.7" per second while 750 RPM is about 13.6" per second. Now, don't get all excited and think that you'll be able to cut at 13.6" per second before loosing steps. There are too many other mechanical factors involved, but anticipating speeds of 5" per second is not unreasonable. Even with my PRT-Alpha and 7.2:1 Alpha motors, I rarely cut faster than 5-ips when cutting cabinet parts. That is the speed that gives me the best quality with the least amount of edge sanding. Taking a little more time on the Shopbot can make a huge difference in the total time to fabricate a part.

Use an 80VDC power supply, which is the maximum that you can use with a G203v stepper driver. Use a 13k-1/4watt current limit resistor on the G203v's terminals 11 & 12 to limit the current to about 1.5A. Wire the motor half-coil (Black wire to terminal 3, Yellow wire to terminal 4, Red wire to terminal 5, White wire to terminal 6. You exchange the Black and the Yellow wires OR the Red and the White wires if the motor runs backwards.

Assuming that you're using 25-tooth spur gears (1.25" pitch diameter), you'll get about 0.0005454 inches of motion per step - which is probably better than your machine can handle mechanically. Your unit values will be about 1833.5773 depending on the tolerances of the mechanical parts.

If you adjust the ramping properly and use the proper cut depth and the proper move speeds, you'll get excellent performance without loosing steps.

You'll need to monitor the case temperature of the Geckodrive G203v stepper drivers. Measure the bottom plate of the G203v. If the temperature exceeds 70C, you'll need to use a heat-sink. Personally, I would always use a heat-sink. It allows the stepper drivers to be mounted off-board and makes maintenance easier.

If you choose to modify your existing controller card, you'll have to do your own design. The G203v uses Active High signals, meaning that the G203v's terminal 10 is connected to ground and terminal 9 expects to see a high level (2.5V to 5V signal) for each pulse step. The G203v draws 2.5mA, so the chip sending the signal needs to be able to Source at least 2.5mA at 2.5V. Most TTL chips are designed to Sink current. I BELIEVE that Shopbot used chips capable of Sinking or Sourcing more than 2.5mA, but I don't have a schematic of your controller card to verify that assumption. Over the years, I've found that it's a fairly simple matter to interface a stepper driver to a controller if the controller produces Step and Direction signals. The most that I've had to do is to invert the signals to match the input requirements of the stepper driver. Step and Direction means that the controller uses one signal line to send pulses to the stepper controller and another signal line to tell the stepper controller whether to turn the motor CW or CCW.

[cutitout]

thanks Mike
I have direct drive now -- I believe the UV are around 509 and some change. I believe my spur gear has 18 teeth. the slow speeds must be coming from the speed of the steps from the PC using the serial port to the stock (mid 2000)shopbot controler box.

Steppers PK299-01AA 2phase -- 1.8deg step -- DC 2a 3.2 ohms


will a USB (from pc) to serial port(on control box) be faster off of a new PC?

[richards]

Eric,

The PK299-01AA motor is rated 14mH inductance when wired half-coil, so 32 X SQRT(14) = 119V; however, the Geckodrives are only rated to 80VDC, so, if you change to the Geckodrives, you should use a 70 to 80VDC power supply. (That motor is rated 56mH when wired bipolar-series, meaning that it is not a good match for a Geckodrive unless it is wired half-coil.)

As far as using a USB to serial converter and a faster PC, I don't know that that would speed things up. You should probably call Shopbot tech support on that one. My 2004 PRT-Alpha had a USB to Serial converter that worked great until last summer, when the controller died. I replaced it with a V201 that Ted Hall sent me to experiment with before the Maker Faire two years ago. The V201 card plugged right into my machine's controller card. I had to enter a negative number for one of the unit values. Other than that, it works very well.

The effective baud rate has tremendous impact on speed. Baud rate (the number of state changes per second) is almost the same as bits per second, depending on how the serial port is set up. If you use eight data bits, one stop bit and one start bit, then baud and bit rate are the basically the same. However, approximately 10,000 pulses per second is not very fast to a Geckodrive. If the entire bandwidth were used just to drive one axis, that would spin a motor at 300 RPM - which is pretty good, but there are three axes on a machine and a lot of other things to monitor. So a baud rate of 230400 (or higher) is a whole lot better than 9600, 19200, 38400, or even 115200.

(To tell you the truth, I have no idea whether the stepper pulse train on a Shopbot is generated by the PC or whether it is generated by the micro-controller on the controller board. I've never taken the time to hook up an oscilloscope. Back when I designed process control systems, I had a micro-processor assigned to each stepper motor. The master computer simply sent a count to the motor control micro-processor and the motor control micro-processor generated the actual pulse train.)

Again, only Shopbot can tell you whether a USB to Serial converter will allow faster operation or whether you'll need to upgrade the controller.

[gc3]

Gene,

If those motors have all six wires, you're in luck because you can wire the motors half-coil. The 4G option can be used, even if you wire the motors half-coil.

Here's the difference. Many of the motors supplied by Shopbot only had four wires, Black, Green, Red, and Blue. The Black wire and the Green wire internally are connected to the motor's 'A' coil. The Red wire and the Blue wire are internally connected to the motor's 'B' coil. Your motors have two extra wires, a Yellow wire and a White wire. The Yellow wire is the center tap for the 'A' coil and the White wire is the center tap of the 'B' coil.

If you wire the motors using the Black wire and the Yellow wire for the 'A' coil, you will have wired the 'A' coil using only half of the coil and that's why it is called half-coil. You do the same for the 'B' coil by using the Red wire and the White wire.

Many people think that if you wire a motor and only use half of a coil that you only get half of the motor's torque. That is not correct; you get 70% of the motor's torque. Sometimes that is a concern. With the PK296A1A-SGxx motors, that has no effect. The gearboxes attached to those motors cannot handle all of the torque that those motors can produce, so wiring the motors half-coil does not reduce the available torque.

What you do get is the possibility of higher speed. The Oriental Motor torque charts show that when wired bipolar-series, a PK296-01AA motor (electrically equivalent to the PK296A1A-SGxx motor) is good for about 150 RPM before torque takes a complete nose-dive. That same motor when wired half-coil has the same amount of torque at 750 RPM.

Depending on what size spur gear you use, 150 RPM is about 2.7" per second while 750 RPM is about 13.6" per second. Now, don't get all excited and think that you'll be able to cut at 13.6" per second before loosing steps. There are too many other mechanical factors involved, but anticipating speeds of 5" per second is not unreasonable. Even with my PRT-Alpha and 7.2:1 Alpha motors, I rarely cut faster than 5-ips when cutting cabinet parts. That is the speed that gives me the best quality with the least amount of edge sanding. Taking a little more time on the Shopbot can make a huge difference in the total time to fabricate a part.

Use an 80VDC power supply, which is the maximum that you can use with a G203v stepper driver. Use a 13k-1/4watt current limit resistor on the G203v's terminals 11 & 12 to limit the current to about 1.5A. Wire the motor half-coil (Black wire to terminal 3, Yellow wire to terminal 4, Red wire to terminal 5, White wire to terminal 6. You exchange the Black and the Yellow wires OR the Red and the White wires if the motor runs backwards.

Assuming that you're using 25-tooth spur gears (1.25" pitch diameter), you'll get about 0.0005454 inches of motion per step - which is probably better than your machine can handle mechanically. Your unit values will be about 1833.5773 depending on the tolerances of the mechanical parts.

If you adjust the ramping properly and use the proper cut depth and the proper move speeds, you'll get excellent performance without loosing steps.

You'll need to monitor the case temperature of the Geckodrive G203v stepper drivers. Measure the bottom plate of the G203v. If the temperature exceeds 70C, you'll need to use a heat-sink. Personally, I would always use a heat-sink. It allows the stepper drivers to be mounted off-board and makes maintenance easier.

If you choose to modify your existing controller card, you'll have to do your own design. The G203v uses Active High signals, meaning that the G203v's terminal 10 is connected to ground and terminal 9 expects to see a high level (2.5V to 5V signal) for each pulse step. The G203v draws 2.5mA, so the chip sending the signal needs to be able to Source at least 2.5mA at 2.5V. Most TTL chips are designed to Sink current. I BELIEVE that Shopbot used chips capable of Sinking or Sourcing more than 2.5mA, but I don't have a schematic of your controller card to verify that assumption. Over the years, I've found that it's a fairly simple matter to interface a stepper driver to a controller if the controller produces Step and Direction signals. The most that I've had to do is to invert the signals to match the input requirements of the stepper driver. Step and Direction means that the controller uses one signal line to send pulses to the stepper controller and another signal line to tell the stepper controller whether to turn the motor CW or CCW.

Mike

My motors do have 6 wires.

In the last paragraph of this post it seems like I may be able to use my pre 4g controller with modifications to the card. Can you be more specific about the modifications?

Thanks

Gene

[richards]

Gene,

When I design an interface, I start from both ends.

One end is the Gecko stepper driver. It requires a step signal and a direction signal. The G201x model does not care if the signal polarity is Active High or Active Low. There is a switch or jumper setting to select the polarity of the signal. The G203v model requires Active High signals. The G201 and G202 models require Active Low signals.

The other end is the Shopbot controller. It generates signals. If it has a step signal line and a direction signal line, then all you have to do is to connect the lines (perhaps using a buffer chip or transistors to amplify or modify the signals so that they have the correct voltage and polarity).

Because I don't have a schematic of the controller board, I can't give you any more help than that.

If none of this makes sense to you, then ask someone who has an oscilloscope to read the signals being generated by the controller board. It will only take a few minutes and then you will know exactly what you have.

Once you know what the controller produces and what the stepper driver requires, the rest is simple. You build a 'glue board' from parts that you can get from Radio Shack to amplify/modify the signals so that things match up. A 'glue board' is simply a proto board which costs about $5.00 and a few chips or transistors which should cost less than $5.00.

Even if the controller board generates CW and CCW pulses instead of Step/Direction signals, you can still build an interface using a 7474 chip and a 7400 or 7408 chip along with any other chips that may be required to change the voltage or polarity. The 7474 chip can be used to generate the Direction signal from the CW/CCW pulses and the 7400 or 7408 chip can be used to combine the CW/CCW pulses to generate the Step signals.

Everything is possible, but you'll need a schematic or some verifiable data get things started.

[rb99]

Gene I bought the G4 controller box for my old pr but I have not hooked it up yet. Mainly because everyone have scared me with the com errors they get. I am tired of troubles...

RIB