Help Me With The Big Picture Re: Drivers

[mannster]

I am working on building a CNC mini mill using surplus components. I have some NEMA 34 unipolar motors that I feel will provide enough power to drive this machine (slowly). I will be using TurboCnc on an old PC driving a very simple phase driver using TIP120's. I am being careful to keep the maximum draw to 4amps since the transistors are limited to 5amps. I built a linear power supply that should provide enough current for this setup. My phase driver is built around this basic design:
SIMPLE PHASE DRIVER

My problem is, why is the discrete driver in these forums so complicated ? Since it is not a chopper ? I am just wondering because I do eventually want to move on to a better and faster machine with better drivers. I am a little confused with all the different designs floating around here, and I feel that my best choice is to just buy Gecko Drives when the time comes. I think that the speeds that these chopper drives achieve are awesome ! I think that the fastest I can crank out is about 11 ipm using a 1/2-10 acme screw.

It would be nice to find a simple chopper design that is a little dumbed down for a beginner like me (the designs I have seen here still blow me away) Just because I like the satisfaction of building something myself.

Right now, the best I can hope for is to mill new simple driver PCBs when my machine is ready, so I can use IRFZ44 and not worry so much about the current.
Also, some microstepping drives say that they can handle 3 - 4 amps, that doesn't seem very powerful to me. What am I missing ?

Can somebody with more experience give me some insight about this ?

[pminmo]

mannster,
Your phase drive circuit places all the stepping load on the PC driving the phase driver. Some software will run phase drive but not all. All that I know off will do step and direction and rely on the driver to do the state sequence. When you get into running 3 or 4 axis, at high step speeds, you start over burdening the PC.
A second aspect is it is good design philosophy. The user can determine if their machine application requires microstepping, voltage and current ratings, via the driver they select.

The simplest of the open source boards, the discrete driver board (http://www.pminmo.com//discrete/osunischematic.pdf) is basically comprised of four building blocks. The interface to the computer which provides a noise filter and electrical isolation, the second which is the state sequencer, the third power drivers and last a 5 volt regulator. I could debate both sides of the necessity of the interface circuitry and the onboard 5v regulator. But a good design would include the interface circuitry for several reasons, and why I include them. Runnning a cnc router with stepper motors and a spindle is an electrically noisy enviornment. With proper wiring and grounding techniques, the noise can be minimized, BUT if your cutting a part and missing a step or two even 1% of the time could be disasterious. The interface circuitry and onboard regulator go a long way to prevent missteps. I consider the target audience as users that may not have the most experience in electronics and try to post something that is as "bullet proof" as possible. I want those users cutting chips, not tweeking wiring, or scratching their heads when their machine doesn't cut repeatably.

Choppers take the user to the next level. Able to acheive higher step rates without bulky expensive power resistors acting as big heaters. But add another couple of elements to the circuit complexity. They must sense the current the motor is pulling and make decisions as to when to interrupt the current flow and how long to hold off before starting current flow again (thousand of times a second).

Microstepping in itself adds a level of complexity that is best handled by onboard intellegence. Getting the intellegence and the high current drive capability starts to get expensive.

Allegro has a uniploar microstepping driver that does three amps, that is the simplest solution if 3 A solves your problem. The SLA706X http://www.allegromicro.com/sf/97060/

Hope this helps,
Phil

[mannster]

Thanks Pminmo,

I didn't know that electrical noise problems needed to be addressed. So, that makes it worthwhile to invest some time in building a proper driver.
It still seems like a daunting task, but I will cross that bridge once I finish
version 1 of my mini mill.
So now, I think I will start a thread showing my project's development.

[pminmo]

It's not as daunting as you think, if you can assemble a phase driver you can do the discrete boards.

Phil

[H500]

Manster: I'm using 1.4 amp nema 34 motor's and get 30-40 ipm rapids on my milling machine. A chopper drive is well worth the effort.

[mannster]

I guess I don't understand how a chopper can use that little current to push the motors so fast....

Using TurboCnc on a 133mhz PC, and a simple phase driver, my motors skipped steps when I pushed them past 300 (not clear if it is Hz or Mhz or Khz) but that is the number I remember from the "configure axis screen"

I decided to use 11 volts for these motors to keep them below the 5amp limit of my TIP120's even thought they are rated for 4.5 volts

I only had problems when I tried to increase the rapids IPM. The motors have no problem with the actual working feedrates that I would expect to use.

I know I need better drivers, but I would also like to know if there is something wrong with what I've done so far.

Is the PC too slow ? Am I hurting these motors ?

(2x) Motor = NEMA34 rated @ 4.5Volt 1.4Amp / phase
(1x) Motor = NEMA34 rated @ 7.4Volt 1.3Amp / phase

Thanks in advance !

[H500]

Your motor have the same spec as mine (sanyo stepsyn 103-820-2). They are rated for 1.4 amp, so you need to keep the current below that value. 5 amps will eventally smoke them.

Your lack of speed is exactly what I would expect with 11 volts. You need more voltage if you want more speed. I'm using about 33 volts.

A chopper drive allow you to use the high voltage without requiring gigantic (expensive) current limiting resistors.

The phase drive circuit is good for learning, but 11 ipm is too slow for real work. Phil's a3977 board is simpler than the discrete board, is very inexpensive and will do microstepping. I'm using a variation of it.

[TinkerDJ]

I'm just starting with building Phill's 3977 board, its an excellent chip, the boards Phill has designed are perfectly suited for the hobbiest and you can get the chips from allegro as samples. I had a ucn5804b 3 axis board running my small router and blew one of the chips when i went to a higher voltage. Those chips are obsolete so i decided to build the 3977 board, i also have plans to build the picstep boards for my next mill. I have played with the sla7051 chips as well as the lmd18245's from national that the picstep boards are based on. Theres lots of options out there. I'm still working on a serial interface rather then using the parallel port, using pic18f452 chip. I even built my 30 volt power supply with help from guys on the site. Basically comes down to what you want to do.

[mannster]

Thanks for your input ! It has helped me alot to guide me in the direction I need to go. I will continue on to finish the mechanical portion of my project, and then attack my driver issues.

I realized that I need to think about my electronics more seriously.

Thanks everyone !

[mannster]

I have really been looking into chopper drivers and am really interested in their inner workings. I understand the concept but am having difficulties with one detail. The Vref and Sense resistors that are used by the comparator circuit have an important relationship which I am trying to understand.

If you have a simple explanation, please share it with me.

Double check my thinking:

If my Vref voltage is 2.5 volts, then regardless of the motor power source 24volts-30volts, if I desire 1.5 amps as the limit for the chopping, then I need a resistance of 1.6 Ohms and 4 watts of dissipation from the sense resistors.

*If this is wrong....please don't think I can explain how i got here...This was the best I could do with the internet and Ohm's law.

I do want to use 2.5volts for the Vref, and my motors are 3.2 ohm rated for 4.5 volts and 1.4 amp per phase UNIPOLAR.

Thanks in advance !

[pminmo]

For a vref of 2.5 your statement is true. 2.5V is pretty high for a Vref. Is there a specific reason you want a Vref that high?

Phil

[mannster]

At this point, there is no specific reason, other than 3 volts was the maximum Vref input voltage.

However, I am glad that I actually made the right conclusion:

The Vref voltage is used as the reference for the resistance of the Sense input.

What I understand is: 1 volt as the Vref and 1 Ohm resistance for the sense = 1 amp current chopping.

After many hours of pulling my hairs out, I realize that it is as simple as:

I = V/R

OMG ! I could kick myself ! You wouldn't believe how much researching it took to get this far

Thanks for your reply Pminmo ! What tips do you have for setting the Vref voltage ? Do you have a general rule as far as this is concerned ?