Hello Guest it is April 29, 2024, 03:38:03 AM

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - HimyKabibble

391
General Mach Discussion / Re: Converting Heidenhain sine wave to TTL
« on: February 26, 2012, 03:12:06 PM »
You can buy a good digital encoder for under $30....

Could you suggest any

Holt

These are the ones I use - about US$23 each:

http://products.cui.com/CUI_AMT102-V_Datasheet.pdf?fileID=7573

Regards,
Ray L.

392
General Mach Discussion / Re: Phantom Limit Switch Events
« on: February 25, 2012, 06:32:12 PM »
There are two basic ways to connect home/limit switches to use a single input:

1) Series - this requires using normally closed switches that open when the home/limit is hit.  A pullup resistor is connected between +5V and the input pin.  The first switch has one terminal connected to the input pin, and the other to one terminal of the second switch.  The second terminal of the second switch is connected to one terminal of the third switch, and so on for however many switches you need.  The second terminal of the last switch is connected to Ground.  When ANY switch opens, the input will go high.

2) Parallel - this requires using normally open switches the close when the home/limit is hit.  A pullup resistor is again connected between +5V and the input pin.  Each switch has one terminal connected to the input, and the other terminal connected to ground.  When ANY switch closes, the input will go low.  Note that this scheme uses considerably more wire than the series case.  As a result, it will also pickup more noise.

Switches are more likely to fail open than closed.  For that reason, the series connection is preferred, since, if a switch fails, you will know BEFORE you need it.

Noise reduction can be performed several ways.  By far the best is to determine WHERE the noise is coming from, and eliminate it at the source.  If you can't, or won't, do that, then you're left with doing what you can to make the other circuits not misbehave as a result of the noise.  The two most common ways of doing that are adding a pullup or pulldown resistor, and/or adding a capacitor.  Both have the effect of attenuating the noise.  In the case of the resistor, there should be little or no change in the signal timing, while the capacitor can delay the transitions of the signal.  So, a resistor is preferred, if it will accomplish what's needed.  Most logic outputs can sink considerable current, but can source very little current.  For that reason, I always prefer to use a pull-up resistor, and let the switch pull the signal low when active.  I've seen very few, if any, cases where a stiff pull-up (100-300 ohms) will not kill any noise.  If the noise is so bad that this does not work, then you REALLY need to figure out where it's coming from, and reduce  or eliminate it at the source.  It should nearly always be possible to do this.

You can also improve noise margin by using higher voltage.  It's common to run limit/home switches using up to 30V.  But this requires some means of reducing the 30V signal to a safe level for a 5V or 3.3V logic input.  This can be done with a simple series current limiting resistor between the switches and the input pin.

Just as an example, I have ALL of the electronics for my mill in a single 12"x24"x24" enclosure - the PC, the servo drives, servo power supplies, a VFD, a bunch of relays, a Modbus board, a SmoothStepper, a KFlop, several breakout boards, and several other things.  I have not had to use ANY capacitors, and I have NEVER had a noise problem, even using all 5V limit/home switches.  The secret is in proper grounding (star connections for all power and ground connections), and making appropriate use of shielded cables.

Regards,
Ray L.

393
General Mach Discussion / Re: Phantom Limit Switch Events
« on: February 25, 2012, 09:55:31 AM »
I would be willing to bet simple adding stiff pullup resistors would have yielded the same result, without the potential delays that sizable capacitors can introduce....

Regards,
Ray L.

394
Who cares how old the computer is, or what kind of DRAM it uses, as long as it runs Mach3?  Put the SmoothStepper on your existing PC, and you'll have all the speed  you need.  I ran Mach3 for several years on a 500MHz PC with a SmoothStebper, and had no problem getting up to 350 IPM.

Regards,
Ray L.

395
General Mach Discussion / Re: Converting Heidenhain sine wave to TTL
« on: February 23, 2012, 12:25:45 AM »
You can buy a good digital encoder for under $30....

396
General Mach Discussion / Re: combining gcode
« on: February 22, 2012, 09:25:06 PM »
It's called a text editor....

397
General Mach Discussion / Re: Converting Heidenhain sine wave to TTL
« on: February 22, 2012, 04:40:41 PM »
What you need goes FAR beyond simply using a Schmitt trigger to convert the sine wave to TTL.  What you have is an anlog resolver, which will output one cycle of sine wave for each full revolution of the resolver.  To get useful position information from that, you need to do an a very precise analog to digital conversion.  You'd be far better off replacing those with digital encoders.

Regards,
Ray L.

398
General Mach Discussion / Re: MDI Feedrate Problem
« on: February 20, 2012, 08:23:34 PM »
If you simply type "X4.00 F6" how fast it goes will depend on what move it made last.  If the last move was a G0, then it will ignore the F6, and move at rapid speed.  If you want the F6 to be respected, you have to MDI "G1 X4.00 F6".

Regards,
Ray L.

399
General Mach Discussion / Re: Problem making my steppers turn
« on: February 15, 2012, 10:38:08 PM »
Your subject line says "steppers", your post says "servos".  Which is it?  What kind of drivers??  Nobody here can provide any help without knowing exactly what the hardware really is...

400
General Mach Discussion / Re: Why use a Motion Controller?
« on: February 12, 2012, 08:27:54 PM »
Motion controllers have several potential benefits.  How much these are worth will vary from one user to the next, depending on his needs, and the characteristics of this machine.

1) Increased speed - the PC is limited to a 100kHz step rate, and relatively few PCs will work even at this speed.  A more practical limit for most PCs is probably 45kHz.  If you have a machine with a high steps/inch (my machine runs 20,000steps/inch on 3 axes, and 96,000 steps/inch on the fourth), then a motion controller will likely give you faster rapids.
2) Smoother motion - Motion controllers generate the step pulses in hardware, which results in a smoother, near jitter-free pulse train than is possible with the software pulse generation when using the Parallel port.  On my machine, this made a noticeable difference.
3) Reduced loading on the PC - Motion controllers move the bulk of the motion control logic to the external device, significantly off-loading the PC.  This can enable, for instance, using a relatively low-power PC, or a laptop, and the machine controller.  I ran my machine for a long time with a 588MHz PC, and it worked just fine.
4) Expanded I/O - many, perhaps most, motion controllers have greatly expanded I/O capability.  For example, the KFlop I now use provides something like 50-60 I/Os, all programmable as inputs or outputs, and the number of I/Os can be greatly expanded by adding additional boards.  The PC parallel port solution is limited to two ports, and about 30 pins, and the split between inputs and outputs is fixed, and not changeable.  A complex machine will invariably run out of either inputs or outputs, and you then have to expand using a completely different technology, like  ModBus, PLCs, etc.
5) Some motion controllers are capable of doing things that would be simple impractical any other way.  For example, my pendant is connected directly to the KFlop, so I don't need a PC at all to operate the machine through the pendant.  I can also connect the Gecko ERR/RES signals from all my G320s to the KFlop to monitor and control servo errors, and stop the G-code program if a servo fault occurs.  It can also do true closed-loop control of stepper motors, and halt the program if a stepper loses position.

Regards,
Ray L.