[kowell]

My first idea was to build some new drivers for my PCB mill.

LiniSteppers , as these have micro stepping modes.

but i was not so keen on the heat that these generate.

After a bit more thought and a bit of hacking ,

I came up with this new design using IRFz 44 Mosfet's

instead of the Tip 122 transistors that were used in the LiniSteppers.

Using the code of the LiniStepper , with a few modifications , a new

PCB was created in eagle , then using the direct print to PCB method

a new board was produced and assembled.

This can handle a lot more amps and very little heat without any

heatsinks at all. Also including on board 5 volt regulators for the

supply for the PIC 16F84a micro that is used.



( due to a brain fade during printing of the PCB the Mosfet's :withstupi
are actually reversed in the photo and the 5 volt reg should
be at the Top of the photo - oh bugger )

Kym

[pminmo]

So how are you determining the phase currents to limit the values? i.e. using a 24V Supply with a 2.2A motor vs using a 30V supply with a 1A motor?

[kowell]

So how are you determining the phase currents to limit the values? i.e. using a 24V Supply with a 2.2A motor vs using a 30V supply with a 1A motor?

You mean limit the current to the motors ?

I an running hp laser jet motors 5.2 volts at 1.4 amps with a constant
current supply at 35 volts:

The voltage being variable while the current is stable:

Kym

[pminmo]

Ok, I thought you were doing something other than full step.

[kowell]

Full step , 1/2 step , 1/6 step , 1/18 step , just as
the Lini-Steppers but without the adjustable drive:

Kym

[pminmo]

The linistepper can control phase current by utilizing the bipolars in the linear region. Your circuit doesn't look like it can do that. Other than contolling duty cycle by phase in which case you would have to readjust your supply current. What am I missing?

[ciccio]

This long discussion started as an "upgrade" to Linistepper, that is:
how to avoid heath-sink required by linistepper, but still have microstepping.
Now we are talking how to build a FULL STEP driver?
I'm sorry, but where is the promised

Full step , 1/2 step , 1/6 step , 1/18 step , just as
the Lini-Steppers but without the adjustable drive:

Kym

?????????

[Steve-Tee]

Hi Paul,

When I built my unit, it was intended to be full step only.
My machine is big and heavy, so fullstep is all it needs.
While these simpler versions are capable of half step, I don't think they can do higher resolution than that.

I am hoping that someone can devise some clever programming to improve upon what we already have. Perhaps it is possible to enhance the resolution of these?

Ideas anyone?

In your case, you are trying to use microstepping to resolve vibration problems with the motors.
Is there not a simpler way to achieve this, using a different mechanical interface between motor and machine?

Ideas are coming to mind for electronic solutions, but I need to think on it some.


Cheers,
Steve.

This long discussion started as an "upgrade" to Linistepper, that is:
how to avoid heath-sink required by linistepper, but still have microstepping.
Now we are talking how to build a FULL STEP driver?
I'm sorry, but where is the promised





?????????

[kowell]

Just Uploaded this new hex File for 628, outputs on Pins
10,11,12,13 .

Simulated ok.

Please try

Instead of posting as a text file , just put it in a zip file.

Kym

[kowell]

I am thinking, you have down loaded the ZIP file, which contains the
much reduced Lini-stepper source code, compiled hex file and eagle board
file :

The PWM loop is the output to the Fet's, the Phasing of which is
set in the Mode-3600, Mode-1200, Mode-400 and Mode-200 loops:
switching the Fet's off or on

The other half of the code , which is removed , is for setting the
bias to the transistors to allow different current levels in different modes:

As seeing is believing ,grab a breadboard and do a quick mock up of the circuit

Kym

[pminmo]

If it is doing pwm duty cycle for current determination, it will work. But a single current limited source would probably limit you to 70% of the available motor power and keep coil currents in spec.

[lovesakhi]

dear fellow cnc'ers using linistepper
I had designed a pcb layout for above.
Additional features are.
1. all floating inputs are tied low
step direction and speed selection inputs
2. clock generator using 555 using full functionality of 555 chip
ie upto 500khz clock speed
3. 1/4 clock indicator of the motor.. beginning pulse detector indicated
by LED.
4. current limiting option for higher voltages marked by A, B, C.
see attached attachment in pdf format.
will try out this in this weekend till then
have fun
dr.sanjeevi rao s from visakhapatnam India

[James Newton]

Just to make sure everyone understands, this is really not a Linistepper anymore. It can't microstep (only half step) and it doesn't have any of the analogue current regulation features that are part of the Linistepper design.

As far as I can see, the only thing this controller has in common with the Linistepper is that PART of the source code was re-used. It cannot do any of the firmware microstepping that the lini uses to get 3600 steps/rev.

As far as the operating temperature and the need (or not) for heatsinks: There has apparently been a huge misunderstanding here. The Linistepper can be operated in Full or Half step modes with the exact same thermal efficiency. The Lini only gets hot when you go up into microstepping modes because that is where the power drivers start to be used in a partially on mode. In Full or Half step, the Linistepper drives the transistors fully on or fully off just as this "hacked" version does.

While the Linistepper source code is open source and we are happy to see anyone getting good use from it, this driver really should NOT be reffered to as a Linistepper in any way at all. I would very much appreciate it if the thread owner can edit the title or at least make it very clear that:

THIS IS NOT A LINISTEPPER.

Thank you!

[kowell]

To James newton

I have never claimed this to be a LINI-STEPPER > other than borrowing
the freely available source code ( thanks very much ).

ANYONE WHO CAN SEE WILL BE ABLE TO TELL THE DIFFERENCE.

THIS IS NOT A LINISTEPPER.

By the way keep up the good work on the website.

Kym

[FunnyNYPD]

I knew there is hardware design doing the micro-steps, are there any software solution for micro-steps?

[sahilkit]

hello kowell

this is my first post n i would like to know if i can replace 16f84a with 16f628a without any change in code ? as this ic is cheaper in india and can i replace the mosfets with
MP102 or SMA4033 darlington qaud transisitors as i have plenty of them lying around !

thanks

sahil

[kiwichris]

You should be able to change the code to work in a 16F628 without too much trouble.

You will need to add some initialization code for the ports on the 628, as on startup some of the pins are configured as analog inputs/outputs for the comprator, which the 16F84 doesn't have.

[sahilkit]

thanks kiwichris , can anyone of you guys make the required changes and post the code as i'm only good at pic basic pro programming and asm. is kinda not my thing .

thanks
sahil

[kiwichris]

thanks kiwichris , can anyone of you guys make the required changes and post the code as i'm only good at pic basic pro programming and asm. is kinda not my thing .

thanks
sahil

Hi-ho,

To disable the comparator you just need to lob the following code in the port before the port initialization code:

movlw 07h

movwf CMCON

ie: Write 07 into the CMCON register.

Aside from that 16F84 code generally works in a 16F628 in my experience...

Cheers, Chris H.

[kiwichris]

Just remembered.....

You need to be in register page 0 to do that, so just to be sure, make that code:

bcf STATUS,RP0 ; select register page 0
bcf STATUS,RP1
movlw 07h ; turns off comparator.
movwf CMCON


Must stop replying to my own posts, rather bad form, what.... :-).