[lietuvys]

Hi all,

recently I found stepper controller Linistepper which called my attention.

What do you think about it?

I am interested in running three unipolar steppers at currents about 3-5A per phase. I will use them for wood router. Currently I have build chopping controller based on L297 with FETs. I would like to drive my steppers at faster speeds with more torque. I think that Linistepper controller would eliminate resonance problems while using microstepping.

I see that it will be necessary to modify Linistepper - place more powerfull transistors, resistors and tune it up... I have skills and work with electronics, and want to do it myself. I have a scope which will be useful there.

Maybe someone allready done this?

Am I going right way? I know gecko drives, but they are too expensive for me now.

[drk]

HI ,

I must say its a great find. Ive been looking for something of that sort myself, and yes , geckos do work out quite expensive..

Also, have you posted gerber files of your drive??

Regards

Dev

[lietuvys]

...
Also, have you posted gerber files of your drive??

No, I wired everything on proto board because I was just experimenting.

[James Newton]

Hi all,

recently I found stepper controller Linistepper which called my attention.



I think that Linistepper controller would eliminate resonance problems while using microstepping.

I see that it will be necessary to modify Linistepper - place more powerfull transistors, resistors and tune it up... I have skills and work with electronics, and want to do it myself. I have a scope which will be useful there.

I sell and support the Linistepper in the USA. Just wanted to note that it IS an open source design: Although we do sell a PCB and a full kit with a pre-programmed micro controller; the circuit, source code, and PCB layout are all on the web site for anyone to see.

Linear drivers are very good at elminiating resonance and related issues. The motors run cooler (although the controller runs hotter) and turn smoothly.

The kit (or standard components) should work for up to about 1.5 amps per phase and 3 is very doable with a little fine tuning and... and REALLY big heat sink on the controller! 5 amps is really pushing it.

If you were just doing full step or half step, the higher power requirements would be easy, but the combination of microstepping and high power is difficult to do at low cost.

One thing I should mention is that we are doing a promotion for a limited time: If you purchase a kit or set of kits, build then hook up to your device (mill, lathe, telescope, etc..) and send me a picture of it (clearly showing the Linistepper) I will refund $10 of your purchase price. I will refund an additional $10 for a picture of any item showing the difference between linear microstepping and any of the standard stepping modes. The second refund is only available after the first has been qualified. It can be a picture of two items made by your device or, if you are using the linistepper to control a telescope mount, it might show the difference in clarity between two identical long exposure photos of the stars. Finally, a third $10 refund, available after the first two have been qualified, is offered for temperature readings of the motor and controller after 1 hour of operation in full step mode and again after an hour in linear microstepping mode.

Then end result is that, at least for one channel, the controller will be totally free.

[H500]

I like to commend the designer for coming up with an interesting and creative solution.

But as his website clearly states, this is not a chopper drive, so it's really not suitable for high voltage and currents. For example, if you need 4 amps per coil with a 55 volt power supply and 3 axis, the driver would generate as much heat as a 900watt heater! You will need a jet-like fan to keep it cool.

So the bottom line is, this is an interesting drive for those looking for a very low noise drive for relatively small motors. But it is not a serious choice for those who are looking to handle more current and voltage.

[James Newton]

Yes, the linistepper is good for up to 1.5 amps per phase in microstepping mode (or up to 3 amps per phase in full or half step ) and does best for small, precise applications or where smoothness is critical.

The heat generation that H500 warns about is generally valid, but perhaps a bit over stated... Not all the power is turned into heat. Not every step results in the driver working in a linear mode. etc... But yes: It does run quite hot in microstepping modes.

[posix]

Hello there James.

C5 and C6 are 2.2uF on the schematic, however you provide 4.7uF in your kits. Also Q5-Q8 are BC337 on the schematic yet you show TO-92 in your kit. Which one is the correct component?

Thanks

P.S. as far as those pictures are concerned, you'll get them from me when I'm done making the PCBs - laying out and routing this tiny board on a single layer takes time with eagle

[James Newton]

Hello there James.

C5 and C6 are 2.2uF on the schematic, however you provide 4.7uF in your kits. Also Q5-Q8 are BC337 on the schematic yet you show TO-92 in your kit. Which one is the correct component?

Those components are not critical and so either is correct. The values shown on the schematic are "ideal" but the components supplied with the kit work perfectly and happen to have been easier to source.

BTW: TO-92 is a case size not a transistor type, and in fact the transistors supplied with the kit are actually BC337 in TO-92 cases. But the transistor is not critical and the standard 2n replacement for a BC337 could be used instead.

P.S. as far as those pictures are concerned, you'll get them from me when I'm done making the PCBs - laying out and routing this tiny board on a single layer takes time with eagle

I'm looking forward to it!

[pminmo]

I added the linistepper link to my webpage if thats ok?

Phil

[posix]

Ok, It's done. But it doesn't spin. I did for a brief moment after a lot of re-wiring attempts (and helping it along by hand). Now it just wobbles again. I also get around 6v on Q1 and around 11v on Q2,3,4.

When on 3600 mode, full power (111) it does 4 steps (as far as I can "feel" under my fingers) to one side then 4 to the other. The time it spun the stepper it was on 200 mode i.e. full step, full power (100). I am still on the default 1ohm set-up. I have also incorporated the two 470ohm resistors found on the assembly page between +5v and r14/15 and r16/17 respectively. Everything is basically as per web page.

I'll experiment some more and report on progress.

P.S. just to get this straight (transistor->coil) Q1->A+ Q2->A- Q3->B+ Q4->B-, right?
P.P.S. my steppers are 1.2A step-syn items (blue label)

[James Newton]

pminmo: Heck yes that's ok, thank you!

posix: Sounds like a wireing problem. Check out http://www.piclist.com/techref/io/st...p/lini_use.htm and especially the section under Motor does not turn??

For help with how to find the wireing of a motor, see the expert script at: http://www.piclist.com/techref/io/stepper/wires.asp

I would recommend you stick to full step (200) until you get the motor turning then try the microstepping modes.

[posix]

no matter what I do it just wobbles and sits there. At some point in the frequency range it begins moving back and forth erratically then stops again and wobbles. I have downloaded the lini zip file, is that the correct software? I compiled it from source rather than re-use the existing hex file.

are you sure the schematic and software on the home page of linistepper are the glitch-free versions?

I am using an atx power supply which worked fine on another tip1220-based driver of mine so that side is ok.

[James Newton]

posix: There has never been a "glitch-present" version on the web site... The controller was well tested before release and has been sold for years now. Something is wrong with your unit, and given some time and tests we can figure out what it is.

Do you have a volt / ohm meter? Can you check the signal on each of the motor leads with the frequency as low as possible?

[posix]

Ok, here is the "voltage report"

	transistor	Q1	Q2	Q3	Q4
step
1		5.5	11	5.5	11
2		11	7.5	5.5	11
3		11	7.5	4	4
4		5.5	11	4	4

phew...haven't coded in html in 6-7 years....

[James Newton]

Very nice table!

Q3 isn't turning on. Trace back from there to the PIC pin checking for signal as you go. E.g. Check from Q3's base (the lead nearest Q4) back to Q7's collector (the lead nearest Q3) and if that is ok, check from Q7's base back to R9 and then from R9 to pin 7 of the PIC.

Actually, in keeping with good troubleshooting technique, find the mid point of that chain (R9 to Q7's base) and check that signal. If it is good, (should be a swing of several volts) then try between Q7 and Q3, otherwise try between PIC pin 7 and R9. That will allow you to find the problem in two or three tests.

There may be a blown component, solder tail short, bad solder connection, or (I hope not!) a defective PCB trace.

Do let us know what you find?

[posix]

Ok all is good. But I found a strange thing.

For example, Q1 base gets 5v and on the emitter I get 5v. Q1 base gets 0v and on the emitter I get 12v. But the Q3 base never gets the 0v. It's allways either 2.5v or 5v and thus the emitter is showing 7.5v and 5v respectively.

[James Newton]

Ok, so again, Q3 isn't turning all the way on and it isn't getting a correct signal at it's base, so something before Q3 isn't driving it correctly. That would be Q7 or R9 or Pin 7 of the PIC. You need to track back along that path and find out which one of those the problem starts at.

[posix]

Oh no, sorry you misunderstood me, the singnal gets from pin 7 all the way to the base of Q3. Can you please measure values on your linistepper from pin 7 all the way to base of Q3 so I can have those for reference? There _IS_ a change on the signal from 2.5v to 5v but just not from 0v to 5v.

[James Newton]

Ah! Well then if the base of Q3 isn't being driven correctly by Q7, and Q7 is being driven correctly at R9 then the problem is likely Q7, or the joint between one of Q7's leads and the PCB, or something like that. It could be Q3 as well, but Q7 is a good bet.

I would suggest swapping Q7 with one of the other Qx transistors of the same type.

Pin 7 should be going from 0 to 5 volts and you should also see that at the collector of Q7 and the base of Q3.

Could you contact me off list or privatly with your real name so I can see about sending you some replacement components to assist in the troubleshooting? I can't seem to find your order from "posix"

[posix]

Ok I have an update. Everything is working now! :banana:

It was a combination of (semi)dead Q7 and having to re-solder R9 (tested it on-board and it had infinite resistance, removed from board showed 1.5k, soldered back in place and got 1.5k, strangest thing). Luckily I had a bag full of bc337s and it was a simple swap. Everything working 100% now!

Thanks for your help and support james, and sorry to have doubted you regarding the correctness of the schematic.