[tachus42]

I don't really think the voltage regulator on the charge pump is necessary either but found space for it just in case it was, I only takes one link or a 15 second edit and its gone. For that matter one 5 volt regulator for both signal and power will probably be ok.
The discreet regulator is for high input voltages ie 50v I'm not aware of any common high voltage 3 term regulators, also it splits the dissipation between two devices. Again it can be bridged out if not required.
The 244 takes up slightly less space than a few transistors would and is a bit less fiddley to put together. Paralleling is necessary to get fast enough switch on and off times. I have arranged it so that more gates are used for turn off, the idea being to have turn off occuring before the upper mosfet turns on to avoid shoot through. Some of the turn on gates will probably be disconnected in the final edit to get reasonable symetrey with the high side drivers.

[pminmo]

Makes sense. Turns out a LM317HVT will go to 50V in with .5A out, I wasn't aware of that either till I just looked at Nationals website.

Phil

[kd6oji]

Success! My three channel board is finished and working properly. Tested using a 555 to generate clock pulses rather than my switch (had to go to its home position anyhow) all three channels drive my motors @ 12vdc from old PC supply. IRF44's cold as can be while operating (7805 is warmer than they are!!) now to tweak my software and I can start working.
Thanks for the designs!!

[slp_prlzys]

congrat's!

[berin]

Soon as I get the schematic built with a more refined design, I'll post my simple step to phase converter circuit. its based on two IC'S. A 7474 and 7486. its just a straight step/dir to phase whole stepping converter. So far I have one thats 98% built. with the 8chip's sitting on 14pin sockets. the board mesures 3"x4". and could EASLY be reduced by building a pcb instead of a rats nest of wires on perf board. The board design I'll post is based on pcb.

[pminmo]

Berin,
7474 and 7486, I suspect it based on the same logic as the 3 axis and single axis unipolar discrete as we have published here:
http://webpages.charter.net/pminmo/cncelectronics.htm ?
Always looking for input, but if these can be of any help just wanted to make sure you knew of them.

Phil

[Greg Fill]

Tachus42 I looked at your schematic using a 74HC244 as a FET driver. You may want to consider using a 74AC244 instead. The drive current of a normal HC part is about 5 to 6 ma. / gate output. A 74AC244 has about 24 ma of drive current / gate output meaning you can either reduce the number of paralled gates used to drive a FET or just get more drive current --> faster switch on/off with the same circuit.

On a second note did you ever consider using a micro as the front end to your controller? Don't get me wrong using discrete logic gates to generate your analogue control voltages etc. is excellent, a micro just makes the job so much easier. Analogue control voltages can be created simply using an I/O port and an R2R ladder or with a PWM output.

[pminmo]

Greg, Tach
Just looking at the atTINY13. At the 20mhz version, I'm thinking utilizing the analog comparitor one could be used to do do the logic for chopped current sensing and drive enable/disable logic to provide the FET drive circuitry. At 1.29 each usd it becomes a pretty flexible part.

Phil

[Greg Fill]

I am not familiar with the Atmel parts but you may find that the comparator may be too slow. You may run into the same problem I have with the Cypress PSOC and the PIC micros that have onboard comparators. The slew rate of the comparator is too slow. A slow slew rate allows a delay in sensing of an overcurrent state and the current overshoots before the chopper can kick in and turn it off. Otherwise using a micro to perform everything else involved in the chopper controller is pretty straightforward. Also if the atTiny13 has hardware PWM outputs it can be used to generate the sinosoidal chopper current drive waveforms. I guess the only way to know is to look at the specs of the comparator in the atTiny13 and give it a try. The faster the comparator the better. The other issue you have to ensure is that the comparator has direct control of the FET drive circuit so that as soon as the overcurrent condition occurs the FET outputs are shut down. Any delay the micro has in shutting down the FET drive will result in a noisy and bad looking chopper waveform. A circuit I use that works very well consists of a 4013 CMOS S/R latch, a 74HC08 and comparator. The 74HC08 is an AND of the low side drive enable produced by the micro and the output of the latch. When the overcurrent condition occurs the rising edge of the comparator output is latched, the FET outputs are turned off and the micro in its own time (< 5 us) sees the overcurrent condition. No interrupt is used. Once the micro is satisfied that the overcurrent condition is over it resets the latch allowing the driver back on. This part of the chopper is all under software control. Glad to answer any questions you might have.

Greg

[pminmo]

Greg,
I would think since the HC08 has no hysteresis, it would be questionable near threshold voltages? Plus the charge time constant cause by the motor winding inductance wouldn't be that fast?

Phil

[tachus42]

Greg are you any closer to presenting you design? I'm sure we would all like to see it.
The intention with the latest design was not to present the most elegant design but instead one which sticks with very common parts. The phase generaor is a good place to use a micro but a logic design maybe less offputing to those with no experience with them. The HC series will output consideribly more current than 6mA around mid rail which is where all the gate charge is, but AC is always a option. I need to look at the drive arrangements closer before finalising the design.

Did you look at my 3 axis Servo design? Driving a servo is much easier and less messy than a stepper.
The AVR comparitor has a propagation delay of about 500nS and i have seen it used in plently of fast applications so that shouldn't be a problem. You should be able to get to a interrupt in less than 2 uS with a 20MHz part even using c if the full working register set has to be pushed. If the interrupt routine is simple and less registers have to be pushed it should be about half this.

[Greg Fill]

Presenting my design - I got sidetracked playing with the Cypress PSOC. If you are into micros the PSOC melds configurable analog and digital blocks with a nice fast processor. I have a working controller using the PSOC but it is not good enough. I will need to use an external comparator on the PSOC design to get a better responding chopper control. I should post my PIC 16F73 design as this is the one I am using but I haven't finished the code for the programmer. If you have a method to program the micro this isn't a problem.
The comparator in the AVR is much better than the one in the PSOC, about 10 times faster. The PSOC has very fast interrupt response as well, typically about 500 ns. BTW your discrete design is elegant and I agree discrete is good for those who have no micro experience. You will find that the 74AC parts are better for driving your logic level FETS but better yet why not just use a LOW side FET driver. 2 drivers in an 8 pin package for roughly the same price.

I did not look at your servo driver yet but the good thing is the stepper H-Bridge will drive a servo. Driving a stepper is "messy" but once it works it works well.

Pminmo
The comparator has the hysterisis. The comparator is comparing the voltage across the H-bridge current sense resistor and a micro generated analog voltage. When the voltage across the sense resistor exceeds the voltage generated by the micro the comparator output will fall acting as a clock edge to the 4013 latch. The 4013's data input is tied low so that when this falling edge occurs the 4013 Q output will go low and stay low, it is latched. This 4013's Q output is tied to one input of the 74HC08 AND gate, the other input of the 74HC08 is tied to one of the LOW side FET enable lines from micro. The ouput of the 74HC08 is tied to the LOW side FET driver. Thus when a overcurrent occurs the hardware will immediately shut off the LOW side drivers. The Q output of the latch is also fed as an input to the micro so it can see that the overcurrent condition occurred and take further appropriate actions.

[pminmo]

Greg,
Gotcha, went back and reread, missed the "and comparator".

Phil

[pigifly]

Hi all.
I suspect there is an error in the copper view for the 3 axis tachus42 Channel Discrete Unipolar Baord.
In effect, running a Design Rule Check it signal an error of clearance near the R27/R27/Probe Point.
This point is connected to GND because the route that goes from C8 to the via under IC7 is wrongly routed.
Is correct that R27/R28/Probe point is connect to GND or I'm right and there is an error ?

In the attachement there is a better ( visual ) explanation of my doubt, as my english is not so good.


Pigi

[tachus42]

Yes you are right there is a error. I actually had a small bridge there on my prototype but assumed it was a problem with the toner transfer or etching rather than the design.
Thanks for the errata.

Phil do you want to fix this in the file on your site?

[pminmo]

Tach,
I've goofed. I have a whole bunch versions of your 3 axis board. Can you update and send me the files I need to upload to the website. My bad. Configuration control is not one of my strong points.

Thanks,
Phil

[tachus42]

Phil you didn't goof it was a problem with the last version. Anyway heres the fixed files.

[pminmo]

New files uploaded, somebody double check.

Phil

[Konstantin]

I have completed almost the whole 3 axis system. BOB, 3 single axis boards, and the chopper adapter. Now when it comes to testing the chopper I dont really sure about how to do it.

Before I start hooking up everything Id like to verify the correct operation of the chopper board.
I use a 25v 30A PSU and I used two 150 Ohm in series 1 Watt supply droppers. Those get HOT.
I am trying to get 3A so I went with 2 x 1.2 Ohm, 2 x 1.8Ohm in parallel 1W resistors.

Maybe I missed something from the beginning, the chopper board will regulate the amount of Amps that my motors are receiving. My motors are 5.2V 1.4 A phase. What about the Voltage.
I thought I could use my 25V supply with the chopper board, maybe I am wrong.

Please help to clear the things for me.


Konstantin.

[tachus42]

Konstantin, the supply droppers will get hot. With 2 * 150 and 25V they will dissipating 3/8 Watt each, which is comfortably inside the 1 Watt rating. The Zener will be dissipating 1/2 watt and will also get hot.

Back a few pages in this thread foxgarden posted part of the schematic with voltages. You can check against this before you hook the motors up. Just remember to subtract the difference between his supply voltage and yours (ie 25 - 13.42) before comparing.

If the static voltages look pretty close then everything should be ok and you can go ahead and hook it up.