Hi Bojan,
Which chip are you using?
I have just ordered a couple of A4988's to experiment with.
What are your thoughts on the Allegro's? It seems some people have been having problems with missed steps at low speeds and low currents while microstepping at x8 and x16, tho apparently the A4988 handles these conditions better than others like the A3977/79's
Audio adapters arrived today - Fake SJ588 (blue transparent case), 4 pieces.
Because it is better to use two adapters of different varieties (so the w' doesn't confuse AR and DEC), I ordered additional 4 (grey cases).
Tomorrow I'lll have a look at waveforms.. however Maciel told me the common is actually at 2.5V respective to PC ground, so it will not be a problem to design the electronics for NJM3771 (full wave rectifier (for input voltage to NNJ3771) with OPAMPs and comparator ( to generate Phase direction signal for NJM3771).
The concept schematic is attached, however there is some work to be done in relation to output level shift and components values need to be finalised.
When I am finished with this, I will post the final schematic.
Maciel, I mean something that show a component layout on a a piece of stripboard like the diagrams on this page: http://www.kpsec.freeuk.com/stripbd.htm That way electronics dummies like me can simply put a component where it has to go rather than having to build a rats nest of wires . I'm sure the electronics parts are what stops lots of people from using your system. some of us need the equivalent of a "paint by numbers picture
I had missed that link....will start building whilst awaiting arrival of USB dongles....
Don't start yet, first check the SJ588 chipset and output voltages. 2 years ago every SJ588 had TP6911 chipset, but now there are TP6911, CM109 and CM119 models, with different output voltages' behaviors.
Hi all,
I checked the waveforms on my "fake" SJ588, and they seem OK (http://www.ebay.com.au/itm/USB-2-0-3...item19c8d67d39).
(BTW, it is not quite what was advertised - it has only one red LED, and on ebay photo I can see two.. but it doesn't really matter).
Reference (2.5V) is at left side of the large ceramic cap placed between audio sockets, as marked on the photo (other side is GND).
Now, based on that I did some simulations of the circuit for 3771 driver, see attached. I will make the protoboard using this circuit and see how this behaves - I am pretty confident it will be OK.
This circuit produces two waveforms: one to modulate the current in one motor coil, another one is logic level for direction of the current though the coil (this signal is required as input into 3771 stepper driver).
The interface schematic for RA (we need another one for DEC) is also attached.
As for step-dir mode of operation, this is straight-forward - just a couple of Schmitt triggers will do, the pulses are already referenced to PC GND.
Waveform straight out of audio device is not really good, you need to pass it to a schmitt trigger, preferably HC-type (if you are using C-MOS ... HC has threshold around 2.5V (for VCC 5V), while HCT, has it much lower - so noise from audio device (there is some) may trigger it and then you will have multiple pulses. Or, you should use comparator (opamp with some positive feedback to have small hysteresis for noise immunity).
You have to have a look at waveform with oscilloscope to be sure what is really going on.
I checked the waveforms on my "fake" SJ588, and they seem OK (http://www.ebay.com.au/itm/USB-2-0-3...item19c8d67d39).
(BTW, it is not quite what was advertised - it has only one red LED, and on ebay photo I can see two.. but it doesn't really matter).
Reference (2.5V) is at left side of the large ceramic cap placed between audio sockets, as marked on the photo (other side is GND).
Yes, seems identical to my "fake" (a CM-109 chipset). But the output voltage swing of mine are from 0.38V to 3.91V, centered at 2.13V
You can measure the impedance of "ref" line? If is high (>100 ohms), is not suitable for my analog circuit "for SJ588", being preferrable use the "for HY544" version (and replacing "common" to "USB ground").
Yes, seems identical to my "fake" (a CM-109 chipset). But the output voltage swing of mine are from 0.38V to 3.91V, centered at 2.13V
You can measure the impedance of "ref" line? If is high (>100 ohms), is not suitable for my analog circuit "for SJ588", being preferrable use the "for HY544" version (and replacing "common" to "USB ground").
Maciel
I will try and let you know as soon as I i can (today is Saturday, and my wife wants me to do other things today and tomorrow - so this will have to wait until I escape to freedom of my lab at work ) .
I connected the driver in the USB dongle using a P2 plug, but this is not the correct way to connect it. I need use the USB ground!
Using the "common" output plug we have a oscillation of 2 volts in this "false ground". This probably cause lost steps ans a lot of noise. Today I will test it using the USB ground.
I will let you know about the tests...
OK, I've got it..
It seems that there is variety of those dongles on the market (after all, Maciel mentioned this on his website, but I fully appreciated what he wrote only after I measured my own specimen and when I saw what I saw).
Attached is the screen shot of step pulses from my dongle, there are some artefacts from (built-in) digital filter visible (ringing), so pulses are not clean - but they will be OK after Schmitt trigger.
BTW, the second batch of dongles arrived from ebay this morning (grey ) and they are the same as the first, with blue case.
Interestingly, the waveform depends on OS running - the second shot is from w7 (the first one is from laptop running xp) - most likely because I didn't configure the audio driver properly in w7, so the OS is interfering with SoundStepper). Please note, the second shot is scaled as 0.5V per division (GND is at the bottom of the screen graticule) and in both cases, the audio level was set to 120%.
Anyway, this is all very good - the whole system - telescope, guiding, ASCOM, star charts- all running on a single desktop. And all of that is freeware.
Of course, driver electronics must still be built, but with those Pololu or StepStick PCB's around, this is not a problem...
Finally, here is my board (with Pololu drivers), at the moment controlled with SoundStepper (This board was originally designed as Bartels interface for EQ-6, that is why there is that 26-pin IDC connector on the left side, plus additional connectors for keypad and focuser).
Motor on the picture is 1.8° step/rev, and it works at maximum speed SoundStepper can deliver with no problems observed.
However, the current adjustment is critical: if too low, microstepping disappears, if too high, then the movement becomes jerky for some reason.
Later I plan to fit the heavy wheel on the motor shaft, this should help to further smoothen the movement.
Very happy !
Hi Rod,
The noise seems to be minimal because the microstepping PWM frequency is higher than I can hear (~50kHz - this depends on inductance of the motor coils, of course).
Max speed from SoundStepper is around 2k pulses per second, resulting speed is similar to scope.exe (I am sure motors can do better than that, since here I am using high current, low voltage bipolar steppers, while on my dob I have steppers from old 8" floppy drives, operating at 36V or so.. and the PWM from scope is audible.
Also,please note, mechanical reduction will be different - much lower so slewing will be faster than my dob - so here I have 15° for 30 sec (half degrees per sec).