Has anyone made bartels or Ek's box or similar

[hikerbob (Bob)]

On the subject of encoders I've found the page for US digital encoders http://www.usdigital.com/products/en...tary/shaft/s6/

The S6-2500-250-I-S-D (2500 CPR) is $75.87US and I think you can get 10K out of that by sensing changes of state in both channels. I don't know how that would work with the Bartel's circuit's.

I've done some play with the predecessor to that encoder and a damaged Arduino which only had one external interrupt working and could read off 5k positions per rev with a simple piece of code. I think that if I ran interrupts on both channels for a change in state 10K resolution could be achieved.

These also come with an index line but I've only played with that briefly and with limited success but I suspect that if I triggered an interrupt from a change in state in the index line it would be work well.

The Arduino code snippet I used to handle an interrupt on channel A from the encoder and alter a step count is as follows.

void StateChange() {
Ch_a_state = digitalRead(Channel_a);
Ch_b_state = digitalRead(Channel_b);
if ( Ch_a_state == Ch_b_state ) ++StepCount;
else --StepCount;
}

[mswhin63 (Malcolm)]

Originally Posted by hikerbob

On the subject of encoders I've found the page for US digital encoders http://www.usdigital.com/products/en...tary/shaft/s6/

The S6-2500-250-I-S-D (2500 CPR) is $75.87US and I think you can get 10K out of that by sensing changes of state in both channels. I don't know how that would work with the Bartel's circuit's.

I've done some play with the predecessor to that encoder and a damaged Arduino which only had one external interrupt working and could read off 5k positions per rev with a simple piece of code. I think that if I ran interrupts on both channels for a change in state 10K resolution could be achieved.

These also come with an index line but I've only played with that briefly and with limited success but I suspect that if I triggered an interrupt from a change in state in the index line it would be work well.

The Arduino code snippet I used to handle an interrupt on channel A from the encoder and alter a step count is as follows.

void StateChange() {
Ch_a_state = digitalRead(Channel_a);
Ch_b_state = digitalRead(Channel_b);
if ( Ch_a_state == Ch_b_state ) ++StepCount;
else --StepCount;
}

There the ones I am using, currently with my gearing I am getting 22,800 res. I am considering increasing the gearing and operating only single channel instead of quad. I hate the fine tuing of the backlash but I am only in experimental stage ATM.

[bobson (Bob)]

I just tested my EK box with Hyper Terminal and it works fine.

What I would like to ask Bojan is what sort of connection did you use to connect encoders, since the order of connectors is not the same as on EK board. I can solder it but I really would like to make some sort of a male/female plug. Went to Jaycar and asked them but the guy said they have nothing like it. And yes I had encoder with me so he could see it.

Could not get pins connections for encoder to solder for the EK board, the holes are too small and again they have only one size which is too big
So direct wire soldering is the only option and then some sort of plug in between.

cheers

[bojan]

Originally Posted by bobson

What I would like to ask Bojan is what sort of connection did you use to connect encoders, since the order of connectors is not the same as on EK board.

I was using crimp connectors, similar to those available from RS components:
http://australia.rs-online.com/web/s...577+4294858891

http://australia.rs-online.com/web/s...duct&R=0334561

http://australia.rs-online.com/web/s...duct&R=0467605

It is easy to swap wires as the wires are crimped (or soldered) individually.

The holes on PCB can be enlarged with drill 0.8mm dia.


Yes, sadly, Jaycar and DSE are not what they used to be... they even do not know what they have in their stores any more..

[mswhin63 (Malcolm)]

Originally Posted by bojan

I was using crimp connectors, available from RS components:
http://australia.rs-online.com/web/s...577+4294858891

http://australia.rs-online.com/web/s...duct&R=0334561

http://australia.rs-online.com/web/s...duct&R=0467605

It is easy to swap wires as the wires are crimped (or soldered) individually.

Jaycar and Altronics have similar:

http://www.jaycar.com.au/productView...4&form=KEYWORD

http://www.altronics.com.au/index.as...=item&id=P5494

[bobson (Bob)]

Bojan,

Old (and not so old) printers and scanners may be a supply for those things..
Timing belts used for moving printing head (30cm in dia) could be found in them.

Can you tell me which printers or scanners have those kind of timing pulleys/belts so I can search for them on auctions or for sale classifieds.

Thanks

[bojan]

I found Epson Stylus 580 on the road the other day (someone throw it away) and it had timing belt on the printer head, ~30cm in diameter - I think ideal for the purpose. The timing pulley was on the stepper motor shaft.
The tricky part was to remove the glue that was holding the belt in place on printer head and not to damage the belt itself.
I think majority of today's printers and scanners will also have them, as this is proven, accurate and cheap solution for linear moving parts (print heads, scan heads etc).
You can also purchase those things at Small Parts (http://www.smallparts.com.au/hobby/.. a bit hard to navigate this website..) or here: http://www.sdp-si.com/Press/New/Belts2_5.htm

[bobson (Bob)]

Sorry mate, I wasn't clear when I asked you a question

I have timing belts from my Canon printer, they are approx 30 cm in diam. What I need is a pulley for that timing belt. What would be the best material for it, any hints would be helpful.

This bit:

The big wheel (or pulley) does not have to have all the teeth, only couple of them are enough (say, every 50th or 100th tooth is enough to keep sync and to prevent micro-slippage). Those teeth could be made by gluing them on the perimeter of the big pulley (appropriately dimensioned of course), using timing belt as template

thanks mate

[hikerbob (Bob)]

I've pulled a couple of Multi-Function printer/scanners to bit's recently for play parts. Both have had a couple of long timing belts, one for the print head and the other for the scanner part. One Canon, one Brother. The belts have been attached to the components they move by clamps which they are threaded through - remove the tension and they slip out. The scanners have had stepper motors but the print head drive's have been conventional motors with optical encoders.

One unit was given to me the other was purchased for a couple or dollars from the recycle market at a regional tip.

Bob

[bobson (Bob)]

Hi guys,

Yes, the belts are not the problem here, they are easy to source out. I have two of them already.

But I need the pulleys for them. Especially the big ones that goes on Alt and Az on my 12" dob. Bojan mentioned he glues teeth using belts as template but on what? What do you guys use to make this pulleys?

cheers

[bojan]

OK..
As a first option, I plan to use a round aluminium plate (wheel), and machine it such that the belt tightly fits around the perimeter, but inverted, with teeth pointing outwards.
This way I will have a gear (with rubber teeth). the other gear (actually timing pulley) that goes to encoder shaft should be taken from stepper motor as it has the same distance between teeth (pitch, module).

The second option is again aluminium plate 4-6mm thick, machined to a smaller diameter than timing belt, with belt attached in conventional way (teeth in contact with the wheel). This should be OK as such, however to prevent any possibility of (micro)slipping, I plan to glue only couple of teeth along the wheel perimeter, which will mate with every 20th or 100th valley on the timing belt. Timing belt will drive the encoder shaft in conventional way (pulley from the same source - mating stepper motor).

I tried this concept on much smaller wheel and it worked: there was no slippage or mis-synchronisation between belt and this type of pulley.
Unfortunately I do not have any pictures..

Originally Posted by bobson

Sorry mate, I wasn't clear when I asked you a question
I have timing belts from my Canon printer, they are approx 30 cm in diam. What I need is a pulley for that timing belt. What would be the best material for it, any hints would be helpful.

[bobson (Bob)]

I made everything and hooked up to the computer and now encoders are not working
I checked them before with Hyper terminal and everything was fine. When pressed Q it responds fine. But when I turn encoders a bit and press Q again nothing happens, the same +00000 appears, its suppose to change the value like it used to when I tested them first time.

I followed EK troubleshooting, measuring voltage on pin 14 U 2 and its all fine, encoders are connected properly, checked it numerous times. Encoders are getting 5 V. Reheated all joints as recommended by EK and still nothing.

Just when I sorted out gearing and belts and now this
What could gone wrong here?

bob

[bojan]

That looks like encoders are not sending anything. Or nothing is being received on processor side.
Do you have a CRO?
If not, try to measure voltage on A & B with multimeter. When encoder shaft is not moving, you should read 0V or 5V, this will depend on shaft position.
When the shaft is moving (reasonably fast), the voltmeter should show ~2.5V.

Also, try to reset processor.


Are you typing "Q" or "q"?

[bobson (Bob)]

Hi Bojan,

I don't have CRO, I have multimeter.

Yes I know difference between q and Q, I do type Q and it used to respond when I was testing encoders first time but now no change.

I dont have reset button, I just unplug and plug the battery, this should reset processor?

I just noticed that supply is 5v but channel A and B are either 0.02V or if turned a bit it changes to 4.95V on both A and B channels? Is this normal, I thought it would be alternating the voltage between A and B channels as it turns?

cheers

bob

[bojan]

I think I had that problem once, after processor reset it was back to OK.
I assumed there was a problem with noise.
Also, it may be that you have grounding issues (RS232 cable).
Something is definitely different from when you tested this before... try to think what it may be.
Are you testing now with Ek mounted on telescope?

[bojan]

Reset should be done after power-up, but if the reset cap (10uF if I remember correctly) or associated resistor is faulty, it will not reset.

0.02V or 4.95V is perfectly OK.

[bobson (Bob)]

NO, I took encoders off and even soldered one of them directly on EK board just to make sure there is no some sort of attenuation in cable or interference but still the same.
There are 5 of 10uF capacitors on the board, it could be one of them holding electricity or discharging when its not suppose to?

I'll try to measure capacitance of them and see what happens.

How to measure capacitance with voltmeter, I do have that option on multimeter but if I try to do that while connected on circuit the voltage regulator get really hot!

[mswhin63 (Malcolm)]

Originally Posted by bobson

I thought it would be alternating the voltage between A and B channels as it turns?

No they alternate in the middle so there is 4 different levels. Quadrature works on level change not level it self, so if you are getting level change then they are working.

00 A and B are low next B will change
01 to produce this,
11 and so on
01

The above is called gray code typically for this type of application allow only 1 level change at any time.

[mswhin63 (Malcolm)]

Originally Posted by bobson

There are 5 of 10uF capacitors on the board.

I can't imagine the need for these capacitors on the A B side of the encoders only power side for decoupling.

[bobson (Bob)]

Now the voltage regulator is getting very hot every time I plug in the battery? It started even smoking a bit so I unplugged it immediately. Could it be voltage regulator or something prior to that?