[xairflyer]

Satisfaction is always the best reason, I have a few nice bipolar motors so might have a go at this myself, it will give me an excuse to make a PCB.

[Garfield2]

Hey Gang,

Just finished the input opto board for the "opto-bob".

Have a lookie, and let me know if you see any errors (chair)

Cheers,
Alan.

[da21]

Hey Gang,

Just finished the input opto board for the "opto-bob".

Have a lookie, and let me know if you see any errors (chair)

Cheers,
Alan.

looks ok , can't see any errors Alan

[santiniuk]

I need some moral support....

In summary No joy with the Opto Board. I have fitted the 'fix' resistors as specified but when testing it the leds blink in a random manner and the steppers dont step although they do 'squeak' but no real movement.

I have checked the board for errors and cannot find any. The 5V rail is stable on all ic's.

I was hoping the problem was due to the use of a laptop but so moved a pc into the garage and still no improvement.

To make matters worse I then tested the original interface board and one channel is not working. It's not the picstep driverboard its somewhere between the pc and the interface.

To be honest I have not had a lot of time to look at it due to other commitments.

Ahhhh.. feel better getting it off my chest...

[Garfield2]

I need some moral support....

To make matters worse I then tested the original interface board and one channel is not working. It's not the picstep driverboard its somewhere between the pc and the interface.

To be honest I have not had a lot of time to look at it due to other commitments.

Ahhhh.. feel better getting it off my chest...

Hey dude, sorry to hear about your bad luck. Seems very strange the old interface has stopped working too... Could it be a cable problem?

Alan.

[da21]

sounds very much like a parralell port to opto or logic level problem , so theirs not a clean signal to fire the opto's , or indead they are no shutting of completely try dropping changing series resistor or leaving it out alltogether , and also check the ground return from the board to the pc and make sure theirs a good ground , also stick a few .1uf caps across the power rails

if you have a scope handy let me know what the voltage level is like at the input to the opto , both when the line is held high and pulled low

or even too much noise floating around from the mains transformer would do it to , theres to many variables at the moment we need to narrow it down .






Dave

[10bulls]

Here is a trace at pin 6 of one of the optos during a step pulse, from a (fingers crossed) working opto board.

As you can see, the pulse is pretty ramped for a 10uS pulse (which is what my mach2 is kicking out). Much shorter than 10uS and the inverter probably won't fire. Interestingly I had a problem with an intermittent 'pulsing' on the second channel. This appears to have been caused by having a Mach2 min pulse width of 15uS (or above)...which it does warn may cause locking.

I *think* the value of the pull up resistor from pin 6 to 8 of the optos will influence the angle of this ramp. I'm still using 470R.
You seem to know yer onions Dave, whats your view on this? I agree with you about filter caps...couldn't hurt!

One more thought. I have been testing mine with an old 6 wire printer stepper. I knocked the steps per unit way down (100) and speed low (10).
The stepper won't hack much more than this (I running it at 9V for testing).
Maybe this slow speed is helping mine work ?!
Oh, and the outputs from mach2 should be inverted.

As a process of elimination, here's what I would try...

1. Feed some test signals into the opto board from a little test jig. (Eliminate
the PC from the loop). Just try putting GND on pin 24 and then with a test lead feed 5V and 0V into pins 2,4,6,8 (the direction pins) of the input connector. At 0V, the leds should light, 5V...go out. This should also take high frequency out the equation. If there is no joy here then there must be something pretty wrong...shorts, dry joints, blown chips etc etc

Try the same with pins 3,5,7,9. Similar to above, but when you high to low the input the step led should pulse once.

2. If all the above works then I suspected it's the frequency that's the problem.

[da21]

Here is a trace at pin 6 of one of the optos during a step pulse, from a (fingers crossed) working opto board.

As you can see, the pulse is pretty ramped for a 10uS pulse (which is what my mach2 is kicking out). Much shorter than 10uS and the inverter probably won't fire. Interestingly I had a problem with an intermittent 'pulsing' on the second channel. This appears to have been caused by having a Mach2 min pulse width of 15uS (or above)...which it does warn may cause locking.

I *think* the value of the pull up resistor from pin 6 to 8 of the optos will influence the angle of this ramp. I'm still using 470R.

You seem to know yer onions Dave, whats your view on this? I agree with you about filter caps...couldn't hurt!

One more thought. I have been testing mine with an old 6 wire printer stepper. I knocked the steps per unit way down (100) and speed low (10).
The stepper won't hack much more than this (I running it at 9V for testing).
Maybe this slow speed is helping mine work ?!
Oh, and the outputs from mach2 should be inverted.

As a process of elimination, here's what I would try...

1. Feed some test signals into the opto board from a little test jig. (Eliminate
the PC from the loop). Just try putting GND on pin 24 and then with a test lead feed 5V and 0V into pins 2,4,6,8 (the direction pins) of the input connector. At 0V, the leds should light, 5V...go out. This should also take high frequency out the equation. If there is no joy here then there must be something pretty wrong...shorts, dry joints, blown chips etc etc

Try the same with pins 3,5,7,9. Similar to above, but when you high to low the input the step led should pulse once.

2. If all the above works then I suspected it's the frequency that's the problem.

I Would have used 330R but 470R is near enough , your quite right with your thoughts as of course the saturation level for the led to be fully on is going to be controlled by the pulse , so i can understand it may need to be changed to suit pc's or even laptops if it comes to that to help in fully turning on the led which in turn will fully saturate the darlington array to give a nice clean signal ,

and yes it may be that a few more decoupling caps would be an idea or at least placment for them ( alan )

dave

[santiniuk]

Really appreciate the details of your post 10bulls. Thanks.

But.... I had about 30 mins tonight to check things and I'm still no further forward.

I checked the pcb out at work today and i'm quite confident that its assembled correctly. I traced through the complete pcb !

My symptoms are still the same.

I have a scope thats in the loft so suspect this may have to be called on as I'm working blind now.

I'll try some of the suggestions 10bulls has posted and bypass the opto's and try manually linking etc.

So I have tried a desktop pc and a laptop.
I'm using Mach2 and tried various pulse widths on motor tuning
I have tried two 470 ohm resistors in parallel on the opto drive side
The pullup resistor pin 6 - 8 have tried 470 Ohm, 560 Ohm and 4.7K Ohm

One good thing is that my original interface now works fine. I had stupidly used a wrong header for a channel.

Anyone else 'lurking' who has this up and running ?

Thanks

[Garfield2]

Anyone else 'lurking' who has this up and running ?

Man I never thought it would be this complex to get this to work! What type of 74x14 are you using? Are they HC? The resistor between 6 and 8 needs to be set to the right value for your part else the loading on the darlington pair inside the opto isn't correct. The 6N139 is a high speed opto and I pick it as a part to deal with the high frequency pulse trains from the PC parallel port. If your parallel port(s) aren't supplying enough power to fully saturate the opto LED then that maybe the big cause of your problem (but I really doubt it).

The output of the opto only has to get vaguely near the upper edge of a '1' signal and the schmitt-triggered inverter should clean up the rest of the signal. Have you tried increasing the pulse width in Mach2 to something supremely huge (1ms or more) and see if that works?

Another question, how long is the cable between the PC and the board? Is it all ribbon cable?

Cheers,
Alan.

[santiniuk]

Yeh, fun isn't it

I'm using 74HC00N and 74HC14N IC's. If I'm reading the datasheet link that you posted I should be using 560 Ohm resistors.

The opto's are 6N139

The parallel port cable is a ribbon cable, length 1 metre.

Except the pulse width trials, I have not changed anything else in Mach2 which works fine on old interface.

I will try a super huge pulse in Mach2.

Back to work.....

[Garfield2]

Yeh, fun isn't it

I'm using 74HC00N and 74HC14N IC's. If I'm reading the datasheet link that you posted I should be using 560 Ohm resistors.

The opto's are 6N139

The parallel port cable is a ribbon cable, length 1 metre.

Except the pulse width trials, I have not changed anything else in Mach2 which works fine on old interface.

I will try a super huge pulse in Mach2.

Back to work.....

Weird.. almost exactly the same setup I have, in TurboCNC I just left the pulse width to the defaults (2us I think from memory). You might have to crack out the CRO and have a look at both sides of the opto's unless something *funky* is goin' on with ya 74HC14's.

Cheers,
Alan.

[10bulls]

Mine stops working with a pulse < 7uS in Mach2.
(Edit) ...just retested and is now working down to 5uS

I'm trying to get my head around the datasheet for timings...

My reading is 'Propagation delay time to logic high at output'
For RL=270R, typical = 1.5uS, Max=7uS
For RL=4K7, typical = 4.0uS, Max=60uS
So seems lower value resistor improves speed.

Speed also dependent on current through led, so lowering the series resistors
may help with speed as well.

[Garfield2]

Mine stops working with a pulse < 7uS in Mach2.
(Edit) ...just retested and is now working down to 5uS

I'm trying to get my head around the datasheet for timings...

My reading is 'Propagation delay time to logic high at output'
For RL=270R, typical = 1.5uS, Max=7uS
For RL=4K7, typical = 4.0uS, Max=60uS
So seems lower value resistor improves speed.

Speed also dependent on current through led, so lowering the series resistors
may help with speed as well.

Indeed that's how I read it too.. I had a 220R in mine, but the series resistors I left a 470R because I had no problems.

Interesting.

I wonder what pulse widths you really need for rapids? Maybe playing it safe and pushing the widths up to 10uS in Mach 2 might be worth doing?

Alan.

[santiniuk]

Well I'm losing the plot now....

No time at home to do much so I worked through my lunch time and scoped up 'Opto-Bob'.
In my typical slack manner I forgot to take the parallel lead with me so could not test when equiped to a pc.

In summary I used a function generator to feed a square wave into each input pin and scoped the output at various points.
All of the opto's are using 560 Ohm pullup's and I'm getting perfect output from pin 6. Then traced through the inverters and output again is perfect.

Using a simple header and LED I checked led outputs and again all work fine. Finally checking the ouput headers I'm getting a good signal.

So I'm now happy that the actual assembly of the board is ok and somewhere it's timing or parallel port related.

Back at home I did a quick test with Mach2 and increased pulse width as high as it can go and looked at the opto outputs. Sure enough pin6 is toggling state depending on step direction. (Typical !) - but it still doesn't move the motors.

I really need to brave the cobwebs in my loft and get my scope. I need to look at the opto output in stepping mode as everything else seems correct.

A couple of pictures attached to show the rats nest in action !

It's all good fun.....

[10bulls]

Bummer!...you must be reading the scope wrong!...Tell you what...Ship that scope down to me and I'll check it out for you....shouldn't take more than six months or so.

I didn't test mine to HUGE PWs in Mach2...this may cause more harm than good. I would have thought 15uS was ample. I would definely lower those 560R pull ups though. Try 470, 330 even 220.

Are those Phillps 74HC14's? I read the threshold voltages are around 0.4V lower than the Fairchilds I'm using...If anything I would have thought that would make them better for higher frequency ?!?!

Could you try connecting a picstep and motor up and pulse it with the frequency generator to see if a) it works and b) what pulse width it will hack.

Good luck!

[JavaDog]

Bummer!...you must be reading the scope wrong!...Tell you what...Ship that scope down to me and I'll check it out for you....shouldn't take more than six months or so.

You really need an O-scope? I have two spares in need of repair (should be simple if you know your electronics). I would give them to you...shipping to UK might be a bit much though...

[10bulls]

You really need an O-scope? I have two spares in need of repair (should be simple if you know your electronics). I would give them to you...shipping to UK might be a bit much though...

You are a gent! Tell you what...if they're 4 channel Tektronix 300MHz jobs I'll fly over and pick them up!

[Garfield2]

You are a gent! Tell you what...if they're 4 channel Tektronix 300MHz jobs I'll fly over and pick them up!

I've already got my bags packed!

[Garfield2]

Well I'm losing the plot now....

No time at home to do much so I worked through my lunch time and scoped up 'Opto-Bob'.
In my typical slack manner I forgot to take the parallel lead with me so could not test when equiped to a pc.

In summary I used a function generator to feed a square wave into each input pin and scoped the output at various points.
All of the opto's are using 560 Ohm pullup's and I'm getting perfect output from pin 6. Then traced through the inverters and output again is perfect.

Using a simple header and LED I checked led outputs and again all work fine. Finally checking the ouput headers I'm getting a good signal.

So it's not a lemon then it works in itself, it just doesn't play well with others!

So I'm now happy that the actual assembly of the board is ok and somewhere it's timing or parallel port related.

Back at home I did a quick test with Mach2 and increased pulse width as high as it can go and looked at the opto outputs. Sure enough pin6 is toggling state depending on step direction. (Typical !) - but it still doesn't move the motors.

So you're seeing pulses from Mach2 on the driver output headers? I don't understand what you mean by "Sure enough pin6 is toggling state depending on step direction". More importantly, do the status LED function correctly?

I really need to brave the cobwebs in my loft and get my scope. I need to look at the opto output in stepping mode as everything else seems correct.

A couple of pictures attached to show the rats nest in action !

It's all good fun.....

Nice setup, I'd kill for a scope like that..

Alan.