Advanced Search

They are related, but not the same. You connect them together in the hal file, using the metalmusings_encoder comp.
Best is to look in the EaserCAT-2000 folder for how it is done there and replicate that, and adapt to your variable naming.
I see these lines in Turner.hal over there

As mentioned before, you compile the module with
Uncomment also this line in main.cpp

Thank you from the bottom of my heart. You are my hero.

tommylight wrote:

Looong thread about lathe macros here
forum.linuxcnc.org/41-guis/26550-lathe-macros
 

Long is good, lots of information.
This looks like exactly what I was hoping for.
LinuxCNC seems to have every possibility covered, just thrilled I decided to go this direction.

Thanks to all the contributors.

7I95T firmware for 7I95T and a 7I74 on P1

(note 7I95T sserial channels are port-0 and 7I74 channels are port-1)

 
 

Whether you can wire your DRO in parallel with the 7I95T encoder inputs depends on 2 things:

Do the encoders have differential outputs?

and if they are differential, are the DRO inputs terminated?

The 7I95T encoder inputs (when in differential mode) are terminated (120 Ohms nominal)
so if the DRO is terminated also this may be a pretty heavy load on the encoders.
If this is the case one option would be to buy an encoder splitter for each axis, like Mesa's ENCY
or make your own with a couple RS-422 driver/receiver IC's

On the encoder, you can use a single "A" input for counting frequency/RPM, you just need to set the
encoder into "counter" mode in the hal file



 

Now I changed the new encoder, it works fine both cw ccw. I use oscilloscope to check ABZ, it's complete.
But I still can't find the index working in Hal.
And another question, the pin name in .comp and the name in pdo are different. Are they related?

I like that. Great idea. I have no idea if I can so I see the DRO coming to bits shortly.

Impedances? The most horrible thing I have had to do recently was to try to explain the Smith Chart display on a cheap VNA (aka  network analyser) to an older radio ham who could just about manage that 50 and 75 are different. It scanned the range of frequencies you gave it as about twenty discrete points and then interpolated a line in between. Great for almost all real problems on a 3" display but if you set in to scan from DC to the speed of light on a piece of wire hanging down the garden you get a bunch of virtually random points with straight lines criss-crossing the display. Very pretty but...

That said I use a nanoVNA to match my piece of wire hanging down the garden.

Looong thread about lathe macros here
forum.linuxcnc.org/41-guis/26550-lathe-macros

Are the encoder outputs differential or single ended?
If differential, i would use one output for whatever it is being used now and the other to Mesa 7i95T.
Be aware that everything connected that way must have the same 0/gnd/be on the same potential.
I have one such system in use for exactly two years!

nigelh wrote:

However being a bit of an electronic engineer (retired industrial physicist) I worry about loads and impedences.
The 7i95t manual section on the high speed encoders ofers me:
   DIFFERENTIAL MODE IMPEDANCE the min=118 max=122 Ohms
and I'm perplexed. I suspect I am thinking of something different.

Have you ever tried to explain to an inexperienced electronic guy, how the old antena cable (RG58) was 75 Ohm, and the security camera cable (RG59) was/is 50 Ohm?
I've gotten some very strange looks when i tried that... It all seems like black magic to them, till they dig deeper into the subject.

A scope is cheaper than any Mesa card, the MiniWare DSO211 is ~40$ and 212 is ~60$ for two channels.
To "safely" use the Mesa field inputs, 7i84 set in Mode 2 should have 4 analog inputs (00-03) at low resolution and low speed, add a resistor to each input. Probably will not be enough for what you need.
Using FPGA inputs as logic analyzer with some resistors is OK, but care must be practiced as those inputs are very sensitive, and adding zener diodes to each input is advisable. These inputs are fast enough for most uses, are 5V tolerant but with 0-3.3V range.
Any of those will not be very usable for finding interference as all above are DC coupled, whereas you would need AC coupling to measure interference.
And to make things worse, interference from cheap switching power supplies and motor drives and VFD's can in some cases be over 60VPP (volts peak to peak), meaning it can easily destroy any of the above setups.

It is only A and B for counting. You should probably verify with an oscilloscope that the signals from the encoder look like a nice quadrature signal.

RotarySMP wrote:

If you install Andy's "Lathe Macros", you get a set of canned cycles to do all the typical lathe ops, including threading with simple conversational interface. I beleive it works on most of the GUI's.
github.com/andypugh/LatheMacros

I of course use Gmoccapy as I like it modal nature.

 I will definitely check this out. I don't suppose there is a demonstration video? 
I'm trying to visualize this, especially WRT the axis GUI ( perhaps that's the problem?), and if I need physical buttons how many?
 If Axis is not the way to go then need to think about touchy vs. Gmocapy.

My hardware can't get here fast enough.

Move it to plasma section or to show your stuff?

If you install Andy's "Lathe Macros", you get a set of canned cycles to do all the typical lathe ops, including threading with simple conversational interface. I beleive it works on most of the GUI's.
github.com/andypugh/LatheMacros

I of course use Gmoccapy as I like it modal nature.

Now I unplug the TXS0108EPWR and connect it directly to mcu with R1.5k pull up to 5V. it work 

I can find the counting up when the motor shaft is rotated, but sometimes when I let go of the rotation the numbers will continue to count until the shaft is rotated again.
Another problem is that the numbers will only count one way even when the motor is rotated in reverse direction. This may be due to a faulty Z phase for the index. I am investigating further.