[ZincBoy]

With my recent lathe CNC conversion, I used G320X drives and wanted to be able to use the fault output. The advantage of the fault output is that the system can be programed to shut down other axis when one gets too far out of range (faults). This has saved me a few times when I had a crash

The attached image is of the circuit I used to connect the Err/Fault line to the controller. The circuit is used to provide electrical isolation as well as level conversion. As the Enc_VCC and Enc_GND on the G320X is tied to the Servo VCC/GND I wanted to isolate the grounds as well as the power.

The circuit also has a provision for connecting the encoder index pulse back to the controller for homing.

J1 is used to connect to the G320X.
J2 is used to connect to the controller. In my case, the controller IO voltage was 12V (Field_VCC). ERR is an output to the controller and will be low when the drive is faulted or not started. RESET is and input from the controller and is driven high for 1 second during startup to initialize the G320X

U1 is an Analog Devices isolation chip. I used it as I had a bunch lying around and I did not want to load the ERR/RESET pin with an opto-coupler. It transmits the signals from the G320X to the controller. The VDD2 side of this IC can operate from a 5V to 18V supply level.

U2 is an opto-coupler used to pull the ERR/RESET line high to start the G320X. R2 limits current through the input of U2 and must be adjusted for your Field_VCC level.

[Craig Smith]

Hi ZincBoy, thanks so much for putting up your Gecko reset circuit 31/12/ 2012, I am looking to achieve the same thing as you did. Might be a good idea to just copy yours I think..
Not being an electonics guy, I struggle with full understanding, so I hope you don't mind if I ask a quetion or two?
R1 seams to have a value of "0" Is that correct and if so, why Zero?
There is no INDEX input on the Gecko, so I assume that the Index Terminal (no.4) on your board is there purely to accept the index pulse output from the encoder?
I have some PC817 optos and was wondering if it would be okay to use that in place of the TLP127 as used by you?
Thanks agian ZincBoy,
Regards,
Craig Smith

[ZincBoy]

Hi ZincBoy, thanks so much for putting up your Gecko reset circuit 31/12/ 2012, I am looking to achieve the same thing as you did. Might be a good idea to just copy yours I think..
Not being an electonics guy, I struggle with full understanding, so I hope you don't mind if I ask a quetion or two?
R1 seams to have a value of "0" Is that correct and if so, why Zero?

Feel free to copy the circuit. That is why I posted it

Yes, R1 is 0ohms. R1 was there in case I needed to limit current through the opto-isolator. It turned out to be not required so it is just set to zero ohms.

One thing I notice now, is that the R1 and R2 labels on the PCB layout are swapped. An error on my part and I never noticed as the PCB I had made does not have the silk screen on it. The schematic is correct and as long as it is followed the circuit will work.

There is no INDEX input on the Gecko, so I assume that the Index Terminal (no.4) on your board is there purely to accept the index pulse output from the encoder?

This is correct. I wanted the index pulse from the encoder available to LinuxCNC so I could have better repeatability when homing.

I have some PC817 optos and was wondering if it would be okay to use that in place of the TLP127 as used by you?

It looks like a PC817 would work as it just has some minor differences in Vf and If. It is also not a darlington photo transistor so it won't switch quite as hard but that is not really a concern in this application.

If you are using a 12V Vfield, I would drop R2 to 1k ohm to compensate for the higher Vf and If in the PC817.

[Craig Smith]

If you are using a 12V Vfield, I would drop R2 to 1k ohm to compensate for the higher Vf and If in the PC817.[/QUOTE]

Thanks so much ZincBoy, you are a champ. Thank you for taking the time. All a bit new for me, so learning curve is set on kill.

What you said all makes sense though.

My CNCRoom breakout board uses 24V, so I will have to get a work around for the ADuM3221.

Once again thanks so much and have a great day.

Craig Smith

[ZincBoy]

Thanks so much ZincBoy, you are a champ. Thank you for taking the time. All a bit new for me, so learning curve is set on kill.

What you said all makes sense though.

My CNCRoom breakout board uses 24V, so I will have to get a work around for the ADuM3221.

Once again thanks so much and have a great day.

Craig Smith

No problem. Here is a revised version to use at 24V. I omitted the index signal as the opto-isolator might not have enough bandwidth.

If there is already a pull-up on the input lines of your break out board then R4 should be omitted and you don't need the connection to the 24V supply.
Q1 is just about any small signal switching NPN transistor.

This circuit should work, but I have not tested it.

[Craig Smith]

Once again I thank you so much. Wow, it is great to know that people like yourself are out there ZincBoy. Unfortunately I can only offer advice about what epoxy resin to use with whatever composite fibre. And how to make moulds, so in the unlikely event that you ever need to do that, you know, make a boat, or model aircraft etc.,……just ask.
Now you will have to excuse my ignorance, but the 5V top left of picture, is that a introduced 5V supply separate from the 5V that exists on Pin 7 (ENC 5VDC)?
And last question, (I think) if I was to put an external momentary RESET switch in, where might it go?
Sorry to use up you time ZincBoy ( I feel I should call you Zinc for short) I am sure you have other stuff to do besides coach a newbie.
Again, thanks so much.
Regards,
Craig Smith

[ZincBoy]

Now you will have to excuse my ignorance, but the 5V top left of picture, is that a introduced 5V supply separate from the 5V that exists on Pin 7 (ENC 5VDC)?

The 5V and 24V symbols are just to show that there should be 5V and 24V on those nets. The Gecko drive provides the actual 5V from the encoder regulator and no external supply is needed. Note that this circuit will draw about 10mA from the encoder supply on the Gecko drive when in reset and very little in normal operation. This supply is only rated for 50mA so if you are using an encoder that draws close to that it might cause issue when coming out of reset. Just check the current consumption on your encoder to see if it will be an issue.

And last question, (I think) if I was to put an external momentary RESET switch in, where might it go?

There are two good places you could put a switch. One the non-isolated drive side, you would connect a normally open switch between pin 1 and pin 2 of J2. When pressed, this will short the 5V to the ERR/RESET pin on the drive and clear the error.

On the isolated side, you can connect a normally open switch between pin 1 and pin 2 of J1. When pressed, it will provide 24V to the opto-isolator and reset the drive in the same way the controller would.

[Craig Smith]

Hi ZinkBoy, I should have sent this earlier.

Before I went ahead and made up the modified Reset Circuit that you so kindly drew up for me, I passed it on to Weerasak from CNCRoom in Thailand, he manufactures the MB2 Break-Out Board which I am using together with the Ethernet Smooth Stepper. Until now his experience is mainly with Industrial controls and so has had no experience with the Geckos. Any help he could give to me was from what he could read in the Gecko manual. He modified your circuit slightly just to suit the MB2 and then I went ahead and made it up on a bit of Vero Board. I will mount it in a small plastic box and then onto a bit of DIN rail in the cabinet.

I will attach his modified version of your circuit (G320 ERR-RESET REV04.png) and my Vero Board set out diagram (Drive Reset Circuit_Rev04.pdf) and a photo of the finished board (Gecko Reset Board.JPG). It all works very well, although I haven't had it working in a real life environment yet. That is still yet to come.

I hope someone else may find this of use if they happen to be using the Gecko 320X together with the MB2 BOB.

If someone does find it of use, I will explain a bit of the Vero Set-Out diagram pdf. The faint vertical purple areas are the breaks in the copper tracks under the Vero Board. I fizzed three of the opto-couplers by forgetting to break the tracks under R1, R3 & R7. So must double check!

The CAD software I used to draw this would only allow me to rotate the components at 90 deg increments, so in fact the transistors should be rotated CW slightly to align with their mounting holes, shown as black dots. The size of the board was determined by the plastic box I had for housing it, so the terminal blocks had to be at a certain distance apart, hence the strange spacing of the vertical links. It all had to fit into a given area.

Once again ZincBoy thank you so much for all you help and have a great Christmas.

Regards,

Craig Smith