Hi Russ,
Such a long delay is not normal. The input itself to KFLOP is sub-microsecond, the Windows Screens update ~10 times per second but occasionally longer depending on Windows.
You neglected to include a sensor spec or link. I Googled and found this: http://www.mouser.com/ds/2/307/ee-sx47_67_ds_csm483-40324.pdf
How did you wire it? It looks like those sensors are NPN open collector (like a switch to GND). Do you have something like a 500ohm pull up resistor connected from the pin to +3.3V to pull the pin to +3.3V when the output goes open? If not it is likely to take a long
time (or forever) for the pin to float high.
HTH Regards TK
| Group: DynoMotion |
Message: 9461 |
From: Russ Larson |
Date: 5/4/2014 |
| Subject: Re: New Kflop question on Digital inputs |
Tom, The sensor was the Omron EE-SX670 http://www.ia.omron.com/product/item/eesx1680r/index.html The load is between + and the OUTPUT wire, they are connected to 5VDC. We used these with MACH3 for years with no issues. Will double check the wiring again tonight. Russ From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com] On Behalf Of Tom Kerekes
Sent: Sunday, May 04, 2014 5:14 PM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] New Kflop question on Digital inputs Such a long delay is not normal. The input itself to KFLOP is sub-microsecond, the Windows Screens update ~10 times per second but occasionally longer depending on Windows. You neglected to include a sensor spec or link. I Googled and found this: How did you wire it? It looks like those sensors are NPN open collector (like a switch to GND). Do you have something like a 500ohm pull up resistor connected from the pin to +3.3V to pull the pin to +3.3V when the output goes open? If not it is likely to take a long time (or forever) for the pin to float high. We were working on a new machine the other day with the KFLOP as the primary controller. We just started the install and updated the unit with the latest software release. It comes up and seems to be working fine. We started testing the IO one homing sensor at a time and discovered which the digital screen does see the EE-SX670 Omron sensors it takes about 2 seconds for the display on the KFLOP software to update is that normal. All of these inputs are on the J5 connector. We tried a touch sensor for tool length, again it sees the sensor change but takes about two seconds to update the screen. |
|
| Group: DynoMotion |
Message: 9462 |
From: Tom Kerekes |
Date: 5/4/2014 |
| Subject: Re: New Kflop question on Digital inputs |
Hi Russ,
A TTL or LVTTL input (KFLOP input or PC Parallel Port) must be driven both high (>2V) and low (<0.8V) to be guaranteed to work reliably. It is somewhat different from a current load like a relay coil or a light bulb those sensors were designed to drive.
The sensor's output will drive low just fine, but doesn't drive high at all.
Some TTL inputs may tend to float high fairly reliably or may even have internal pull ups which is why it may have worked for you in the past.
KFLOP's inputs are MOSFETS with high input impedance (~10M Ohms) and take almost no current.
Regards TK
| Group: DynoMotion |
Message: 9463 |
From: Russ Larson |
Date: 5/4/2014 |
| Subject: Re: New Kflop question on Digital inputs |
Thanks Tom, I will take a brand new sensor and test it on my bench KFLOP and see if I can get it to work reliably without delay. Sounds like you are suggesting I need a pull up resistor for this sensor, never used one with the standard parallel port but I will give this a try and see what happens. Thanks Russ From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com] On Behalf Of Tom Kerekes
Sent: Sunday, May 04, 2014 6:59 PM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] New Kflop question on Digital inputs A TTL or LVTTL input (KFLOP input or PC Parallel Port) must be driven both high (>2V) and low (<0.8V) to be guaranteed to work reliably. It is somewhat different from a current load like a relay coil or a light bulb those sensors were designed to drive. The sensor's output will drive low just fine, but doesn't drive high at all. Some TTL inputs may tend to float high fairly reliably or may even have internal pull ups which is why it may have worked for you in the past. KFLOP's inputs are MOSFETS with high input impedance (~10M Ohms) and take almost no current. The sensor was the Omron EE-SX670 The load is between + and the OUTPUT wire, they are connected to 5VDC. We used these with MACH3 for years with no issues. Will double check the wiring again tonight. Such a long delay is not normal. The input itself to KFLOP is sub-microsecond, the Windows Screens update ~10 times per second but occasionally longer depending on Windows. You neglected to include a sensor spec or link. I Googled and found this: How did you wire it? It looks like those sensors are NPN open collector (like a switch to GND). Do you have something like a 500ohm pull up resistor connected from the pin to +3.3V to pull the pin to +3.3V when the output goes open? If not it is likely to take a long time (or forever) for the pin to float high. We were working on a new machine the other day with the KFLOP as the primary controller. We just started the install and updated the unit with the latest software release. It comes up and seems to be working fine. We started testing the IO one homing sensor at a time and discovered which the digital screen does see the EE-SX670 Omron sensors it takes about 2 seconds for the display on the KFLOP software to update is that normal. All of these inputs are on the J5 connector. We tried a touch sensor for tool length, again it sees the sensor change but takes about two seconds to update the screen. |
|
| | | |
