Dumb question relating do negative vacuum pressure measurements.

[Maple_Dude]

Finishing up my PnP - for the vacuum measurement, can my vacuum readings be negative, or will that screw it up? 

[Maple_Dude]

for instance, using the 0816 vacuum sensing approach - I get -29kpa as my pressure measurement. 

[Mike M.]

What do you get out in Actuators -  Read Vacuum ?

[Iceturf]

I'm getting 26 as my reading. I noticed I could use a negative scaling factor which has allowed me to convert my pressure in to a positive value. (I would screenprint, but I don't know how on linux)

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/iMFuzA5t6A0/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/aaf4fe4f-7079-40c5-85ce-1b61e137f3bf%40googlegroups.com.

[Maple_Dude]

[Mike M.]

https://www.wikihow.com/Take-a-Screenshot-in-Linux

On Friday, 15 November 2019 22:44:31 UTC+1, Iceturf wrote:

I'm getting 26 as my reading. I noticed I could use a negative scaling factor which has allowed me to convert my pressure in to a positive value. (I would screenprint, but I don't know how on linux)

On Fri, Nov 15, 2019 at 2:03 PM Mike M. <mike...@gmail.com> wrote:

What do you get out in Actuators -  Read Vacuum ?

On Friday, 15 November 2019 21:58:35 UTC+1, Maple_Dude wrote:

for instance, using the 0816 vacuum sensing approach - I get -29kpa as my pressure measurement. 

On Friday, November 15, 2019 at 1:49:04 PM UTC-7, Maple_Dude wrote:

Finishing up my PnP - for the vacuum measurement, can my vacuum readings be negative, or will that screw it up? 

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/iMFuzA5t6A0/unsubscribe.

To unsubscribe from this group and all its topics, send an email to ope...@googlegroups.com.

[Mike M.]

Do you know ? bar or PA ?

[Maple_Dude]

Oh sorry, kpa. About 26 kpa below room pressure. 

[Maple_Dude]

It was showing up as a negative number, -26kpa. Now I have it showing up as in the screenshot - as a positive number. I'm hoping that solves the problem. 

[Mike M.]

Dude - seams OK - You have two nozzles ? two Vacuum pumps?

[Maple_Dude]

One Vacuum Pump - 1/8th Hp pump. Should be sufficient.

[Mike M.]

I am having as well 1 vacuum pump - 0.8bar vacuum but I am experiancing problems with vacuum readings when one nozzle is not loaded. readings go out of normal limits and part is considered by PnP as not picked up...

Let me know how it will go on your side.

Mike

[Iceturf]

Wow I'm 0/2 today. 2 nozzles, 1 vacuum pump.

--

You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/iMFuzA5t6A0/unsubscribe.

To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/7af7b62c-21bf-46c9-9b16-9445bac4f139%40googlegroups.com.

[Mike M.]

Do the readings when one nozzle is loaded and compare with both nozzles loaded or empty - let me know what you get ?

[Maple_Dude]

Vacuum on Nozzle 1, no Part on Nozzle 1, No Vacuum on Nozzle 2                                 ~9kPa (Only Nozzle 1 has vacuum and is sucking room pressure)

Vacuum on Nozzle 1, and Part on Nozzle 1, No Vacuum on Nozzle 2                               ~25kPa  (solenoid to nozzle 2 is closed, so no vacuum or pressure is applied to nozzle 2)

Vacuum on Nozzle 1, Part on Nozzle 1, Vacuum on Nozzle 2, No Part on Nozzle 2          ~20kPa (Nozzle 1 has a part its holding with vacuum, nozzle 2 is sucking back room pressure)

Vacuum on Nozzle 1, Part on Nozzle 1, Vacuum on Nozzle 2, Part on Nozzle 2                 ~25 kPA          (each nozzle has a part its holding with vacuum at the same time)

On Monday, November 18, 2019 at 3:15:03 PM UTC-7, Maple_Dude wrote:

                                                             

Vacuum on Nozzle 1, Part on Nozzle 1, No Vacuum on Nozzle 2                                      9kPa                  ~25kPa          (solenoid to nozzle 2 is closed, so no vacuum or pressure is applied to nozzle 2)

Vacuum on Nozzle 1, Part on Nozzle 1, Vacuum on Nozzle 2, No Part on Nozzle 2    Not Measured        ~20kPa           (nozzle 2 is just sucking back room pressure while a part is held on nozzle 1)

Vacuum on Nozzle 1, Part on Nozzle 1, Vacuum on Nozzle 2, Part on Nozzle 2             ~25 kPA              ~25 kPA          (each nozzle has a part its holding with vacuum at the same time)

[Mike M.]

Hi Dude,

This seams better to what I have here....

- You use Open PnP 1.0 or 2.0?

did you place the values in the nozzle locations? - see enclosed ?

[Maple_Dude]

Hi Mike, 

    OpenPnP 1.0 for now. I'm guessing those pictures are from OpenPnP 2.0?

Yes, I placed the values in the same sort of location though I think my window was slightly different. Is your vacuum actuator under your "Heads" catagory? 

Are those pressured in mmHg or ?? Are you not getting enough vacuum pressure or is it not detecting properly? 

[Mike M.]

Oh Ok - I am on Open PnP 2.0

Yes under Head - see enclosed.

But I have a small manifold with two solenoids for vacuum and for exshoust as well one. This might be my problem...

Mike

[Mike M.]

I can change the units on display - SMC has analog output - to smoothie:

[Mike M.]

I have an issue that if one nozzle is not loaded - vacuum falls on secound nozzle as well.. 

[Iceturf]

Hi Mike, 

  while my vacuum setup leaves something to be desired, my setup goes something like this

Nozzle 1 1: 

    Solenoid 1:

         Output: Nozzle

         Input 1: Vacuum Pump

         Input 2: Solenoid 2 output

    Solenoid 2:

        Output: Solenoid 1 Input 2

        Input 1: Pressure Source

        Input 2: Ambient Air Pressure

Same for Nozzle 2, except Solenoids 3 and 4.

So.... My solenoids default to input (1) when off. So I setup one actuator for nozzle 1. 

I tell that one actuator to do the following:

1. If vacuum desired (pickup): Solenoid 1 Off, Solenoid 2 Off (solenoid 2 state doesn't matter right now)

2. If Place Desired: Solenoid 1 On (Turn the Vacuum Off, so that the pressure doesn't compromise the vacuum)

                               Solenoid 2 Off (pressure triggered)

                               Dwell (G4 P0.25)

                               Solenoid 2 On (Ambient Triggered

Finito

It sounds to me like your system isn't properly shutting off vacuum to the nozzle not currently in use? I assume you have one vacuum pump for both nozzles as well? 

In OpenPnP 1, the Pick / Place operations seem to only be able to use a single software "actuator" but that single software actuator I have trigger multiple actual actuators. Does that make sense? 

I am having a problem with fiducial alignment and uplooking camera in mine though. It doesn't seem to properly center in on the fiducials - I have to manually center it. And it incorrectly interprets part position about 40% of the time, when a part is brought above the uplooking camera. I may have to ask Jason about this. Though I don't see other people complaining of the same problem so I am confused. 

--

You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/iMFuzA5t6A0/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/cce9ae41-a841-4bb7-8b57-76944bc5e828%40googlegroups.com.

[Iceturf]

I think I am having a problem with the vacuum seal though. I originally was getting 40 to 55 kpa of vacuum with two solenoids in parallel for solenoid 1 and two in parallel for solenoid 2 (my solenoids are tiny compact ones, a bit of a mistake on my part). Then I was getting 30-36kpa of vacuum with just one solenoid for each stage. Now I'm getting 15 kpa to 25 kpa. I think I'll have to make a new vacuum adapter backpack for my motors. I think my current design must be failing. I know from the get go I loose about 20 kpa just from the seal where the hose goes in. (would be about 70-77kpa vacuum then)

[Marek T.]

@Iceturf, set settle time -200 in uplooking camera settings and take care whether you have selected "settle time" ON in the pipeline first stage.

(-200 not 200, it's not mistake, it works totally other than positive values and great, but probably only in Openpnp 1.0).

W dniu środa, 20 listopada 2019 17:49:10 UTC+1 użytkownik Iceturf napisał:

Hi Mike, 

  while my vacuum setup leaves something to be desired, my setup goes something like this

Nozzle 1 1: 

    Solenoid 1:

         Output: Nozzle

         Input 1: Vacuum Pump

         Input 2: Solenoid 2 output

    Solenoid 2:

        Output: Solenoid 1 Input 2

        Input 1: Pressure Source

        Input 2: Ambient Air Pressure

Same for Nozzle 2, except Solenoids 3 and 4.

So.... My solenoids default to input (1) when off. So I setup one actuator for nozzle 1. 

I tell that one actuator to do the following:

1. If vacuum desired (pickup): Solenoid 1 Off, Solenoid 2 Off (solenoid 2 state doesn't matter right now)

2. If Place Desired: Solenoid 1 On (Turn the Vacuum Off, so that the pressure doesn't compromise the vacuum)

                               Solenoid 2 Off (pressure triggered)

                               Dwell (G4 P0.25)

                               Solenoid 2 On (Ambient Triggered

Finito

It sounds to me like your system isn't properly shutting off vacuum to the nozzle not currently in use? I assume you have one vacuum pump for both nozzles as well? 

In OpenPnP 1, the Pick / Place operations seem to only be able to use a single software "actuator" but that single software actuator I have trigger multiple actual actuators. Does that make sense? 

I am having a problem with fiducial alignment and uplooking camera in mine though. It doesn't seem to properly center in on the fiducials - I have to manually center it. And it incorrectly interprets part position about 40% of the time, when a part is brought above the uplooking camera. I may have to ask Jason about this. Though I don't see other people complaining of the same problem so I am confused. 

On Wed, Nov 20, 2019 at 9:26 AM Mike M. <mike...@gmail.com> wrote:

I have an issue that if one nozzle is not loaded - vacuum falls on secound nozzle as well.. 

On Wednesday, 20 November 2019 16:52:48 UTC+1, Mike M. wrote:

I can change the units on display - SMC has analog output - to smoothie:

On Friday, 15 November 2019 21:49:04 UTC+1, Maple_Dude wrote:

Finishing up my PnP - for the vacuum measurement, can my vacuum readings be negative, or will that screw it up? 

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/iMFuzA5t6A0/unsubscribe.

To unsubscribe from this group and all its topics, send an email to ope...@googlegroups.com.

[Iceturf]

Hi Marek, 

    So I'm sure I understand you - settling time should be a negative number in OpenPnP 1.0? Also, Is this improved in OpenPnP 2.0? Should I switch across ASAP? 

To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/16a42a30-1b85-41a5-afbd-4fb4ebf20249%40googlegroups.com.

[Marek T.]

Normally it is positive and means ms delay. In Openpnp Jason has added experimental feature using negative values and then it is not ms delay but comparing next picture frame with the previous one, and checking how much they may differ. Value is kind of limit telling how much the pictures may differ. So as lower value as more identical two following pictures have to be. If differ more than the limit - another picture is taken to compare. If differ under the limit - it means the picture is stable.

I don't know if Jason applied it to 2.0 - I don't use 2.0 for some reasons.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/16a42a30-1b85-41a5-afbd-4fb4ebf20249%40googlegroups.com.

[Marek T.]

If you use 1.0 you can try this. Rather no sense to switch from 2.0 to 1.0 for this. "-" is very convinient option in my oppinion and worth to try if exists.

Just pay attention whether the picture used to align is not taken to quick after achiewing the camera position. Turn on DEBUG in log and observe the pictures created by the Openpnp during the alignment, you will find them in Openpnp user directory.

W dniu środa, 20 listopada 2019 18:01:17 UTC+1 użytkownik Iceturf napisał:

Hi Marek, 

[Mike M.]

Hi Iceturf,

Thanks for your comments but I found what I was doing wrong - I was setting my Pick up value on value when both nozzles are loaded which is not correct.. as the pressure-vacuum is lower than when bouth nozzles are loaded- see the chart enclosed:

for others to lern... might help ... so by setting lower value it started working now - 2 or 4 test runs But now I have to battle my way trugh after pick dwell time - ( nozzle is empty but I get part on nozzle error from PnP).

For your fiducial can you post your image from pipeline and pipline text - I can try to see if it works for me...

Mike

To unsubscribe from this group and all its topics, send an email to ope...@googlegroups.com.

[Mike M.]

I am on Open PnP 2 for months now - I see no reason why not switching to v2. Do the Smoothie-firmware update as well before going to Open PnPV2.
Mike

[Marek T.]

Mike, my reasons are some automated features that we added with Bert to the 1.0 code and I can't imagine to live without them.

Just have no time to apply it to 2.0 what is not very trivial (for me) as there are many changes in jobProcesor required, while it is very different in 2.0.

For every "normal" people who were happy with options that official 1.0 brings - agree with you: there are no reasons to avoid 2.0, specially that it has many things made really much better (including the mentioned jobProcessor :-)).

What the Smoothie update you mean which is related to switch 1.0 into 2.0???

[Mike M.]

As I stated - firmware from Smoothie - http://smoothieware.org/flashing-smoothie-firmware

Mike

[Marek T.]

I can hear you mean the Smoothie but don't understand why need to upgrade it when upgradingg Openpnp to 2.0...

[Jim]

I think it was because of some side effects in NT calibration (c-axis steppers loosing steps) that occured in v2 with old smoothieware. Thanks to swiss precision it has been repaired finally.

[Marek T.]

I know this issue.
And I don't negate that it's worth to upgrade Smoothie with "Swiss precise", but it is not so that in case of Openpnp1 steppers steps are not lost but in Openpnp2 only.

[Maple_Dude]

I have set the settle time to -200. About 50% of the parts were placed properly with uplooking vision. About 100% of the parts are placed properly without vision. The fiducials are off with automatic alignment - the parts are usually in the area of the pads but not on them.

[Maple_Dude]

Hm, well for the uplooking camera it may be a lighting issue. It seems that sometimes the camera is looking before the LEDs have turned on. Still haven't solved the downlooking camera yet.

[Marek T.]

Which script calling do you use to turn on the bottom light?
The snapshot you have linked is it case that you suspect the issue with the lightning?

[Maple_Dude]

Spoke too soon. I notice there is a much higher chance of failure if pre-rotate is turned on.

[Maple_Dude]

yes, because the snapshot is not lit - I suspect a lighting issue. I am using the stock scripts from Jason's tutorial. They do work, the lights turn on and off.

[Iceturf]

Maybe I'm completely wrong. 

--

You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/iMFuzA5t6A0/unsubscribe.

To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/4cde9c1d-f35b-4f5f-84c0-ba5b89da0883%40googlegroups.com.

[Marek T.]

I don't ask about the code but scriot name that you use.
However, I'd advice to forget about the bottom light control and turn it on constantly. And test it to be sure what is your issue.

[Mike M.]

Hi

No wonder - Centre of Part and centre of Vision Square difference is your mistake in placement !

See blue and red centre - there is for 1 part difference.

You need to change your settings in vision to get the squre around your part!! see picture enclosed.

[Mike M.]

Error - distance between two cercles

[ma...@makr.zone]

To block light in the background you should use a dark shade behind the nozzle tip. Can be as simple as a piece of black paper:

Plus your HSV mask is off, the green nozzle tip should turn black, not yellow. Try resetting your pipeline (after having backed up your machine.xml).

Plus you seem to have a nasty side highlights from your top light, you have to block that or switch it off using scripts.

Plus your threshold is not right, it seems. It must be much higher to only include the very bright part.

_Mark

[Mike M.]

Dude - your image as well has to many dark and bright areas - you need to mask top of needle/nozzle in order to get better image and easy work with vision settings...

On Thursday, 21 November 2019 22:16:39 UTC+1, Maple_Dude wrote: