Blowoff Valve Actuator only works in manual mode

70 views
Skip to first unread message

Christian Liebhart

unread,
Sep 7, 2022, 5:39:03 AMSep 7
to OpenPnP
So I've installed a recently delivered valve for my pump to release vacuum and set up an actuator in OpenPnP. 

Ive tried setting up the gCode commands both manually and with help of I&S. Both methods allow me to actuate the valve manually but once I let machine operate automatically within the "confetti-calibration" cycle, the valve is not controlled by the machine. 

(Sidenote) In the nozzles "Vacuum"-tab I selected the valve actuator as my blow off actuator assuming this is the right choice here.

Do you guys have any suggestions how to fix the problem? 
I read through older posts and tried setting up an double-actuator instead of a boolean one, but this didnt do the trick..

I'm deeply gratefull for any help from you guys!

-Chris 

mark maker

unread,
Sep 7, 2022, 6:22:34 AMSep 7
to ope...@googlegroups.com

Note that the blow-off actuator is actuated with a Double value, not Boolean. Set the actuator to Double:

https://github.com/openpnp/openpnp/wiki/Setup-and-Calibration:-Actuators#actuator-value-type

Then I&S should propose setting the right command type.

Note, as already discussed on this list, the Double value can also determine a time, not only a strength. Some blow off valves might be Boolean in nature, but you can pulse them with the Double value duration. Use a dwell command in G-code (G4).

Blow-off only happens, if your Package has a non-zero Blow Off Level defined.

Frankly, this is not well implemented. There should be a more central default for the blow off level, maybe on the nozzle tip, as I guess blow strength might actually be more related to the nozzle tip air bore than to the package. So if the package level is zero it should take the nozzle tip level.

Do you blow-off-ers (😂) agree?

Furthermore, I don't understand the code. If you have a blow-off it does no longer actuate the vacuum valve off?!

https://github.com/openpnp/openpnp/blob/af43024e03f6d9e98c738cff8fb1424be52c487b/src/main/java/org/openpnp/machine/reference/ReferenceNozzle.java#L414-L425

I get it that some blow-off actuators might actually be the same device/valve, as the vacuum valve, but this cannot be assumed as universally true. And if it were universally true, one would not need a separate actuator.

Similarly strange: if the Package has a Vacuum Level, it actuates as Double in the pick...

https://github.com/openpnp/openpnp/blob/af43024e03f6d9e98c738cff8fb1424be52c487b/src/main/java/org/openpnp/machine/reference/ReferenceNozzle.java#L374-L380

but then actuates as Boolean (false) in the place?! (see upper link again)

_Mark

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/5b689cd7-5f44-4136-9df5-35c95c240b87n%40googlegroups.com.

bert shivaan

unread,
Sep 7, 2022, 6:31:13 AMSep 7
to ope...@googlegroups.com
In my old commercial machine the blow off duration was a global setting that did not change with tip or part.


Christian Liebhart

unread,
Sep 7, 2022, 10:52:55 AMSep 7
to OpenPnP
Thanks for the quick response!
I think I might have to be more precise: I already tried implementing it as a double value. The attachment shows how I did it. 
This solution works perfectly fine when using the machine control actuator buttons manually. I can open and close the valve by clicking the correspondent buttons. 

But once I let the machine operate by itself, in my case letting it calibrate the precise camera <--> Nozzle N offset with help of the suggested "confetti" pieces, the machine does control the valve so the confetti pieces still stick to the nozzle even with the vacuum pump being turned off. 

I assume the calibration process would use the valve to put down the confetti pieces if everything works as its supposed to.
Do you guys have any suggestions on what I might have done wrong?

As to your follow up questions: I am afraid I do not have the expertise to give my opinion on any "blow-off-implementation-code" whatsoever. 

Thanks for your time!
Chris
Screenshot 2022-09-07 164248.jpg

mark maker

unread,
Sep 7, 2022, 11:02:07 AMSep 7
to ope...@googlegroups.com

Ah, I see. The problem is that the confetti Part and Package is just created "on the fly" and you can't set the blow-off-value there.

It is actually also a consequential problem of the missing nozzle tip default.

I will look into it. Stay tuned.

_Mark

Jan

unread,
Sep 7, 2022, 12:32:18 PMSep 7
to ope...@googlegroups.com
Hi!
I'd like to share my experience with blow-off and the confetti
calibration: On my CHM-T36VA there is a blower and a vacuum pump linked
to the nozzle using a single valve. So either the vacuum or the blower
is connected to the nozzle tip. If I operate the blower at the
configuration of the original Chinese software (16kHz PWM @2%) the
confetti is blown away after the nozzle as put it down. I archived good
results with blower switched off and temporarily increased place dwell
time. I usually run without place dwell time but for the calibration I
used IIRC 500ms. I'd suggest to visually inspect the entire calibration
process as a single bad placement will distort the final results.

Jan

On 07.09.2022 17:02, mark maker wrote:
> Ah, I see. The problem is that the confetti Part and Package is just
> created "on the fly" and you can't set the blow-off-value there.
>
> It is actually also a /consequential /problem of the missing nozzle tip
>> Blow-off only happens, if your Package has a non-zero *Blow Off
>> Level* defined.
>>
>> Frankly, this is not well implemented. There should be a more
>> central default for the blow off level, maybe on the nozzle tip,
>> as I guess blow strength might actually be more related to the
>> nozzle tip air bore than to the package. So if the package level
>> is zero it should take the nozzle tip level.
>>
>> Do you blow-off-ers (😂) agree?
>>
>> Furthermore, I don't understand the code. If you /have/ a blow-off
>> it does no longer actuate the vacuum valve off?!
>>
>> https://github.com/openpnp/openpnp/blob/af43024e03f6d9e98c738cff8fb1424be52c487b/src/main/java/org/openpnp/machine/reference/ReferenceNozzle.java#L414-L425
>>
>> I get it that some blow-off actuators might actually be the same
>> device/valve, as the vacuum valve, but this cannot be assumed as
>> universally true. And if it /were/ universally true, one would not
>> need a separate actuator.
>>
>> Similarly strange: if the Package has a Vacuum Level, it actuates
>> as Double in the /pick/...
>>
>> https://github.com/openpnp/openpnp/blob/af43024e03f6d9e98c738cff8fb1424be52c487b/src/main/java/org/openpnp/machine/reference/ReferenceNozzle.java#L374-L380
>> <https://github.com/openpnp/openpnp/blob/af43024e03f6d9e98c738cff8fb1424be52c487b/src/main/java/org/openpnp/machine/reference/ReferenceNozzle.java#L374-L380>
>>
>> but then actuates as Boolean (false) in the /place/?! (see upper
>>> <https://groups.google.com/d/msgid/openpnp/5b689cd7-5f44-4136-9df5-35c95c240b87n%40googlegroups.com?utm_medium=email&utm_source=footer>.
>>
>> --
>> You received this message because you are subscribed to the Google
>> Groups "OpenPnP" group.
>> To unsubscribe from this group and stop receiving emails from it, send
>> an email to openpnp+u...@googlegroups.com.
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/openpnp/64916899-fb22-4e5b-b402-3b50b7062febn%40googlegroups.com
>> <https://groups.google.com/d/msgid/openpnp/64916899-fb22-4e5b-b402-3b50b7062febn%40googlegroups.com?utm_medium=email&utm_source=footer>.
>
> --
> You received this message because you are subscribed to the Google
> Groups "OpenPnP" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to openpnp+u...@googlegroups.com
> <mailto:openpnp+u...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/openpnp/43deb4a7-74ae-804b-b245-8ccb7354a2e2%40makr.zone
> <https://groups.google.com/d/msgid/openpnp/43deb4a7-74ae-804b-b245-8ccb7354a2e2%40makr.zone?utm_medium=email&utm_source=footer>.

mark maker

unread,
Sep 7, 2022, 1:08:33 PMSep 7
to ope...@googlegroups.com

Hi Chris, Jan,

I made a fix, and hope you, @Chris, can test this quickly, in the testing version, because I can't on my machine (no blower).

See the PR:

https://github.com/openpnp/openpnp/pull/1484

And the Wiki about the "confetti" calibration, quote:

The calibration is executed using the regular pick and place cycle, so all the vacuum control settings etc. apply. If you need to control the properties of the part to be picked, like the part height or package blow-off level, etc. you can create a part with the ID "TEST-OBJECT". It will be used automatically.

https://github.com/openpnp/openpnp/wiki/Calibration-Solutions#calibrating-precision-camera-to-nozzle-offsets

@Jan, thanks for the insight, I guess you will now have more options, which I hope would make the temporary changes unnecessary, assuming you can adjust blow strength or pulse time on the CHMT (I know it was discussed but I don't really remember). I hope you could give it a try too 😁

If you can't adjust the blower on the fly, then maybe a better test-object can be created. I'm thinking a heavier one perhaps made from strong aluminium foil (the baking oven floor protection kind). Maybe even glue several layers together then hole-punch it. The underside could perhaps also be treated to increase friction, something a bit rubbery e.g. coat with acrylic glue and let dry.

Thanks for the on-going testing!💯

_Mark

Christian Liebhart

unread,
Sep 8, 2022, 9:52:38 AMSep 8
to OpenPnP
Great stuff Mark! 
I just downloaded the test version and after playing around a bit, the actuators work perfectly fine while performing the confetti-calibration.
Thanks for the implementation. I am sure future OpenPnP novices are thankful for the update and especially for noticing it in the Wiki.

-Chris

mark maker

unread,
Sep 8, 2022, 11:01:48 AMSep 8
to ope...@googlegroups.com
Reply all
Reply to author
Forward
0 new messages