New Ctrl Sys Early Ship - Benchtop Test

[rrossbach]

Hi -
We received our early shipment of the new control system today....just like Christmas a month early for the team!

We ran through the benchtop test procedure this afternoon, and all worked fine except for Auton mode - following the steps in the benchtop test doc, nothing happened when we enabled Auton mode. This is with the out-of-the-box firmware, image, etc on the cRIO and DS.

Has anyone else run into a similar problem? Trying to determine if this indicates a hardware defect (and we have to get a replacement from FIRST) or simply a problem with the default code that ships on the controller....

Thanks,
Ron

Team 2607 - Programming Mentor

[Mark McLeod]

Ours didn't follow the bench test expected outputs in all cases either.
I don't yet know if that's a real problem or an OTB version issue though.
We saw two discrepancies. One joystick didn't behave during teleop, and the first time we tried Auto nothing happened, but it did run for two seconds when we retried it.

We haven't dropped our own test program in yet to run a full verifiable test.
That's tomorrow for us.

[willson.thomas]

If everything worked up to that point and you're feeling lucky, I would just move on to flashing the DS and cRIO, then trying your own autonomous code. If that works, you should be just fine.

[Jon236]

Ours didn't follow the bench test expected outputs in all cases either.
I don't yet know if that's a real problem or an OTB version issue though.
We saw two discrepancies. One joystick didn't behave during teleop, and the first time we tried Auto nothing happened, but it did run for two seconds when we retried it.

We haven't dropped our own test program in yet to run a full verifiable test.
That's tomorrow for us.

Mark,

Our system behaved just as yours did out of the box. Once the software updates are downloaded to both cRio and DS, everything seemed to work well.

[rees2001]

Quick question.

Did auton run once & then not again?

[rrossbach]

No, the OTB auton didn't even run once. Tried it multiple times (rebooting both DS and cRIO) and even hooked up two Jaguars instead of Victors in case it was an "operable range" issue, to no avail.

The Jaguars are nice, btw. Control seems to be much smoother and a much wider operable range.

Thanks,
Ron

Team 2607 - Programming Mentor

[Greg McKaskle]

I don't believe that the autonomous code uses the motors at all. It simply strobes the solenoid module outputs something like a certain black car on TV.

Greg McKaskle

[rfrank]

Mark,

Our system behaved just as yours did out of the box. Once the software updates are downloaded to both cRio and DS, everything seemed to work well.

Is there a software update for the cRio? If so, where is it located?

Thanks!

[Mark McLeod]

2009 Control System page at http://www.usfirst.org/community/frc/content.aspx?id=10934
Under Software Updates

[Alan Anderson]

http://www.usfirst.org/frccontrolsystem is easier to remember. :)

[rfrank]

I see an update for the Driver Station, for LabView, and for WPILib. I do not see one for the cRio. Is it included in the LabView update?

My team has this exact problem:
http://www.chiefdelphi.com/forums/showthread.php?t=70175
But none of the posts on the above thread (nor the one that it links to on the beta forums) are any help.

[stevenlu]

I think the cRIO update is found on the NI website:
http://joule.ni.com/nidu/cds/view/p/id/1162/lang/en

[Abrakadabra]

Yes - the whole update process is a little confusing. If you piece together the various instructions found in various places all in the right order, you might get it right...

Here's what I figured out by reading everything very carefully before doing anything, and my system is working now, so I think I got it right...:rolleyes:

Before updating the cRIO firmware, you need to do the LabView update (this fact is only mentioned on the WPI site where you get the C++ update... go figure)

After you have done the LabView update, you will have an icon on your Windows Desktop that says "FRC cRIO Imaging Tool". Once you have that, you can start the cRIO updating process that is documented in section 5.1.2 (the Configuration chapter, not the Software Installation chapter, and not section 5.2.2 as mentioned in section 1.6.5 (5.2.2 doesn't exist!))

When you do the cRIO imaging, I found it easiest to connect the PC directly to the cRIO using the orange crossover cable provided in the kit (although you can also follow the directions and use the Driver Station as a LAN switch). If this is the first time you are updating the cRIO, make sure you set the IP on your computer to the static address 10.0.0.6, with a subnet of 255.0.0.0. (For subsequent updates, you will use the IP address with the team number embedded).

Then instead of following the menu path detailed in section 5.1.2.4.1, I just clicked on the above mentioned Desktop icon, and everything went smoothly from there.

Also - Please note that for the latest C++ software update found on the WPI site (Beta update 5), they say that you must reimage the cRIO before you install it, which directly contradicts the steps outlined in sections 1.6.4 and 1.6.5.

YMMV, but good luck anyway :)

[Tom Line]

Is anyone currently doing this with 2 nics in a computer so you don't lose net connectivity when you connect to the Crio?

Is there an option to select which network adapter Labview should be using, if you do have 2 nics in a computer?

[Mark McLeod]

I don't believe that the autonomous code uses the motors at all. It simply strobes the solenoid module outputs something like a certain black car on TV.
P.S. Is it possible that you're thinking of the new software patch in FRCLabVIEWUpdate1.0, rather than the OTB?
That does cause a chase light sequence on the solenoid LEDs in autonomous.
-----

Then I think it's possible that at least two versions of the OTB shipped.
1) Autonomously strobe the solenoid lights
2) Autonomously work somewhat in accordance with the bench test description

We evidently got version 2, because there was no indicator light strobing, but autonomous did drive the pwm outputs several different ways.
With our version when the DS was booted into autonomous we saw one of these three different autonomous behaviors occur randomly intermixed.
pwm 2 goes full forward effectively forever all by it's lonesome self
Nothing happens and joystick 1 still controls the motors as in arcade teleop
pwm1-4 go full forward for 2 secondsI'm just trying to figure out what to tell people at the local workshops.

[Nate Smith]

Is there an option to select which network adapter Labview should be using, if you do have 2 nics in a computer?

Depending on how your computer is set up, it should be able to handle this automatically based on the routing table, not having to bother Labview with the details at all.

For those who have no real idea what I'm talking about, basically, your computer keeps track of which gateway(normally, a server at your ISP, or your own router if you are using one) is the quickest route to get to any particular address. Normally, it's as simple as 192.168.0.anything is considered your local network, and can directly connect from one computer to the other, and everything else is an outside network that goes through your gateway. Your computer should be able to figure out that 10.anything can be accessed through the second NIC.

[Alan Anderson]

Then I think it's possible that at least two versions of the OTB shipped...

I'm just trying to figure out what to tell people at the local workshops.

I suggest you tell them that there's a good reason for the software updates (i.e. the OTB code doesn't work exactly as desired) and that they shouldn't worry too much if the behavior before an update doesn't match the expected results perfectly.

[Greg McKaskle]

Sorry, my description was of what they are calling the default code for an installed image, not the OTB code. I don't have access to the code for OTB, so I can't tell you what it was supposed to do, but your description sounds familiar.

There were not two versions. OTB was what was burned onto the controllers at the factory. Default is what you'll have after a cRIO image update.

Greg McKaskle

[Mark McLeod]

You're just dedicated and overworked.

Take the day off, go eat some turkey and wait the parades.

You deserve it.

Thank you

[Bob Steele]

We received our kit yesterday... tried bench test last night.

I was wondering if anyone had a problem with tank drive vs arcade drive.
We could never seem to get tank drive (teleop and enabled)..no matter what we did we got arcade drive.

Is this an artifact of using the out of box software?

Anyone else see this happen?

Happy Thanksgiving...

[Mark McLeod]

If you get any response I'd consider it a successful bench test.
It just means all the parts from the DS to the motor are working together.

In OTB Teleop we only saw Joystick on USB 1 arcade action (w/button pressed) on pwms 1-4, neutral on pwms 5-10, no other button effects noted.

No response to a usb 2 joystick.

The jumpers in the prescribed positions did not affect results, but we didn't try all the other possible jumper positions.

If you add an Allen Bradley light to the hardware setup it won't come on either (but it works with the non-OTB default software update).

P.S.
Send my regards to Nataloia

[David Fort]

....

pwm 2 goes full forward effectively forever all by it's lonesome self
Nothing happens and joystick 1 still controls the motors as in arcade teleop
pwm1-4 go full forward for 2 secondsI'm just trying to figure out what to tell people at the local workshops.

That summarizes our experience too.
(except we didn't check PWM 2 and 3, and I'm willing to believe when it went for 2 seconds, it was only 50%, not full, which would be as "advertised". in the "full speed without any sign of stopping" case, I found that wiggling the joystick stopped it. But those details aside, I saw think I saw this also. )

[jaustin]

For what it's worth, we could not get Autonomous to work either. We did not understand what step 5 of 2.3.1.4 ("connnect a 2 pin jumper... to the sixth set of pins") was trying to tell us so we went back to the arcade drive pin config (ie jumper on the 5th set of pins). After we did that, auton mode worked exactly as adversized. Does any one know what the 6th set of jumper pins is supposed to do? Maybe it's a type-o in step 5?

[Bill Trump]

I don't believe that the autonomous code uses the motors at all. It simply strobes the solenoid module outputs something like a certain black car on TV.

Greg McKaskle


Greg,

The FRC Control System manual said that in OTB autonomous mode, for benchtop testing section 2.3.1.4

Step 9. "Immediately upon enabling the system, the motors connected to PWM#1 and PWM#2 should both be activated for exactly two seconds. The motor connected to PWM #1 should turn forward at about half speed, and the motor connected to PWM #2 should turn backward at about half speed."

Our system did exactly that... so far, so good!

Bill

Team 801
Edgewood Jr/Sr High

[Mark McLeod]

See Greg's post #18

[JohnBoucher]

It's getting close to kickoff, so I am curious where we stand.

Is there a happy story for every horror story? As the problems localized, earlier shipped vs later shipped?

Is the documentation correct? Once the six weeks starts, will the teams that have waited to get the electronics be able to get them working easily?

[Alan Anderson]

Is the documentation correct? Once the six weeks starts, will the teams that have waited to get the electronics be able to get them working easily?

The documentation worked great for us. It all went smoothly, as soon as the students got comfortable with using the wire strippers and developing a good feel for the Wago connectors.

However, note that I got special training (and I'm probably unusually attentive to detail besides). The TechnoKats programming group spent an hour last week doing the electrical work to put everything together, after which the bench test went exactly as expected. We'll be performing the software updates and wireless setup today, and I don't think there will be any issues.

[rwood359]

Is there a happy story for every horror story? As the problems localized, earlier shipped vs later shipped?

Is the documentation correct? Once the six weeks starts, will the teams that have waited to get the electronics be able to get them working easily?
Put us in the happy story file.
We received the shipment a week ago. We went through the benchtop tests slowly and deliberately.
The floating inputs on the DS caused some problems, but by grounding the input we got tank mode. Jumpering to +5 put us in arcade mode.
It took one of our students about an hour to setup the wireless components. He had some problems with security. Unfortunately, he didn't document the problems and solutions.

Once we completed the benchtop tests, we added components one at a time and used the examples to test operation. We connected the encoders from super shifters started the encoder example, spun the output shaft, and observed the display. The same with the camera servos, compressor, etc.

We now have the full drive system (2 super shifters with CIMs, full pneumatic system and 2 Jaguars) plus the camera fully operational. We are programming a basic robot for bench test. In a week or so, we'll integrate into a robot.

We first tested with Victors then with Jaguars. As noted in other threads, the low speed performance of the Jaguars is impressive.

The early shipment is really an extension of the beta test, so the documentation will be much better by kickoff and should help the teams seeing it for the first time.

[Phalanx]

Add us to the fully successful list.

We physically set up everything yesterday, today we did the OTB without issue.

We've successfully re-imaged the CRIO and Updated the Driver Station flash as well.

We followed the most current version of the instructions to the letter and everything worked as documented except...

On small snafu was that we needed to turn off Wireless security on the Gaming Adapter as well as the Wireless Router. We had turned it off on the Gaming Adapter as per the documentation, but either missed it or it wasn't documented for the wireless router. Once that was done, everything worked fine.

So it looks like we are good to go.

[Alan Anderson]

We'll be performing the software updates and wireless setup today, and I don't think there will be any issues.

I was wrong -- there were two issues.

First, the manual was apparently revised a week ago with no obvious announcement. It was just an accident that I decided to download a copy onto my own computer today before going to the shop to work with the students. It was less easy following the steps on screen than it would have been on paper.

Second, the Driver Station seemed to stop forwarding packets between its Ethernet ports for a while after we set the team number. We had to plug the laptop directly into the router in order to complete the wireless configuration properly. The problem did fix itself for no apparent reason, and in the end we had a working Driver Station communicating by radio with a working robot chassis. If the chassis hadn't been missing wheels on one side, we'd have driven it around the shop for a while.

All the messing around with IP addresses on the laptop was unexpectedly tedious, but now that everything is set it shouldn't trouble us further.

[Steve Ketron]

If you get any response I'd consider it a successful bench test.
It just means all the parts from the DS to the motor are working together.

In OTB Teleop we only saw Joystick on USB 1 arcade action (w/button pressed) on pwms 1-4, neutral on pwms 5-10, no other button effects noted.

No response to a usb 2 joystick.

The jumpers in the prescribed positions did not affect results, but we didn't try all the other possible jumper positions.

If you add an Allen Bradley light to the hardware setup it won't come on either (but it works with the non-OTB default software update).

P.S.
Send my regards to Nataloia


I talked with a FIRST Staff Engineer today about this problem, she said that she just found out today that the joystick adjustment knob (the control switch that looks like last year joystick calibrator) also dictates if a joystick is to be used for arcade or tank drive. I have been at a conference and I am unable to test this information. Try to adjust the this and see if it works.

[rwood359]

I talked with a FIRST Staff Engineer today about this problem, she said that she just found out today that the joystick adjustment knob (the control switch that looks like last year joystick calibrator) also dictates if a joystick is to be used for arcade or tank drive. I have been at a conference and I am unable to test this information. Try to adjust the this and see if it works.
This is true after you have updated the firmware in the DS and reimaged the cRIO.
The OTB program uses Drivers Station Digital Input 1 to select between tank and arcade. The documentation says to install the jumper from signal to +5 to select arcade mode and no jumpers for tank mode. This works for some but not all systems. The DS DI inputs are floating so they can be in any state without jumpers. Install a jumper between Ground and Signal on DS DI 1 to select tank mode in OTB mode. We used a PWM cable with black and white tied together.

[Tank2010]

We are having problems with the bench top tank drive test. When the controller in USB port 1 is moved both motors spin and the jaguars indicate in opposite directions. The controller in USB port 2 does nothing. We have tried flipping the controllers and USB ports. We switched sidecars and modules. To no avail. Can we get some help?

[rwood359]

We are having problems with the bench top tank drive test. When the controller in USB port 1 is moved both motors spin and the jaguars indicate in opposite directions. The controller in USB port 2 does nothing. We have tried flipping the controllers and USB ports. We switched sidecars and modules. To no avail. Can we get some help?
The documentation says to install a jumper from signal to +5 for arcade mode and no jumper for tank mode. The inputs on the DS are floating, so you need to install a jumper between signal and ground on DS DI 1 to select tank mode. We used a PWM cable and tied black to white and installed the three pin cable on the DS.

[Ziaholic]

Team 2449 is also on the list of A-OK outta' the box testing.

Our tank and arcade tele-op setups worked just fine, and we didn't need to tie down any floating inputs. Both joysticks performed as advertised. And the Jaguar controllers performed great. (my first time w/ Jag's, looking forward to seeing what they can do)

Our Autonomous mode operated both motors for 2 seconds.

We haven't progressed to the wireless and/or firmware updates yet, but will tackle that (and more) this weekend.

Our only "hiccup" to report at the moment is a 5v status LED on our PD that doesn't want to illuminate. But I've read that we are not alone and the 5v bus is working, it's just the LED that isn't. (we haven't proven that yet, but I'm optimistic)

My only complaint was that I had to steal the 120amp main breaker from our '08 bot to get our test up and running. I think they should've tossed one of those into our "early" kit of parts.

NOTE: When instructed to add the 2nd jumper for Auton mode(on Digi-In2), we did NOT remove the 1st jumper (on Digi-In1). I'm not sure if it's relevent, but some of the failed auton folks might try that.

[blhenze]

When doing the bench top the 12 V LED and the 24 V LED lit, the 5 V LED did not. Does something requiring 5 V need to be attached to make it light or are we doing something incorrectly?

[Mark McLeod]

When doing the bench top the 12 V LED and the 24 V LED lit, the 5 V LED did not. Does something requiring 5 V need to be attached to make it light or are we doing something incorrectly?
You're probably okay.
Many teams have reported that just the 5v LED isn't working, but the WAGO outlet actually does have 5v power.
Your camera gets wired into that.

[billbo911]

I just heard on my way home from work that the kit was delivered to the home of the teams adviser. He will be bringing it to the school tomorrow. I asked him to have the kids inventory it first, then start following the documentation to do the OBT. I will be tied up at a management luncheon tomorrow until mid afternoon, so much for sneaking out early. :rolleyes:

I only have one question now, how am I going to sleep tonight? :eek: