RoboteQ GB Project

Power wheelchair board for REAL info!

POWERCHAIR MENU! www.wheelchairdriver.com/powerchair-stuff.htm

Re: RoboteQ GB Project

Postby Burgerman » 05 Oct 2019, 12:54

A configuration setting, and or a script setting, but not a clue where as its complicated!
User avatar
Burgerman
Site Admin
 
Posts: 65050
Joined: 27 May 2008, 21:24
Location: United Kingdom

Re: RoboteQ GB Project

Postby stevelawiw » 05 Oct 2019, 13:11

It is! I'm leaving that for a bit and connecting the r/c today
stevelawiw
 
Posts: 664
Joined: 21 Jul 2012, 20:55
Location: Isle of Wight, UK

Re: RoboteQ GB Project

Postby LROBBINS » 05 Oct 2019, 14:37

(1) Was this with a script running or not? If yes, pause the script and see if both joystick axes work. If still doesn't, then not in script.

(2) Pause any script, un mute and see how it responds to Roborun sliders. Make sure to plot MotorCommand1 and MotorCommand2 in the Run tab graph. Does one slider move both commands in the same direction, and the other slider move them opposite? If not, there's probably a configuration problem. Remember to click mute again before running any script.
LROBBINS
 
Posts: 5543
Joined: 27 Aug 2010, 09:36
Location: Siena, Italy

Re: RoboteQ GB Project

Postby stevelawiw » 05 Oct 2019, 15:24

Hi Lenny no script running although it is onboard I'm keeping things as simple as possible to just get the basics functioning. It's all still a bench project so far. I just tried the sliders and they perform as you describe with mixing mode 2 set
stevelawiw
 
Posts: 664
Joined: 21 Jul 2012, 20:55
Location: Isle of Wight, UK

Re: RoboteQ GB Project

Postby LROBBINS » 05 Oct 2019, 15:57

Next step, carefully check calibration of your two joystick axes using the calibration tool in the configuration tab set to "Auto Center". If necessary do this with only one axis connected at a time. Now to they behave better?

If not, double check your wiring of the joystick to Roboteq connector. Do remember that the physical pins and logical pins are not numbered the same. Are there any shorts? Any cross connections between the two joystick outputs? The Joystick 5V wire connected to a Roboteq 5V pin (if it doesn't have its own 5V supply) and the joystick ground wire connected a Roboteq ground pin?

Instead of the joystick, temporarily wire two 5k ohm pots to the joystick inputs. Do they work properly?

It's hard to troubleshoot at a distance, so try to keep your eyes and mind open as you're doing these things as I've almost certainly not thought of all possibilities. I'm not infrequently flummoxed by an obvious mistake has been staring me in the face over and over before I spot it. Just as true for hardware as it's been for software.
LROBBINS
 
Posts: 5543
Joined: 27 Aug 2010, 09:36
Location: Siena, Italy

Re: RoboteQ GB Project

Postby stevelawiw » 05 Oct 2019, 19:55

Thanks Lenny.
I've already calibrated the inputs many times. I rechecked my soldering while I was attempting to connect some r/c equipment it was all ok. One thing I think rules out it being any external connection fault is that on the run tab I can see throttle and steering values changing on Ain1 and Ain2 so the controller is seeing the joystick movements.

I haven't been able to see if the r/c control worked any better as the wires I had turned out to have the wrong connectors for the receiver
stevelawiw
 
Posts: 664
Joined: 21 Jul 2012, 20:55
Location: Isle of Wight, UK

Re: RoboteQ GB Project

Postby Burgerman » 05 Oct 2019, 20:04

On RC what does it do?
If analog inputs are reversed, or swapped what does it do?
Does each work individually on its own?
Can you zip and send your settings file so I can take a look for anything obvious?
Does the calibration look normal? And how are you doing it? For e.g. guard band settings. Absolute or relative reading volts?

>>> I can see throttle and steering values changing on Ain1 and Ain2 so the controller is seeing the joystick movements.
What are the channels, and the readings at neutral, and at full stick in each direction?
User avatar
Burgerman
Site Admin
 
Posts: 65050
Joined: 27 May 2008, 21:24
Location: United Kingdom

Re: RoboteQ GB Project

Postby stevelawiw » 05 Oct 2019, 20:48

20191005_202526[1].jpg.png
20191005_202350[1].jpg.png

Two inputs calibration details and the profile
Haven't got rc working yet to try it
Thanks for looking 99.9% I'm doing something really stupid :?
Attachments
Profilea.xml
(40.9 KiB) Downloaded 127 times
stevelawiw
 
Posts: 664
Joined: 21 Jul 2012, 20:55
Location: Isle of Wight, UK

Re: RoboteQ GB Project

Postby stevelawiw » 05 Oct 2019, 21:01

If analog inputs are reversed, or swapped what does it do?
Does each work individually on its own?

Not tried reversing or swapping them, but if I disable steering them throttle works, when I enable steering throttle stops working
stevelawiw
 
Posts: 664
Joined: 21 Jul 2012, 20:55
Location: Isle of Wight, UK

Re: RoboteQ GB Project

Postby Burgerman » 05 Oct 2019, 21:45

Cant open with my version 1.7.
User avatar
Burgerman
Site Admin
 
Posts: 65050
Joined: 27 May 2008, 21:24
Location: United Kingdom

Re: RoboteQ GB Project

Postby LROBBINS » 05 Oct 2019, 22:06

Bingo. I may not have specified the right brushless controller in order to open your config file off line, but for this they're all the same. Take a look at this pic and think about what you see before you read the answer below it:
analog input config.jpg
analog input config.jpg (16.65 KiB) Viewed 4297 times
Look at the Input Use lines. You have both inputs set to control both MotorCommand1 AND MotorCommand2! They will decidedly conflict with each other. Just as with the sliders, with mixing=2 one input should control MotorCommand1 (Throttle) and the other input should control MotorCommand2 (Steering). You've probably stared at this many, many times without it registering. Some psychologist or other has probably had lots to say about this very human tendency.
LROBBINS
 
Posts: 5543
Joined: 27 Aug 2010, 09:36
Location: Siena, Italy

Re: RoboteQ GB Project

Postby stevelawiw » 05 Oct 2019, 22:42

It works now :oops:
Thank you!
RTFM! I thought it was referring to left and right motors and not throttle and steering :shock: thank you both and sorry for wasting your time!
stevelawiw
 
Posts: 664
Joined: 21 Jul 2012, 20:55
Location: Isle of Wight, UK

Re: RoboteQ GB Project

Postby LROBBINS » 06 Oct 2019, 08:04

If I can help solve a problem, you've not wasted my time!
LROBBINS
 
Posts: 5543
Joined: 27 Aug 2010, 09:36
Location: Siena, Italy

Re: RoboteQ GB Project

Postby Burgerman » 06 Oct 2019, 10:52

Some of those roboteq settings are not intuitive or described well. Lots of lateral thinking needed to get it to behave as you want. And you really need to name everything, and disable or configure all the unused options too. Or some unexpected stuff can happen. Esp with shared input/output options. And end points on inputs. I always set those around 10 to 20 higher and lower than actually needed at full stick. To prevent some guardband issues. And 101 other things can complicate stuff.

e.g. on this input, I would set the lowest at 10 to 20 less than it displays here download/file.php?id=12112&mode=view to avoid noise causing it to clip at the ends.
So set minimum to 755, max to say 3500. And I would watch the centre by eye, and see the highest, and lowest the noise, and temperaure, and other variations caused it to go to. And pick a number in between, for a true centre. It will vary a little due to centering errors and temperature, keep checking it and set to the centre of what you see over a day or two. Manually. In this way you will not clip or exceed the edges, and will have a better more accurate reliable centre. It may also allow you to reduce deadband. I use 3%. And set ACTION ON MAX and ACTION ON MIN to safety stop. In case of wiring or joystick faults. With a 10 to 20 headroom at each end. Weather you leave 10, 20 or 40 or more, depends on how good your joystick and sheilding is.
User avatar
Burgerman
Site Admin
 
Posts: 65050
Joined: 27 May 2008, 21:24
Location: United Kingdom

Re: RoboteQ GB Project

Postby stevelawiw » 06 Oct 2019, 16:54

Thanks for the advice in the above post.
rc now. Can I ask if you power thr receiver from the controller? and Do you have a switch on the joystick to turn it on/off? Oh and do you use a relay for this?
sorry I have searched for this info, I know it'll be somewhere but I can't find it.

Thanks in advance
stevelawiw
 
Posts: 664
Joined: 21 Jul 2012, 20:55
Location: Isle of Wight, UK

Re: RoboteQ GB Project

Postby Burgerman » 06 Oct 2019, 17:21

Yes, and yes. I use a seperate 5v for this so it does not affect the joystick input via any voltage drop when turned on. And a 10 channel DSMX spectrum reciever with 4 satelites, and a PROGRAMMABLE failsafe set to go to zero (centre) on the relevant channels on signal loss.
User avatar
Burgerman
Site Admin
 
Posts: 65050
Joined: 27 May 2008, 21:24
Location: United Kingdom

Re: RoboteQ GB Project

Postby stevelawiw » 10 Oct 2019, 14:26

Thanks! Got the rc working, except not with the posh spektrum gear and satelites :shock: but it's working :thumbup:

Next problem! So I bought this cheap contactor off ebay :lol:
conntactor.jpg

And when I apply 14v to the small silver contacts the main contacts do not close (I've tried in my circuit and with a 12v battery and a multimeter on the bench) :problem:
So it's faulty isn't it?
It seems obviously faulty to me but I've been wrong before so I'm just looking for confirmation please!
stevelawiw
 
Posts: 664
Joined: 21 Jul 2012, 20:55
Location: Isle of Wight, UK

Re: RoboteQ GB Project

Postby Irving » 10 Oct 2019, 15:15

stevelawiw wrote:Thanks! Got the rc working, except not with the posh spektrum gear and satelites :shock: but it's working :thumbup:

Next problem! So I bought this cheap contactor off ebay :lol:
conntactor.jpg

And when I apply 14v to the small silver contacts the main contacts do not close (I've tried in my circuit and with a 12v battery and a multimeter on the bench) :problem:
So it's faulty isn't it?
It seems obviously faulty to me but I've been wrong before so I'm just looking for confirmation please!


Got a link?

I'd hazard a guess its a 48v coil, not a 12v one... The voltage is irrelevant to the contacts, it is almost always the coil pull-in voltage.
C5/6 A (complete)
Puma 40, 75Ah LiFePO4 (pic is on tour @ Whistler, BC)
Puma 40 backup, 73Ah MK (for now)
Spectra Plus (weedy 40Ah MK)
User avatar
Irving
 
Posts: 2114
Joined: 04 Dec 2012, 11:51
Location: NW London


Re: RoboteQ GB Project

Postby Irving » 10 Oct 2019, 16:49

Yep, 12 to 14v should pull in, drop out around 9v.

Use your multi-meter to measure resistance between small terminals...

Though, having looked at other vendor examples, the voltage in the part # is the coil volts. So this could be a 48v coil...

This is the 12v version....

Image

This is what happens when you buy from clueless vendors
C5/6 A (complete)
Puma 40, 75Ah LiFePO4 (pic is on tour @ Whistler, BC)
Puma 40 backup, 73Ah MK (for now)
Spectra Plus (weedy 40Ah MK)
User avatar
Irving
 
Posts: 2114
Joined: 04 Dec 2012, 11:51
Location: NW London

Re: RoboteQ GB Project

Postby stevelawiw » 10 Oct 2019, 17:04

Coil resistance 375 :problem:
stevelawiw
 
Posts: 664
Joined: 21 Jul 2012, 20:55
Location: Isle of Wight, UK

Re: RoboteQ GB Project

Postby stevelawiw » 10 Oct 2019, 17:11

Thanks Irving but I think it's me!
This is a very steep learning curve I'm on, but I'm undaunted, I'm making progress slowly and that mistake only cost £13 :roll:
I just tried it at 48v. Guess what? It works.
stevelawiw
 
Posts: 664
Joined: 21 Jul 2012, 20:55
Location: Isle of Wight, UK

Re: RoboteQ GB Project

Postby Irving » 10 Oct 2019, 17:48

stevelawiw wrote:Thanks Irving but I think it's me!
This is a very steep learning curve I'm on, but I'm undaunted, I'm making progress slowly and that mistake only cost £13 :roll:
I just tried it at 48v. Guess what? It works.

well on the basis you shouldn't use part of a pack, then why not use 48v?

48v/375ohm = 0.128A. The 12v version will need roughly the same ampere-turns so will need 4x the current = 0.5A, 6W

Are you planning to drive this from the Roboteq, hence needing 12v? Can Roboteq source/sink 0.5A?
C5/6 A (complete)
Puma 40, 75Ah LiFePO4 (pic is on tour @ Whistler, BC)
Puma 40 backup, 73Ah MK (for now)
Spectra Plus (weedy 40Ah MK)
User avatar
Irving
 
Posts: 2114
Joined: 04 Dec 2012, 11:51
Location: NW London

Re: RoboteQ GB Project

Postby stevelawiw » 10 Oct 2019, 17:58

I waas never going to use part of a pack, I intended to fire it up through a 12v inverter but I could use one of my SSRs to control it at 48v. Do you think the main contact will be ok at 48v then? I tought the main contacter was rated for 48v, not the coil. I could still use it if it will cope.
stevelawiw
 
Posts: 664
Joined: 21 Jul 2012, 20:55
Location: Isle of Wight, UK

Re: RoboteQ GB Project

Postby Irving » 10 Oct 2019, 18:53

stevelawiw wrote:I waas never going to use part of a pack, I intended to fire it up through a 12v inverter but I could use one of my SSRs to control it at 48v. Do you think the main contact will be ok at 48v then? I tought the main contacter was rated for 48v, not the coil. I could still use it if it will cope.


That will work. The main contactor is rated for 200A but strangely at 14v. Basically if you open the contacts at any serious current you'll get a spark. If the contacts don't open far enough and/or quick enough that spark has sufficient energy to create a plasma maintaining the current flow across the small gap - essentially turning the contacts into a mini-welder. The result will be some badly melted contacts and other collateral damage. What that seems to be saying, in the absence of any real data sheet, is that at 200A, 14v is the maximum voltage across the contacts before problems may arise. Now, switching at a lower or zero current will allow a higher voltage. If this is intended as an isolator and/or emergency cutout the only time you're likely to be switching 200A is at stall when motor back-emf = 0 so the full 48v is across those contacts. It'll probably be ok, but they won't guarantee it. Your call or get them to swap it over when the 12v one comes back in stock.

I personally wouldn't use an SSR to drive it unless the SSR is intended for low-current DC usage. Most SSR are intended to switch AC and while they will switch DC there are limitations on the minimal current allowed. Depends on yr SSR, got a spec sheet or linky?
C5/6 A (complete)
Puma 40, 75Ah LiFePO4 (pic is on tour @ Whistler, BC)
Puma 40 backup, 73Ah MK (for now)
Spectra Plus (weedy 40Ah MK)
User avatar
Irving
 
Posts: 2114
Joined: 04 Dec 2012, 11:51
Location: NW London

Re: RoboteQ GB Project

Postby Burgerman » 10 Oct 2019, 19:12

SSRS are available as AC/AC AC/DC DC/AC DC/DC and pretty cheap. Protect with a diode across the in and out. Mine are fine, just cheap 20A max eBay ones.
User avatar
Burgerman
Site Admin
 
Posts: 65050
Joined: 27 May 2008, 21:24
Location: United Kingdom

Re: RoboteQ GB Project

Postby stevelawiw » 14 Oct 2019, 11:50

If you didn't have a PL8, and you were tight and didn't mind a bit of a faff come charge time, could you use Burgermans original idea for charging a chair in a car at single cell voltage?
batsc.JPG

Thanks to woody for original pic
So for 12v read 3.2v and rather than two cells there would be 8 LifePO4s, so a bank of 8 andersons.
And charge at 3.5 - 3.6v? So no need for any ballancing.
And then remove the 8 andersons in your charging loom and plug in the anderson jumpers as in the pic only there would be four, not one to plug in probably use 8awg silicon.
stevelawiw
 
Posts: 664
Joined: 21 Jul 2012, 20:55
Location: Isle of Wight, UK

Re: RoboteQ GB Project

Postby Irving » 14 Oct 2019, 13:33

stevelawiw wrote:If you didn't have a PL8, and you were tight and didn't mind a bit of a faff come charge time, could you use Burgermans original idea for charging a chair in a car at single cell voltage?
batsc.JPG

Thanks to woody for original pic
So for 12v read 3.2v and rather than two cells there would be 8 LifePO4s, so a bank of 8 andersons.
And charge at 3.5 - 3.6v? So no need for any ballancing.
And then remove the 8 andersons in your charging loom and plug in the anderson jumpers as in the pic only there would be four, not one to plug in probably use 8awg silicon.

In theory yes, you'd be charging all 16 cells in parallel so would need a 3.6v high-current charger. I've been toying with the idea of an onboard charger that has an isolated 3.6v 100w charger for each cell group (8 in total for 24v). Then no balancing required and one simple off chair connection at either 240v or 12v (to be decided).
C5/6 A (complete)
Puma 40, 75Ah LiFePO4 (pic is on tour @ Whistler, BC)
Puma 40 backup, 73Ah MK (for now)
Spectra Plus (weedy 40Ah MK)
User avatar
Irving
 
Posts: 2114
Joined: 04 Dec 2012, 11:51
Location: NW London

Re: RoboteQ GB Project

Postby LROBBINS » 14 Oct 2019, 19:25

How about 8 of these:
https://www.digikey.com/product-detail/en/texas-instruments/TPS53355DQPT/296-29519-1-ND/2751282
Board design is a bit tricky (3-layer with vias in pad and lots of heat sink area needed) and input can't be > 15V, so at 12 V you'd need near 70Amps for the 800 Watts total output. Everyone seems to have these, and RS has a much lower price (but is evidently discontinuing it as it's only "until supply exhausted").
LROBBINS
 
Posts: 5543
Joined: 27 Aug 2010, 09:36
Location: Siena, Italy

Re: RoboteQ GB Project

Postby Irving » 14 Oct 2019, 19:56

LROBBINS wrote:How about 8 of these:
https://www.digikey.com/product-detail/en/texas-instruments/TPS53355DQPT/296-29519-1-ND/2751282
Board design is a bit tricky (3-layer with vias in pad and lots of heat sink area needed) and input can't be > 15V, so at 12 V you'd need near 70Amps for the 800 Watts total output. Everyone seems to have these, and RS has a much lower price (but is evidently discontinuing it as it's only "until supply exhausted").

No good as not isolated output. You need them isolated if they have a common input supply. Yes, the 70A (prob nearer 75A) @ 12v is an issue but for on-board charger its either 12v for direct plug in to car or 120/240v. 24v makes no sense 'cos if home/indoors you have 120/240 and if on the road you're unlikely to have 24v. Also those units are not easy to control CC/CV transition (they don't really do CC).
C5/6 A (complete)
Puma 40, 75Ah LiFePO4 (pic is on tour @ Whistler, BC)
Puma 40 backup, 73Ah MK (for now)
Spectra Plus (weedy 40Ah MK)
User avatar
Irving
 
Posts: 2114
Joined: 04 Dec 2012, 11:51
Location: NW London

PreviousNext

Return to Everything Powerchair

Who is online

Users browsing this forum: No registered users and 279 guests

cron

 

  eXTReMe Tracker