ixxat USB-to-CAN compact & Sevcon DLD flashing

If you don't already have a tool to sniff the CAN communications here is one I used. It is cheap, open source and works really well (tested on a car): http://www.fischl.de/usbtin

I don't know much about CANOpen but here are some informations https://stackoverflow.com/questions/1232624/how-to-program-a-simple-canopen-layer
And CANopenSocket is a stack that can be used on Linux (PC, Raspberry, ...) with the USB interface above (and others that support SocketCAN)
 
Hello,
I have tested the DVT version "Customer Version 12.568" with IXXAT USB-to-CAN 2, and dont get a CAN connection.
Anybody can share a DVT version that works with my IXXAT?

best regards, Michael
 
Not up to date on this thread:

FYI: We had some DCF files land in the Schindler Engineering Repository.
http://www.schindlerengineering.com/public/repository/

We are not in a position to distribute anything proprietary like DVT software... but if you find yourself with a motor that you are trying to get running we may have some motor params that will help.

75-7
75-5
Other
Motenergy PMAC's as seen on 2012 era Zero's
Markacycle Motorcycle Hub Motor

Still nothing for anything Joby - no idea if they will work.

We have some experience now...
Honest truth?
If you are an enthusiast looking to get on the road just buy a Kelly.

http://kellycontroller.com/kls12301-8080ips24v-120v300asinusoidal-bldc-motor-controller-p-1397.html

They come in UVW and Sin
Yea... they might blow up... but look....

Sevcon is F'ING COMPLICATED beyond your wildest speculation. Really.... Seriously... this is not a toy that you will tinker with and get right.
Plan on 3-9mo of lessons learned if you want to run a Sevcon
Plan on 3-9 hours if you want to run a Kelly

Kelly = PLUG AND PLAY with reasonable performance

Sevcon = Infinite possibilities... which you better be ready for... as there is NO BASELINE what so ever... unless you are lucky enough to understand PID tuning and have lots of spare time.

You have been warned.

Until next time... brush your teeth with gasoline and build something illegal... especially if it flies in the face of the ICE paradigm.
Spread the truth
Celebrate EV
and... Get ready... to have energy flowing both into your EV... and from your EV.... for next generation micro-grid.

Anyone who says it is not going to happen is a pure hater. :roll:

-methods
 
Idiot.png

I was trapped in IXXAT hell for a couple days....
I think that was me....
I went FULL RETARD and...
Bugged smart guys with a problem.... before I painstakingly eliminated every possibility of it being my fault.

How it happened... (mindflowdump)

(With regard to IXXAT dongles going bad, isolated vs non-isolated, counterfeit vs genuine)

When the dongles start going they become inconsistent. Very hard to diagnose (especially) when you are using them to diagnose. :x

When the knock-off went... it first required USB unplug resets... more and more... then eventually quit completely to where 10 re-plugs would not bring it online. Pretty sure it died while I was operating the Dyno. Probably when something stopped suddenly. Big shoot-thru of some sort.

SubEdit -> I wonder if Terminating more aggressively would sink noise spikes and protect the device from Transients????

Now I have an genuine IXXAT giving up the ghost. Compact V2

I change the baud (reset...) see a few values pop up in DVT... glitch (frozen)... then it traps out.

Anyone reading should be thinking:
1st) His CAN termination is off
2nd) Is there even CAN traffic?


I have the dongle attached to a Dsub-9 connector populated with only CAN_HI, CAN_LO, and GND
Twisted pair for CAN, ground not twisted in
Custom cable - short run - to the 35pin Sevcon Connector

I have the loop-back installed... terminating the loop with 120ohms in one cable... (thats the Sevcon internal termination)
I have a ClearView installed in another... (replacing the loop back) where termination happens out at the ClearView (with its loop back)

I have confirmed traffic by connecting ClearView and monitoring voltage.
I have confirmed no stale data by varying the voltage and monitoring it on the bar graph
No pack voltage goes to the ClearView... only GND, 24V, CANhi, CANlo... so there is a 100.0% chance good data is on the CAN bus

I dial in the ClearView settings to view the Auto-Baud and confirm 500khz...

So... Unknown controller working... spitting out CAN to a known good ClearView... (termination should be reviewed here... hrm...)

I attach a Was-Known-Good IXXAT Compact V2...
Buss gets fouled sometimes - locks up

Sometimes a few bytes hit DVT if I cycle thru different bauds to reset the buffer in the box

Anything from No CAN light, to Solid CAN light, to blinking CAN light.

Pretty sure there are programmable loads in the IXXAT boxes... you can turn termination on and off... (Found my Fail... these are not NI CAN boxes...)
Thinking some nefarious Lukeman may have changed the settings in my tool?
I was gone for like... a month... but usually nobody plays with the dongle but me :lol:

CANterm.png

Bah... no internal termination that can be turned on and off...


Ok - So I have two CAN networks.

One goes ClearView to Gen4 to Dongle
One goes Gen4 to Dongle

In the Gen4 to Dongle... Termination is at the Sevcon via Loopback
Termination at the Dongle is... eh... eh hem... uh... well you see boss... :oops:

CANterm.jpg

Oh... Yea... eh... I see...

On the setup where the ClearView is attached... we have Termination at the Clearview, semi-short run to the Sevcon, then super short run to the Dongle
Single ended again

Wow failing.

Anybody see how hard I failed?

TAKE HOME:
This is one of the biggest take-homes you should take home from any of my blatherings...

You can develop a process..
A detailed and disciplined procedure...
You can follow it for days, weeks, months, or even years.......

Then one day something changes.
Maybe the weather
Maybe the ... position of the moon relative to your work space (almost serious here...)

And that luck suddenly runs out.

Yep... luck.
It runs out sometimes.
Best to recognize it when it does...

In my case I was running without CAN termination for EVAR... I mean for a real long time.
It just worked
Out of luck... and my good looks...

Then something changed. (du know... maybe blew out some protection diodes or something...)

Now... Termination seems mission critical.

I suspect... that when I go back in to work today... that when I plug in the DB-9 that I will pick up at Santa Cruz Electronics (with the built in Termination)... that my problems... will dissolve away.

Ok - how do we write it up?
Pfffftttt... Lessons learned
Expensive Lessons... on the importance of CAN termination.

NOw... if I go in and that termination DOES NOT get me rolling again... and I DO have another jaked Dongle... Grrrr..... That will really suck :mrgreen:

Place your bets

-methods
 
Turns out it was the outlier

Not one, but two blown IXXAT dongles.
Both Isolated
Compact and V2 Compact

V2 Compact blew the digital isolator. Turned it into a heater. Threw me off because it would work for 3 or 10 seconds... then overheat and shut down.
Upon clear examination the dongle was actually warm... so...

No warm dongles :idea:

I diagnosed it by finding 6 ohms from CAN HI to GND

The V1 was also blown (held down the CAN bus) but did not present in any easily measurable way.

Root Cause
I had been using both dongles on 2 different bikes. Both were working first half of the day
Second half of the day we were working on a 2012 Zero that was half assembled
On that frame... to get the 35pin off the Sevcon easily.. you need to drop the controller a bit.
Dropping the controller (on this particular semi-custom bike) puts tension on the primary Anderson

With the dongle attached... if the Anderson pops out of the battery... a measurable HV spike hits the 3 CAN pins.
Doh... Expensive lesson.

As for the termination... I did go ahead with making a terminator (sub 9 pass-thru) and it did bring the bus from 120ohms single ended to 60ohms at the ends... but it was a bit late.

On the particular bike we were working on, the terminator was not required, because either the MBB or BMS was creating the matching termination for the Sevcon... so... adding another 120 ohms would have just loaded down the bus beyond spec.

-methods
 
Good day sir.

I have already described in excruciating detail the answer to your simple query. Please use the search function :idea:

To inspire you - here are some pictures of my YESTERDAY

Surfing_003.jpg

Beautiful girls in bikinis, powerful surf, warm sunlight

Surfing_014.jpg

Marinating tapas in various flavors... next to DIY projects

Surfing_017.jpg

Sunset...

Surfing_021.jpg

Sunset with a beautiful lady :shock:

Surfing_025.jpg

kids doing back-flips off of 2 story sand hills

Surfing_027.jpg

and... LANDING THEM

Surfing_028.jpg


To inspire you further - here are pictures of how it works
Hard work - getting your hands dirty - and GIVING BACK more than you TAKE

Surfing_044.jpg

Get your tools brother (The Search Button)

Surfing_045.jpg

Dig into that confusing mess of data (my raw postings)

Surfing_046.jpg

And catch yourself a fish!

-methods
 
De-Motivational aside:

The Gen4 manual describes this in minutia
You can find this manual at ThunderStruck EV

The Size two is over priced and under performs
The Size 6 is under priced and over performs
I suggest making room for the extra few square inches and pounds - as its an order of magnitude upgrade

There is a clear programming connector on the Gen2 - but I have never played with one - and am told that they are "meh"

Good job posting (I see you have 33) keep it up.
We highly value contributing members.

Most of them get job offers at Google with starting pay around $300k :)

-methods

(I am snippy today... ExWife is cock blocking on custody again... unbelievably frustrating... but SSDD)
 
I was approached by an enthusiast to solve a problem.
I think I set him on the right course - but we will see.
Here is a clarification around "bricking" the sevcon


1) I do not believe you can actually brick one - its just a matter of - digging your way back out.

2) I previously posted some descriptors:

New Terminology:
Black-Brick = Sevcon in Bootloader mode (solution, get out of bootloader)
Green-Brick = DLD disrupted during transfer, use command line to force into bootloader at next restart, flash DLD

Where black means no LED
Where Green means solid LED but controller does not work or respond
Where blinking LED means all good but with errors
Where solid LED with comms means you are golden

To clarify...
(and this is from memory - I am not typing live while I do it - so there may be a mistake or two)

Black-Brick
* Should you command the sevcon to go into bootloader using DVT the LED will go out. This just means that you are in boot loader and you can definitely recover from that. Tell it to exit boot loader and power cycle

* Remember to power cycle...

Green-Brick
* Should you disrupt a DLD transfer the controller will get pissed and lock up in a solid green light. You can easily recover from this and I have posted the solution elsewhere. I have tested the solution as have others and it works.

Think of Green Brick as being caught in some no-mans land.
Since we were trying (previously) to load up a DLD... lets get back there
You queue a command set, power cycle, and get the controller to boot back into boot-loader

NOTE

You can reason out that this makes sense using the following logic:

1) Everything has a base bootloader - it must - otherwise how could you load a DLD (So the DLD is not the boot loader)

2) We have a bootloader, takes a DLD, which takes a DCF - where DCF is settings and DLD is algorythems

3) We can always get something to boot into boot-loader... but you have to know the magic sequence (Like F2 or F10)... or a button ... and this usually happens in a small time window when booting* (hence why you load DVT with bd then turn on the controller...)

4) CAN has a buffer - hardware and then firmware

5) So - even if you brick out into black or green - you can always get the fundamental hardware to get back to SOME known good... if there is not hardware damage.

Thats what I have seen
I have not yet seen a bricked controller that was not physically damaged due to over-voltage, heat, etc.

-methods
 
Hi methods
Do you still have the files for Sevcon controllers?

I need a DLD for my size 6, as it seems it has a faulty firmware or something.

Thank you
 
john61ct said:
4 years old thread...
Yes I know. I can see the date.
I was just wondering if there is a new place that methos use to keep the files.
I'm needing a dld for a size6 controller.
 
Back
Top