10S custom skate ESC: testers wanted!

[lizard]

From what I gather from this thread and the pictures posted in it, the VESCs from enertion seem not to be made with great care and quality. Different parts used, solder joints that look like cold (or maybe "lukewarm") solder joints to me. Parts missing. Parts not aligned properly. Leg bent on one of the FETs. Solder sprinkles. Different software/config on them. Sorry if it's not nice to say that, but that's simply how it looks like to me.

The fact that Flier told onloop they didn't make any money on the order and tried to ask higher prices seems to support that. Seems what they delivered was like "We don't care anymore, let's just get these things out quickly and be done with it ..."

I wouldn't just put it on the board and give it a go, it might break because of something that would have been easy to fix beforehand.

My recommendation to people who have a VESC by onloop is:
- Don't power it up
- Remove the shrinkwrap
- check the whole PCB with a magnifier thouroughly, use a photo of an ESC by Benjamin for comparison. If there's something not right, fix that first.
- Use a fuse. If the guy who developed this ESC tells you that, don't question it. Do it!
- Now power it up
- Don't use settings copied from other motors, follow the measuring procedure of the BLDC tool

 

[jacobbloy]

Just need know some one with a VESC and a PC with windows to do me a fav!

I need some one to download this software.
http://www.flexihub.com/download/flexihub.exe

Then plug your VESC into you PC and share the USB device with my developer using these account logins
jacob.bloy@gmail.com
password: 98CeXOCQcvK2tvu0

This will allow him to debug the windows serial code we have been working on. Pm me if your in!
This software only remove accesses the USB port you allow it to and not access your desktop.

I would do this but I'm at a funeral in the UK ATM and don't have a VESC with me!

 

[vedder]

My recommendation to people who have a VESC by onloop is:
- Don't power it up
- Remove the shrinkwrap
- check the whole PCB with a magnifier thouroughly, use a photo of an ESC by Benjamin for comparison. If there's something not right, fix that first.
- Use a fuse. If the guy who developed this ESC tells you that, don't question it. Do it!
- Now power it up
- Don't use settings copied from other motors, follow the measuring procedure of the BLDC tool

Good recommendation. If you have a lab power supply, use that first when you test (and make sure to connect minus first to not fry the USB port with an unisolated supply).

Here are some pictures of my beta batch that I'm going to start shipping out today that you can use for reference:
http://home.vedder.se/public/VESC/VESC_batch1_front.JPG
http://home.vedder.se/public/VESC/VESC_batch1_back.JPG
The photos on my blog are hand soldered by me, so they don't have as high quality as this beta batch.

Also, since the some of the enertion BETA VESCs have an undefined configuration, I recommend that you first load the turnigy_sk3_168kv xml file and then to the customization according to my tutorial.

 

[benj]

Who solved the usb connection issue, I can't get mine to connect no matter what I do.

Did you connect VESC to a power suppy? It needs USB and the power leads plugged in - preferably to a lab power supply (so you can provide a low current)

 

[lizard]

Is there a way to limit current for people without lab power supplys? I read on an R/C forum sometime ago, that car lamps can be used as current limiters for testing purposes, is this okay with the VESC? Or maybe just a fuse with some low Amp value like 5A for the first tests?

Just to get things summed up, this would be a safe way?

- Don't power it up immediately
- Remove the shrinkwrap
- check the whole PCB with a magnifier thouroughly, use a photo of an ESC by Benjamin for comparison. If there's something not right, fix that first.
- For first powering up and motor configuration, use a lab power supply with current limit or find some other way to limit current
- Connect USB
- Connect minus battery cable first, then plus
- Start BLDC tool
- Load turnigy_sk3_168kv xml file
- Follow measurement / motor detection procedure on Benjamin's blog
- When everything is setup and runs, use a fuse, Benjamin recommends around 40A

 

[vedder]

Is there a way to limit current for people without lab power supplys? I read on an R/C forum sometime ago, that car lamps can be used as current limiters for testing purposes, is this okay with the VESC? Or maybe just a fuse with some low Amp value like 5A for the first tests?

Just to get things summed up, this would be a safe way?

- Don't power it up immediately
- Remove the shrinkwrap
- check the whole PCB with a magnifier thouroughly, use a photo of an ESC by Benjamin for comparison. If there's something not right, fix that first.
- For first powering up and motor configuration, use a lab power supply with current limit or find some other way to limit current
- Connect USB
- Connect minus battery cable first, then plus
- Start BLDC tool
- Load turnigy_sk3_168kv xml file
- Follow measurement / motor detection procedure on Benjamin's blog
- When everything is setup and runs, use a fuse, Benjamin recommends around 40A

Good summary. I didn't think about using car lamps before, but that should work perfectly. One or two parallel 12V 100w car lamps should allow enough current to finish the configuration.

 

[chuttney1]

Who solved the usb connection issue, I can't get mine to connect no matter what I do.

Did you connect VESC to a power suppy? It needs USB and the power leads plugged in - preferably to a lab power supply (so you can provide a low current)

I did that and everything Vedder said on his blog post. Found out the reason is never connected was due to a solder bridge on the MCU and it fixed that, but now I got the DRV8302 error code. So close, yet so far away. I did assemble the board according to Vedder's specification including omitting the resistor for R16, R6, and capacitor C26. The DRV8302 is solder with a hot air gun and I checked for solder bridges to which there are none.

 

[okp]

Who solved the usb connection issue, I can't get mine to connect no matter what I do.

I think okp solved it by using a different USB cable. Do you have a virtual machine or a "real" install?

Confirmed. New/other USB cable make it work.

if quality checks are done as expected (and I remember that quality checks have been discussed quite hardly lately on my eboard thread)... we should not have VESC BETA released with different configuration (and even one - especially on a small batch). As already said, one of my VESC had weird values everywhere when I read the configuration. I don't really care about Flier or whatever (another company could have been picked); this is part of the BETA choice, and I am a customer of the BETA. If at the end people have to spent more money to make this BETA product released by Flier 100% reliable... this would be quite a shame (even if it's BETA)?

Nevertheless, for me BETA, means BETA management and BETA feedback. How many of the BETA have it working perfectly? Maybe there are only a few units who have problems. For me BETA management, means also feedback to the supplier, to the BETA testers (top down communication), to improve stuff and create a medium term relationship based on trust. Not sure going immediately to as second supplier will ensure a full BETA support from end to end (or it means that the choice of Flier was an error).

So how many have the VESC BETA working... without theses values ... nothing can be managed, because as I know we cannot manage what we can't measure. There should be a questionnaire (down on google or wherever) to have the feedback. This is the minimum of the minimum.

For me, the feedback is quite positive at this time (with tests videos I posted, on 6S) - I will go riding next week and provide additional feedback. I will also post photos of my VESC PCB as I am not an electrical engineer. Only remark is about the conf which was not the good one (which has to be clarified as it's a feedback from a BETA tester).

I am no electrical engineer neither a company making business, I am a just a skateboarder/customer ... and I also know as the baseline that BETA without BETA management is just a big mess of everyone trying to run in the same direction... with no clear instructions.

 

[onloop]

Let me enlighten you all with some facts/details of the beta run.

》50 units ordered from flier.... so far most conversation has been about this supplier not the device???
》46 total sold to people all over the world.
》24 sold from enertionboards.com
》22 sold by jacob
》Jacob had about 20 sold initially but to get the qty high enough to get lowest prices I teamed up with jacob.
》known faults reported so far is, One from benj. "Drv8xxx fault"
》don't know if fault from hardware, firmware or user error.
》either way, counting this, we have a 2.18% fail rate.
》however not all the 46 beta testers are reporting data. Maybe just 6 or 8 people. Will try to gather more info via emailing directly.
》firmware (from feb 19th) came pre loaded From flier.
》settings / configuration that are loaded where not defined or setup during manufacture. Up to user to configure.
》Once stock arrived at enertion - motor config / detection was done using enertion motors, settings entered in 4 vescs... no problems.
》bench testing carried out, two vesc connected with servo splitter, on 10S battery..no problems. See video below.
》real world testing carried out, two vesc connected using canbus, no problems. See video below.
》 vedder released new BLDC version day after all enertion units shipped making all BETA incompatible with pre installed firmware.
》work around discovered. Download old version of BLDC TOOL from different location On github.
》Jacob has now develop the apple compatible version of BLDC TOOL. Also offering old firmware compatibility.

So far no other real world testing has been done.
no other news can be given, people need to test.

We are waiting the feedback

[youtube]h_UEydrXhak[/youtube]
[youtube]59d9JLXF2TI[/youtube]

 

[onloop]

Oh I forgot to mention I blew those two up on the bench because they shorted against each other. The usb from one touch a capacitor on the other.

So a few hundred $$ disappeared quickly.

Hence the shrink wrapped vesc.

 

[777arc]

Thanks for reminding everyone that the beta run is going pretty well overall. My beta VESC is still running great!

 

[erwincoumans]

The VESC 4.7 hardware and BLDC tool is working very well for me too. It is great Benjamin Vedder and various providers of hardware put so much effort in this motor controller. I sourced my components directly from Vedder's BOM from Mouser and PCBs from Hackvana, and support Benjamin with some money too.

My only issue is related to dual VESC CAN bus, but that is not related to any particular BETA hardware provider, it is discussed in a different thread.

Congrats to all involved supporting this new motor controller!

 

[chuttney1]

So far no other real world testing has been done. No other news can be given, people need to test.

One test I can suggest is definitely an overload situation such as 200lbs on the ebaord to simulate something close to towing. I believe this will really push the VESC as hard as it can.

 

[benj]

Okp - you are spot on. The Beta feedback is important across the whole chain.

Onloop - perhaps an email out to the VESC customers would get them talking?

Also Onloop you told me that all VESC units had been tested before being sent out. Is that true?

 

[torqueboards]

So far no other real world testing has been done. No other news can be given, people need to test.

One test I can suggest is definitely an overload situation such as 200lbs on the ebaord to simulate something close to towing. I believe this will really push the VESC as hard as it can.

Don't think that's an issue. I'm sure most people are at about 190-200 lbs. I'm close to 200 lbs with a backpack and batteries and VESC or any other ESC which I've used has had no issues with that. I ride up pretty steep hills 20-35% hills although I'm using a dual motor setup which I would recommend for anyone going up steep hills as well as the correct gearing ratio for torque which doesn't overstress your escs and motors.

 

[chuttney1]

Who solved the usb connection issue, I can't get mine to connect no matter what I do.

Did you connect VESC to a power suppy? It needs USB and the power leads plugged in - preferably to a lab power supply (so you can provide a low current)

I do not have a lab power supply. I'm thinking of may getting a cheap, but good one othe wise use either a 9V or 12V battery and a trimpot with some resistors to control the flow of current, but I know this is not the same with a lab power supply for it can deliver a set, desired voltage with such small current of 0.1 amp or even smaller. It means I got to improve my standards on making some VESC by hand with better quality control on the soldering of the MCU and DRV8302 for any solder bridges.

 

[onloop]

Who solved the usb connection issue, I can't get mine to connect no matter what I do.

Did you connect VESC to a power suppy? It needs USB and the power leads plugged in - preferably to a lab power supply (so you can provide a low current)

I do not have a lab power supply. I'm thinking of may getting a cheap, but good one othe wise use either a 9V or 12V battery and a trimpot with some resistors to control the flow of current, but I know this is not the same with a lab power supply for it can deliver a set, desired voltage with such small current of 0.1 amp or even smaller. It means I got to improve my standards on making some VESC by hand with better quality control on the soldering of the MCU and DRV8302 for any solder bridges.

I used an old laptop power supply. 18v at 2A.

 

[onloop]

Okp - you are spot on. The Beta feedback is important across the whole chain.

Onloop - perhaps an email out to the VESC customers would get them talking?

Also Onloop you told me that all VESC units had been tested before being sent out. Is that true?

Yeah. I will definitely email my 24 customers and ask them to post about it. Or at least write a review on my website about it.

Regarding testing, yes I was told they would be tested. I have no idea what testing they did. I doubt they where tested with a motor and a load. I imagine some electronics test to ensure the components where working. And they whwre powered up to install the firmware. You/I need to speak with jacob for more detail as He was the contact with supplier, he managed the technical stuff and production so could possible give more detail.

 

[okp]

hey onloop, great knowledge base initiative

btw,

What to do with your vesc before using

My recommendation to people who have a VESC by onloop is:
- Don't power it up
- Remove the shrinkwrap
- check the whole PCB with a magnifier thouroughly, use a photo of an ESC by Benjamin for comparison. If there's something not right, fix that first.
- Use a fuse. If the guy who developed this ESC tells you that, don't question it. Do it!
- Now power it up
- Don't use settings copied from other motors, follow the measuring procedure of the BLDC tool
lizard

is incompatible with this :

Let me ....
》settings / configuration that are loaded where not defined or setup during manufacture. Up to user to configure.
....

You need to use a baseline setting from other motors, otherwise you will only get two parameters modification from a possible totally inconsistent default configuration (which was my case). I recommend as Vedder stated to use the SKxxx conf, and then run auto detect, then modify the two values by hand, rounding them as explained by Vedder

hey Jacob,

As far as I know, the first guidelines I read here where to run the auto detect with the motor connected and then modify the two values by hand, then GO ride... this can only work if you are sure that all the configurations are at least the same on VESCs when sold to customers... by reading onloop message, it seems that the VESCs from BETA onloop has only the firmware and the config is not defined... what kind of quality checks where done by the supplier ? how come a VESCs can come with a different configuration than another one if this is done "industrially" on such a low batch ?

thanks

 

[vedder]

I did that and everything Vedder said on his blog post. Found out the reason is never connected was due to a solder bridge on the MCU and it fixed that, but now I got the DRV8302 error code. So close, yet so far away. I did assemble the board according to Vedder's specification including omitting the resistor for R16, R6, and capacitor C26. The DRV8302 is solder with a hot air gun and I checked for solder bridges to which there are none.

Check the 5V rail and the 3.3V rail. Do you have the drv fault code permanently or just when you try to run the motor? If it is permanent, you can try replacing the drv8302. The drv8302 is quite sensitive and the chip that I had to replace the most by far. Silviasol had a VESC where the drv8302 fault code was present from the beginning, and fixed it by replacing the drv8302.

Let me enlighten you all with some facts/details of the beta run.
》 vedder released new BLDC version day after all enertion units shipped making all BETA incompatible with pre installed firmware.

BLDC Tool has been incompatible with the enertion beta vescs for the past 10 or so versions. All I did was send an email where I suggested to update the firmware to the latest version before shipping the beta VESCs because that would enable upgrading the firmware from BLDC Tool in the future. I had been working on the bootloader and the support around it many hours every evening after work that week hoping to finish it on time for my beta batch, and sent the email to onloop right after pushing the final updates to github. I also implemented version checks in BLDC Tool so that old firmware is detected and configuration is refused until the firmware is upgraded. With previous versions of BLDC Tool and the firmware, there was no compatibility check and incompatible versions could upload a completely misaligned configuration and mess many things up (I think this is what happened with one okps beta VESCs). I knew this was an issue and warned about that in the beginning of the beta batch because I felt that I need more time to implement checks like this before letting many other people test the VESCs, but I didn't have time to implement it in time. Therefore I provided a version of BLDC Tool and the firmware for Jacob on my server that should be used with the enertion beta batch:
http://home.vedder.se/public/VESC_Beta_Jacob/
but since there are different instructions around and there is no tutorial for the enertion BETA VESCs that states that this version of BLDC Tool (and not the one from my github page) should be used with the BETA VESCs people got confused. I don't want to blame anyone, I think this happened because things went a bit too fast and there was a lack of communication between me and onloop + Jacob. I'm not very patient myself, so I understand why everyone was in a hurry.

 

[okp]

update : http://endless-sphere.com/forums/viewtopic.php?f=35&t=64332&p=1054588#p1054588

 

[lizard]

I have a question regarding the "connect minus first to prevent the ESC or USB port blowing up" thing. Isn't that already taken care of by the USB plug? I think it's designed in such a way that ground always gets connected first. So if USB is connected before power, it shouldn't matter, or?

I am asking because I use EC-5 plugs for the battery connection, that makes it impossible for me to plug minus first.

BTW: My VESC from vedder arrived and it looks good

 

[okp]

I am using an Antispark/Powerswitch from APS, no issues with minus/positive; using also EC5

 

[vedder]

I have a question regarding the "connect minus first to prevent the ESC or USB port blowing up" thing. Isn't that already taken care of by the USB plug? I think it's designed in such a way that ground always gets connected first. So if USB is connected before power, it shouldn't matter, or?

I am asking because I use EC-5 plugs for the battery connection, that makes it impossible for me to plug minus first.

BTW: My VESC from vedder arrived and it looks good

If you have a battery or an isolated power supply it doesn't matter since that gives you galvanic isolation. However, if you use an unisolated power supply where minus is connected to ground and have the computer on the same power outlet you can get problems. If you have just USB plugged in in that case, minus will be connected to the power supply via the usb cable to ground and minus to the power supply. If you then plug in plus, the VESC will get power via the usb connector and plus, which is really bad since that will overload the logic ground. I have killed two microcontrollers on VESCs in that way, so that's why I mention it.

Nice that you received the VESC Please give some updates when you have tested it.

 

[lizard]

Thanks for the explanation. Yes, I'll give some updates when I got around driving it. But it could still take a while, don't have much time and the weather is not very good, hope that's going to change soon.

Here's the vehicle the VESC will go into
https://www.youtube.com/watch?v=EVJkGJpTDKI