Posts made by Vinny

Hi @VictorG
You should RMA the FlightCore and we should replace it. Just mention this thread or place a link to it in the form:
https://www.modalai.com/pages/rma?_pos=1&_sid=190e46152&_ss=r

We do not yet have a replacement set of cables, under normal circumstances, it takes us 6-8 weeks to get more, so we are trying to see if we have any small batch in our inventory to hold us over and replace every bad cable that we shipped.

If you want to try to re-pin your cable for the interim (you will not need to send it back, we can just replace it when we get more), try this link:
https://docs.modalai.com/cable-userguides/ and at the bottom are tips for re-pinning

or you can just search youtube for s JST GH re-pin tutorial and I'm sure you'll find plenty of good techniques. All that is needed is tweezers.

@Martin-Lukac , @VictorG
Yes, we are working a replacement solution.
Exactly what @Martin-Lukac said is a temporary solution by swapping the conductors until we can source replacement cables. We might need a few weeks honestly.
Any HW that was compromised by this issue is an no-brainer replacement by my standards, and I will help ensure our production and operations team replaces any HW at jeopardy of this issue promptly. We have the HW in stock to replace, just not the USB cables as of yet.

We apologize for this issue. We actually had that batch of cables in stock for over 2 years and our sample lot from a different sealed bag in the same shipment that was tested did not experience this issue, so we were caught flat footed and we will put in better inspection processes to ensure this does not happen again.

Thanks!
Vinny

Hi @hula
YOU ARE CORRECT!
I found a whole bag of mis-wired MCBL-00010's in our inventory.
I'll work with our @Moderator to figure out a plan.
We will likely need to setup an email where any customer can let us know if they have offending cables and we will send replacements promptly.

Hi @hula
Interesting. I will investigate. The cable appears as it should according to our tech docs and every batch is sample tested but I will inspect further and respond tomorrow.
https://docs.modalai.com/cable-datasheets/#mcbl-00010
Thanks
Vinny

HI @Jetson-Nano
Your welcome. Apologies I cannot find anything blatantly wrong with your design at the document and layout level.
If your PCB still does not work after all these reviews and positive affirmations, then you may very well have a fabrication issue with your PCB.
The only way to verify that is to try your own net-by-net buzz-out using a DMM to ensure your netlist matches the expected schematic. If that all matches up, then you may be suffering from poor assembly/soldering. The socket for the lepton does require a reflow oven/line and cannot be hand soldered, unless you have an IR tray/heater and a way to deposit solder paste on the pads.

Any further debugging I could offer would require physical samples of the hardware to validate everything.

Hope that info helps!

Hi @Jetson-Nano I have reviewed your schematic again and your layout files.
I cannot see any reason why your design will not work.
The only suggestion I have is to increase your capacitors here to 1uF:

But, otherwise, your layout is OK (some improvements can be made, but nothing is detrimental):

• Pin 1 correct on connector, socket, LDO, other ICs
• Vias are picking up planes correctly
• Voids where expected
• Generous copper trace thickness

The one issue I think I might have found is your reference designators do not match the schematic between the layout. So, I'm wondering if you are installing parts from the BOM in incorrect locations...
For example, the cap on the output of the LDO is "C6" in the layout, but "C7" in your schematic:

And, your resistors are also not in sync. R1 and R2 in the layout are for the Ic2 expander, but in the SCH, it is for your FLIR module and not R5/R6 as in the SCH:

and there are a few other examples. So, I just think perhaps your layout reference designators are not synced to your design.

J1 should be J2, and R3/R4 on the layout does not exist on the SCH.

So, maybe this is where your issue is. I'd recheck the annotation between SCH and Layout, and be sure you are installing all the correct parts in the correct places.

Hope this helped!
Keep us posted.
Thanks!

Hi @Jetson-Nano
I received your data package and will review it this evening.
Thanks!
Vinny

Hi @Michael-Mortensen
Please see the post to Victor about using the power module instead of just USB.
Let us know if that helps.
Thanks!

Hi @Martin-Lukac
Well that's not good. Your board will not work with that chip busted off like that.
Please use our RMA process and we'll get this assessed for for you:
https://www.modalai.com/pages/rma

Hi @VictorG
It is not unexpected to see some windows machines complain.
Our MCU does not enumerate as a normal USB device with current requirements reporting. So, some PCs will correctly complain.

Our goal with USB based powering is for ease of use with configuration. However, if this is posing a problem in your specific system, connect it as it would be required for a sUAS by using our power module and using an MCBL-00062 into the power port.
https://docs.modalai.com/flight-core-v2-power/#j13---voxl-power-module-v3-input
https://docs.modalai.com/cable-datasheets/#mcbl-00062
This will prevent pretty much all current draw on the USB port: https://docs.modalai.com/flight-core-datasheets-v2-functional-description/#power-supply-and-power-source-mux-details

Now as for the heating, we do know the MCU can get hot. It has no sleep modes or throttling in it when running PX4/etc, so it is not uncommon, but it will still work and is within specs. Some folks have used heatsinks or small fans with success to minimize the effect, such as: https://www.modalai.com/products/voxl-cooling-fan?_pos=1&_psq=fan&_ss=e&_v=1.0

Hope this helps explain things a bit.

Hi @Jetson-Nano
I do not have access to the link you have above.
Please reach out to us using the contact form https://www.modalai.com/pages/contact-us and we can get a more detailed review scheduled using your native cad tools which is more effective for us.

Just mention in the form you wish to reach Vinny and then in the response from our team, you can reply with your data and they will forward it to me.
Thanks!

Hi @Jetson-Nano
I've been on vacation and just catching up. It will be a few days before I can get to this.
Thanks for your patience.

@Peter-Milani
However, I am on vacation for a week starting in a few hours.
So, I will get to the review the week of 8/4 when I return (just to set expectations).
Thanks!

Hi @Peter-Milani
Yes, I will gladly review it!
Please reach out to contact@modalai.com (or this https://www.modalai.com/pages/contact-us) and let them know to please forward your email to me.

Thanks!!
Vinny

Hi @Jordyn-Heil
We try to keep this regularly updated, but sometimes miss some... in this case, M0173 is there https://docs.modalai.com/pcb-catalog/

Hi @jared
Presently they are integrated within certain drones...
https://docs.modalai.com/stinger-vision-datasheet/

I can poke the team to see if we can get it listed separately in the near future.

@KLindgren
OK, M0188 will not work with anything other than our M0187.
Thanks for clarifying.

Hi @KLindgren
You raise a concern and something I will run by the team, but if you are powering the M0187 at 3.3V directly, you are violating the FLIR specs for the Lepton and our system design constraints.

How are you connecting M0187 to M0188? Maybe your cable is introducing some extra losses giving you a lower than 2.8V supply?

Hi @Nate-The-great
Please do not delay any of your efforts or programs in anticipation of a VOXL 3 platform.
Our development roadmap is company confidential, but you can imagine we are working on many new and exciting things. One of the goals we have is to minimize disruptions should someone wish to "upgrade" platforms in the future. Similar to how we transitioned from VOXL to VOXL 2 by keeping mounting patterns the same, expansion compatibility (J13 -> J3), and plug-in board and sensor compatibility (M0048, M0030, M0084 & M0135 Camera adapters, etc), we would mimic that approach, but strive to make it even more seamless than before.

Therefore, there is very little to no risk to your system if you start developing on VOXL 2 today.

Hope that helped.

Hi @KLindgren
You must not hot plug cameras.
Think of this as a phone and when power is applied, all items must be securely connected. USB is the only architectural feature we include with hot-plug functionality.
All tests with hot plug are void and cannot be considered a solution, cause, or method.
Thanks!