Hi @Mihir-Bala
You will need a 2S minimum to satisfy all the requirements of our power module: https://docs.modalai.com/power-module-v3-datasheet/
Then, you need to do the math to support a 30Watt transient on power up... which equates to a ~5amp burst at the low end of the 2S cell ~6.4V.
So, if you have a 1,000mAh cell at 2S, that will require a 5C rating.
If you have a 2,000mAh cell, you need ~2.5C rating, and so on...
Hope this helped!
Hi @jacobcayetano1
Apologies for not updating some of the tech docs content about that test board we developed.
Here are some links:
https://docs.modalai.com/voxl2-dev-test-board/
That describes the board and provides schematics and other useful user info.
Since the test board is a very easy way to destroy your VOXL 2, we keep the sale of it on our Beta page to alert customers that it is an "at-risk" design since you can easily void your VOXL 2 warranties by over-stressing the VOXL2 and introducing ESD and other signal shorting risks.
That said, here is the purchase link:
https://www.modalai.com/pages/beta-voxl-2-b2b-breakout-board
It comes as a kit with many goodies for it's use including cables, a mounting case, and an SD card.
Now, as far as being able to use Microhard AND J5, we recommend using the standalone version ("dongle") of the Microhard board: https://www.modalai.com/collections/modems/products/mdk-m0048-2 since that will allow you to cascade it off any USB port we have (or one that you include in your design).
Here are some tips for designing a good USB port with our Voxl ecosystem: https://docs.modalai.com/expansion-design-guide/#usb-expansion-over-j3--j5
Should you go down the route of designing a custom VOXL 2 plug-in board, we will offer a courtesy review of your schematic to help you along! Feel free to tag me again when you are ready, and we will start an email exchange to share your data so you are not posting it on our forum (unless you wanted to).
Hope this helped.
Vinny
Hi @serviceberry
From a HW viewpoint (I am HW team) that sounds reasonable to use the hub the way you suggest. But I know for a fact from personal experience the Boson's are very dependent on VBUS rise times, and a very fast rising 5V supply may not allow it to enumerate properly. We have had an instance in the past where one particular VBUS port rose really fast, and the Boson did not enumerate until we slowed it down. It's one of those situations that may require testing to check. Our J3 10-pin host port has a nice moderate rise time that allows Boson to enumerate. So, when using a Hub, check the VBUS rise times compared to J3.
I do NOT expect J9 USB-C OTG port to work with Boson.
The Snapdragon SOC is very smart, and when in OTG mode, it will not provide more than 500mA to any device that enumerates as USB2 which I am lead to believe how the Boson enumerates. if the Boson could enumerate as a USB3 device and the drivers were all present in the Kernel for this, the Snapdragon will give it 900mA over VBUS letting it boot but I think this is a non-starter without the Boson being able to do that.
Our Boson kits use USB2 only and we provide 1A support to enable them to work without requiring enumeration to negotiate more current, hence why they work on the J3 ports and other expansion board ports that have more than 500mA: https://docs.modalai.com/expansion-design-guide/#usb-expansion-over-j3--j5
Hi Nikhil,
Sorry for our delayed response on this. I just recently joined our forum to help out with more HW centric question.
As Chad mentioned, this does require HW mods if you do not want the SW driver timestamp method (which would be our recommendation/preference).
The Lepton signal VSYNC (on GPIO3 most likely is what you are referring to) is running at the VIO voltage of 2.8V to 3.1V per Lepton specs. Our camera sensors are MIPI based, and as such all of those I/Os are at 1.8V, so a level translator would be required which makes connecting these a bit tricky, especially if you need bi-directional support with another sync controller/snoop device. This would need to be something done on a custom interface between the Lepton and the stereo sensors. We do not have such module but would be happy to give further guidance if you need it.
Thanks!
Vinny
@Ansel-Misfeldt Hi,
The approach we will take for CANBUS right now is similar to what we have. That is, to add CANBUS, it requires the use of a FlightCore Flight Controller. So, with Voxl1, you have Voxl-Flight which integrates the FlightCore, or you add on Flight Core independantly.
For Voxl2, you can still use a FlightCore, but we have some exciting news coming soon for a newer improved FlightCore, we will likely call it FlightCore V2 
It will retain CANBUS functionality but has a slew of other upgrades and user improvements added in, all in the same form-factor as FCv1.
Our Docs portal is just about ready for launch and I think we did a better job up-front on this one to help our customers get running quickly.
Stay tuned, we should be posting it for sale in the up-coming days!!
Hi @dlee
All of our Doodle configs are only using USB2, not UART.
Hope that answers your questions.
Thanks!
Hi @dougmiller
If your VOXL 2 Mini is TrustZone Configured for UART over J10, then yes, there are 2 pins available as inputs.
https://docs.modalai.com/voxl2-mini-connectors/#j10---external-uart
You can see GPIO_40, GPIO_46 and GPIO_64 are input capable.
GPIO_40 and GPIO_64 are able to be configured to do wake events in the SoC so they have extra configuration to route to a power manager. Depending on how you write your IRQ routine, that may or may not help.
Hope this helps!
Hi @Gary-Holmgren
Yes, we did design VOXL 2 Mini to be powered by 1S capable source. But, we'll need @Alex-Kushleyev to respond regarding the MINI ESC and how that may need to be tuned for 1S operations.
For VOXL 2 Mini, here is the way it can work, and the few caveats to explain why we have not offered it yet in exactly that way. If I go in too deep on any electronics jargon, let me know and I'll try to clarify as needed
The local 3.3V regulator for the UART, GNSS, and all other ports will start to drop-out at around 3.4V. So, even though the core electronics (Snapdragon, Memory, Cameras, USB) will work down to 3.2V, some of our extra circuits will not be happy. If you are powering modules from our 3.3V supply, and they can operate at slightly lower, like 3.2V or 3.1V, and you use our 3.3V as the VDDIO ref, then all is good there.
Our 5V boost circuit likes higher voltages. At 3.8V VBATT in, we can supply ~900mA of VBUS (to mimic USB3 VBUS specs). However, as that input VBATT voltage goes down, so will the capacity of that supply. Also, I have a UVLO setting on that rail to trigger the 5V boost to turn OFF if VBAT drops below 2.5V to prevent 1S battery continual drainage (deep discharge) in the event you disable the flight control stuff, but forget to disconnect the battery. However, the ramp ON has a very large hysteresis and it triggers at ~3.5V. So, if you try to start a power ON with less than 3.6V, that 5V boost may never turn ON. Now, if you are not using VBUS, or the 5V on the J19 connector, then it's a do-not-care. But, your ToF use case will require it....
TOF is not easy on VOXL 2 Mini, especially the new V2 one. Since we only have that 900mA of 5V (less at lower voltages), the ToF circuit may get starved of it's needed 5V if you start to use other stuff, like USB or lower input VBATT. So, we would only caution you to use 1 ToF module at max, and be prepared to lower exposure and frame rates at VBAT < 3.8V. Our ToF modules use that 5V to generate a local 3.3V, so we do get some power conversion gains, but we just have not done it yet on V2 Mini to capture all the other side effects.
Thermal, if you are using our Fan connector, that is powered by that same 5V supply IN mentioned above. So, if that does start to shut down, you'll get a fan OFF condition (possibly ON/OFF throttling like a forced PWM if the 5V comes back on), triggering thermal rise compared to the current 100% duty cycle ON as it is now.
Unplanned/Untested Software responses at VBAT <3.8V. As stated, we have not tested this config. The Snapdragon parts are very complex SoCs. When power input starts to go below a certain threshold, the underlying Power MGMT IC's (PMICs) will start to throw interrupts. I think the first one by default is around 3.3V VBATT, but they change based on platforms and chips. So, as we start to approach the "dead battery" SOC%, the system itself may start to react differently and we have not characterized that at all for our builds. It is electrically spec'd to operate, but SW may start nagging and doing weird stuff, like turning off functions such as cameras for example to conserve power.
Motor noise!!! This is the biggest one... 1S motors will draw a lot of current to equal similar power demands of higher 2S+ rated motors (there is some V*I correlation here to motor power, so the lower the V, the higher the I). This will result in a lot of noise on the VBATT power rail. These transients can go not only low (droop, triggering resets, 5V OFF, etc) but also too HIGH (back EMF, energy dumps, etc...) potentially damaging components on V2 Mini's front end power system. V2 Mini ABS MAX voltage in is 6V, at which point the PMICs will be damaged. This is the single biggest reason we have not tried to connect a battery directly to our electronics, and always opt to have some type of nice DC/DC to provide solid line-regulation protection. So, if you try to use a direct 1S VBAT, you will need a really good motor-noise transient protection, maybe even a fast acting Zener diode, set to ~5.6V or so. On a side note, we just started including a 5.6V Zener on a new Spin of V2 mini we are working on, but existing M0104 V2 Minis do NOT have that 5.6V zener. Details on a new V2 Mini will follow in the up-coming months, as teased here: https://docs.modalai.com/voxl2-mini-connectors/#j3-usb-3-10-pin
Those are my first thoughts... If I think of anything else, I will let you know.
I think it is doable.... V2 Mini was designed with micro-quads in mind, but there is some planning and design considerations that need to be done. We can help you along with the HW side of stuff, but I can't promise we'll have any SW support for issues possibly related to low VBAT in since we have no system to check against or develop with. That is why we only use 2S+ in our platforms (currently).
Hope this helped!
Thanks @aheyne for alerting us.
I had made the updates to our site and the 3D models are now there.
Let us know if you need any more help.
Thanks!
Hi @John-Nomikos
Have you confirmed the issue we pointed to in the other forum post?
https://forum.modalai.com/topic/1356/sentinel-drone-randomly-reboots?_=1673383475709
Do you have Kapton or another insulating tape you can place on the J5 120-pin B2B connector on Voxl2 to help confirm the Microhard is not touching J5? We've seen it first hand that it triggers resets.
I doubt all 3 MCBL-00001 are all bad on you (although we do know they go bad fast from users pulling them out by the cable wires instead of the plastic shroud, but that's another issue).
Let us know.
Thanks!
Hi @Nando75
A shielded, impedance-controlled flex was our go-to solution in the early days, and we still offer a few legacy adapters that use it. These provide excellent signal integrity and EMI shielding. However, they’re limited in deployment flexibility; each flex must be meticulously designed for a specific routing path.
Coax, on the other hand, greatly improves routing versatility and can be deployed across many platforms. The trade-off is reduced signal distance and slightly degraded signal integrity compared to a tuned flex.
Ultimately, it’s a system-level tradeoff. We’ve transitioned to coax as our mainstream solution in recent years, driven by customer demand for easier installation without the constraints of custom flex circuits.
Only your team can decide what's more important and hopefully we can enable you either way.
Hi @Nando75
If it was less than 200mm, yes. 200mm is at your own risk, and make sure your SW team knows how to reach out and request slower MIPI CLK speeds in the event of signal integrity failures (camera failures).
Keep us posted.
Thanks!
Hi @SKA
Apologies no one responded promptly.
From what I gather, these are best to be sent back as an RMA and we can try to recover them.
D1 (green) is an indication of the local 3.3V power being Good. So, that implies you have 3.8V input, and the 3.3V reg is OK.
DS2 (RGB) is controlled by Qty-3 GPIOs from the SoC. This can briefly blink on boot until the GPIOs are configured by apps software.
D2 (green) is the 5V USB VBUS USB1 (10-pin) port power indicator, which only comes ON after the QRB5165 boots and Apps controls the GPIO to turn the VBUS ON.
On failed SW loads, it is common to diagnose that by seeing DS2 or D2 blink, which indicates boot loops. D1 should NEVER go off, if so, something is wrong in HW or cabling.
We do caution any customer that sets up their own CI or internal validation automation to include a nice pause before repeated power cycles. A common 5 second rule is good to let the large caps discharge, or certain ICs will never fully reset. This is the most common cause of "bricking" devices.
The RMA is your best bet here and we can try to recover them for you, or diagnose for any other issues: https://www.modalai.com/pages/rma
Hi @QSL
OK, keep us posted!
Any, for any new issues, please be sure to start a new forum post so it is seen by the team.
Thanks!
Hi @QSL
Is your hub powered externally? You might be starving the VBUS if you are not powering the hub externally. The USB/UART board adapter does not have enough current to power a hub and additional downstream peripherals as well.
Hi @Nando75
Yes, that is a fundamental limit of ALL uCoax that is not Twinax Impedance controlled conductors. Therefor all IPEX, Hirose, Molex, TE, etc (you name it) coax will fundamentally be limited in performance compared to true 100-ohm differential controlled impedance you can get from flex circuits or Samtec's Twinax. It is one of the many tradeoffs in our design space we need to make. As such, that is why we are firm on keeping the coax short and we do not sell any option longer than 180mm since at some point that small impedance offset and tolerances between two 50-ohm conductors compared to a true coupled 100-ohm may play a role.
Hope this helps you.
Thanks!
Thanks @joseph-vale for the kind words.
We will continue to try our best
Our success only comes with your success.
Sorry there is not much else I can offer. I'd need to have the HW in my hands to diagnose.
I'd start with the basics, Power. It is possible you blew the VBUS protection circuit that feeds J3 VBUS. It is also possible that GPIO_157 which enables VBUS on J3 is not correctly set in SW.
However, I still ask you discontinue use of M0151 for Doodle as you will have problems in the future and we cannot continue to support that.
Please follow the guidance listed in our cable page for Doodle. Those boards listed do not use the "VBUS" from J3 (rather we power switch the 5V DC Local with proper Hub monitoring), so you can likely recover proper functionality and continue with your efforts.
Thanks!
Hi @joseph-vale
Great, thanks for clarifying.
Apologies our website is not always perfect. It is maintained by us Engineers, so we often take some info for granted. I am a few years delinquent in adding image sensor design guidance to our tech docs
All our MIPI_CSI lines are indeed 100-Ohm Differential, so the IPEX mate will match.
We include ESD protection on all MIPI CSI and I2C and GPIO/MCLK signals at the VOXL 2 J6, J7, and J8 connectors, so no need to double protect (or you can impact the bus signal quality with the added capacitance).
When you get to the point of generating a schematic, I will offer a courtesy review. Ping me on this thread again and then I can email you privately to share if you wish.
One thing we did on our Coax design was plan on always using the same 26-pin cable, so it is our adapters that vary in pinouts, but our cable is always the fully populate 26 conductor DF56 series.
For Boson/Hadron, we use multiple conductors of 5V. We set that based on the DCR spec of our coax cable, so with your IPIEX cables, you can do something similar to achieve success.
We were also advised by Hirose to NOT use the outer shield as current carrying, so for every +DC path, we have a return GND conductor.
Keep us posted!
@joseph-vale I think a lot of FLIR's docs are public domain, but we did not receive them that way, hence our reluctance. If you can be more specific as to what you need, I'll try my best.