Developer

mRo mid 2017 updates

3689719326?profile=original

Hi there! It has been almost a year since we re-opened our DIY store and I can report a lot is going on inside mRo as I write this. Our mission to be the best drone component manufacturer in the USA (World?) is no easy task!
In January, our sales increased more than we have anticipated, rather than compromise on our mission, I decided to stay quiet and not promote anything until we had stabilized our production line, sadly that was not possible (duh!). I was afraid that if we continued to grow that fast, we would start having problems and... I was right, we did. We have been able to iron out all those issues. Luckily, one of our first steps was a proper ERP/MRP system, which helped us keep production details very accurate and we added several new automated inspection machines.
I would like to thanks all our customers who supported us and gave us the chance to correct all our initial mistakes. We will not let you down, we never did!
I also wanted to share my gratitude to all mRo staff who have done unbelievable things to keep everything running smoothly and help implement all the foundations needed in our quest to be a World Class, NASA grade company.


Important news...

I'm glad to announce the merging of AUAV and mRo.
After working together with Phillip Kocmoud & Nick Arsov, we realized we had numerous "parallel" projects, so it didn't make sense to compete and work double. On the other side mRo didn't have the bandwidth to layout a new autopilot (at the moment), and AUAV already had cool autopilots that when combined with mRo manufacturing capabilities in California will allow us both to have such fantastic systems available sooner and with constant high quality supply (& support). That means more and better DIY options for everyone. ;-) AUAV continues to focus on research and development as well as outsourced custom flight controller designs, if your hobby needs become more serious. This can cut a design solution time-frame to just a fraction.
With this merging, mRo will become one of the most experienced DIY Drone companies in the world (Sounds cool huh?).
Putting under our belts the experience of creating/debugging products like Ardupilot 1, APM1, APM2.5, ArduIMU 1&2, PX4 Autopilot, PixHawk 1, The Solo Cube, Solo Link, AUAV3 (UAV Dev Board. Hi Bill we love you and your DCM!), PixRacer, AUAV X2, X2, X2.1, AUAV X3, AUAV X4, Pixhawk Pro (The FMUv4 Standard). Among other innovative products (at their time) like the MinimOSD, optimized PPM encoder, uBlox GPS to the masses, differential I2C and the first budget power module with hall sensor. Just to mention a few. =P


Clarification: In most of those projects we were involved 100% and when I say we, I mean AUAV team and ex-3DR employees that now some of them work for mRo, not just designing, but also in the complex production lines, from ex-Tijuana factory and China (when fabricated Solo).

3689719295?profile=original

[Picture for sizing reference, plus a few Easter eggs, except for PH2 or PixHawk Mini, we've made everything else]


Here comes the "meat," our new products for you:

-uGPS, uBlox SAM8. The tiniest GPS+Glonass GPS out there, it also has a built-in magnetometer. uBlox entirely designed the antenna+RF and we trust them with our lives! Well, the life of our Drones.
-ACSP7 is ready! First micro (17mm x 17mm) power module with hall sensor, no more invasive, noisy readings (especially below 5-8 Amperes).
-Next generation AirSpeed sensor. MS5525, Temperature compensated. Wuju!
-FTDI to JST-GH. Easily attach your ports or GPSs for debugging. 
-Military grade Shielding for your GPS.

-Don't forget our list of Fantastic Cables and where to find them

Coming soon:
-X2.1 Rev B. Which is a redux version of the robust and popular PixHawk1, but with newer/better sensors, and some standard JST-GH connections and incredibly flexible when building your drone. We are not adopting any known names like PixHawk 4000 or Ardupilot 7000; sadly there are a lot of politics and interests everywhere right now that can potentially delay our products. We just want to focus on designing good electronics and make them available to you as soon as possible, so you can get the work done or get fun.

3689719410?profile=original

Note: Pins can be either right angle, straight or both. 

The X2.1 has a few advantages, with ArduPilot 3.5, dampening is no longer required, which reduces price and removes failing points. I can tell you a lot of the issues I saw on the Cube used on Solo (We have the official Solo repair center in the same roof, Current UAS). Is not that bad, you can easily fix most of those issues, for example, a common problem the cube has is the foam getting stuck, with X2.1 you don't have to worry about this.
Besides, the dampening was calibrated just for SOLO so that it won't be as useful in other vehicles with a substantial difference in vibrating frequencies. We proposed sorting the cubes in different colors, so each had a different damping frequency range. For this and other reasons, as 3DR, we decided not to make our own version of the Cube and rather give an actual DIY attribute to X2.1.

So, X2.1 is the most DIY friendly and capable Autopilot we have made so far. You can adapt and solder your custom shields with ease in just minutes, or you can get fancy and layout your PCB to populate all you want on board, no complex, difficult to align connectors required. If you are not interested in that, you can just use standard 0.1" pins/cables to plug into your system. Every pin, almost every signal is the same as PixHawk 1, expect an excellent integration with your "actual" drone.


3689719425?profile=original

-Lastly but not least, our Solo Link, now mRo Link, based on the i.MX6.
Believe it or not, we are still working on this project as I mentioned here. Now that 3DR has officially open sourced the SOLO; we feel more comfortable than ever to put full weight on this project. We already have everything working; we just need to remove what is not required and make it "smaller." Check our picture up here with all of the boards. You might see it. 
We are bringing a top engineer from the other side of the world to dedicate to this project during August 2017 fully. So expect more news early September 2017. BTW we know that somebody else already released something (Fox), which we fully support and that is precisely what we love about open source. We hope to work together in the future.

Enjoy my wife's lazy cat: 

3689719343?profile=original

E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • Hi Jordi.

    Your product range is looking good and growing every day!

    I'm particularly interested in the Solo Link/Mro link wifi setup for a range of projects that operate LOS. Do you have a rough estimate when it will become available, and if by including a iMX6 processor will it gain a CSI connector to run some basic imaging out of the box with a CSI based camera as well? The HDMI will work, but seems to be a fairly large and clunky solution that is expensive seeing it needs a an external dedicated camera, and larger gimbal. It's also more orientated towards video and not stills, or basic CV operations.

    I think a longer range (even if a bit high latency) wifi based FPV setup would be ideal provided it does the whole range of telemetry and companion computer type communications as well for the FC. Maybe the HDMI in can be optional instead? 

    Lastly what is the expected range on the link?

    Regards

  • Great news, Jordi!

    Rob, spot on with "DJI figured this out long ago". Out of curiosity I cracked open a Phantom 3 Pro a while back, the IMU is inside a fairly large plastic cubic plastic case, about  2” sides iirc. Inside is fully foamed, with at the center  2 hollow ½ blocks of aluminium about ½”x1/2”x1/4 screwed together. The IMU and circuitry is inside that. They definitely  thought very seriously  about damping.

    FWIW, pics of the main board, (plastic case is in the middle), aluminium ½ block, and close up of IMU (MP65xx?).

    3702391367?profile=original

    3702391235?profile=original

    3702391279?profile=original

  • Yes, choice is always good.  I just caution about broad comments that isolation is not required.

    I have an AUAV X2 that is nearly hard mounted on a helicopter, and it's working fine.  And I have a gas powered heli, which needs the entire autopilot to be isolated anyway, due to extreme vibration levels.

    But IMO, a vibration isolated IMU is really good for many applications.  The problem is that most users cannot design their own vibration isolation system.  And most of the aftermarket solutions are pretty bad too.  DJI figured this out long ago, and I'd been begging 3DR to make one 4-5 years ago.

  • Developer

    Guy, 

    The price is going to be below the $300's, all depends on the features. The nice about the i.MX6 is the ability to upgrade to dual or quad core, we prepared the board for such upgrade. Also, the GPU is not used at all and can be potentially used for something else, ie. video processing. Also, it will still have PCIe slot, so we can use a new Wifi card, with MiMO, 4 Antennas, and the 802.11ac. 

  • Developer

    Rob, in extreme cases you can still add damping foams which we offer and we will soon offer more options for different vibration frequencies, giving it more flexibility. Now, I'm referring to Solo Cube, which doesn't have sensors capable to sample at 8Khz. PH2.1 has 8Khz capabilities plus damped imu, which is a nice feature to have for a premium. There is some DIY that is willing to save $60 dlls and a lot of space, in exchange the internal damping for $5 foams, if required. 

    In the other hand, we have countless hours in PixRacer (which is capable to sample at 8Mhz), mounted directly and never had a vibrating problem. 

    Ultimately customers have a lot of options to choose from. 

  • Ok, this is what I suspected.  I would say that stating that "vibration damping foam is not required" is a gross over-simplification of the situation, and will not do you or your customers well.

    It can work in some cases.  And it can work very badly in others.  Your comments are based on the idea that there is not much vibration above 4000 Hz.  It's a bad assumption. Even with something you may have tested like an Iris, will work when it's new.  But when the bearings start to get notchy, it will generate high frequency vibration again.

    The other problem, is simply the amplitude.  If you saturate the accelerometers, you're done.  And again, a new Iris, might be fine.  An Iris with some prop damage, might not.

    I have done some testing on this myself.  I put a PH2.1 on my Procyon helicopter, which is quite possibly the smoothest, lowest vibration helicopter UAV made.  (and lower than most multirotors, before anybody asks)  The 3rd IMU which is not isolated, show signs of aliasing.

  • Sampling a 4kHz signal at 8kHz will resolve issues with aliasing, but you also need to consider how MEMs devices actually work. There’s a harmonic response at a hardware level that can distort measurements in a way that can’t easily or accurately be restored by software.

    Hard mounting (gluing the controller directly to the frame) is a form of dampening that works by increasing the moving mass, which in turn attenuates high frequency vibrations. It’s viable for some platforms, but it usually involves some compromises (See Forrest Frantz’s work for example). I think most robust solutions will still require some form of dampening based on my understanding of the physics.

    The mRo Link has the potential to be a game changer for researchers and small commercial operators if the final price is competitive. Will there be any changes to the technology since it was introduced in the SOLO?
  • Developer

    Sandro, good to hear from you. You were the master mind behind the software! 

    Cala, Thank you! We will announce pre-orders soon. We have all the parts in-house ready for full production. We just need to finish assembling a big order and X2.1 will be next. 

    Jeff, sup man! Still in SD? We are always here to help the good drone makers! 

    Rob, mRo Link sounds better? As engineers, we are bad in marketing. =)  Autopilots like PixHawk1 have old sensors that are sampled at 4Khz, this frequency is affected by motor vibrations, that is why you should use damping foams. The cube was developed to include the foam inside and make the integration simpler. But technology advances fast and the latest sensors are capable to be sampled at 8Khz, this complies with the anti-aliasing theorem, making the sensor immune to the motor vibrations because now is way higher than the vibrating frequency caused by most motors. 

    For example, Iris + with PixHawk1, if you remove the foams, it becomes unflyable. We were able to successfully fly the Iris+ with X2.1 with no foams at all, just glued directly to the frame. 8Khz sampling destroys the purpose of the foam, so there is no need for complex multi IMU, damped boards anymore, that simplifies the design, size and most importantly to everyone, price. 

    Anti-aliasing filter
    An anti-aliasing filter (AAF) is a filter used before a signal sampler to restrict the bandwidth of a signal to approximately or completely satisfy t…
  • Really looking forward to the SoloLink.  CUAV is basically already doing this with the HackLink, but I would prefer to buy from a more reputable source.  I strongly advise you to come up with a name other than SoloLink though.

    But I'm really confused by your comments about Ardupilot 3.5 not requiring dampening anymore?  Can you explain in more detail?

  • I'm excited to see all the new products! It's great to have a reliable and increasingly innovative supplier for all the core drone components again. I can vouch for the good customer support too.

This reply was deleted.