Refresh

This website oscarliang.com/serialshot/ is currently offline. Cloudflare's Always Online™ shows a snapshot of this web page from the Internet Archive's Wayback Machine. To check for the live version, click Refresh.

SerialShot – New ESC Protocol (1 CPU for 4 ESC)

by Oscar

Matek posted something interesting today – a demonstration of a possible new ESC protocol, Serialshot! It’s going to completely change how your ESC is wired to the flight controller.

Some of the links on this page are affiliate links. I receive a commission (at no extra cost to you) if you make a purchase after clicking on one of these affiliate links. This helps support the free content for the community on this website. Please read our Affiliate Link Policy for more information.

What’s Serialshot?

Serialshot is a new ESC protocol developed by Matek, capable of bidirectional communication.

The cool thing about Serialshot is that, instead of using one CPU (micro controller) for each ESC, it only uses one CPU for all four ESC’s in a mini quad. This is going to make wiring simpler, and possibly ESC’s could be made cheaper as well.

 

View this post on Instagram

 

A post shared by Sampson (@mateksys) on


Looking for new ESC? Don’t miss this spreadsheet:

I compiled the specification of all ESC’s for mini quad in this spreadsheet so you can compare them more closely.

Simpler Wiring

As you can see in Matek’s Instagram post, with Serial-shot, the cable from FC to 4in1 ESC only has 4 wires.

FC on the left – 4in1 ESC on the right – running Serialshot

I am guessing the wires are VBAT, GND, and Serial UART (RX and TX), so really only the last 3 wires are needed for Serialshot to work. Normally, with all other ESC protocols like DShot, Multishot and Oneshot, we need at least 5 wires – one wire per ESC, plus GND.

Cheaper 4in1 ESC

Matek posted a teaser just the day before, showing a possible 4in1 ESC designed for Serialshot. It appears to only have one CPU – an STM32 F4.

The labels next to the header connector confirms my guess about the connection (V, G, R6, T6).

Traditionally, there is a CPU driving each ESC. With Serialshot, one CPU would be able to drive all four ESC’s in a quad.

Hopefully ESC’s running Serialshot could be made cheaper because they only use one CPU instead of four? But that only seems to make sense on 4in1 ESC’s, it’s going to get a bit complicated for single ESC’s.

Do you put the processor in one of the four ESC? Or do you integrate that processor in the flight controller… And you would have to run wires to each individual ESC as well, so it would be the same as non-Serialshot…

Maybe Serialshot is only going to be used on 4in1 ESC?

Update (Dec 2019)

Matek posted a new picture of their latest ESC prototype with Serialshot. They stressed that they have not abanoned this project, but in fact making some great progress, one of the main change is that they’ve upgraded the MCU to F7 (F722).

More Info?

Good or bad, I am extremely excited to see more innovation in this hobby :) I will keep this post up to date with more info as I learn more.

As far as I know, Matek is still trying to optimize the protocol so it’s not yet ready.

Leave a Comment

By using this form, you agree with the storage and handling of your data by this website. Note that all comments are held for moderation before appearing.

11 comments

Andrew Blanche 30th August 2019 - 8:34 am

Is there any new info on serial shot?

Reply
David Sherman 18th April 2019 - 1:35 am

It looks like it runs super smooth without the jerkiness of Dshot protocols. it seems like noise wont play such a factor and the installation errors will be at a minimum. All around a great thing!

Reply
Tom 15th April 2019 - 7:54 pm

Similar to SBUS protocol no doubt(see dualsky.com). Only real advantage is makes the link completely digital and eliminates inherent noise in a CPPM based controller, since STM32 probably feeds PWM for each motor into a pre-driver; not enough outputs on one STM32F4xx to drive three phases of four motors directly, so probably will require FOUR pre-drivers like DRV8307. Only software then is the translation of SBus to setting four PWM outputs to the STM32. With 115kbaud UART, anticipate 12byte packet from host(flight controller) would give about a maximum 900hz refresh rate on motor speed.

Reply
Moritz Beyer 15th April 2019 - 11:35 am

– Not being an engineer I would estimate that with a single ESC solution you could have one master esc with the cpu and 3 slave ones. As every esc is wired over the serialshot UART they could communicate easily. So single or 4-1 esc would profit.
– Les wires is nice but it is all about performance. I do not see any advantage based on the information provided. One scenario could be a better motor control as those are controlled by one cpu which might have the possibility to use some sort of feedback mechanism and combined with the FC further improve overall performance. Any knowledgable around who could comment on this?

Reply
ryq1212 14th March 2019 - 2:29 am

Wow, this idea has already been realized! I came up with this idea last autumn, but now it seems that someone is one step ahead of me…

Reply
Andrew_J 11th March 2019 - 12:43 pm

Does anybody know what frame is that. It looks like a 4inch frame, and I’d be interested in it. Thanks

Reply
Datt Mtn 8th March 2019 - 10:20 pm

I’d be interested in how this compares or is affiliated with the new dual direction Dshot that is part of Betaflight’s RPM package with 4.0.

Reply
Darren Louw 7th March 2019 - 6:36 am

Seeing this is exciting for silly people like me who ALWAYS wire the motor outputs incorrectly and get exciting back-flip take offs! :)

Reply
Sampson 7th March 2019 - 4:40 am

Sounds like logic is reversed.
Serialshot or future SPIShot is by-product of this product. Serialshot can’t help driving 4 motors with one MCU. 1MCU driving 4 motors makes serialshot/SPIShot implement convenient.
1pc F405RGT6 is expensive than 4pcs F051K8U6, one MCU can’t make ESC cheaper so far.

Reply
Peter Newman 10th March 2019 - 2:58 am

Yes that’s a good point and having only four wires compared to the normal 5 or 6 (with an onboard BEC) is really neither here nor there is it? There really would have to be some performance improvements with the protocol that make this worthy of being too excited about ?

Reply
Daniel Ribeiro 5th March 2019 - 11:54 pm

“I am guessing the wires are VBAT, GND, and Serial UART (RX and TX)”

This is a Matek F722-STD Flight Controller, this wires are soldered to the UART1, so the pins are 5V, GND, TX and RX. It’s not VBAT as this FC doesn’t has an integrated 5V regulator and would burn if fed by VBAT.

Reply