@@Cmdrbzrd Just use a megasquirt. At least their code has been made to be sold. That way it has to be quality and supported. I have had a long play with speedy and ran it on one of my own personal vehicles as i like the idea. The problem is the core code is broken. No matter what features anyone works on and no matter what circuit board they make it will still always be broken. It takes assembling one, installing into a car, setting fixed ignition timing and revving the engine to see which is a big shame for alot of people that get that far.
In the late 80s GM introduced the GMP4 ECM. It boasted a 2MHz 68HC11-based MCU running out of a 27256 EPROM (something like half of which was calibration data, the rest program)and had 256 bytes of RAM, some of it battery-backed (for learned parameters and diagnostics.) The ECM, along with the previous generation C3 (even slower) basically ran GM cars & trucks from the Cavalier to the Corvette until the OBD-II era. It controlled fuel injection, spark timing, EGR, closed loop idle speed and had a pretty sophisticated suite of sensor diagnostics (GM's "ODB-I" including an 8192bps serial stream.) It used 2- and 3D tables with interpolations, pioneered short and long-term fuel trim strategies etc. I had a great time reverse engineering the P4 out of my Z24 back in the early 90s. I'm glad Josh mentioned the fact that 600MHz wouldn't really get you a lot in the big picture. The memory & speed of an ATMega2560 dwarfs that of the P4's MCU and it would (and does) obviously make a great core around which a full-featured ECU could be based. What would you do with all those cycles when controlling a Victorian-esque reciprocating machine that could be controlled by a 2MHz MCU?
this. im making my own little single cylinder ECU. 4MHz chip. as soon as i run it at say, 8MHz... timers start to overflow. i have to deal with extra bits of data. have to consider more loops, as i now have 12 bits of data rather than 10. and on 8-bit, thats already more than enough of a hassle. and even then, 90% of the time, the chip is simply "twiddling its thumbs" waiting for interrupts as certain processes take place. 1MHz would work just as well. FFS, ive had an engine running off of a 555..., with the TPS adjusting pulse width. not terribly hard. speed of device solves no issues, in my experience it just adds MORE of them!
inb4 'Making a 5GHz liquid cooled/phase changed cooled ECU' :D I've also just discovered Speediuno. Pretty much combines my love for mechanics/electronics/programming all in one and I immediately decided that I absolutely had to throw one together. Just waiting on parts and then some free time. I LOVE that this is a thing now! Great work!
That would be the dumbest thing ever if the code it was running wasnt any good to start with. Lol a 8mhz cpu runs a shitload of cars on the road and better than most oem ecus of their time.
Thank you sir for creating speeduino brother, I have 2 speeduino based ecus, I really love the updates and bug fixes that keep making the Ecu more up to date as things progress. keep up the excellent work brother.
Just subscribed to you after finding your channel in an quite unusual way. I was browsing a Chinese Import website and saw a new empty ECU Case. That made me wonder if I would be able to build a ECU from a Micro controller. Just 2 clicks later I found what appears to be a clone of an early prototype from your project for sale on said Chinese import website 😁 That made me google the name of the product and that's how I found your channel.
@@MrJermbob you've posted several similar comments but haven't explained why you think the code is junk or what alternative you recommend. Are you planning to make a video about your specific speeduino concerns?
Wait for microcontrollers on i.MX RT1170 with ARM Cortex-M7 core running at 1 GHz with M4 core running at 400 MHz... But teensy can be overclocked (from what I heard, even up to 1 GHz mark), so...
Yeah wild they forgot about fixing the fundamental code and made a ecu on a larger mhz chip to try market the ecu as being better lolol. No way does a ecu need more mhz to do a better job. Tools.
I wonder if something like this could actually do something like traction control and some of the stuff like ABS and the Yonder.? well I think this board allows is that you can use a hybrid setup and that's probably where this board is really going to shine. this includes something like stability control and active role control things like that can really bring out the best in a vehicle especially in a heavy car.
Sure just buy a ecu with those features. Really simple. Hybrid lol. It cant even run a full fuel based setup properly yet. Its still in development and never was a stable released code.
It'd be interesting to see an MCU chip based on a SiFive X280 or similar core (maybe some ARM thing with SVE). 512-bit vector processing in a chip with good RT, timers, and IO would make some mighty interesting things explorable (finer nonlinear predictions? ML training and inferencing? microphones in the intake manifold?). I read a paper the other day on using neural networks in combination with a conventional physical model and conventional sensors to explore and optimize control signals, they evaluated it for predicting exhaust temperature (which was apparently useful for an EGR system), it seems like it could be something to look at. non-linear model predictive control in general is something to look at.
That would be pretty neat! I was taking a look on those SiPeed boards embedded with RISCV processors and a AI acceleration IC. Looks a pretty decent for this kind of application too.
You have an estimate of when the speeduino dropbear will be available, I am very interested since I have a v8 engine that I want to run in full sequential? Thanks you for your grate work.
I am also interested in it, I have a toyota 1uz engine and I also bought that same aluminum box thinking that I could put the board of the speeduino 0.4
Im assembling a speeduino board, and I think that others share my happiness to see new videos on week basis. A tip, open the channel to permit paid members.
Thanks for your comment! I've been trying to do a weekly video, either one like this or a live stream, to see if people enjoy them, so great to hear that feedback :) I have to reach a certain level of watch hours (4000) before I can turn on subscriptions, but I'm very close to that now
I know I would pay, I am just finding you and considering jumping into the challenge on a custom motorcycle build I’m doing. Would love to join a community on this
Are there any speeduino configurations that can run 4 OCV valves, 4 cam position sensors? Subaru dual AVCS here so still for 4 spark plugs and coils. Been looking at the website and only see support for 1?? camshaft sensor and 2 OCVS Thank you for the work this is amazing for car enthusiasts
i hope any board printed in 3 layer and made it only 1 1/2 credit card size.. and I really interested to try this ecu in my motorcycle, just 130 cc but convert from carburator to fuel injection
Seeing the unavailability of the Teensy 3.5 in the near (or even far) future, any plans on boosting support for the 4.x series? I'm aware that the processing power isn't necessary, but for managing more than 4 cylinders in full sequential, the Arduino Mega ain't enough!
Not really. He will go down as a legend in history as making the least capable most engine breaking ecu on the market. Time will show and idiots will buy into marketing every time.
Evening Josh. Looking at installing (hopefully) a 400+ cubic inch gen3 Hemi into a Plymouth Duster soon. Will this future ecm change also have the ability to fire all 16 spark spark plugs at the same time, do an exhaust stroke wasted spark or only fire single sided coils ? I was looking at the MSD wiring harnesses to adapt to the engine , just looking for a decent controller. I just came across you guys while watching some how to videos on a couple tuning programs. Thank you for the information on the site. Looks like an amazing piece.
With the 3.5 board being delayed until April 2023 and the 4.1 shipping shortly what are the chances of being able to get a dropbear bare or with a 4.1. I am waiting to build my first fuel injected motor and was hoping to use a dropbear but I understand the supply chain issue throwing a wrench in things.
i wish there was an option for my old bosch edc landrover ecu and the stupid needle lift injector issues gogin to be easier to put a BMW n57 swap in it
Hello, I have a Speeduino powered Scg-ecu 1.0 board. When I install the Speeduino v202103 version, there is 2.4 volts continuously on pin 31. This pin is the ignition pin. When I install the Speeduino v20208 version, 2.4 volts is not always available, but when I test it in Tunerstudio program, it comes to 2.4 volts on pin 31. 3.2 volts comes standard on other ignition pins.
This is the first time I heard of Speeduino. I played with Arduino a few years ago for an auto application. If I have time I would love to give this a try for one of my track cars.
@@ulissesmuniz Just buy a megasquirt product and be happy. No one will show you the videos of it not working properly as they dont want to be proven wrong. I wont stick a speeduino in any vehicle ever again to do it myself.
@@MrJermbob Thank's a lot, man did my search here the only problem for the megasquirt I found is the hardware here in Brazil. I could import but the temperature here is higher so some components don't work properly
Interesting, how it will compare to Portenta H7... That thing got ARM M7 core running at 480 MHz coupled with M4 core running at 240 MHz, which could be great for parallel stuff like oh, I don't know, computing on one core and logging/transmitting info to dash on another?
In terms of hardware changes? There will likely be small revisions going forward, but nothing major planned. The 0.4.4 line of SMD boards has had a few more things added with the extra board space, but nothing major Is there anything specific you're after?
Here are some suggestions. Sp721app isn't recommended for newer designs, the onboard BT on 4.4 is very user friendly and we need a way to do the 5 and 6 injectors work right on since they use the same extra mosfets in the board. Do you think there are any chances to implement a SW way to route those injectors? Thanks in advance, Josh. Big fan here.
Just started looking into speeduino and my first thought was about using a teensy instead of an atmega. Funny you should happen to have a video on this.
Am I right in assuming that the Arduino mega can be configured for 1 channel of spark and 8 channels of fuel? Or was the 9th (if I remember right there 9 timers, and apparently that's what matters for that) timer used for something else? I'm thinking locked distributor and fully sequential fuel.
I have a V6 with coil on plug engine. Trying to figure out how to make a speeduino work on it. Really I can double up injectors but can I use 2 injector outputs for fuel and remap 2 injector outputs to control spark instead? Or does the teenie give you more outputs so I can run sequencial fuel and spark on 6 injectors and coils? I've always been a diy guy! Heavy into electronics and build fpv freestyle drones that have stm f405 f722 h722 processors and learned how to re map pins to do a different function. Is it possible to do with speeduino or is there a board and parts you reccomend? It's a 06 honda pilot 3.5 j35z1 engine and automatic trans. The reason I want to do this is the engine is strong even with 300k on it! Want to run forced induction on it and remapping the ecu has been quite a daunting task! Hondata doesn't support my year ecu and just bought a scanner tool that's bidirectional and claims to do ecu programming and coding! Knew speeduino existed but always drove gm vehicles or vehicles with tons of aftermarket support! I have a ton of arduinos and esp32 chips kicking around for thinks like wifi deauthing my sister or controlling leds. Discovered a can hat for arduino and building a can reader to maybe try and do a ecu dump and poke around in the code! My old bt dongle was able to rip my ecu wirelessly and create a dump file and could use pcm hammer to edit the code but no way to reflash the ecu! Also do you know of a good resource for what vehicle use what protocol? Thanks you in advance! And thank you for doing all this work! I know how much goes into making custom hardware and then write firmware and software for it to do what you want! So thank you and if I had a better attention span than I do now I would be doing the same thing!
The original intent was never to use this as the Teensy 3.5 was the better suited MCU, however with that going end of life the Dropbear V2 will switch over to using the 600Mhz Teensy 4.1. The first batch of these are in production at the moment.
Actually, now that I think about it, shouldn't they all use the same fuel map but per-cylinder VE tables? It's the difference in air intake that causes most differences if I'm not mistaken. I mean you could do with with fuel, but to have 8 widebands... lol.
@@CreeperOnYourHouse Well sure, but you don't want to mess with your AFR target table, do you? If the variance in physical VE between cylinders causes them to produce a different lambda measurement, you don't want to cheat by changing the AFR for that cylinder. I'm approaching this from a physical perspective here, correct me if I'm wrong on this. Let's say you want to run it at 12.7:1 but measure 13.4:1 at your rearmost cylinder, making the AFR target 12.0:1 to compensate so the measured value ends up at 12.7:1 causes misinformation. If a cylinder gets more or less air due to manifold design or something, that information has to go into the appropriate table to make sure all the tables remain accurate. A table with a 12.0:1 value to make sure it runs 12.7:0 sounds absolutely nonsensical to me. Now for a given engine those two tables can be combined to a single table in which case it's effectively a fuel flow or injector duty cycle table, in which yes you're changing the fuel table/map, but this approach seems rather opaque to the tuner imo, since it doesn't reference any physical or desired measurements.
If anything the T3.5 is probably better suited. It may have a lower clock speed, but it has more available timers. The firmware though hasn't been updated for the extra 4 channels as there hasn't been a call for it at all.
@@Speeduino thank u for answering my question although i had made a mistake. I wanted to say v12 with wasted spark. I would like to add that you are the hero of us all and never stop developing the speeduino even until it surpasses motec and suvecs although its considered a stretch at this point in time.
This is interesting but can you tell me if i could use this ecu as a piggy back or as the ecu it self to run itbs on a N/A 2014 mitsubishi lancer ? If i can can you tell me where can i get this fully assembled or if i have to assemble it and if you have a parts list for this where can i get it thanks for the videos this opens a los of Doors for tuning
You still have that old mazda, R&D car :) Any way excellent work Sir and thank you for upadate . I can't wait for dropbear to show up. Greetings from Bosnia and Herzegovina
@@mayk3lll They have been in development since july~october time though. And the dropbear teensy 3.5 chip has been discontinued. There is supposed to be news on this imminently. But that appears to have been stated for a while now. SoooooonTM! 😬
could you guys make one for the ae86? or at least one with pinout diagrams so i can make my own harness. Id love to get one just to do a simple supercharger setup with one
I have a question on how to wire up a idle PWM two wire controller (ford mustang idle pwm) on speeduino .4 board. you have pin #37 and what other pin from speeduino board do you connect to? ecu ground? thanks for your help.
Hi Shawn, I’m a new speeduino user and while looking at some additions for my application I discovered the Hella OPS+T sensor, which is an oil+pressure sensor outputting data in a way somewhat similar to the SENT protocol you mentioned. I wrote an interfacing implementation of this sensor to speeduino which is currently being reviewed: github.com/noisymime/speeduino/pull/494 Actually the code works but there’s room for optimization and more elegant ways of interfacing/reacting to the sensor’s signal. It may be useful as a starting point for the SENT protocol, do you have some particular sensor in mind ? It should work on the mega as well, I have plans of doing some benchmarking on this application in the next weeks.
Potentially, but support would need to be added for the H7 MCUs. It also has a serious lack of IO unless you use the high density connectors, which are a bit of a pain.
Hey man, love your work, keep it up! quick question, i have an arduino due and have some trouble uploading the firmware to it using the speedyloader, it keeps giving me that avrdude: stk500v2 time out message, could it be that the speedyloader only works for the mega 2560 or am i missing something? since I've seen yoyr previous videos using the due instead of the mega, I'm new to the arduino and coding world so definitely looking forward to your reply, thanks in advance! (Sorry for my bad english)
I think this has been a long time coming: faster processors allow for greater sampling rates on analog signals/sensors, so by theoretical extension you can get more finely tuned in terms of overall vehicle performance. This should be the way forward as far as I’m concerned
Once you exceed the slew rates of the sensors themselves, faster sampling just introduces noise to your dataset. Even running a motor at 20,000 RPM means things happen in the motor at 20KHz, so only sample rates of 20KHz and harmonics of that are essentially real-time. Trust me, I've spent way too much time implementing advanced digital filtering algorithms, eating up gobs of memory and processing power to try to keep a PID stable sampling at every code cycle, just to ultimately end up reducing the sample rate to 1Sample/sec
@@Koumajutsu I accept that, but believe noise can (should!) be filtered or otherwise compensated for prior to a signal being digitized…the fidelity of the input transducer isn’t often questioned.
@sourand jaded When i learned what "Sequential Injection/Ignition" was, my first thought was "wait - you're telling me manifold injection wasn't timed with the opening of the valve this whole time?!" Top 10 anime betrayals. And yet the engine still runs. So whilst 600Mhz doesn't make a lot of sense if you're only target is to get it to run, because it doesn't need to be accurate, I can see people using such CPUs to do the math needed for alternative combustion cycles, like HCCI, that do need to be accurate, and see a change from 20mpg to 45+ mpg. Could be a lifesaver for the combustion engine frankly
I'm in love with this ecu and I don't even own one YET, once they're back in stock i will be purchasing one. I'm also laughing because of the thousands of songs available in the youtube library i used the exact same songs in one of my recent videos.
@speeduino hey i have a project coming up which includes dual engines and i would like to control both engines simultaneously on one ECU i would also like to have D.B.W. on both as well as a strain gauge to shift both transmissions sequentially is this something a 8 cylinder capable ECU could handle ?
Hello, I am an engineeringstudent in France, studying electronics/programming. I am about to start a very interesting school project and I need your guidance! I own a Honda CRF250L 2018 engine (single cyl 4 stroke) and a go kart chassis, the plan is design all the rest ourselves: ECU/Harness design, touchscreen dashboard, flappy paddle gear box and servo for the clutch (previous project but swapping from an arduino for a custom circuit), full lights (including brake lights/indicators) and possibly a pivoting rear wing. The project is replacing work experience so it is supposed to last 12 weeks (35hr/week). Do you have an email to contact you directly? Thanks in advance!
Probably easiest to jump on Slack and we can have a chat about it if you'd like: speeduino.slack.com/join/shared_invite/enQtNzE0MjU2ODE5MTM5LWZhOWMwMWFhZTVjYTg5NmYwNzhmMGNhNDE4NmJiNDc0NDgyYmZiMjg1YjJjZmNjNDk2YTYzNzk1ZWEyNGJhNDI#/
Not really... It's more like barely even 1 order of magnitude. And 1 order means 10x, 2 orders mean 100x, 10 orders - 10000000000x, etc. Which still makes it really impressive.
@Corey MacPherson I don't, but there are engines in junkyards sometimes for some silly money. It was some time ago but I think I saw some car. It was bmw or jaguar for sale but there was sth wrong with it and firing it up on standalone cheaper than that car would be fix for me XD
@@lildude6906 No it hasnt. Its been successful to get into the broke customer market sure. Success in making horsepower no it has not. Not bugs that cost you irregular ignition timing. Big NO NO
@@MrJermbob there are countless pnp examples of speeduino in bmw, miatas, dsm, and Hondas lately. I would say that's quite successful for an open source stand alone ecu
@@lildude6906 Sure nothing that makes power or can show a decent dyno run. I challenge them all to set fixed timing, rev the motor and watch the mark with a timing light. I mark my word it will jump around like crazy.
No matter how fast it is still backdated. If there is DBW function I would say yes it does make some forward moves but just adding more drivers I think it is pretty useless.
Thank you for modernizing aging ECU's into opensource soft and hardware!
They didnt. Its broken Speedunio will never take off as its junk code. The physical hardware is so simple. Noobs.
@@MrJermbob What do you recommend in the Speeduino's place, then? Can you make and alternative yourself?
@@Cmdrbzrd Just use a megasquirt. At least their code has been made to be sold. That way it has to be quality and supported. I have had a long play with speedy and ran it on one of my own personal vehicles as i like the idea. The problem is the core code is broken. No matter what features anyone works on and no matter what circuit board they make it will still always be broken. It takes assembling one, installing into a car, setting fixed ignition timing and revving the engine to see which is a big shame for alot of people that get that far.
@@MrJermbob At least spell it correctly, noob.
@@d1e6om Spell what right munter ?
In the late 80s GM introduced the GMP4 ECM. It boasted a 2MHz 68HC11-based MCU running out of a 27256 EPROM (something like half of which was calibration data, the rest program)and had 256 bytes of RAM, some of it battery-backed (for learned parameters and diagnostics.) The ECM, along with the previous generation C3 (even slower) basically ran GM cars & trucks from the Cavalier to the Corvette until the OBD-II era. It controlled fuel injection, spark timing, EGR, closed loop idle speed and had a pretty sophisticated suite of sensor diagnostics (GM's "ODB-I" including an 8192bps serial stream.) It used 2- and 3D tables with interpolations, pioneered short and long-term fuel trim strategies etc. I had a great time reverse engineering the P4 out of my Z24 back in the early 90s. I'm glad Josh mentioned the fact that 600MHz wouldn't really get you a lot in the big picture. The memory & speed of an ATMega2560 dwarfs that of the P4's MCU and it would (and does) obviously make a great core around which a full-featured ECU could be based. What would you do with all those cycles when controlling a Victorian-esque reciprocating machine that could be controlled by a 2MHz MCU?
As you know mhz does not matter. More speed does not mean a better more capable ecu.
@@MrJermbob lol are you just in everybody's thread dropping your opinion?
@@lildude6906 Yup haha
Impressive coding for such a slow chip. Assembler was used without a doubt.
this.
im making my own little single cylinder ECU.
4MHz chip.
as soon as i run it at say, 8MHz... timers start to overflow. i have to deal with extra bits of data. have to consider more loops, as i now have 12 bits of data rather than 10. and on 8-bit, thats already more than enough of a hassle. and even then, 90% of the time, the chip is simply "twiddling its thumbs" waiting for interrupts as certain processes take place.
1MHz would work just as well.
FFS, ive had an engine running off of a 555..., with the TPS adjusting pulse width. not terribly hard.
speed of device solves no issues, in my experience it just adds MORE of them!
I cant wait buddy. these are going to be great ecu's. We are going to make alot of people very happy in my area.
inb4 'Making a 5GHz liquid cooled/phase changed cooled ECU' :D
I've also just discovered Speediuno. Pretty much combines my love for mechanics/electronics/programming all in one and I immediately decided that I absolutely had to throw one together. Just waiting on parts and then some free time. I LOVE that this is a thing now! Great work!
That would be the dumbest thing ever if the code it was running wasnt any good to start with. Lol a 8mhz cpu runs a shitload of cars on the road and better than most oem ecus of their time.
@@MrJermbob Aint that the truth, my old Megasquirt 1 is only 8 MHz and it runs my Honda B-series to 11000 RPM and does not miss a beat.
@@Mike-ry4ti Megasquirt for life. Anyone trying to find cheaper is just playing the car game wrong.
Super video! I applauded for $2.00 👏
Thank you sir for creating speeduino brother, I have 2 speeduino based ecus, I really love the updates and bug fixes that keep making the Ecu more up to date as things progress. keep up the excellent work brother.
Hows your fixed ignition timing hold on revving with a timing light ? Lol
@@MrJermbob i really dont know i dont have a speedy ecu currently in stalled i used a chipped ecu in my civic
@@turboimport95 this guy is going around replying in comments that are years old.. lol
Just subscribed to you after finding your channel in an quite unusual way.
I was browsing a Chinese Import website and saw a new empty ECU Case.
That made me wonder if I would be able to build a ECU from a Micro controller.
Just 2 clicks later I found what appears to be a clone of an early prototype from your project for sale on said Chinese import website 😁
That made me google the name of the product and that's how I found your channel.
Probably want to stay away from those cheap as hell crap ecus in your car just fyi.
@@MrJermbob As someone with quite broad knowledge of electronics I don't see wy not.
As long as the materials used are ok it will be fine.
@@NinoJoel Its the code. Not the hardware. Hardware is easy as you know.
@@MrJermbob you've posted several similar comments but haven't explained why you think the code is junk or what alternative you recommend. Are you planning to make a video about your specific speeduino concerns?
@@adamgarlow5347 Wouldn't waste my time. I have said why and about the fixed timing.
"There it is. A 600 MHz ECU."
Absolutely wild.
Wait for microcontrollers on i.MX RT1170 with ARM Cortex-M7 core running at 1 GHz with M4 core running at 400 MHz... But teensy can be overclocked (from what I heard, even up to 1 GHz mark), so...
@@spakeschannel So... MHZ does not make a ecu better lol. Get with it.
Yeah wild they forgot about fixing the fundamental code and made a ecu on a larger mhz chip to try market the ecu as being better lolol. No way does a ecu need more mhz to do a better job. Tools.
I wonder if something like this could actually do something like traction control and some of the stuff like ABS and the Yonder.?
well I think this board allows is that you can use a hybrid setup and that's probably where this board is really going to shine.
this includes something like stability control and active role control things like that can really bring out the best in a vehicle especially in a heavy car.
Really only limited to your coding abilities as far as I've looked into
@@wesker89 What about the timers available inside the chip seeing you seem to know how it all works ?
Sure just buy a ecu with those features. Really simple. Hybrid lol. It cant even run a full fuel based setup properly yet. Its still in development and never was a stable released code.
It'd be interesting to see an MCU chip based on a SiFive X280 or similar core (maybe some ARM thing with SVE). 512-bit vector processing in a chip with good RT, timers, and IO would make some mighty interesting things explorable (finer nonlinear predictions? ML training and inferencing? microphones in the intake manifold?).
I read a paper the other day on using neural networks in combination with a conventional physical model and conventional sensors to explore and optimize control signals, they evaluated it for predicting exhaust temperature (which was apparently useful for an EGR system), it seems like it could be something to look at. non-linear model predictive control in general is something to look at.
That would be pretty neat! I was taking a look on those SiPeed boards embedded with RISCV processors and a AI acceleration IC. Looks a pretty decent for this kind of application too.
No need. MHZ does not make a better driving car.
You have an estimate of when the speeduino dropbear will be available, I am very interested since I have a v8 engine that I want to run in full sequential?
Thanks you for your grate work.
I am also interested in it, I have a toyota 1uz engine and I also bought that same aluminum box thinking that I could put the board of the speeduino 0.4
Me too.
@@omardelarosa4899 I buy the same case thinking I can put the 0.4 xD
Dont do it. get a better ecu.
Im assembling a speeduino board, and I think that others share my happiness to see new videos on week basis. A tip, open the channel to permit paid members.
Thanks for your comment! I've been trying to do a weekly video, either one like this or a live stream, to see if people enjoy them, so great to hear that feedback :)
I have to reach a certain level of watch hours (4000) before I can turn on subscriptions, but I'm very close to that now
I know I would pay, I am just finding you and considering jumping into the challenge on a custom motorcycle build I’m doing. Would love to join a community on this
@@Speeduino 2 years of weekly videos owing there Josh. Your code is still broken.
@@redbeard4671 Paying to get screwed over. Sounds fun.
Are there any speeduino configurations that can run 4 OCV valves, 4 cam position sensors? Subaru dual AVCS here so still for 4 spark plugs and coils. Been looking at the website and only see support for 1?? camshaft sensor and 2 OCVS
Thank you for the work this is amazing for car enthusiasts
i hope any board printed in 3 layer and made it only 1 1/2 credit card size.. and I really interested to try this ecu in my motorcycle, just 130 cc but convert from carburator to fuel injection
Seeing the unavailability of the Teensy 3.5 in the near (or even far) future, any plans on boosting support for the 4.x series? I'm aware that the processing power isn't necessary, but for managing more than 4 cylinders in full sequential, the Arduino Mega ain't enough!
There are no words. Except what an absolute legend this bloke is 👏🏼👏🏼👏🏼
Not really. He will go down as a legend in history as making the least capable most engine breaking ecu on the market. Time will show and idiots will buy into marketing every time.
Evening Josh.
Looking at installing (hopefully) a 400+ cubic inch gen3 Hemi into a Plymouth Duster soon. Will this future ecm change also have the ability to fire all 16 spark spark plugs at the same time, do an exhaust stroke wasted spark or only fire single sided coils ?
I was looking at the MSD wiring harnesses to adapt to the engine , just looking for a decent controller. I just came across you guys while watching some how to videos on a couple tuning programs.
Thank you for the information on the site. Looks like an amazing piece.
No it wont do 16 channels of injector. Do NOT use this ecu. Buy something better. Buy once cry once.
Ever plan on making more of the drop bear?
With the 3.5 board being delayed until April 2023 and the 4.1 shipping shortly what are the chances of being able to get a dropbear bare or with a 4.1. I am waiting to build my first fuel injected motor and was hoping to use a dropbear but I understand the supply chain issue throwing a wrench in things.
Buy something better. Consider yourself lucky you cant get a speeduino. They are broken junk. The code is fundamentally flawed.
@@8SecSleeper No thanks. Its rubbish. Ill stick to quality code.
would have been nice for you to list why you want to stay with 3.6. you just said there is a couple reasons and left it.
i wish there was an option for my old bosch edc landrover ecu and the stupid needle lift injector issues gogin to be easier to put a BMW n57 swap in it
Hello and thank you man! I don't know if Ur going to answer to my question! What's the newest bord for Teenesy 4.1 adapter?
Hello, I have a Speeduino powered Scg-ecu 1.0 board. When I install the Speeduino v202103 version, there is 2.4 volts continuously on pin 31. This pin is the ignition pin. When I install the Speeduino v20208 version, 2.4 volts is not always available, but when I test it in Tunerstudio program, it comes to 2.4 volts on pin 31. 3.2 volts comes standard on other ignition pins.
Old answer, but you need to change the pinouts on Speeduino firmware. It's easy.
This is the first time I heard of Speeduino. I played with Arduino a few years ago for an auto application. If I have time I would love to give this a try for one of my track cars.
DONT! its rubbish broken code. I know from experience and its a joke. I can prove all very clearly. Speedy is rubbish.
@@MrJermbob Prove me, i'm interested in trying but scared about the problems.
@@ulissesmuniz Just buy a megasquirt product and be happy. No one will show you the videos of it not working properly as they dont want to be proven wrong. I wont stick a speeduino in any vehicle ever again to do it myself.
@@MrJermbob Thank's a lot, man did my search here the only problem for the megasquirt I found is the hardware here in Brazil. I could import but the temperature here is higher so some components don't work properly
@@ulissesmuniz Megasquirt 2 unit would be no issues with the temps you see.
Interesting, how it will compare to Portenta H7... That thing got ARM M7 core running at 480 MHz coupled with M4 core running at 240 MHz, which could be great for parallel stuff like oh, I don't know, computing on one core and logging/transmitting info to dash on another?
But MHZ does not mean a better running engine.
Hi, why speeduino dont have updates for remplace the 2560??
is there any plans to make a 6 cylinder universal ecu? if not which do you recommended to someone whose on a budget?
Buy something better. CHEAPER DOES NOT MEAN BETTER.
@Speeduino Will there be any more updates/upgrades for the old 0.4.3 models, Josh?
In terms of hardware changes? There will likely be small revisions going forward, but nothing major planned. The 0.4.4 line of SMD boards has had a few more things added with the extra board space, but nothing major
Is there anything specific you're after?
Here are some suggestions. Sp721app isn't recommended for newer designs, the onboard BT on 4.4 is very user friendly and we need a way to do the 5 and 6 injectors work right on since they use the same extra mosfets in the board. Do you think there are any chances to implement a SW way to route those injectors? Thanks in advance, Josh. Big fan here.
Just started looking into speeduino and my first thought was about using a teensy instead of an atmega. Funny you should happen to have a video on this.
Holy shit the loop speed 🤣 amazing work.
Next: ryzen 9 18 core 5 GHz liquid-cooled ECU (may need to upgrade your alternator to power this badass)
I wonder how a Parallax Propeller/Propeller 2 would work for this
Am I right in assuming that the Arduino mega can be configured for 1 channel of spark and 8 channels of fuel? Or was the 9th (if I remember right there 9 timers, and apparently that's what matters for that) timer used for something else? I'm thinking locked distributor and fully sequential fuel.
Yo very far off topic but slightly related. Any thoughts on a Framework mainboard being usable as both an ECU and in car media center?
It's not a bad shout, thought about it myself, however, if one crashes, so does the other. This will cause some issues.
I have a V6 with coil on plug engine. Trying to figure out how to make a speeduino work on it. Really I can double up injectors but can I use 2 injector outputs for fuel and remap 2 injector outputs to control spark instead? Or does the teenie give you more outputs so I can run sequencial fuel and spark on 6 injectors and coils? I've always been a diy guy! Heavy into electronics and build fpv freestyle drones that have stm f405 f722 h722 processors and learned how to re map pins to do a different function. Is it possible to do with speeduino or is there a board and parts you reccomend? It's a 06 honda pilot 3.5 j35z1 engine and automatic trans. The reason I want to do this is the engine is strong even with 300k on it! Want to run forced induction on it and remapping the ecu has been quite a daunting task! Hondata doesn't support my year ecu and just bought a scanner tool that's bidirectional and claims to do ecu programming and coding! Knew speeduino existed but always drove gm vehicles or vehicles with tons of aftermarket support! I have a ton of arduinos and esp32 chips kicking around for thinks like wifi deauthing my sister or controlling leds. Discovered a can hat for arduino and building a can reader to maybe try and do a ecu dump and poke around in the code! My old bt dongle was able to rip my ecu wirelessly and create a dump file and could use pcm hammer to edit the code but no way to reflash the ecu! Also do you know of a good resource for what vehicle use what protocol? Thanks you in advance! And thank you for doing all this work! I know how much goes into making custom hardware and then write firmware and software for it to do what you want! So thank you and if I had a better attention span than I do now I would be doing the same thing!
Just to think, they used to do the same thing with tiny toilet valves and adjustment screws.
It's been 3 years. Ik hope this is going to be released.
The original intent was never to use this as the Teensy 3.5 was the better suited MCU, however with that going end of life the Dropbear V2 will switch over to using the 600Mhz Teensy 4.1. The first batch of these are in production at the moment.
@@Speeduino wooow thanks looking forward
With greater speed and larger storage, will per-cylinder, as well as adaptive, fuel mapping become available?
Per cylinder fuel mapping would be amazing.
Actually, now that I think about it, shouldn't they all use the same fuel map but per-cylinder VE tables? It's the difference in air intake that causes most differences if I'm not mistaken.
I mean you could do with with fuel, but to have 8 widebands... lol.
@@ChielScape AFR table is a form of adaptive fuel trim. They're both a type of fuel map.
@@CreeperOnYourHouse Well sure, but you don't want to mess with your AFR target table, do you? If the variance in physical VE between cylinders causes them to produce a different lambda measurement, you don't want to cheat by changing the AFR for that cylinder. I'm approaching this from a physical perspective here, correct me if I'm wrong on this.
Let's say you want to run it at 12.7:1 but measure 13.4:1 at your rearmost cylinder, making the AFR target 12.0:1 to compensate so the measured value ends up at 12.7:1 causes misinformation. If a cylinder gets more or less air due to manifold design or something, that information has to go into the appropriate table to make sure all the tables remain accurate. A table with a 12.0:1 value to make sure it runs 12.7:0 sounds absolutely nonsensical to me.
Now for a given engine those two tables can be combined to a single table in which case it's effectively a fuel flow or injector duty cycle table, in which yes you're changing the fuel table/map, but this approach seems rather opaque to the tuner imo, since it doesn't reference any physical or desired measurements.
Any plan to release the board as an open source ?
?
i mean just buy one and look at it?
Is the 3.5 capable of running v12 full sequential if we really wanted to or you would choose 4.0 for that??
If anything the T3.5 is probably better suited. It may have a lower clock speed, but it has more available timers. The firmware though hasn't been updated for the extra 4 channels as there hasn't been a call for it at all.
@@Speeduino thank u for answering my question although i had made a mistake. I wanted to say v12 with wasted spark.
I would like to add that you are the hero of us all and never stop developing the speeduino even until it surpasses motec and suvecs although its considered a stretch at this point in time.
@@FirstLast-tx3yj Ohh then definitely, v12 wasted support will be there on this board.
@@joshstewart2131 cool man, i was about to buy ms3 pro. Now i am just sitting around waiting!!🤣
@@joshstewart2131 cool man, i was about to buy ms3 pro. Now i am just sitting around waiting!!🤣
Hi. Is there knock sensor input? It need a madule? Is there a cheap way to have the knock sensor module?
Do you have any model of plate that makes ignition and injection sequence 5cil fiat?
This is interesting but can you tell me if i could use this ecu as a piggy back or as the ecu it self to run itbs on a N/A 2014 mitsubishi lancer ? If i can can you tell me where can i get this fully assembled or if i have to assemble it and if you have a parts list for this where can i get it thanks for the videos this opens a los of Doors for tuning
You still have that old mazda, R&D car :) Any way excellent work Sir and thank you for upadate . I can't wait for dropbear to show up. Greetings from Bosnia and Herzegovina
The MX5 cannot be killed, no matter how many test ECUs and half baked code I throw at it 😂
@@Speeduino Its still half baked mate.
Can I use speeduino for a honda k24? Vtec and vtc?
The cam/crank trigger patterns are in development right now.
@@RuneRelic awesome! Thank you for the info
@@mayk3lll They have been in development since july~october time though. And the dropbear teensy 3.5 chip has been discontinued. There is supposed to be news on this imminently. But that appears to have been stated for a while now.
SoooooonTM!
😬
@@RuneRelic is this all on speediuno website? I seen there it multiple website for different regions
Why don't you increase the size of the injection table and ignition table, since now there's more power available?
Have you even tuned cars to say that ? Lol. Interpolation my friend.
Would this do 16,000rpm for a cbr600 motor?
could you guys make one for the ae86? or at least one with pinout diagrams so i can make my own harness. Id love to get one just to do a simple supercharger setup with one
The speeduino forum/wiki has all you need to know
So is this available yet?
Are there any downsides to using that board and what are the advantages just faster response times and stuff or what
Great job my respect 🫡
im running an on/off switch to my supercharger, will the speeduino be able to correct itself on the fly or will i have to change the tune manually ?
Of course, whilst off boost the tune will simply run in the NA portion of the fuel and ignition maps.
@@Speeduino thanks for the fast response mate :)
Thanks for your contribution.
probably could run one on a rp2040 [dualcore arm mcu from RPF] as well, probably might need some external ADCs though
have you used simulink at all to help you with the ecu??
Καλησπέρα τα βίντεο που εχεις κανεις με τον αγραφο στo crypton x ειναι με speeduino ?
Awesome work
I have a question on how to wire up a idle PWM two wire controller (ford mustang idle pwm) on speeduino .4 board. you have pin #37 and what other pin from speeduino board do you connect to? ecu ground? thanks for your help.
Connect one side of the idle valve to a switched 12v source and the other side of it to the output on the Speeduino.
@@Speeduino hello, is that a positive 12v source and pin #37 on speeduino? thank you
Is it possible with this fast cpu to change the int values from the ve table?
SO IS IT MADE FOR MY OBD1 HONDA IF SO WHERE CAN I BUY ONE
can we have 100hz logs now?
Once the SD logging is written I'll have a good idea how fast this can run, but 100Hz should easily be possible on the Dropbear.
@@Speeduino can't wait
100hz ? :DD now 500-1000hz isnt surprise with modern ecu's
What are the advantages of the 3.5? I'd have thought standardise on the latest, balancing cost & speed.
As stated above, it has more available timers.
More timers is certainly the big one but I also have concerns about the temperature that the 4.1 runs at
@@Speeduino Great thank you for that
@@Speeduino Nothing a tiny heatsink and some thermal epoxy can't fix, no?
@@ChielScape No. that does nothing. Where is the heat energy getting sent to ?
would this fit in my link ecu?
600 Mhz at them! Love the growth I see of Speeduino over the years!
It hasnt grown at all. Still cant hold fixed timing lol.
Can these boards run i-vtec ?? K20 engines
Thank you this is the coolest open-source thing I have seen
When do you think the dropbear will be available?
Hoping to do a beta in about a month or so, production run will depend on the results of that
@@Speeduino looking forward to hearing on the results. Got a 1uzfe with a supercharger on that needs an ECU
I don’t see those boards you were using for diy or purchase
What's the ticking noise when you start the car? Sounds like a misfire.
Classic MX5 lifter tick :)
Do you think the 4.1 will enable use of SENT sensors or do you think the other processors are capable?
Hi Shawn, I’m a new speeduino user and while looking at some additions for my application I discovered the Hella OPS+T sensor, which is an oil+pressure sensor outputting data in a way somewhat similar to the SENT protocol you mentioned.
I wrote an interfacing implementation of this sensor to speeduino which is currently being reviewed: github.com/noisymime/speeduino/pull/494
Actually the code works but there’s room for optimization and more elegant ways of interfacing/reacting to the sensor’s signal.
It may be useful as a starting point for the SENT protocol, do you have some particular sensor in mind ?
It should work on the mega as well, I have plans of doing some benchmarking on this application in the next weeks.
@@deffie83 I feel sorry for your engine.
Thank you so much!!!!
Would this work with an automatic gearbox? Planning on using this with a dq500. Originally planned to use an MS but then discovered this amazing ECU
Its not a amazing ecu at all. Stick to megasquirt. I bet you cant find anything amazing a speedy is doing.
Do you make a PnP foe the 99-00 NB Miata?
What is the cost???
🤔 Hows abouts a TCU (Transmission Control Unit)
Is this thing for sale?
This fast enough to use on a rotary application?
Why wouldn't it be fast enough? I mean it can run normal piston engines to pretty high rpms, and i see no issues in running speeduino on rotary...
Also there is a video of a guy on TH-cam running FC Na engine on speeduino
There are at least 5 rotaries running on the 16Mhz Arduino so.... Yes
Ladislav Daníček my concern was that they run with half the cycle time of a conventional engine. Thank you
Speeduino FD RX-7 is still under development?
Tanks friend god bless 😅
Isn't the Arduino Portenta H7 a viable option?
Potentially, but support would need to be added for the H7 MCUs. It also has a serious lack of IO unless you use the high density connectors, which are a bit of a pain.
put a link to speeduino in the description of your vids
i dont know much about ecu but im assuming faster processing,more accurate data?
Faster just means a more powerful processor... more computation power
Hey man, love your work, keep it up! quick question, i have an arduino due and have some trouble uploading the firmware to it using the speedyloader, it keeps giving me that avrdude: stk500v2 time out message, could it be that the speedyloader only works for the mega 2560 or am i missing something? since I've seen yoyr previous videos using the due instead of the mega, I'm new to the arduino and coding world so definitely looking forward to your reply, thanks in advance! (Sorry for my bad english)
He wont reply. He needs to help himself with his broken code.
Waiting for RGB on the board ;)
The added processing power is only useless until some nutter finds a use for it :)
I think this has been a long time coming: faster processors allow for greater sampling rates on analog signals/sensors, so by theoretical extension you can get more finely tuned in terms of overall vehicle performance. This should be the way forward as far as I’m concerned
Once you exceed the slew rates of the sensors themselves, faster sampling just introduces noise to your dataset. Even running a motor at 20,000 RPM means things happen in the motor at 20KHz, so only sample rates of 20KHz and harmonics of that are essentially real-time. Trust me, I've spent way too much time implementing advanced digital filtering algorithms, eating up gobs of memory and processing power to try to keep a PID stable sampling at every code cycle, just to ultimately end up reducing the sample rate to 1Sample/sec
@@Koumajutsu I accept that, but believe noise can (should!) be filtered or otherwise compensated for prior to a signal being digitized…the fidelity of the input transducer isn’t often questioned.
20KHz is 20,000 times a second. 20,000RPM is 20,000 times a minute. 2KHz. This is 600Mhz. That’s 300000x faster. Computers are mad fast bruh 😎
@sourand jaded When i learned what "Sequential Injection/Ignition" was, my first thought was "wait - you're telling me manifold injection wasn't timed with the opening of the valve this whole time?!" Top 10 anime betrayals. And yet the engine still runs. So whilst 600Mhz doesn't make a lot of sense if you're only target is to get it to run, because it doesn't need to be accurate, I can see people using such CPUs to do the math needed for alternative combustion cycles, like HCCI, that do need to be accurate, and see a change from 20mpg to 45+ mpg. Could be a lifesaver for the combustion engine frankly
No MHZ does not translate into a better running engine lol. Look at all the engine failures speedy has caused already lol.
Why stop uploading bro😢
hmm.. i think it would be cool to have a free ecu memory gauge on the dash
Why no need. Ecus before this work fine with only 8mhz and a tiny bit of ram.
Does the 600 mhz mean 20000 rpm rev limit
No.
Am I crazy to assume you can use a faster chip to run closer to lean making more power in a safer manner?
The ecu processor speed has nothing to do on how lean you can run engine, how much it makes hp, etc.
maybe if you use a wideband o2 sensor. with the traditional one for sure that is not going to happen
Not at all. The code the ecu is based on is broken and does not hold solid ignition timing. Run for the hills away from speeduino.
I'm in love with this ecu and I don't even own one YET, once they're back in stock i will be purchasing one. I'm also laughing because of the thousands of songs available in the youtube library i used the exact same songs in one of my recent videos.
Dont bother get a better ecu. Speeduino is broken code junk.
@speeduino hey i have a project coming up which includes dual engines and i would like to control both engines simultaneously on one ECU i would also like to have D.B.W. on both as well as a strain gauge to shift both transmissions sequentially is this something a 8 cylinder capable ECU could handle ?
Hello, I am an engineeringstudent in France, studying electronics/programming. I am about to start a very interesting school project and I need your guidance!
I own a Honda CRF250L 2018 engine (single cyl 4 stroke) and a go kart chassis, the plan is design all the rest ourselves: ECU/Harness design, touchscreen dashboard, flappy paddle gear box and servo for the clutch (previous project but swapping from an arduino for a custom circuit), full lights (including brake lights/indicators) and possibly a pivoting rear wing.
The project is replacing work experience so it is supposed to last 12 weeks (35hr/week).
Do you have an email to contact you directly?
Thanks in advance!
Probably easiest to jump on Slack and we can have a chat about it if you'd like: speeduino.slack.com/join/shared_invite/enQtNzE0MjU2ODE5MTM5LWZhOWMwMWFhZTVjYTg5NmYwNzhmMGNhNDE4NmJiNDc0NDgyYmZiMjg1YjJjZmNjNDk2YTYzNzk1ZWEyNGJhNDI#/
that is orders of magnitude... DAMN...
Not really... It's more like barely even 1 order of magnitude. And 1 order means 10x, 2 orders mean 100x, 10 orders - 10000000000x, etc. Which still makes it really impressive.
@@spakeschannel 2 orders of magnitude is most surely 100x not 20x
@@IIIRotor Oh, yeah, my bad, sorry...
Are you going to make it now up to 12 cylinders ?
@Corey MacPherson If there are possibilities why to limit it?
@Corey MacPherson I don't, but there are engines in junkyards sometimes for some silly money. It was some time ago but I think I saw some car. It was bmw or jaguar for sale but there was sth wrong with it and firing it up on standalone cheaper than that car would be fix for me XD
threadripper ecu when
Just found out about your product and look forward to learning more!
Run for the hills. Buy something better. Please dont use these broken code ecus.
@OOFSPEC speeduino has been quite successful....bugs always exist in firmware even with big name companies.
@@lildude6906 No it hasnt. Its been successful to get into the broke customer market sure. Success in making horsepower no it has not. Not bugs that cost you irregular ignition timing. Big NO NO
@@MrJermbob there are countless pnp examples of speeduino in bmw, miatas, dsm, and Hondas lately. I would say that's quite successful for an open source stand alone ecu
@@lildude6906 Sure nothing that makes power or can show a decent dyno run. I challenge them all to set fixed timing, rev the motor and watch the mark with a timing light. I mark my word it will jump around like crazy.
but stillwithout knock control
cant wait to get Teensy 80.1 and 5ghz
And still wont be able to hold solid ignition timing. Lol.
this is so stupidly awesome :D
well..... my first ibm pc was 4.77 MHz/16 bit, so ...
No matter how fast it is still backdated. If there is DBW function I would say yes it does make some forward moves but just adding more drivers I think it is pretty useless.