K.I.S.S. Flight Computer Development.

[Mason T]

The majority of flights would be below 10000ft. That's where the market is for a KISS computer. The scope discussed here is already well beyond KISS.
So KISS or build a full blown 100k capable flight computer. What is it you want to make?

That is a great question. This thread turned into an interesting thread on high altitude measurement. but because as you pointed out the vast majority of flyers don't even come close to 100k feet. I am using a barometer that is good to ~30k. I believe this captures ~ 95% of the market. This allows me to use a much cheaper barometer and will allow the price to stay as affordable as possible. The high altitude stuff was brought up as a potential expansion in the future if there was demand.

 

[Spacedog49Krell]

What data rate are you using for your accelerometer and gyro's with the Rev 3? I found that 200-250 Hz was a usable minimum data rate. The resolution is adequate for launch rod or rail velocity studies.

 

[SolarYellow]

For me, the current benchmarks are Eggtimer Quantum, Proton and Quasar, and the Blue Raven (both form factors). I would not see the point of something that didn't improve on the envelope defined by those with some combination of performance/capability, fitting in smaller tubes (BT-50 or even BT-20), or price.

 

[Mason T]

What data rate are you using for your accelerometer and gyro's with the Rev 3? I found that 200-250 Hz was a usable minimum data rate. The resolution is adequate for launch rod or rail velocity studies.

My current data rate on the Accelerometer and Gyro is 1000Hz. Once I get the FIFO working I will be able to to to 6.66KHz. Those are some data points I want to look at as well as detection of when the ejection charge fires. The forces on the rocket as the parachute opens and impact forces upon landing.

 

[Mason T]

For me, the current benchmarks are Eggtimer Quantum, Proton and Quasar, and the Blue Raven (both form factors). I would not see the point of something that didn't improve on the envelope defined by those with some combination of performance/capability, fitting in smaller tubes (BT-50 or even BT-20), or price.

Those are all great options. The KISS is certainly not a replacement for them. It will be closer to a more modern StratoLoggerCF with faster sampling and an added accelerometer/Gyro, no custom cables needed as it has USBC, and it speaks to you rather than beeping information out.

 

[Spacedog49Krell]

My current data rate on the Accelerometer and Gyro is 1000Hz. Once I get the FIFO working I will be able to to to 6.66KHz. Those are some data points I want to look at as well as detection of when the ejection charge fires. The forces on the rocket as the parachute opens and impact forces upon landing.

I don't classify 1000Hz as a KISS flight computer. To record 7 channels at 1000Hz data throughput speed, I use a Teensy 4.1 to a QSPI Flash. I see you have a STM32G0x. I sometimes use the STM32F407 in applications.

 

[Mason T]

I don't classify 1000Hz as a KISS flight computer. To record 7 channels at 1000Hz data throughput speed, I use a Teensy 4.1 to a QSPI Flash. I see you have a STM32G0x. I sometimes use the STM32F407 in applications.

It's not KISS from a HW prospective. There is a lot going into the HW to make is robust and as simple as possible for the user. The KISS is from the users perspective. All the user has to interact with the flight computer is 2 buttons. It will tell the user how its configured so the user doesn't have to have a manual to reference to decode how its configured. I am not following how it having 1000Hz data excludes it from being a simple. If you are referring to simple HW sure, I can see that. A Teensy 4.1 running at 600MHz would be way overkill for this project. I already am already capturing data at 1000Hz with a 500uS of buffer before implementing any DMA. Yes it is a little G0 but it has enough horsepower to get the job done.

 

[Spacedog49Krell]

It's not KISS from a HW prospective. There is a lot going into the HW to make is robust and as simple as possible for the user. The KISS is from the users perspective. All the user has to interact with the flight computer is 2 buttons. It will tell the user how its configured so the user doesn't have to have a manual to reference to decode how its configured. I am not following how it having 1000Hz data excludes it from being a simple. If you are referring to simple HW sure, I can see that. A Teensy 4.1 running at 600MHz would be way overkill for this project. I already am already capturing data at 1000Hz with a 500uS of buffer before implementing any DMA. Yes it is a little G0 but it has enough horsepower to get the job done.

At 1000Hz, the vast amount of data requires SW for the user to manipulate and display the data. Even at 500Hz, a typical flight data file exceeds what the average user can comfortably manipulate within Excel. Unless you are providing the display and analysis SW it crosses the line, in my perspective, as a KISS development.

I'm not implying that the STM32G0x doesn't have horsepower, I know it does have the power. I had supply issues with STM parts during COVID. I moved to suppliers of processor chips that were readily available with similar processing power and never went back to STM.

Keep up the drive for faster data rates.

 

[cerving]

Suggestion: In the spirit of KISS, ditch the gyro and accel. The vast majority of hobbyists don't need them and won't use them, and it just jacks up the cost and complexity. If you just have it doing dual-deploy with the push-button setting selections and the voice setting/flight summary call-outs, that's all you need. After you've sold a few thousand of them and gotten your manufacturing and customer support processes tuned, you can create a "deluxe" version with the extra bells and whistles.

 

[Spacedog49Krell]

Suggestion: In the spirit of KISS, ditch the gyro and accel. The vast majority of hobbyists don't need them and won't use them, and it just jacks up the cost and complexity. If you just have it doing dual-deploy with the push-button setting selections and the voice setting/flight summary call-outs, that's all you need. After you've sold a few thousand of them and gotten your manufacturing and customer support processes tuned, you can create a "deluxe" version with the extra bells and whistles.

That is a very good suggestion. A friend developed a low cost flight computer a few years ago, but could not sell enough to recoup his costs. He now sells a 1.3 gram altimeter. The altimeter is rugged and can survive a ground impact. He is getting it approved for competition altitude accuracy.

 

[Mason T]

Suggestion: In the spirit of KISS, ditch the gyro and accel. The vast majority of hobbyists don't need them and won't use them, and it just jacks up the cost and complexity. If you just have it doing dual-deploy with the push-button setting selections and the voice setting/flight summary call-outs, that's all you need. After you've sold a few thousand of them and gotten your manufacturing and customer support processes tuned, you can create a "deluxe" version with the extra bells and whistles.

That really is a great suggestion. If I wasn't one of the few that really wanted the accel/gyro and that being a major driver of me building this it would be a no brainer to go down that path. It would certainly cut down on cost. It would certainly make for a great lower cost version. I will certainly have to make a version like that as well eventually. Thank you for the great idea.

 

[Tommie D]

The accelerometer and gyro is what caught my interest in this post.

 

[Mason T]

The accelerometer and gyro is what caught my interest in this post.

Great feedback, thank you.

 

[Spacedog49Krell]

That really is a great suggestion. If I wasn't one of the few that really wanted the accel/gyro and that being a major driver of me building this it would be a no brainer to go down that path. It would certainly cut down on cost. It would certainly make for a great lower cost version. I will certainly have to make a version like that as well eventually. Thank you for the great idea.

You and I are very much alike. I've been striving for faster data rates for 45 years. My first flight recorder, 1979, was one 8-bit channel @ 50Hz, no CPU. The first CPU system, 1998, was 12-bit accel/altimeter @ 37Hz. The main feature in the SW was accelerometer launch detection. Today, I fly 10 channels @ 500Hz with 3-6 channels at slower rates. I have bench tested a 1600Hz system, but I have no plans on flying it.

SolarYellow, the 500Hz flight computer can be configured to fit in a BT-20 at ~28 grams with a 100mAh LiPo.