Better and faster Arduino analog sensor sampling - improve Stirling engine pressure wave measurement

[@2700kelvinandhobbes]

I know this is rather unrelated, but wow, what a pleasant voice

[@myengines2443]

Thank you very much!

[@e7yu]

Your project has been very enlightening and interesting to watch. Looking forward to seeing this project continue. 👍😎

[@myengines2443]

Thanks!

[@Lechoslowianin]

You approached the issue very professionally

[@myengines2443]

Thanks!

[@dancollins1012]

Looks like a great project - enjoy. A few thoughts for your consideration: a) I was surprised that the bandwidth of the sensor relative to the system frequency wasn't discussed. The sensor datasheet lists its response time (which can be converted to bandwidth) and that needs to be compared to the frequency of the system, allowing for Nyqist Sampling Theorem. There's several ways to measure system speed: a timing gun, bouncing a laser off a moving part and picking up the reflection with a phototransistor, or measure frequency of the sound via a smartphone app. You may need to buy a faster pressure sensor. Following that, the low pass RC filter should be specifically designed so as not to materially impact the desired signal frequency. I agree with other comments that a long air tube may be problematic. b) I prefer to use absolute pressure sensors. I believe the two port sensor you have may be a differential air pressure sensor. In which case, you're actually measuring pressure relative to ambient air pressure. Even if you block one port with a membrane, that membrane would transmit ambient air pressure. Wikipedia's 'Pressure System' page notes that ambient air pressure can vary up to approx +-10% with weather systems. While it won't impact your high frequency measurements, it could impact repeatibility across the day, and between days. c) I urge you to show actual values in your videos (rather than use words like 'more' or 'less'), and equations used, as this will help others contribute to, and learn from, your work. Looking forward to the next video!

[@myengines2443]

Thank you for your qualified and informed comment. I can't show everything I do in the videos, and it's also very important to decide what's really important and what can be left out to save time and money. I will look at your suggestions very carefully and then implement the most important ones for me! Your help is exactly why I make these videos! Thank very much you and best regards!

[@electricAB]

Thanks for the update and for sharing you process.

[@myengines2443]

You are welcome and thanks for your nice comment!

[@mattw7949]

Ooo! After watching your other vids, I was hoping you'd do one on exactly this. I haven't even watched it yet, but I'm already excited.

[@myengines2443]

Thanks!

[@markgeurts258]

Very nice, good progress already!

[@myengines2443]

Thank you very much!

[@wayne851]

A lot to take in as layman but interesting nontheless. Sending good vibes in your direction!

[@myengines2443]

Thank you very much!

[@pixmooo1]

Good job.... Greetings from iraq

[@myengines2443]

Thanks!

[@jgbstirling6938]

Congratulations you engines are similar to the mines but better builder

[@myengines2443]

Nice to hear of you, i tried to contact you but it did not work, sorry for that! You are the inventor of this type of engine as Attila Blade wrote recently to me and I must apologize to you for not making it clear enough that the engine is built according to your design. I'll make up for that now and wrote it in every older description and will show it clearly in every new video! I hope it is ok to make this videos of my replication and to try to develop it further! If I should do something or to not do please let my know and we surely will find a solution! I thought it was a older, common design but Attila Blade just corrected me and I can not apologise enough for that! I don't want to make any money with your Ideas I just want to generate electricity for myself and spread the idea of the Sirling engine. I would be very happy if you can write your opinion about this as I really feel sad about my fault! Now I just want to say how much I like your design and that you are a great developer. Thank you very much! Ralf

[@janmachacek4866]

Great project. Arduino's ADC is incredibly slow, it can take even miliseconds. You will have to avoid using analogread() function and instead make your own by following atmega datasheet. I still think it might be too slow. I would recommend using different mcu, such as stm or esp.

[@myengines2443]

You are right, using registers to avoid analogRead() is an additional way to speed things up I forgot that. At the moment I am very happy with the speed and I can read accurate and much faster than I need to. Many thanks for your help!

[@yasirrakhurrafat1142]

Hey bro, can you help me to the right direction? I need a method to collect and store 100ms of (megasamples) of digital (don't require analog) data over a relatively long runtime. I've heard that dedicated logic analysers can take simple inputs of up to a few hundred mhz or megasamples of data, and log and store a short period of that high frequency data. I want to be able to log and sotre possibly.. a few megahertz/megasamples of data over a good few minutes.. optimally hours. All while being able to respond to those inputs and track their fluctuations and respond again accordingly. All.. at those high frequencies. But I haven't found any solutions, and even seen high end scopes can only see a few megasamples and may only store the log of a very short range of time. Can you provide me a solution.. that'd allow me to do the long period, high frequency logging and response (like a microcontroller). I'd not mind if I can find a convoluted solution to daisy chain multiple microcontrollers and use some method to synchronise them to get a high frequency input output solution.. just as long as it'd be accurate. Also, any tricks to get more sample rate out of the existing popular microcontrollers like you were suggesting in this comment? Very long comment, sorry for all this. I know I'm asking a lot. Would love to get a solution though. Thank you very much for even reading all this. I'm not very efficient at communicating complex things.

[@felipevellasco6526]

@@yasirrakhurrafat1142That is a really complicated problem. First things first, an arduino won’t help you with that. You need a very high end controller, such as an esp32 or maybe even some kind of microprocessor board such as the raspberry pi. Second, you’d need an ADC with decent precision and high sampling rate. L Third and most important, you need an extremely fast storing method, which would be capable of storing data continuously at such a high speed. I suggest you look into professional cameras. They usually deal with high speed storage, since photos really need to be stored stupidly fast. A Jpeg file might be small, but the RAW file that outputs from the sensor and stuff is way bigger, ranging from 2-6 times the size. Anyways, I think one of these two boards might be the way to go. Especially bare metal raspberry pi, boasting gigabytes of ram memory. It really depends on what you want to do though. Your explanation was a little lacking on what you want to do with the data, so I can’t really pinpoint the best development board to use. Not that it’s an easy question though. Best of luck on your adventure.

[@myengines2443]

If I understand you right, you want to sample with several MHz ? That's beyond my experience, maybe you look out for the high end external ADCs and special information about that. A very fast data storage system may also be needed... Greetings and good luck!

[@yasirrakhurrafat1142]

@@myengines2443 I've thought about using an ssd as the storage and downloading a library or sigh doing the tedious task of learning to manipulate each address on the ssd manually. I'd also need to use the inbuilt ram as efficiently as possible, and also perhaps try to add more.. somehow. The thing is.. I'm trying to do the same thing as you are, but I want the data at increments of less than a 1/10th of a degree of rotation (one tenth out of 1/360. So more like 1/3600 or optimally 1/36000) at a minimum of 5000 rpm, but at over 12000rpm if possible. To see the fluctuations in speed and torque between the power stroke and the compression stroke. Then build either a gear with dynamic profile that'd help me extract all the energy.. per each compression and power stroke.. rather than waste all the peaks of torque and speeds during the strokes. Or some other mosfet driven method or something. I'd say though honestly, this is just about an unreasonably ambitious project,.. and pursuing it would probably leave me with major frustration and resentment for even attempting such a thing, with my intelligence level. By the way brother, I've made another comment in your video, check that out. Love your videos! Love that I just learnt a little about microcontroller recently and was thinking that I'd need the same setup to optimise engines, the same way you're probably going to try to. Sorry for such a long comment, you've made me geek the hel out, with doing this at such a coincidental timing.

[@breaker6598]

The tube between the engine and the sensor may seriously effect results. The mass of air in a long tube acts like an electrical inductor. It will attenuate the pressure signal at high frequencies. How high? You can check the natural frequency of that tube with a little math. 343 meters (speed of sound) divided by tube length (~.3!?). Thats around 1000 and on the same order as your signal (500 hz). Longer wires, shorter tube. Ok - at the end the tube WAS a lot shorter. It needed to be.

[@myengines2443]

You are right the tube has to be as short as possible. It is still a very provisional setup and in the next step I will attach the pressure sensor direct at the T-branch. I am just testing everything before I can make some serious measurements. Thanks for your tips and suggestions!!

[@reinhardtburger7108]

Form my understanding of these systems (travelling wave thermoacoustic engines). You get max efficiency with 3 stage system and max power with 4 stage system? Also the second cold heat exchanger after the buffer tube in some cases seems to be replaced with a bypass tube that connects to the cold side. I'm curious if the bypass method works better than the second heat exchanger. Also the resonator tube sometimes is liquid.

[@arnavpawar256]

What maximum sampling rate did you achieve? I think a faster board like ESP32 can sample ar a better speed 😃

[@myengines2443]

I did not tested the maximum I can get with accurate results but it's already faster than I need it. I have also a few esp32s and 8266s for other tasks but isn't their ADC not as good as that of the arduino?

[@johgude5045]

nice but why dont you use your other, much better measurement equipment? I see no use in choosing a cheap 3€ arduino over a 1000€ oscilloscope as long as you do not want to control anything with those measurements.

[@myengines2443]

You are right, the oscilloscope is very good to show exact voltages and the waveform but I want to measure many other parameters like power, temperature, speed, measure time, timestamping, save the results in a file for further processing, and show them in clear readable form on a display. Later I want to control and monitor the engine also over the net and many other points as you can see in my older videos of the rhombic Stirling test bench. Also the oscilloscope is very bulky and impractical. This is just the beginning of the building of my test bench!

[@hosseinpeste9213]

Hello 🤗, would you like to work together to compile an article?

[@myengines2443]

Hi, what kind of article you mean?

[@hosseinpeste9213]

@myengines2443 research article that can publish in journals. If yes, tell your affiliation please.

[@Psrj-ad]

the lack of info about thermoacoustic engines is really irritating to me. Seeing someone trying to make an actual Thermoacoustic engine for power output, and not just some dinky model toy is really pleasing. I hope you'll be able to add some more information about your findings to the currently lacking materials about these engines.

[@myengines2443]

Thanks!

[@tibsim]

Big likes! What do you think This how much more power do compared to a standing wave engine? For example:kzbin.info/www/bejne/fWSnc5Kwq9R-odk

[@myengines2443]

I don't know as I just start to make serious performance tests when the test bench is ready! Thanks for your help and the presentation of your great engine!! Many greetings!

[@tibsim]

@@myengines2443 I ask because I have not found this TASHE designed to be better than the standing wave if it is as a generator only as a refrigerator. I'm curious what others experience...

[@myengines2443]

I haven't heard of any high power standing wave engines but there are many high power (up to 5 KW) TASLiCE or TASHE engines. That's the reason I try this concept. It would be very interesting to scale a standing wave engine for high power!

[@tibsim]

@@myengines2443 I have found that thermoacoustic engines are not running similar well if its are generators or refrigerators. The standing wave is a weak cooling machine, and stronger generator, and the other way around, my classic Tashe models are much weaker generator and stronger refrigerator. Thus I am interesting and strange that it is high-performance old standing wave engies cannot be found as generator, but gas liquefier yes. Maybe, I have only this experience because tashe requires more expertise, that it to be better but I do not think it. Traveling wave came into vogue because the thermoacoustics, as well as Stirling technology, are mostly used as coolers rather than generators. Nowadays Maybe scientists wrong start from the assumption that standing wave is bad as a cooler, and thus for generator also, I think they undervalue standing wave machines and they do not deal with them because the travelling wave is the vogue.

[@ad2040]

Forget efficiency, its a rabbithole. Make a big, inefficient system that lasts and produces enough power.