วีดีโอ

When I discovered that one about a year ago I think I was quite surprised that this is not being mentioned more often. The receive to idle makes it much easier to work with serial receive. Happy you enjoyed the video! Check out my other videos in the serious - you might find another surprise or two (bit banding is a life safer too).

Well, I know some things ;) But you would need to be a bit more specific what you have in mind.

USB will come later. I am struggling a bit coming up with ideas which will be possible to squeeze into < 30 min videos. The bootloader - well - first you should check the earlier video on jumping to the built-in bootloader. With enough storage on-board it _would_ be possible to do this more elegantly so that video would come after I have been showing SPI flash ;) Your comments and suggestions are greatly appreciated!

That is a pretty good idea. I don't have a STM32World board designed yet but I might throw one together. Alternative I do have a couple of STM32F405 based boards with a SD card slot on them. The sd card interface is quite simple. The hard part is to use the FAT library actually.

Ok, point taken. In this case a few typedefs

I think you will find an answer to your questions in Tutorial video #17 th-cam.com/video/j-GaEZKrkbQ/w-d-xo.html In short, you can use a st-link v2 if it is an original. The Chinese clones do not include a SWO pin by default.

Sorry, I missed the "live expression" part of your question. And no, no need for Trace - Serial Wire is enough and that will work with any st-link - even the Chinese clones.

Right - but I lifted the formulas straight from the datasheet. Easier to toubleshoot later.

That's a tricky one. Ultimately it will probably boil down to how much current will run into all the pins on the stm32 when going low and that could become a bad thing. You would have to read the datasheet for your specific mcu and schematics of the "shield". 5V tolerant pins is a nice thing but they probably shouldn't be relied on without careful consideration. I have used it a few times - for example to drive ws2812 leds (which require 5V data).

@@stm32world Thank you, to be on safe side use level shifters 3v3 to 5v for each pin, can you recommend which shifters for analog pins and digital pins.Thank you very much.

Sorry, no - the recommendation would be the same - look at the schematics and the datasheets for the individual stuff.

Hmmm, yeah - you are probably right about that, but nitpicking back - IF the initialisation is run at each reset then that would technically also be the same. I am going to test this, which _can_ be done the moment "main" is called before the HAL/NVIC etc. get executed.

I hope it is useful. I am doing a deeper dive into the HAL library in the next few videos.

There was one declaration different - which was so far fetched I didn't even bother to check.

And I know it's embarrassing - hence the video title ;)

The clone ST-Link devices does not have serial or swo (swo _can_ be added with a bit of tweaking). So, if you want to use that you have to use a separate usb-serial adaptor. You don't need both swo and serial. Both can be used to print debug statements and the swo can do tons of other stuff - like monitoring variables "live". I made my own st-link which can do it all ( stm32world.com/wiki/DIY_STM32_Programmer_(ST-Link/V2-1) )

If you read the evaluation part of the wiki article (here: stm32world.com/wiki/STM32_Jump_to_System_Memory_Bootloader#Evaluation) yes - it could potentially go wrong and I guess disabling interrupts would improve that.

And yeah - took me a while to notice the missing assignment - long enough that I had to pause the recording while I figured it out ;) Wait until you see the solution to the Crazy Mystery video I posted some days back.

Thanks, glad you liked it.

I expected to be shot down for messing with the bloody stack ;) But thanx.

Let's put it this way - normally the stack is better left to the compiler ;(

That is very cool! I am at least 12-14 hours ahead of you timezone wise, so I had to leave the live stream before you were done - sorry about that.

I do have a few comments on your live stream but I'll post that on your channel. Any comments on my videos is greatly appreciated. I am still in an early learning phase when it comes to videos.

Thanx. I actually didn't think this particular video would be well received - might make more like this.

Your comment is much appreciated ;) From the limited experience I got so far, 10 minutes or less is hard though. There's simply not enough time to explain. But I will try to make them as short as possible.

Check the first video in the series, which covers installation and setup as well as hardware needed. It's ST's STM32CubeIDE - which is Eclipse based. The entire series is described here: stm32world.com/wiki/STM32_Getting_Started_Tutorial_Video_Series_(by_STM32World)

You are most welcome. Will dig deeper in the following 2-3 videos although I might get bored and squeeze something else in.

That is interesting. Does STM32CubeIDE and HAL work with those. I have done some work with non-st mcu's - mostly risc-v ones but this particular series of videos is focused on STM32CubeIDE and HAL.

I will continue to make these videos when I have the time. Please continue commenting and ask questions down below.

If you watch the video to the end, I enable the input capture interrupt and the callback is executed. Would be quite trivial to establish the speed either by looking at the uwTick OR resetting a faster running timer in that callback. It might even be possible to cascade two timers.

If you run say TIM2 at 1 MHz (prescaler of 83 in this case) you can detect the space between the interrupts on TIM3. Quite simple.

Yeah that makes sense.

Glad you like them. Good to get some response ;)

Right - I missed that. Not going to redo the video because of that detail ;) Let it be a learning experience for the viewers.

I've emailed you on stm32world domain.. hope that reaches you.

The Facebook group is probably the best place to ask. Quite active.

@@stm32world got it. will do.

Yeah - I am quite sure this is the most likely explanation - looking at the assembler code.

It is probably something along that line - except, local variables are allocated on the stack which is towards the top of the memory. This is definitely something I need to figure out - it annoys me I don't understand it and it is definitely worth knowing. I appreciate you taking the time to respond.

@@stm32world maybe there is something going on with the compiler trying to optimise the stuff but ruins a thing or two... wonder if you try compile a hex with different settings and then see if it is the same?

Your (NEW TERMS) have me somewhat confused to be honest but I am not overly familiar with C#.

Glad you enjoyed it

Well - just to be clear - I do love the nitpicks - that is the ONLY way I can improve this. And yes - I can easily imagine situations where interrupt - or even polling would work fine but I _think_ the vast majority of cases uarts are used it is to receive strings of sorts. So - audio quality if ok-ish now?

@@stm32world Audio is a magnitude better than previously.

Compression enabled and level boosted

Curious , what software do you use for editing the video/audio?

@@bubaks2 I mentioned in the first video that I _really_ want to do these videos unscripted. I record them with OBS studio and two external cams. I use a noise cancelling Logitech headset for the audio (the noise cancellation is essential as I am sitting about 1 1/2 m from a fan). The audio compression was done inside OBS studio while recording. IF I edit anything I do that in an application called kdenlive, but many of the videos are one shot unedited.

Running a bldc motor is probably a bit too advanced for half an hour tutorial videos. ST themselves have some pretty detailed application notes covering that topic. I am not sure I understand what you mean by sin look up table? This video does not use a look up table but rather a quick algorithm to calculate individual values.

Well, the DAC of the STM32F4xx MCU I was using in this example can actually do triangle wave without any calculation at all - you just need to feed the frequencies. Let me know if you can make that work - if not I could do a short video on it.

​@@stm32world actually that sounds awesome.. i'll try that today.. i'm more worried about triggering the 2nd dac after the completion of the triangule wave.. i'm hoping that i can stop the wave generation somehow with the complete callback and then switching to generation of it again once the 2nd dac is incremented? i'm not sure, i'll give it a go... also i'll record a video of what i'm trying to do. i'm sure you'll love it..

​@@stm32world I tried but am unable to work with dma interrupts specifically i'm not sure how to look for interrupt when dac is finished outputting a triangle wave. Also i'm trying to start the triangle wave through built in settings but i'ts not working.. i would appreciate it if you could create a video on how how to enable and look out for transfercomplete interrupts and do some work in the callback.

@@capistor1 I am not sure I understand what it is you want to achieve. If you look at my DAC example I calculate a sine wave. Nothing would prevent you from doing a triangle wave instead - in fact the math is simpler. Then you can trigger whatever you want in that callback once the wave is completed. I can have a look at the triangle generation later, there's quite a lot about it in the reference manual. But right now I am preparing some I2C videos and I need to get them out of the way first ;)

@@capistor1 you might want to have a look at AN3126 here: www.st.com/resource/en/application_note/an3126-audio-and-waveform-generation-using-the-dac-in-stm32-products-stmicroelectronics.pdf

Well, they (the videos) are meant to be useful for newcomers but will get more advanced as we move along ;) I appreciate the comment.

I did that in the last video #14. In this case however, it is a bit more complicated. I guess I _could_ calculate the value from the prescaler and counter values of the timer. But put that down as pure laziness ;)

Well, IF you ask yourself a question like that, I reckon I have done a pretty good job ;) THAT's the primary purpose of these videos. Now, go ahead and try that and please let me and everybody else know ;) It's a good question.

I just quickly tried with the dac and it made no difference. Need to dig into this a bit further.

That would almost be another video. Try it yourself - the Debug optimization is the fastests - always. -O1 -O2 and even the one that should be the fastests are all slower. I have NOT figured out why and I guess the only way is to analyze the resulting assembler code.

Hi

Yeah

Yeah