Finally someone explains an upgrade to 1.46 I can digest!
Enjoy
* Post edit *
Famous YouTuber DanyK.cz added his overview, so I might as well add it here (hard believing he's never owned one till now..)
It's actually rather humorous watching him try 3 components simultaneously to see the results..
I suspect he was "playing dumb" but managing to demonstrate theory and measurement limits, it worked out well.
Famous YouTuber DanyK.cz added his overview, so I might as well add it here (hard believing he's never owned one till now..)
Everything that is shown in this video, you can easy find in the documentation of Karl-Heinz in detail with diagrams,tables and test results, a long time ago!
Hi.
More fixes for fonts. Sides of resistors were swapped in file font_8x12t_iso8859-2_vfp, so in my converted hf version too.
I've been using Karl-Heinz firmware on my tester.
But today I felt I should try Markus firmware on it, however I can't get the encoder to behave properly.
I have tried setting "ENCODER_PULSES" to 1,2 & 4 but it moves erratically.
If I am very careful moving one indentation while in PWM mode it will count 2 steps no matter what I set "ENCODER_PULSES" to.
That is to say one indentation seems to correspond to two pulses.
If I however turn it at a "normal slow" speed it will go very fast, probably due to the double signals?
My encoder has 24 physical indentations or steps which I've set in "ENCODER_STEPS"
What can I do to troubleshoot this further, any more info I can provide?
Out of curiosity I bought a new Bourns PEC11R-4215F-S0024 encoder, the securing legs where a little bit too wide but I snipped them down and then it fits well.
I have attached a couple screenshots, first the original Chinese encoder that was included with the kit for my unit.
Then I tried adding 100pF 0603 SMD NP0 capacitors between the legs of the A/B signals and ground to it as this is recommended in the datasheet of the Bourns encoder.
Then I changed to the Bourns encoder but I did not add the capacitors to it.
You can see that the Chinese one is just crazy: some of the indentation steps are not 90 degrees apart but more like 0 degrees or even maybe negative!
So no wonder it did not work well, I had kept the firmware I compiled back when I made the above post and it now behaves perfectly with this new encoder!
Madires: would you like that I send you this encoder? Because I have used it with Karl-Heinz firmware all this time with very little issues...
If not it will go in the trash because that is where my oscilloscope says it belongs!
The screenshots of the dodgy rotary encoder are interesting. Back then I bought a bunch of the cheapest encoders I could find on aliexpress to optimize the algorithm for reading the rotary encoder. They cause some glitches, but yours is truly worse.
There's a huge difference beween good struff from ALPS or Bourns and Chinese cheapies. I agree, your encoder should go into the e-junk container. It wouldn't help to improve the encoder algorithm, as it detects glitches and drops them, i.e. nothing happens (intended design).
This is an interesting capture of the encoder; my tester is quite erratic and I will see if an old spare that I bought for my oscilloscope will fit and work.
Moving to 1.45m on an AY-AT clone I have started to experience phantom selections when scrolling through items in the menu. This doesn't happen with k firmware or the original 1.12k Chinese firmware that came with the tester. I'm using the internal 8Mhz oscillator if that matters at all.
Chinese rotary encoders are of poor quality and may cause firmware instability. This was noticed by many users of AY-AT and other clones. If changing the m-firmware encoder settings in config.h
#define ENCODER_PULSES
#define ENCODER_STEPS
does not lead to the elimination of unpleasant effects, then you should simply replace the rotary encoder with a similar one of better quality.
I mentioned the poor quality of Chinese encoders several times before!
Hello everyone.
I am expanding the GM328 with the different Madires mounts.
E built the optocoupler tester which works perfectly.
Today and tried to mount the "sound emitter", as the schematic and as you can see in the photo, using a buzzer, a 2N2222 and a 1K5ohm resistor.
The base of the transistor connected through the resistor that is connected to PD5 (pin 11 of the Atmega), because when you connect the power supply of the GM328 and put it to work, the buzzer is always sounding.
I've checked everything, I've even reloaded the firmware again, and everything is still the same. What could be the problem?
In the config_328.h file, you must additionally disable the PD5 port for the ST7735 display
//#define LCD_CS PD5 /* port pin used for /CSX (optional) */
Indman, I followed your instructions and all the same, constantly ringing.
I've cut the track and jumpered, and disabled //#define LCD_CS PD5, I'm using version 1.46.
Any more suggestion.
Thank you very much, greetings
You can program my Mod1 1.45m from the link I gave earlier and test the Buzzer.
indman with your firm Mod1, perfect.
I'm going to compare my config_328.h with yours to see if I find the problem.
I liked some details of your firm, such as the battery issue, the display of messages, etc.
If you don't mind, you can send me a complete signature to copy your ideas.
Thank you very much for everything and greetings.
PD: Solved the continuity problem, it already works, my mistake
Indman, I told you if you could send me the files to modify some things that I like about your firm, among them:
1 - That the messages appear in the center of the screen and not at the top.
2 - The battery indicator is very good and I would like to change it because it gives it a very professional touch.
3 - Between the changes I would like to change the color of the menu (white better green, the same as the font (thicker), both changes are better seen with a lot of lighting
If you don't mind, can you tell me what would be the things to modify or files I should look at to get it.
Thank you and regards
carrascoso,sorry,my firmware modification involves significant changes to the madires code in a very large number of files.
You probably missed or misunderstood what I wrote in the ReadmeModEn file:
I do not publish or distribute the mod source files!
You can change some colors in colors.h (scroll down to the end).
You can change some colors in colors.h (scroll down to the end).
Thanks, Madires.
I already make what changes I can in colors.h, but I like the font for my view that Indman uses.
I also liked the different messages, warnings, etc. in the center of the screen, this can be done in your firm or at least I haven't found it.
And the battery icon and result is also very good, although this surely takes up a lot of memory.
Cheers
I like the font for my view that Indman uses.
You can create/modify fonts (I've shared here some improvements to existing fonts/symbols, with some cues on how I did it). For the m-firmware, there are several fonts to choose from. Think on those as bitmaps; if you understand the encoding, you can modify or create one "by hand". Or you can use one of the few AVR font tools available on the net.
The character bitmaps aren't specific to AVRs, any font tool will do. Be aware that based on the display controller a particular bit order is needed.
I also liked the different messages, warnings, etc. in the center of the screen, this can be done in your firm or at least I haven't found it.
And the battery icon and result is also very good, although this surely takes up a lot of memory.
Maybe I'll add such features as options in the future. The next firmware version will come with several new UI options.
Maybe I'll add such features as options in the future. The next firmware version will come with several new UI options.
On the previous page, I gave an example of characters from my font. Here is an example of these images on the display...
I'll add from another thread. To obtain high accuracy when measuring small inductances on k firmware. I'm applying a little tweak, highlighted in red on the schematic fragment. On the 1st picture, the measurement of this inductance with the values indicated on the diagram. The range of possible deviation of the inductance L1 is shown in pictures 2.3.
The readings are different, different measurement methods, different frequencies. MS5308 shows within its error at this frequency. But I determined a long time ago that his readings were underestimated. After purchasing the NJ100S, I checked its readings on the standards available to me, its readings on the fuel are more accurate. Below are the readings without L1 and with L1, and error table MS5308.
P.S. The quality factor readings differ from those previously given, it took time to normalize the parameters of the parts after soldering.
To obtain high accuracy when measuring small inductances on k firmware. I'm applying a little tweak, highlighted in red on the schematic fragment. On the 1st picture, the measurement of this inductance with the values indicated on the diagram. The range of possible deviation of the inductance L1 is shown in pictures 2.3.
I don't see any high accuracy in measuring the inductance of R36,if you look at the readings of your 3 instruments? They are different and I can't see comparing readings without tweaking upgrade the circuitry.
Hello everyone, I tried to complie the newest version of m-firmware, I got everything under (100%). However,when I uploaded the firmware, there is nothing displayed in the screen.
I’m using ay at transitor tester with some modification, the Vref of mine is 2.234V, the resistor I’m using 680R 470K with 0.1%
I also write fuses as this
fuses_lo = 0xf7
fuses_hi = 0xd9
fuses_ext = 0xfc
lock_byte = 0xff
please guide me out of this problem. Many thanks!
Likely either the wrong display is selected in the config_328h or the pinout to drive it is not standard. I think you need LCD_ST7735 to be active. But then, my hardware version used totally different pins to drive the LCD which I had to discover and redefine. In the hope yours may be similar, you can try my config_328h which you can download from here:
https://github.com/TheHWcave/TransistorTester My changes are marked with "theHWcave" in the file