In my experience, an LCD display that has been broken in one corner will be completely damaged in a relatively short time, on the order of months.
I concurr.
Screens cost more than a new T4; no sense to try to replace the screen.
Then I would buy another component tester with a 328p, and keep the one you have for spares.
...Screens cost more than a new T4; no sense to try to replace the screen.
Maybe a 2x16 LCD display would be easier and cheaper to obtain?
FWIW here's a proposal of a sample reversed_numbers-font I was talking about, that could be used on graphical c_testers -with enough free flash- for drawing tiled_pin_numbers next to the symbols. (note: as discussed on the previous page, this is a low-priority request).
The best solution would be to add the three inversed digits to each font, use ASCII values for control characters (e.g. 8-10) for the mapping and simply change the base character in Display_ProbeNumber(). That way the reversed digits would be also used for other pinouts, not just for the semiconductor symbols. And, of course, as an option.
Sounds good to me, but you know best, so I leave it to you (as an option for equipments with enough free flash, and when you have some spare time).
Shall you want me modify some other font let me know.
If you like you can do the three inversed digits for all other fonts (not for HD44780 and ST7036).
But as you told me for compatibility reasons we rather no reasign special characters for this, do you want me to insert some tile_numbers bitmaps without code (to be called by some special method you'll design), or you want an additional font bitmap file (with the three numbers each) and then you select one fontset or the other according to what is going to be draw?
If we go with the later, on the main code we select font_x_hf.h for drawing measured values, and when we're about to write the first line icon or the font symbols we select font_x_inv_hf.h, like the .h I attached on my last post. (note: atm I have only have developed how to render horizontal_flipped fonts).
Please confirm.
Please do them as in your example above. I'll add them to the normal font files afterwards.
About a year ago I also had an idea to make an interface similar to the Chinese version of LCR-T4 firmware. I made new graphics for the elements as well as inverted pin numbers for the 8x8 font and m-firmware. But later I was convinced that such interface looks good, but it takes a lot of memory space in the 328th ATMega and for the basic functionality is not enough space. I gave up on that option.
I see. Which font did you use, 8x8? If you still have the bitmaps, and if you are willing to share it, it will save us some time for this test, and it would be appreciated.
On the other hand, do you recall how much extra memory took your approach?
No, I've put that idea aside for the future and I'm not going to publish it just yet.
Random thought - Rather than take space making full numerals, could you use 1,2,3 pixels arranged to indicate pin number instead? Something like this using one pixel per dot, or could be expanded to 2x2 pixels per dot.
Well, we could reduce (the size of) symbols and/or fonts to save memory (that options already exist), but what I'm proposing goes in the opposite direction, i.e. improving readability for older eyes, perhaps using a bigger font for the symbols and/or title bar, and/or by inverting the colors of the 1,2,3 numbers actually being used for that (like some clones do).
We all know this would use more lines of code, but only the developers or somebody who already did it can assess how many Bytes it would be. If I understand this correctly, the bitmaps of 1,2,3 are going to consume a few Bytes itself (for instance 24B+ on 8px_tall numbers and 96B+ on 16px_tall numbers). What it's going to use more Bytes would it be the routine for selecting one font/character or the other.
Another alternative I'm working on is to create bold fonts 10x16 and 12x16. That wouldn't have impact code-wise.
Anyways, whatever we conclude for this low-priority enhancement, it's going to be optional, so each one can make the best-suiting combo of features
For those with an HD44780 based display (2x16) and missing characters at the beginning of the first line (m-firmware):
- caused by a slow LCD controller which needs more time for clearing the display
- edit function LCD_Clear() in HD44780.c and increase the delay to 3-5 ms until the output is fine
LCD_Cmd(CMD_CLEAR_DISPLAY); /* send clear command */
MilliSleep(2); /* LCD needs some time for processing */
What's the minimum wattage that can be used for the 470k and 680 ohm resistors? I want to use smd precision resistors and 1/8 watt seems to be the most economical.
The AVR uses 5V, so any resistor of 1/16W or more should work fine (for the I/O).
If you like to beta-test the ringtester (requires ringtester frontend) or logic probe (requires dedicated ADC pin with 4:1 voltage divider and rotary encoder) please drop me an email.
Edit: Latest additions are continuity check with buzzer and third profile for adjustment values.
Had some issues with my TC1 not shutting down properly. Last time this happened I reflashed the U4 IC and that worked. But this time that would not work so I replaced it with the transistor version (can be found here in this thread). Just though I'd show some pictures of it to inspire others with same problem. Basically pads on bottom that solder to the pcb and then through vias to top.
Had some issues with my TC1 not shutting down properly. Last time this happened I reflashed the U4 IC and that worked. But this time that would not work so I replaced it with the transistor version (can be found here in this thread). Just though I'd show some pictures of it to inspire others with same problem. Basically pads on bottom that solder to the pcb and then through vias to top.
What issue did you have with the TC1 not shutting down? I have one "original" TC1 that the battery will go flat on its own when left for a few days, but I have a clone of a clone TC1 that works perfectly and the battery is never discharged.
I quite fancy modifying the problematic TC1 I have.
My ~4 yr old TC-7 (blue panel, yellow button, wider screen) had the same issue with eating the battery. My solution was to just install a DPDT 8mm x 8mm push latch switch inside and exposed it on the bottom edge near the micro USB charge port. Just do a physical on before starting use then use the normal button until you're done then hard shut if off.
Attached is a pic of all 7 of my testers.
...Comments are welcome ...
Hello Yuriy,
... your Color-Testversion looks very nice, bat at my TTester GM328A I have to change some items in the Makefile befor I can test it.
For that I need the changed source-files from you so I can compile it again.
Many thanks and best regards
Horst
Do you have english version? i like this firmware and need to try
Below you will see a summary of the main issue with the TC1 (and similars), and it's workarounds:
Whoever manufactures the TC-1 has made some changes (control MCU U4) which are incompatible with k and m-firmware. For running one of the OSHW firmwares the changes need to be undone. One way is to reprogram U4 (https://github.com/atar-axis/tc1-u4), and the other way is to replace U4 by a small circuit (https://github.com/madires/Transistortester-Warehouse/blob/master/Hardware/TC1-Mod.kicad.tgz). The TC-1 also has a few issues with poorly chosen components, especially around the voltage booster for the Zener test. I'd recommend to replace MLCCs C11 and C1 with a 10 or 22µF low-ESR electrolytic rated for 100V.
For more details you can search the related posts on this thread starting around june 2018.
and I don't see the VOLTAGE menu, can you add
thank
After optimizing the source code, it became possible to add this to the menu.
Changed archive
Who repeats the situation described below?
Everything works fine for me.
and I don't see the VOLTAGE menu, can you add
thank
After optimizing the source code, it became possible to add this to the menu.
Changed archive
Hi Yuriy_K
This firmware works, but I don't understand why I have to press push 2 to 3 times to start, if only pressing the button 1 times clone doesn't turn on.
thank