$20 LCR ESR Transistor checker project

[madires]

Most likely you can keep current fuse settings (otherwise: make fuses). Maybe you have to clear the lock bits as some clone manufacturers like to set them.

[LarryS]

This is my first message; so if I do something wrong, please be....

My V 2.68 came with the proper screws and I had no trouble assembling it.

[jakeisprobably]

1.34m
config_644.h
ST7920 4 bit parallel
change line:
#define LCD_EN1  PB3
to
#define LCD_EN   PB3



It looks like my 7920 has seen better days. I triple checked my wiring, but nope. I managed to make and upload the code though. AVRDude shot out an error for an unrecognized function or something like that. Just removing the "1" fixed it. Without the screen working I just tested an LED and it runs the whole test routine (on top of the fact the MCU power control still works). The same adapter board also has a Nokia header and that one works fine too.

[madires]

The new m-firmware version is progressing, including several bug fixes and improvements. I'm considering a change of the capacitance zero offset (for PCB, probe leads, etc.). Currently the firmware uses an average value for all probe pairs, and the idea is to go for probe pair specific values. The reason for this idea is that I've seen some small differences of a few pF for some hardware. What's your opinion on this?

The second topic is about displays. Some support an SPI 9 bit mode, i.e. the D/C (data/command) signal is part of the SPI data (1st bit) instead of a dedicated line. Since the ATmega's hardware SPI does only bytes, the 9 bit mode will be available just for bit-bang SPI. I already have added the code and extended the SSD1306 driver. Some users requested a driver for the Nokia 1202 display (STE2007/HX1230) which seems to require the 9 bit mode, and also a driver for the ST7036. If you like to support the project by sponsoring display samples please contact me.

[jakeisprobably]

..."the idea is to go for probe pair specific values. The reason for this idea is that I've seen some small differences of a few pF for some hardware. What's your opinion on this?

  I think it's a great idea.
   I haven't tried to calibrate my new design yet with either fw, but I'm looking forward to exploring this down to the code and functions level as a learning exercise. I also would like to try to test and understand the resistance of the internal port and the comparator offset figures. I'm sure I need to go read an app note or two about that before going into detail.

  I picked up an old Keithley 197 5.5 digit DMM. I still need to get a proper LCR meter to really contribute useful info though.

  The more displays the better. I'm trying to test out one that was sold as an ILI9341 right now but I think it's a fake as I've read more. I ordered a different one last night along with a couple other options.

[jakeisprobably]

1.34m
config_644.h
The ST7920 SPI configuration is not pulling the Reset pin high. I have it set to the default PB2 and it just stays low. I checked ST7920.c and the function LCD_BusSetup has the lines:

#ifdef LCD_RESET
  Bits |= (1<< LCD_RESET);
#endif

  The same routine is applied to LCD_CS. CS is assigned to pin PB4 and it is being pulled high like it should be. Neither pin is actually required. Both of them can connect to a 10k pullup resistor to Vcc. The only required MCU connections are Data and Clock. Power, Ground, Backlight, Chip Select, Reset, and the Parallel or Serial Bit (PSB) can get connected directly.
  I would recommend adding this in the comments. 'The user can add a 10k pullup to CS, RST and comment out these connections to the MCU in the configuration.' You could also make a note that the PSB pin must be tied low for serial mode and high for parallel. IMO it would be a helpful reminder for troubleshooting.

 BTW thanks for getting me motivated to try out Sigrok software and a little 24MHz 8 channel clone logic analyzer
  I took apart the broken ST7920 and managed to get it kinda working...enough to troubleshoot this stuff, and figure out what was going wrong. I even quadruple checked my wiring by bypassing my little adapter board just to be sure. I'm not sure if this is also a problem with the parallel version but it wasn't working for me either yesterday. I figured it was just my crappy LCD but I got the LCD working with an Arduino test setup and SPI so I know the screen mostly works, there are just a few dead lines in the matrix. (Short story: don't take these things apart...me=big dummy)
-Jake

[madires]

That's just for setting the pin to output mode. A few lines below you'll find:

Code: [Select]

  #ifdef LCD_RESET
    /* disable reset */
    LCD_PORT |= (1 << LCD_RESET);       /* set /RESET high */
  #endif
... which sets the pin high (no reset).

The reset is performed in the function for initializing the display:

Code: [Select]

  #ifdef LCD_RESET
  /* reset display */
  LCD_PORT &= ~(1 << LCD_RESET);        /* set /RES low */
  wait10us();                           /* wait 10µs */
  LCD_PORT |= (1 << LCD_RESET);         /* set /RES high */
  MilliSleep(1);                        /* wait 1ms */
  #endif
This creates a low pulse of 10µs to reset the display. It's done only once after powering on.

The comments for LCD_RESET and LCD_CS include an "optional" as hint already. However, I could add a note about using pull-up/down resistors in the README as help for beginners.

[jakeisprobably]

The comments for LCD_RESET and LCD_CS include an "optional" as hint already. However, I could add a note about using pull-up/down resistors in the README as help for beginners.

Thanks. I wasn't sure what the line I quoted was for exactly. Now it makes more sense.

...and after looking at the logic analyzer I see I was making a mistake reading it... (me=big dummy)...sorry...

  On a whole different note...

AY-AT ST7735 LCD Fix
   If anyone has the same junky 1.8" TFT ST7735 display that comes with the AY-AT clones, let's fix a problem here...
Do you have problems with the screen? Text appears in different locations, in multiple colors, with random junk on the screen, and or sometimes it just doesn't work at all or white screens? Then here is your fix.

If you look at the back of your display module you will find it has a CD4050 and a small voltage regulator. What you will not see is any power supply filter capacitors. This is why your display sucks. It (probably) isn't the firmware causing this. It's just a terrible LCD module design. All you need to do is scrape off the solder mask on the regulator's input trace along with the ground plane directly adjacent. I added a 4.7uF and a 100nF ceramic capacitor just beside the input for the regulator. I also added a 100nF ceramic capacitor to pin 1 of the CD4050 (and ground plane beside it). All of these are 0805. I also added a 0603 100nF capacitor directly to the 2 power pins of the LCD flat flex. That fixed the problem as shown below.

*This is the screen from my AY-AT clone but tested with Markus's firmware and on my board design. The fix is still applicable. I haven't used my AY-AT much because of these issues with the screen. This should greatly improve the use of this screen anywhere.

[jakeisprobably]

Hack a T3 Clone Display
   So I've been playing around with a bunch of different stuff... Basically, I had a little T3 tester for awhile (along with the two kit clones). Of the 3 clones the T3 has the best SMD component test pad configuration by far. I have found myself using it the most because of that test pad. However the LCD got damaged. I'm not even sure how. I just noticed a tiny chip in the glass near the flat flex and the display quit working. Okay, no problem, a new one is cheap....A month and a half later I get the replacement... The display doesn't work  I tested the programming by updating it. Checked connections and while trying to take it apart to resolder the flat flex, I discovered it was already pre-broken for me.
  ...anyways... I took the display off of my Little t tester clone. That's the smaller version of the ST7565 LCD. I mapped the connections and tried the T3 firmware just to see what would happen. Nothing of course. After switching the firmware over the KH`s "mega328_st7565_kit" it worked great. I didn't make changes to the tester option settings. I always have to recompile everything to use the USBasp, but that's all I changed. I chose this display because it is cheaply available if I want to just get a spare for each tester.

  You'll see a few mods here both past and present.
1.) I try to swap out the ZIF socket for headers on all clones, because I've learned the ZIF sockets get worn out quickly when all the slots are not used or you test a lot of stuff with thicker pins.
2.) I have added a 2p3t switch that I have configured for quick manual calibration. This can be in an open configuration, can short out all 3 test pins, or connect a film cap to 1 and 3.
3.) I added a stereo audio jack and made my own test clip leads that plug into it,
4.) I added 4x 10mm standoffs to the OEM locations, but also added a 5th standoff near the SMD test pad. This is probably the most significant mod worth doing on this tester. That makes the SMD pad much nicer to use.

  The last bit of rambling I'll do here is just to add the reminder, the LCD backlight for the T3 tester is the required LED for the power management control circuit. If you take things apart for doing mods be sure you add an LED in place of the backlight to test anything. Here are a few pics of mods

[jakeisprobably]

Spare LCR T3 Clone Spec Test
 ...because I have a spare and you must remove stuff to measure the 470K sense resistors.
Test Voltage   
    9.369V (battery)
Voltage Regulator
    5.032V (78L05)
Voltage Reference
    2.500V (TL431)
LCD Power (could easily be the problem killing a lot of LCDs)
    Open Circuit
        4.535V
    With 10k Resistor Dummy Load
        3.889V
Sense Resistors
  R1  681.27R
  R2  468,470R
  R3  679.43R
  R4  474,810R
  R5  679.05R
  R6  472,250R

Deviation
  R1  +0.19%
  R2   -0.33%
  R3   -0.08%
  R4  +1.02%
  R5   -0.14%
  R6  +0.48%

[timelessbeing]

Maybe I missed something, but why are you posting all this?

Why do we need images of your measurements?

[jakeisprobably]

Maybe I missed something, but why are you posting all this?

Why do we need images of your measurements?

Are you having a bad day?

  Personally I was curious about the tolerance of the resistors used on the LCR T3 given the price range and the 3 digit markings that generally indicate a low tolerance. If you try to test these in circuit you will find it impossible to test the 470k's on any device with the microcontroller permanently mounted. I haven't seen anyone post this. Perhaps I missed it, who knows. These specs are better than both of my other kit clones. I certainly didn't expect the cheapest one to have the better specs. Did you? The one schematic that exists for this clone is also a different hardware revision with different connections. There have been many people that have had screens fail. I haven't seen anyone mention the open circuit voltage of the LCD power supply. I find it interesting that the 78L05 I tested here is within the same specs as the MCP1702-5002. The TL431 here is spot on and, in this instance, more accurate than my lm4040's. If I had known all of this I probably would have just gotten a T3 in the first place.
  I have a much better version of the project to play with. I'm just sharing the kind of info I would have liked to have known. Perhaps I'm wasting my time, but those are a few reasons why.

[timelessbeing]

Are you having a bad day?

Nope. Simple question. Are you?

[timelessbeing]

I find it interesting that the 78L05 I tested here is within the same specs as the MCP1702-5002

Voltage Regulator
    5.032V (78L05)

This is measured with your AN8009? 0.5% +3 meter.
so actual voltage could be up to 5.060V. That is a tolerance of 1.2%

The MCP1702 has a tolerance of 0.4%

[spiderb]

i myself think that what makes this topic interesting is the vast amount of details available
if something was not of interset to me then i wont read it
i certainly would not complain as it could just what someone else needs

bill

[jakeisprobably]

   I'm no expert, just a self proclaimed student of this project. I try to show my work so that others can (kindly) correct me when I inevitably write something stupid.

@Madires
   Is it possible to build an ATmega328 version with HW SPI available for the display if the test resistor port is swapped in your firmware? Would that speed up the display refresh rate significantly for the tft options?

[madires]

That should be possible when you also move the power control to the other port. And yes, the hardware SPI is much faster and will speed up the display output, especially for color displays.

[rocknroller]

Can someone recommend a specific pre-made tester that I can just purchase now that has the latest firmware, or close. I really don't have the time to build from a kit and I don't want to have to download firmware etc. Just want to buy one and start using it. Color is nicer, but prefer more flexibility in features/accuracy over that. Mostly using for testing transistors, including JFETs and related (maybe an occasional germanium or darlington). Also ESR for caps, though I have other esr testers could be used so less important. Transistir the focus. thanks for any advice and/or link to something I can get now. I see TC1 and T6 and T7 on ebay, not clear which is later/better or if there are better choices.

[timelessbeing]

Sellers don't tell you the firmware version, but you can bet it's old.

This is a community project, not a commercial product. You might be more interested in Bob Parker products or something similar.

[beanflying]

Not that I follow this topic that closely but sometimes people lose sight of what these devices are. An excellent Cheap, Simple and reasonably accurate toy to have in your collection of gear. If you think they will become a miracle device by getting the latest firmware upgrade to a $7-15 item you are wrong.

The extreme lengths and work that continues to go on with these is way OTT but in a good way

Upgrading firmware on mine is a non issue it just works and it sits in my on site bag powered by a 2S LiPo with a few other meters that won't do what it does. I have posted photos of mine ages ago but not with the 2S300 LiPo fitted into the HDD case it is shoehorned into.

Buy one and be happy 

[bitseeker]

External drive case. That's nifty, bean.

[jakeisprobably]

@rocknroller
  I don't have the more expensive clones. I recall there are a few problems with some of the more advanced clones as far as their ability to be flashed with newer firmware. All clones are pirate hardware. No one is too upset about it, but I know it's a minor irritation to some. The firmwares are usually modified for whatever reason. The Devs have reversed most of the common clones well enough to support the hardware various configuration.
   All that said, I think Nick L was selling some of his version 2 testers over on the Russian forum at some point. I don't know if that's still happening. As far as I onow that's the only simi-official complete hardware that has ever been sold. All the clones I have came with their own issues. The T3/T4 at $5-$7 is about as disposable as most of the clone designs. The actual OS project is centered around creating your own hardware from scratch. I had no intention of making my own a year ago, but it's the only way to get the proper specs and hardware configuration.

[timelessbeing]

making my own a year ago, but it's the only way to get the proper specs and hardware configuration.

It's ONE way. Most people find it easier to start with a clone and modify it.

[beanflying]

'Proper Specs' is questionable given the components typically used on these. They are not well defined typically and even improving on the absolute 'accuracy' or tolerances and quality of the components without known equipment and standards to compare them to makes this fairly pointless. If you have those standards very doubtful you will be using one of these to get 'accurate' measurements 

The use of the word 'clone' in this discussion is also pointlessly diminutive of available options and designs as there is so many deviations since the original hardware and software first came into being. What is a clone and what isn't a clone of what actual version exactly?

If you want 'proper specs' buy a 'proper meter' otherwise be happy with what they are

[jakeisprobably]

Read the documentation.