$20 LCR ESR Transistor checker project

[radioFlash]

Thanks!

Here are a couple more images of the layout.

[madires]

I made my version of the project based on the ttester_eng110k.pdf schematic. It implements the 2.5 voltage reference, external crystal, TVS diode and relay protection of the inputs, but not the zener measurement. I'm able to get it to work with the 1.10k firmware with the relay removed. However, with the 1.12m version of the firmware, I either get no component found or some very large capacitance when no capacitor is attached. I'm assuming I haven't built the firmware correctly--any hints on what I need to adjust in the config.h or makefile? (Attached are the makefile and config.h I'm using when trying to build the 1.12m firware.)

Which frequency does your crystal have? Based on the Makefile it's 16MHz. I haven't tested the m-firmware with the new input protection IC and TVS yet (lack of prototype with that IC). So it could be possible that it may cause some trouble, but mostly the crystal frequency and the fuse settings are misset. Regarding the relay problem I'd suggest to try increasing the delay time. The boards with relay protection I got use a quite small relay (Karl-Heinz got the same). A large relay might simply need some more time.

[radioFlash]

Which frequency does your crystal have? Based on the Makefile it's 16MHz. I haven't tested the m-firmware with the new input protection IC and TVS yet (lack of prototype with that IC). So it could be possible that it may cause some trouble, but mostly the crystal frequency and the fuse settings are misset. Regarding the relay problem I'd suggest to try increasing the delay time. The boards with relay protection I got use a quite small relay (Karl-Heinz got the same). A large relay might simply need some more time.

Thanks to Karl-Heinz and you for all your work on this project--it's a very nice contribution.

About my build:
* It's a 16 Mhz crystal
* The fuse values are L:f7 H:d9 E:fC
  * I'm using ProgISP with a cheap ebay programmer
* I get the same behavior with and without the relay installed--if I physically remove the relay, the TPs should all be disconnected, so it should have no effect, correct?

For the K version of the firmware, I used the built in RC oscillator (because I couldn't figure out if it supported a 16Mhz crystal) and didn't see a place to specify a relay in the makefile, so I had the relay removed. The K version works as expected.

[madires]

About my build:
* It's a 16 Mhz crystal
* The fuse values are L:f7 H:d9 E:fC
  * I'm using ProgISP with a cheap ebay programmer
* I get the same behavior with and without the relay installed--if I physically remove the relay, the TPs should all be disconnected, so it should have no effect, correct?

For the K version of the firmware, I used the built in RC oscillator (because I couldn't figure out if it supported a 16Mhz crystal) and didn't see a place to specify a relay in the makefile, so I had the relay removed. The K version works as expected.

The fuse settings look fine and the missing relay shouldn't have any impact besides a small delay for the measurements. The k-firmware supports 16MHz too. I think, the next step is to check if the crystal oscillator is running. If it's not running please verify if the two caps between the crystal and ground got the correct value (18 or 22pF).

[radioFlash]

The fuse settings look fine and the missing relay shouldn't have any impact besides a small delay for the measurements. The k-firmware supports 16MHz too. I think, the next step is to check if the crystal oscillator is running. If it's not running please verify if the two caps between the crystal and ground got the correct value (18 or 22pF).

The crystal is oscillating (see attached capture of the waveform). I can see the tester going through the test and updating the display.

[Wh1sper]

Hi,
I did build one at my own with the great PCB made by Moritz A.
down side: expensive to build
plus: I have build it myself, by using the best components I could get. i.e. the high precision Reststors.
I understand the design, I have the schematics
But I bought a Chinese clone with graphic Display recently as second device.
Plus: CHEAP! and very nice build.
down side: no schematics, no precision resistors.
The measuring experience is by far not so exact as the German home build .
But for a quick qheck it is very useful.
So my thumb goes up for having 2, at least :-)
Also important is to have 2 different software versions.

Overall I like this project a lot.
Whenever one of the developers comes at my City, I own them a beer, or two :-)

[madires]

The crystal is oscillating (see attached capture of the waveform). I can see the tester going through the test and updating the display.

Ok. Have you run the self adjustment? Prior to that please measure a film cap around 220nF up to 2µF three or four times. Does the tester identify resistors, diodes and transistors?

[radioFlash]

The crystal is oscillating (see attached capture of the waveform). I can see the tester going through the test and updating the display.

Ok. Have you run the self adjustment? Prior to that please measure a film cap around 220nF up to 2µF three or four times. Does the tester identify resistors, diodes and transistors?

It doesn't recognize any components attached. While testing, the relay chatters (rapid on/off). Using the K firmware, there are just a few clicks of the relay during testing and it reports reasonable values.

As a side note: I originally used a BC547  to control the relay, but the transistor failed. The BC547 is rated to 100 mA and the relay is rated at 80 mA, so I didn't expect a problem. I replaced it with a 2n2222 which should hopefully be sufficient.

[madires]

It doesn't recognize any components attached. While testing, the relay chatters (rapid on/off). Using the K firmware, there are just a few clicks of the relay during testing and it reports reasonable values.

That's strange! Could you please set the fuses to select the internal RC oscilllator (8MHz), recompile the m-firmware for 8MHz RC oscillator and try it again? BTW, could you also please sketch the circuit part driving the relay (PC4 and the voltage reference).

[bob808]

I've been working on this project and I managed to get a single sided pcb with 5 wire-passes on the topside.
It has the Zener diode test and frequency test. Also it includes the left-right encoder with push-button.  There's a ZIF socket, pcb banana connectors and a daughter-board for the SMD stuff. You can just connect that if needed. Also there's a protection relay for charged caps. Also I've used 2 regulators, one for the IC and one for powering the extras. There's an ICSP connector as well. There is no pot for adjusting the contrast/light on the LCD, you will have to figure out the values for those things.
I still have some work to do on it, need to match holes for LCD support so it can be fixed with spacers. I will need to move some parts around to do that. Also I need to double check the footprint of some parts, and see what's available. The LCD should clear most parts.
This is a nice diy board that can be made at home. I will post the files after I'm finished with it.
Also I want to thank Mr. Madires as he guided me through this project.

[bob808]

I made the holes for the LCD. I used this datasheet:
http://www.adafruit.com/datasheets/TC1602A-01T.pdf
Seemed like a generic 2 line LCD.
Also now there are only 4 wire-passes on the top side. Rearranged the SMD pads a bit.
Any suggestions?

[radioFlash]

I made the holes for the LCD. I used this datasheet:
http://www.adafruit.com/datasheets/TC1602A-01T.pdf
Seemed like a generic 2 line LCD.
Also now there are only 4 wire-passes on the top side. Rearranged the SMD pads a bit.
Any suggestions?

Nice layout. You may want to try to re-position the programming header so it's not under the LCD. With the LCD in place, it may interfere with the header. Also, it would be nice to reprogram it without having to remove the LCD. Your TO-92 packages seem to have different sized pads--you may want to make all of them the same with the larger pad sizes. You could also probably increase the pad sizes for the film capacitors.

[radioFlash]

It doesn't recognize any components attached. While testing, the relay chatters (rapid on/off). Using the K firmware, there are just a few clicks of the relay during testing and it reports reasonable values.

That's strange! Could you please set the fuses to select the internal RC oscilllator (8MHz), recompile the m-firmware for 8MHz RC oscillator and try it again? BTW, could you also please sketch the circuit part driving the relay (PC4 and the voltage reference).

I switched it to use the RC oscillator with fuses  L:e2 H:d9 E:fc and adjusted the makefile to use OSCILLATOR = RC, but get the same behavior. Attached is the schematic.

[nessatse]

I made the holes for the LCD. I used this datasheet:
http://www.adafruit.com/datasheets/TC1602A-01T.pdf
Seemed like a generic 2 line LCD.
Also now there are only 4 wire-passes on the top side. Rearranged the SMD pads a bit.
Any suggestions?

• Your zener and frequency labels appear to switched.
• You probably need a lot more clearance around the lcd mounting holes if you don't want screws/nuts shorting out your traces.
• If this board is designed for home etching, I would suggest increasing your ground plane clearance a little
• As radioFlash suggested, try increase some of your pad sizes and get the size consistent

[madires]

I switched it to use the RC oscillator with fuses  L:e2 H:d9 E:fc and adjusted the makefile to use OSCILLATOR = RC, but get the same behavior. Attached is the schematic.

Thanks! The schematic looks fine, my prototype with the relay protection has exactly the same driver. The only difference is the input protection IC and it got a 8MHz crystal, and a smaller relay. Since you mentioned the Makefile, have you also set FREQ = 8? What puzzles me, is that there's no flapping of the relay with the k-firmware. My next idea would have been to check Vcc in case the current draw of the relay causes the issue. Both firmwares drive the relay the same way, just with a slightly different timing regarding the measurement procedure. The MCU clock shouldn't have any impact on the relay anyway. I think we're overlooking something.

[radioFlash]

I made the holes for the LCD. I used this datasheet:
http://www.adafruit.com/datasheets/TC1602A-01T.pdf
Seemed like a generic 2 line LCD.
Also now there are only 4 wire-passes on the top side. Rearranged the SMD pads a bit.
Any suggestions?

• Your zener and frequency labels appear to switched.
• You probably need a lot more clearance around the lcd mounting holes if you don't want screws/nuts shorting out your traces.
• If this board is designed for home etching, I would suggest increasing your ground plane clearance a little
• As radioFlash suggested, try increase some of your pad sizes and get the size consistent

* I used a 130 thou radius circle for the keep-out region around the LCD mounting holes. This gives just barely enough clearance for a 4-40 nut.
* I think you will have a clearance problem with the trim pot positioned under the LCD. To avoid that and allow adjustment with the LCD in place, you may wish to move the trim pot down a bit.
* The LCD has some metal tabs on the back which are bent to keep the LCD attached to the PCB. They project out about 1/10 inch. The LCD PCB will be about 430 thou above your PCB when mounted, so keep that in mind when thinking about clearance.

Attached is a photo of a 4-40 nut with the 130 tho radius keep-out region and an image of the backside of the LCD.

[bob808]

I guess this is the best I could do.
It's hard as everything is tight, and as soon as I move some parts/traces I loose ground in some places and I don't want to add more wire passes. I think one could mount some non-conducting spacers on the back of the PCB for electrical isolation between the tracks and the nut. Also I think you could mount some taller spacers for the LCD but I think it should clear most of the parts.
If you wish I could post the Eagle files.

[radioFlash]

I switched it to use the RC oscillator with fuses  L:e2 H:d9 E:fc and adjusted the makefile to use OSCILLATOR = RC, but get the same behavior. Attached is the schematic.

Thanks! The schematic looks fine, my prototype with the relay protection has exactly the same driver. The only difference is the input protection IC and it got a 8MHz crystal, and a smaller relay. Since you mentioned the Makefile, have you also set FREQ = 8? What puzzles me, is that there's no flapping of the relay with the k-firmware. My next idea would have been to check Vcc in case the current draw of the relay causes the issue. Both firmwares drive the relay the same way, just with a slightly different timing regarding the measurement procedure. The MCU clock shouldn't have any impact on the relay anyway. I think we're overlooking something.

Good news--I was able to get the M firmware working. I switched from the cheap ebay "USB ISP" programmer to a genuine Atmel AVR ISP MkII programmer. The relay is working properly and values look reasonable. I may have been using the programmer software for the cheap programmer incorrectly. Using avrdude from the makefile to program the MCU has cleared up my problems.

Does the M version of the firmware support a rotary encoder yet?

[madires]

Good news--I was able to get the M firmware working. I switched from the cheap ebay "USB ISP" programmer to a genuine Atmel AVR ISP MkII programmer. The relay is working properly and values look reasonable. I may have been using the programmer software for the cheap programmer incorrectly. Using avrdude from the makefile to program the MCU has cleared up my problems.

That's great! One thing less I have to worry about

Does the M version of the firmware support a rotary encoder yet?

The next version will, I'm currently adding support for rotary encoders. Reading the encoder already works nicely using an ALPS EC11, but other encoders should also be fine (you just need to set the pulses per step/detent in config.h). Right now I'm working on using the extra user input to improve the UI. When running the PWM tool you can use the encoder to increase or decrease the PWM ratio in steps of 1% additional to the 5% steps of the push button. And I also have to get more encoders for testing the algorithm for reading the encoder signals, some crappy ones

[madires]

Just a quick update, you'll find 1.13m in the SVN now. And I'd appreciate any feedback on the rotary encoder's usability.

[MrJohnnny]

  Hello everyone! Built this tester, works great, huge thanks to authors  Currently I use Mega8 with v1.04k and ordered Mega328, but encountered a problem - no compiled software available. I need version from Karl Heinz because i need to measure inductance. Can someone please post link to compiled version v1.11k ? 

[madires]

Hello everyone! Built this tester, works great, huge thanks to authors  Currently I use Mega8 with v1.04k and ordered Mega328, but encountered a problem - no compiled software available. I need version from Karl Heinz because i need to measure inductance. Can someone please post link to compiled version v1.11k ? 

IIRC, the inductance measurement requires an ATmega with 16kB flash or more.

[retrolefty]

Hello everyone! Built this tester, works great, huge thanks to authors  Currently I use Mega8 with v1.04k and ordered Mega328, but encountered a problem - no compiled software available. I need version from Karl Heinz because i need to measure inductance. Can someone please post link to compiled version v1.11k ? 

IIRC, the inductance measurement requires an ATmega with 16kB flash or more.

Hence why he ordered a mega328?

[madires]

Hence why he ordered a mega328?

Got it. It's http://www.mikrocontroller.net/svnbrowser/transistortester/Software/trunk/mega328/ for the standard circuit.

[MrJohnnny]

Thanks  Now I can test this great gadget