Did you check the troubleshooting guide ?
Having an oscilloscope (25MHz and up) will be useful for some of the tests. Any parts getting hotter than 40°C ?
i didn’t knew about that! 
thanks alot!
so…every voltage is nearly the values expected other than on the display pin #3
which is about 1,6V and on pin #18 which is about 0.6V
yes there are parts that are a bit hot… the regulator gets a a little hot and the atmega64 but i thought it’s normal… isn’t it? 
the troubleshoot guide says: “To verify MCU state check voltage at its pin 5 (measuring at C17 is easier). This voltage should be about 1.9V.”
when i measure at the C17 i see close to 1.9V
The voltage regulator gets quite hot (so it hurts to touch it) if you feed it with more than 9V. I use 7V and it is fine. As long as the red led is fully on, it still works. The mcu may get warm as it is overclocked to 20MHz, but not really hot.
if i go for 12V power supply then the regulator gets quite hot and it’s not easy to touch…i believe this is reasonable since the input-output voltage difference gets high…
now about the MCU it gets a little hot but you can touch it… also the 1N4001 diodes get a bit hot…
so… we can safely assume there is display issue or there’s something else i should try?!
Hmmm. As the cpu drives the negative voltage generation for the opamps using the pin labeled VGEN, it should still work if you’ve confirmed the negative voltages are ok. If the cpu is frozen, no negative voltages. You could also check for activity on the serial port, as suggested in the manual.
Here you can find a schematic:
jyetech.com/Products/LcdScope/eDSO062K.html
I’d check all connections to the display for bad solder joints or shorts and the parts near LCD1 shown in the schematic.
sorry… where is the VGEN pin? when you say about the negative voltages you mean on the AV- pad?!
this is hard to do since i don’t have a serial level converter…
Yes. If you measure about -7V on the AV- pad, the cpu (or at least one of its timers) is still running, as it drives the generation circuit with one of its IO pins (square wave).
VGEN in the schematic is ‘PE3’ (port E, pin 3) on the atmega64 chip. This is pin #5 of the chip. As it outputs a square wave at 5V, you might only measure about 2.5V with a voltmeter. But if AV- is OK, you don’t need that.
Here’s a photo of it:
I also didn’t know about the troubleshooting guide. Thanks for posting the link madworm. The guide seems to point that my problem is in the analog part, and suggests the following: (supply voltage used was 9V)
A. Check AV+
AV+ should be the same as or slightly lower than VRAW+. If it is significantly lower then it is
possible that somewhre are shorted. If it is 0V L3 may be open.
AV+ = 6.2V, VRAW = 6.4V, so seems ok.
B. Check AV-
AV- should be -6V or lower if you use a power supply of +9V or higher. If AV- is close to 0V then
you would not be able to see any traces.
AV- is 0.3V on my board, so that’s a problem.
[b]For AV- problem please check:
- Is D7 placed correct? - D7 should be placed with its striped end pointing 7805 (see Fig. 1). If
D7 is placed wrong Q1 is likely burnt. Q1, if burnt, can be replaced with any general purpose
PNP transistor (even a through-hole one). Q1 and Q2 polariy are shown in Fig. 1.[/b]
Yes, D7 was placed correctly.
2) Is L2 open? - Please check it with multimeter. It should read about 3 ohms.
Checked, reads 3 Ohms as it should.
3) Any possible short/open of components around Q1 and Q2?
Checked it over for possible shorts, though the soldering isn’t the best, it doesn’t appear to have any broken connections.
Checked the resistors:
R17 supposed to be 2K, reads 0.2K, checked that the component was the correct value
R20 supposed to be 1K, reads 1.2K
R22 supposed to be 10K, reads 0.4K, checked that the component was the correct value
So these resistors are likely the cause of the problem. Could I just replace them and be done? Or should I expect them to burn again and there something else I should check before replacing them? Any good way to check Q1 and Q2?
Added image with the voltages I measured and the expected values. Assumed the 8.1V was measured on the pad of R17, not on the Q1 base pad.
indeed it was quite helpful link…
playing around with the scope suddenly it worked but only for a little… i could only see that
the firmware runs… then the display faded and i could only see the backlight…
i searched again for bad joints around the display but i couldn’t find any, i also used an eye loop…
i believe the two resistors appearing in the photo are responsible for the VO and VEE… but i don’t see
anything wrong with the joints. Then i was pushing the display in many places in case of internal bad contact but nothing…
turning it off and on many times the display started to show the pattern you can see on the following photo… after a while if i was keep turning it on and off i could see the firmware version and then
the letters were fading… sometimes i could see the main screen with some random waveforms but they
were rapidly disappeared.
i don’t know what’s going on… any other opinions?
@wjt:
Did you measure the resistors in circuit ? That may give wrong values. For that to work you usually need a special ohm meter with an ultra low voltage (so diodes, transistors and other gates don’t conduct).
R20 with 1.2k seem OK.
The value of R17 depends on how you measure it. If you apply + to VRAW and - to the other side of R17, the base-emitter-diode of Q1 will conduct (effectively a parallel resistor to R17) and give a wrong value! It should show the correct value with reversed polarization.
It’s the same game for R22. Correct values for the right polarization, too low values if reversed.
If the scope seems to work from time to time and you’ve seen the boot screen, it could be a bad connection somewhere. Have you tried carefully twisting and bending the board a bit (not really much, approx 0.2mm), maybe there is a broken trace somewhere. You could also try to warm it up with a hair dryer. Other than that, I can only recommend further inspection and re-soldering of suspicious joints. If you can measure the current that is taken by the board, it should be somewhere between 190mA and 250mA, if I remember right.
no it didn’t worked… last resort is what the seeed guys proposed me… i should reburn the firmware…
Hi Nti30, could you please post the voltages at VO and VEE here? It looks these voltages were not right since the screen was totally dark.
- A dedicated discussion group for JYE Tech oscilloscope DIY has been created. Please visit jyetech.com/en/default.html to join it.
though i’m not sure… i believe the voltages on these pins change depending the behaviour of the
display…sometimes on past measurements the voltages where higher or lower… unfortunately i don’t have a measurement of these pins while the display was dark.
currently the voltages are:
+5V is 4.97
VO is -3.85V
VEE is -8.63V
regulator temperature ~55C
MCU temperature ~42C
the display is blank only backlight is on.
done!..
one more update…
multiple turn on and off proved more beneficial than any other thing i tried…
sometimes when i turn it on the display shows the boot screen but then everything freezes
or it shows the next screen or messed up graphics or just a blank display with the backlight.
now about the VO and VEE voltages, on a previous post i stated 0.6V for the VEE this
must be due to a bad contact of my multimeter probe with the VEE pad since now i measure
the voltages as i say on my last post with minor fluctuations.
here’s some photos…
It looks the LCD module is not good. VEE should be about -9.7V. Let’s do some further check before conclusion. Please isolate Pin 18 (VEE) of the module by de-soldering R27 (and clear anything that might touch the pin) and check VEE again. If same result is found then there may be something wrong with its internal power supply of the module. In that case please contact us at [email protected]. We will send you a replacement module.
the voltage on VEE is now -9.43V
Did you try re-soldering those M64 pins that connect to LCD? Use some flux if you have. Some times there are bad soldering (opens or shorts) in these pins that is hard to uncover.
Now it looks M64 is running. Display is not stable. I guess we should first find out the problem with display. Possible causes:
1 ) Bad connection
2 ) Bad LCD module
Since the LCD can display it is likely some bad connections between M64 and LCD.
almost every joint on the pcb has been re-soldered and examined with an eye loop… but i haven’t
touched some specific areas such as the atmega64.
if you check few pages back i have uploaded a photo which is vissible all the pcb area… right now
the scope does absolutely nothing no matter how many times i turn it on and off… only the backlight works… and i’m afraid if it totaly got damaged with all this manipulation.
P.S forgot to add that i have done continuety check on the display pins and i haven’t found any open circuit.
as far as i can see the rails between the MCU and the display are very short and there are no
components other than the VO/VEE resistors between them… checking again with an eye loop proved no broken joints neither on the display pins nor on the MCU pins that drive the display, though i have already uploaded too many photos if you want i can upload one more close one.
of course i could have missed something anywhere on the board… and that’s what i’m trying to avoid by checking again and again so many days.
to me initially looked like a display problem but since the voltages are ok then maybe it’s not? do you think i should remove the display from the board and try to drive it on a breadboard with an 16F628 and a simple program? i believe it’s a good idea though difficult because i don’t have a de-soldering station with hot air so i can safely remove the display without messing the pins.
MCU fault is out of question?
Since bootloader version was displayed so at least the bootloader was running.
When checking connection did you use a multimeter(ohmmeter) or vision only? It is recommended to use ohmmeter with a pair of sharp tips and check from IC pins instead of pads. This is very important for checking high pitch SMD ICs because very often bad soldering can not be identified by eyes even under powerful magnifier.