Warm Tube Clock sat v2

Sve vezano za mikrokontrolere

Moderators: pedja089, stojke369, [eDo], trax

Post Reply
hetoosten
Pocetnik na forumu
Pocetnik na forumu
Posts: 25
Joined: 30-07-2016, 09:42

Re: Warm Tube Clock sat v2

Post by hetoosten »

Unfortunately no luck :-(

I did the following tests:
* 3v battery to ground: 3v ok
* 5v checks all ok
* Voltage between IC2 and C14 (Test 3) strange --> 10 Volts (without Atmel and HV chip)
* HV check not ok, also 10 Volts (with all IC's)

Does this have anything to do with the fuses on the avr, or the circuit?

I also mounted the shield (without nixies). Once one led lighted up blue, then nothing.
hetoosten
Pocetnik na forumu
Pocetnik na forumu
Posts: 25
Joined: 30-07-2016, 09:42

Re: Warm Tube Clock sat v2

Post by hetoosten »

Sorry for spamming this forum :-)

I made some progress with the Avr programming:

* I got the Usbasp programmer working. It turned out that the connections on the flatcable are mirrored, and only one of the three grounds pins is connected. I also added a crystal to the breadboard.

* I found a forum topichere abount fuse programming. It turns out that efuse:w:0xfd results in an error, and should be changed into efuse:w:0x05

So the commands for programming the AVR with Usbasp on Linux are:

Code: Select all

sudo avrdude -c usbasp -p m328p -v -e -U lfuse:w:0x96:m -U hfuse:w:0xd9:m -U efuse:w:0x05:m

sudo avrdude -c usbasp -p m328p -v -e -U flash:w:wt2_IN14_shield.hex
I know now that the Atmel is programmed right, but unfortunately the hv still doesn't work.Could this have anything to do with the smd connections on the backside of the mainboard?
mainboard.jpg
mainboard.jpg (24.53 KiB) Viewed 12026 times
hetoosten
Pocetnik na forumu
Pocetnik na forumu
Posts: 25
Joined: 30-07-2016, 09:42

Re: Warm Tube Clock sat v2

Post by hetoosten »

Sorry for another post. I wasn't able to edit the previous one.

I flashed the Avr with the firmware from folder 31d-08m-2015y

I assumed I had to solder two 4k7 resistors here too (this part was undocumented as well):
pcb-nixie.jpg
pcb-nixie.jpg (32.17 KiB) Viewed 12007 times
User avatar
trax
Administrator sajta
Administrator sajta
Posts: 3508
Joined: 08-01-2005, 18:04
Location: 75k, BA
Contact:

Re: Warm Tube Clock sat v2

Post by trax »

Sorry for late reply.
It is normal that LEDs on shield PCB blink on startup, it happens the same on my clock.

You don't need 4k7 resistors there, that is actually a pin header in case you want to connect some additional I2C device to the AVR (for your own modifications) :)

The SMD pads on the bottom side of PCB - you don't need to connect those. Those are the pullup resistors for I2C but they are not required since AVR has internal pullups. BUT you need to solder the SQW/INT SMD jumper that is close to them. Just put some solder on that SQW/INT to connect it.

As your HV supply is not working, that might be the problem with IRFD220 MOSFET transistor. Is that OK? Can you check it?
hetoosten
Pocetnik na forumu
Pocetnik na forumu
Posts: 25
Joined: 30-07-2016, 09:42

Re: Warm Tube Clock sat v2

Post by hetoosten »

trax wrote:Sorry for late reply.
It is normal that LEDs on shield PCB blink on startup, it happens the same on my clock.
No problem at all, I really appreciate your help!

I have time this evening to check values of the mosfet (and the solder the smd jumper).

To check the direction of the mosfet: the mosfet has one bridged side (d), pointing to the 1uF cap, and two seperate legs (S,G) pointing to the speaker. Is this placement correct?

On the shield one time one led lighted up, but now all leds stays dark.
User avatar
trax
Administrator sajta
Administrator sajta
Posts: 3508
Joined: 08-01-2005, 18:04
Location: 75k, BA
Contact:

Re: Warm Tube Clock sat v2

Post by trax »

For the Mosfet, I don't have it here with me, but the side of Mosfet where two pins are actually connected together - this faces the diode in the PCB: http://www.elektronika.ba/img-cont/c/c_ ... oard_4.jpg
I hope I am not telling you wrong because I am at work now :-)
hetoosten
Pocetnik na forumu
Pocetnik na forumu
Posts: 25
Joined: 30-07-2016, 09:42

Re: Warm Tube Clock sat v2

Post by hetoosten »

According to your description the mosfet is placed right. The bridged side has a diode next to it, and the other end (the two seperate legs) pointing in the direction of the speaker (the Atmel side)

What (voltage) values should the mosfet have?
User avatar
trax
Administrator sajta
Administrator sajta
Posts: 3508
Joined: 08-01-2005, 18:04
Location: 75k, BA
Contact:

Re: Warm Tube Clock sat v2

Post by trax »

Hello,

AVR should be pulsing the gate of mosfet. This would cause mosfet to short-circuit the coil to GND which in turn would induce high voltage on the Drain of mosfet once it turns off. This is directed through the fast high-voltage diode BA159 and stored in high voltage capacitor C13 = 1uF/250V.

If you are measuring only ~11V at the drain of mosfet, then AVR is not pulsing the transistor. That is why you are not seeing high voltage on drain but only the supply voltage.

You need to make sure that AVR is pulsing the transistor. Also, check the feedback resistors R3 and R4 because through these resistors AVR measures what it generated and "re-configures itself" to generate the required voltage for nixie tubes.

You probably didn't program the AVR properly :shock:
FYI: check the last two pages of this old assembly manual: http://www.elektronika.ba/misc/main_board_v1.pdf (it is for older version, but you can still read it)

Regards!
hetoosten
Pocetnik na forumu
Pocetnik na forumu
Posts: 25
Joined: 30-07-2016, 09:42

Re: Warm Tube Clock sat v2

Post by hetoosten »

trax wrote:You probably didn't program the AVR properly :shock:
FYI: check the last two pages of this old assembly manual: http://www.elektronika.ba/misc/main_board_v1.pdf (it is for older version, but you can still read it)
Regards!
I flashed the Atmel with the firmware wt2_IN14_shield.hex from the folder 31d-08m-2015y
There were no errors. Is there a way I can check if the Atmel is working at all?

I assembled the pcb over a long time, and I noticed a big difference with you picture. Under the mosfet I installed a diode, and a capacitor. On your picture these are left empty. Should I remove these parts?

I also noticed two 22 capacitors missing next to the crystal. Should I remove these as well.

.....and thanks again for your response!
User avatar
trax
Administrator sajta
Administrator sajta
Posts: 3508
Joined: 08-01-2005, 18:04
Location: 75k, BA
Contact:

Re: Warm Tube Clock sat v2

Post by trax »

Hi,

You should place 22pF capacitors on board.
You do not need zener protection diode and a capacitor, but you can leave them. Just in case they are faulty, remove them until you fix the original problem.

If you are looking at the picture from the PDF (assembly manual), there is a difference because in manual it is version v1. You are assembling version v2. I gave you the PDF link so you can read last two pages, to perform HV tests.

:D
hetoosten
Pocetnik na forumu
Pocetnik na forumu
Posts: 25
Joined: 30-07-2016, 09:42

Re: Warm Tube Clock sat v2

Post by hetoosten »

@trax after your comment that the Atmel was not programmed right, I reprogrammed the Atmel (using the same procedure as last time). After reinstalling one nixie dot came alive. And yes 200V!

As I also found out after removing the powerplug :ohno: :D

Now I have to investigate further because the leds should turn on blue after power on, and so far only one nixie dot turns on.
hetoosten
Pocetnik na forumu
Pocetnik na forumu
Posts: 25
Joined: 30-07-2016, 09:42

Re: Warm Tube Clock sat v2

Post by hetoosten »

Waited too long too edit. When I power on the lower nixie dot switches on for a second and then off, the upper stays on. Is this the correct behaviour?
User avatar
trax
Administrator sajta
Administrator sajta
Posts: 3508
Joined: 08-01-2005, 18:04
Location: 75k, BA
Contact:

Re: Warm Tube Clock sat v2

Post by trax »

If I can remember correctly, the nixie dots should start blinking as if the clock is fully assembled. They should start blinking after couple of seconds. At first, they should stay ON for some time.
LEDs should also start to blink aftre few seconds.
Check the two NPN MSP42 transistors that power those two nixie dots. If one dot stays ON, it could be that the NPN transistor is always open (broken), or that there are some soldering problems on the PCB.
Check everything :D

Send some close-up photos here if you like...
hetoosten
Pocetnik na forumu
Pocetnik na forumu
Posts: 25
Joined: 30-07-2016, 09:42

Re: Warm Tube Clock sat v2

Post by hetoosten »

I checked the two transistors. The base of both are connected to the right pin of the Atmel with 10k resistance. The emitters both have zero resistance to ground. And the collectors zero resistance to the right female header pin.

I found a more disturbing issue however. The TLC chip has a dot for the right placement on the shield. In the shield construction manual image this dot is pointed to the inside. When I measure ground (pin1) the dot should be placed towards the outside of the pcb

I it correct that the chip is placed wrong?
tlc-chip.jpg
tlc-chip.jpg (20.03 KiB) Viewed 11946 times
User avatar
trax
Administrator sajta
Administrator sajta
Posts: 3508
Joined: 08-01-2005, 18:04
Location: 75k, BA
Contact:

Re: Warm Tube Clock sat v2

Post by trax »

Yes, it is pointed to the inside. The marking on your chip should match the marking on the PCB. I can't see clearly your chip marking, where is it?
TLC marking.jpg
(58.16 KiB) Not downloaded yet
hetoosten
Pocetnik na forumu
Pocetnik na forumu
Posts: 25
Joined: 30-07-2016, 09:42

Re: Warm Tube Clock sat v2

Post by hetoosten »

The marking was hardly visible on the image, so I placed a red dot (it is on the inside)

I think I am ok with the placement of the TLC chip :-)

Now I have to find out why the led's won't turn on, and the upper nixie dot keeps stays on.
User avatar
trax
Administrator sajta
Administrator sajta
Posts: 3508
Joined: 08-01-2005, 18:04
Location: 75k, BA
Contact:

Re: Warm Tube Clock sat v2

Post by trax »

Check if you soldered LEDs properly, they also have markings :)

Did you solder the bottom GND pad of the TLC chip properly? I once didn't and it behaved strange - when I touched it with finger it started working, when I moved away it stopped. That was all because I didn't solder the TLC ground pad (underneath it).

Also, make sure you program your AVR with fuse CLKOUT enabled! This provides a clock for the TLC chip to operate.
hetoosten
Pocetnik na forumu
Pocetnik na forumu
Posts: 25
Joined: 30-07-2016, 09:42

Re: Warm Tube Clock sat v2

Post by hetoosten »

I noticed something strange in the output of Avrdude:

avrdude: safemode: lfuse reads as 96
avrdude: safemode: hfuse reads as D9
avrdude: safemode: efuse reads as 5
avrdude: safemode: Fuses OK (H:05, E:D9, L:96)

The values hfuse and efuse are switched. Very strange.

I programmed efuse as 05 (should be the same as fd).
hetoosten
Pocetnik na forumu
Pocetnik na forumu
Posts: 25
Joined: 30-07-2016, 09:42

Re: Warm Tube Clock sat v2

Post by hetoosten »

Nevermind the CKOUT is in the low fuse. This value is correct. Time to fire up my multimeter :-)

@trax thank you for being so patient with me

Edit: I noticed the mark on the leds. They are all pointing bottom left (opto coupler and two resistors side)
hetoosten
Pocetnik na forumu
Pocetnik na forumu
Posts: 25
Joined: 30-07-2016, 09:42

Re: Warm Tube Clock sat v2

Post by hetoosten »

I am considering to start over from scratch. I was testing the clock and turned the clock. The lower nixie dot started to light up. Moving the pcb back it went off again. When I test the voltage at the header pin and the resistor, the dot switches on. Touching the nixie dot itself makes no difference.

Also after some light pressing on the pcb half of the leds switched on. One bright blue, and one only half bright (also blue).

I fear there must be a grounding issue. I checked al solder points.
Post Reply