After making a clock based on the MSF60Khz receiver and it working well on a breadboard for a year or so I am looking to design a PCB that uses a barebones Atmega and only the bare components needed to complete the job. Its works well so functionality isn't the issue.
So far I have drawn the schematic on Eagle and sort of completed a PCB with layout but would really appreciate it is someone could check it over and tell me if I've made any obvious errors. I have never used Eagle before and have no idea if I have done it right.
What I'd like from the output is a PCB that is stand alone. Powered from a micro USB phone charger type plug on an easy (read cheap) to make board that I can give to novice ham radio people to use as a shack clock that they will be able to explain what it going on in the circuit, hence the barebones nature.
If anyone can help I would really appreciate it. I can put the files in dropbox, here or somewhere similar if needed.
You can export your Eagle schematic as a PNG graphic and post it here for starters. Plenty of us can help you verify it and help build a board, myself included.
Are R1 and R3 meant as rotary pot resistors? If so, you're missing a connection to the third leg. If they're meant as regular resistors, then you're using the wrong symbol and package.
The next question is whether you want this as a through hole board or SMD board. SMD will make it much smaller.
No decoupling capacitors for the ATmega chip (100nF between VCC & GND near the chip)
No caps on the 7805 (0.33uF & 0.1uF)
ATmega reset pin left floating (Link with 10K resistor to VCC and optionally a switch to GND to reset it)
Are all of those absolutely necessary as it works without them? I'm guessing they are from your response but if they are not included what is the consequence?
Just because you need that particular header for the MSF, does not mean you can't still make the rest of the board/components small. All of my custom boards are SMD with the exception of the headers for FTDI and SPI.
And as Riva pointed out, you're missing several pieces still. Yes, they are necessary for proper operation.
I'll put in the parts you mention, well worth posting to get guidance. I take it the 7805 caps are input and output respectively, I've seen some where its 2 off 100nf. I suppose they both do the same thing
I'm ok with SMD soldering but quite a few people aren't so I wanted avoid them. Its likely that people going from a novice ham radio licence to an intermediate licence will make one of these as part of the exams. Most of them buy a kit from Maplin, make it then put it in a drawer for ever. Hopefully this could be useful.
It would be nice to put it on a board which fits the LCD and is deep to keep it small but SMD's may be a step too far, perhaps 2 options need to be made up
I'll put in the parts you mention, well worth posting to get guidance. I take it the 7805 caps are input and output respectively, I've seen some where its 2 off 100nf. I suppose they both do the same thing
Yes, I use 0.33 on input and 0.1 on output.
Another thing to consider is how your going to program the ATmega. It is supplied pre-programmed or will the builders need to upload the sketch themselves? Will it have a bootloader on board? If yes then consider putting a 10K resistor between RX pin and ground. I built a ATmega8 board some time ago and left RX floating. Sometimes after a reset it would delay several seconds in the bootloader as the floating RX pin looked like some form of serial data.
I was planning on pre programming the Atmega using my Uno, hopefully saving the guinea pigs from that part of the job. I have bought one from hobbytronics for a few quid that seemed to work ok although the longevity of the uno may be reduced but we're only talking about 10 max.
Not sure about connections of the the variable resistors though, have made changes attached
R2 is connected to wrong side of 7805, should be connected to VO side not VI
Contrast POT is not right, along with the VSS & VDD feed to the LCD (assuming standard LCD header).
VSS should be ground (as it is), VDD should be connected directly to 5V (VCC) and not connected through R1.
The wiper (S) of R1 should be connected to VO
R1 (A) should be connected to ground.
See this as example http://arduino.cc/en/Tutorial/LiquidCrystal
Your trying to drive the backlight through the mega pin but you should use a transistor as the current will probably be to high and damage the mega.
I had confused myself with the pots etc (starting to look like spaghetti on the page :)). I think this should be the last attempt today before I get my head spinning too much. btw I couldn't find much on the current drain of the LCD apart from a 60 pager that probably had the data in it somewhere.
What is the expected drain from an LCD for future reference?
Can't thank you enough for your time today Riva & KirAsh4. More progress today than in several days on my own.
What is the expected drain from an LCD for future reference?
Depends on what display you have but a quick look at the Sparkfun one here Basic 16x2 Character LCD - Red on Black 5V - LCD-00791 - SparkFun Electronics the GMD1602K datasheet shows the backlight draws about 120mA. As the Mega pins are rated at maximum (non continuous) 40mA this is to much.
Here is a simple circuit to do PWM dimming or on/off control of backlight
Top end of R2 would be connected to pin 19 as you had in previous images for simple on/off or you could connect to pin 15 to give you a PWM dimmable backlight. The bottom end of R2 connects to base of a NPN transistor (something like a 2N2222). Notice that VCC is fed straight to LCD backlight Anode and the transistor sits in the Ground/cathode side. Unlike your drawing that has the pot on the anode side and cathode goes direct to ground.
If your not bothered about Mega controllable backlight then R3's A to ground connection is not needed.
Mines an eBay special so it'd undoubtedly be around that figure.
I found something similar which gave a figure of 160mA for the LED forward current and a supply current of 4mA. Is forward current the amps required at full brightness and supply current that which is needed to power on.
Thanks for the circuit, I'll add that in to the spaghetti mess I'm making
By the way, it's perfectly okay to create your schematic without drawing every single trace. Just connect them by name. Draw part of the trace on one side and name it. Then draw the connecting trace on the other part and name it the same. Eagle knows to connect the two in board mode. See attached.
That makes a lot of sense doing it that way. I'm going to have to look at some other tutorials I think.
I've only picked up the software over the last week or so and am getting used to designing circuits with it. I haven't really tackled the board side of it yet, I had a sneaky look and it's not looking like a pcb at the moment
It won't till you place the components and route the nets (either auto routing or manual.) Eagle simply groups all the components next to the board and leave it up to you to place things and create your final design.
