I am testing some cheap adxl345 modules with an Atmega4808, and I have blown the i2c bus out of two 4808s already. I'm not sure what is going on.
It starts out where I let them sit for a couple of hours, and the chip stops responding. This is something that seems to happen regularly. On two occasions though, when the module stops responding, cycling the power doesn't seem to fix it, but the module works fine using an uno, and when I replace the 4808 chip, it starts working again.
Looking at the module, it appears that from the output pins of the ADXL345 chip to the SDA and SCL lines, there are 4.7k pullup resistors to the 3.3V rail. There is a voltage regulator on the module, which should allow it to work with 5V, but I am only feeding it 3.3V, so overvoltage shouldn't be the cause. I am going directly from the module to the 4808. I have not inserted any passives between the two, as the module appears to have all it needs.
I considered bypassing the voltage regulator and running straight from my own regulator, to see if that solves the communication problem, thinking that running it from 3.3V through a regulator might be the issue with that, but it doesn't seem likely that too low a voltage would kill the i2c bus.
First, what could be causing this? Second, is there any way to protect the i2c bus from strange things happening?
Sometimes people say they need a picture, so here is a picture of what I've got setup. Since I wrote the above post, I have added a 10uf and a 0.1uf capacitor on the power input to the ADXL345 module. Just in case it helps.
Nice picture but not of much use to me. A schematic, not a frizzy picture would be a big help especially if it is how you wired it including all connections including power and ground. Links to each of the hardware items showing technical information not sales stuff as azon has.
My schematic is not quite the same as this. I have my own circuit with the adxl345 that doesn't use the module.
I didn't do anything strange. Vcc is connected to the 3v rail directly. Ground is connected to ground directly. The sda and clk pins are connected directly to the 4808 sda and clk pins. I now have the capacitors on there, which seems redundant, but i did it anyway.
I can only suspect there is some excess voltage hitting the pins, and I have little way to control that other than bypassing the regulator on the module. Even then, it's only a guess.
We need to know what you actually have. A picture is not a substitute for a schematic, and a schematic must show exactly what you have got, otherwise it is useless.
Do you have a capacitor on the sda and clk pins? That is what it reads as.
On the face of it you appearer to be doing everything right. But I have thought that on every project where I made a mistake, that is human nature.
I know a schematic is best, but I am mocking things up on a breadboard using some prebuilt modules, and some through hole parts that have similar characteristics to the end product, such as mosfets, regulators, and passives. The portion of my schematic that deals with the ADXL345 module portion doesn't deal with the module, but a different circuit. My circuit doesn't have a separate voltage regulator on it, for instance. It is missing a tantalum capacitor, and a ceramic capacitor that the module has. It's a slightly different circuit, and I don't know that it will be helpful. I suppose I can find a schematic for the module, but they are so common, I didn't think I needed to do that.
As for the capacitors i added, i didn't add them to the i2c lines. I added them to the vcc line going to the module. I don't imagine they would do good things for the sda or clk signals.
Anyway, I don't think I'm doing anything wrong either, but two fried chips tells me I'm probably incorrect. I left it turned on at my bench, and I'll see if I've fried another one tonight. If so, I'm ripping off the voltage regulator.
No but's about it. It is the only way we can see enough detail to help.
Never ever put a tantalum capacitor direct,y across a power supply. When they fail they fail short circuit and will cause a fire. That is why UL forbid their use in that position.
Yes I would say so. But without a schematic we can't tell what it is.
Are you telling us that the link you posted (I haven't read it) contains the schematic of the whole project you are trying to make? You know, the one that keeps blowing up processors?
The schematic I have, as I have already explained, although similar, is not the same as what I have on the breadboard. My schematic would be of little use in trying to figure this problem out, and your continued insistence on having a schematic that doesn't exist is aggravating and ridiculous.
I shared a link that explains the workings the module in question, along with a schematic for that module. If you can't understand that I have connected that module to power, and the sda/sck lines of the chip, and only just prior to taking the picture did I have even capacitors in front of the module, you are probably not the person that should be trying to help understand the problem.
In what way does that make sense? A schematic would be exactly the same as what is on your bread board. I don't know what you think a schematic is.
Either you don't understand what a schematic is, or you don't know how to fault find on a circuit, probably both.
However I do see your reluctance to actually even try and learn anything. After repeated requests for a real schematic and your frankly pathetic response to this.
Therefore I will no longer respond to this thread and will put the reporting of this thread to muted, so I will not see any more of your muddled excuses.
Best of luck with your project.
And best of luck to anyone else tempted to try and help you.
HI,
PLEASE reverse engineer your PCB, it is essential to troubleshooting your project.
If it is different to the protoboard circuit, then it as a VERY GOOD reason to supply a revised schematic.
An image of a pen(cil) and paper drawing will be fine.
Thanks.. Tom....
PS. Reverse engineering may show you your possible wiring fault.
