I've come across some devices that state that power supply must be 3.3V but data pins are 5V tolerant and no level shifter is needed, e.g. ESP8266 and nrf24L01+.
When attaching an nrf24L01+ to a 5V Nano, I connect the 3.3V pin to V+ and GND and pin 9-13 to their respective counterpart of the nrf. The data pins get 5V through the Nano directly and all works fine.
Then I soldered a little PCB "breakout" board to attach the Nano and nrf24L01+. I checked the continuity between Nano and nrf24L01+ for all used pins, i.e. GND, V+, 9-13. On both devices I attached the DMM to the solder point on top of a pin. All pins are fine, no shorts/cross connections.
However, when I power the Nano--after it finishes its boot process--the measured power on the 3.3V pin rises to 5V. I suspect that's the moment the data pins were activated and the 5V of the data pin raises the power.
Some notes that might be relevant. First, the power rail on the PCB is attached to some pullup resistors that are connected to yet unused pins. Second, I added capacitors between power and GND rail. Third, there are no other components attached to the circuit, such as transistors, LED.
I'm puzzled because I cannot find any short/cross connections. I also checked/exchanged the Nano and the nrf24L01+ but all modules are fine.
Any idea what I may overlook here? Maybe it's not an unwanted connection between wires but a missing connection? It wasn't that hard to solder 9 pin pairs together through wires.
Sorry that I do not have a picture of the soldered PCB but the wires are not distinguishable.
That the 3V3 device like most other devices have ESD diodes connected from the inputs to the rail. This is a well known effect and is called parasitic powering. It is a bad thing, do not let it happen.
the measured power on the 3.3V pin rises to 5V.
the measured voltage on the 3.3V pin rises to 5V. Power and voltage are not the same thing.
Grumpy_Mike:
That the 3V3 device like most other devices have ESD diodes connected from the inputs to the rail. This is a well known effect and is called parasitic powering. It is a bad thing, do not let it happen.
the measured voltage on the 3.3V pin rises to 5V. Power and voltage are not the same thing.
But the input pins on the nRF are 5V tolerant, so thus do not have ESD diodes to Vcc...
Thanks for the replies. And sorry for the wrong word usage, I was talking about voltage. As an ESL-user I appreciate that kind of feedback.
I've attached a circuit diagram that should be easy to follow. It's still Fritzing in lack of an easy to use offline alternative. The upper circuit is the PCB breakout (not working), the lower is the version with direct wire connection (working). I've switched back and forth between the two settings, PCB not working, wires still working. Note that there is no sensor attached to the 4-pin female header but if all is working there will be.
There must be a self-made failure in the PCB soldering but it's intriguing that I cannot find it with my DMM. One thing that came into my mind right now: I have set my DMM to 200k Ohm, hence, I would have missed higher resistances. I will measure again ASAP.
In regards to the ESD I've no idea what the consequences for my design are. Because when I use wires directly, all is fine--and the data pins are still 5V compared to 3.3V power rail.
I had some time to solder a new PCB board from scratch. This time I added nothing but a HT7333 power regulator, which is fed by the 5V from the Nano. First, the nrf24L01+ did not work. But after I changed the wires that connected the nrf and the PCB pin-header I could receive data.
At least it proves that my basic layout is working and I can solder such a board correctly
Will get back with resistance measures of the other board ...
jremington:
Pencil and paper are offline, and work much better than Fritzing.
Is it really hard to follow the Fritzing diagram I attached?
And I wonder if PowerPoint would have been really easier to follow.
However, I've installed ExpressPCB but I do not find it easy and intuitive. Have to take more time to test it.
Is it really hard to follow the Fritzing diagram I attached?
Yes - I think you fail to realise this because you can't read a schematic and have no idea what that involves.
However, I've installed ExpressPCB but I do not find it easy and intuitive.
It is not. You should not be trying to drive that if you can't draw a schematic. Just draw it by hand on paper and post a photograph. Only when you know what you are doing is it worth getting a program to help you.
Grumpy_Mike:
Yes - I think you fail to realise this because you can't read a schematic and have no idea what that involves.
I'd like to understand what information are missing in the Fritzing diagram that are relevant to understand the setup. Of course it is no electronic schematic but I thought it's about visualizing what circuit someone build.
It is only a physical layout diagram. A schematic has:
Pins with pin names on ICs
a signal flow normally from right to left
Ground at the bottom High +ve voltage on the top
Components grouped as to their function not their physical location
Simple interconnections, fritzing schematics look like a bowl of spaghetti
You can have any component you can draw, you don't have to use substitutes parts that are misleading.
But finally
7) In order to "see" the circuit you have to convert it to a schematic in your head. That is simply too complex and error prone to cope with in all but the most trivial of situations.
Thanks for the explanation.
Obviously pen and paper wouldn't help me because I do not know the rules a schematic is build upon.
Nonetheless, I find it really interesting that, for experts like you, it seems to be difficult to convert a Fritzing diagram like mine in your head to a schematic.
Nonetheless, I find it really interesting that, for experts like you, it seems to be difficult to convert a Fritzing diagram like mine in your head to a schematic.
Once you can read a schematic you will see why.
It is simple enough to go in the other direction, from schematic to physical layout in your head. We do that all the time when we make stuff. This is because you only have to do it one node at a time. A node is a point where two or more components meet. With a layout you need to find the nodes and then hold that in your head while you find the next node.
The normal design process is to draw a schematic then make it. When giving help we need first a schematic to see if their are design errors, then a fritzing layout or better still a photograph will allow us to check if the schematic has been realised correctly.
