Go Down

Topic: Can ground be shared amongst 2 voltages 3.3v and 5v (Read 313 times) previous topic - next topic

junebug12851

So quick question, Can a microprocessor which specifically cannot tolerate voltages above 3.6v tolerate a shared ground between a 3.3v and 5v system. I imagine not but I've read online in a few places that maybe so but online is somewhat vague so I want to be double sure before I blow up a microprocessor trying it.

From what I can gather online, ground is different and ok to share as long as the voltages aren't in the extreme differences but I can't find a clear answer and it doesn't feel right to me for obvious reasons so I want to ask here and hopefully get a clear cut answer that which I can't find elsewhere online because destroying a PCB board, components, and microprocessors isn't on my agenda of things to do today unfortunately.

larryd

If you are saying you have a 3.3v Arduino the GND pin on the board can be connected to a 5v sower supply GND, the 5v line going to Vin of the Arduino to power it.

The same 5v line can the be used to drive a 5v relay which is driven by a 3.3v Arduino out pins via a transistor.

5v signals from one system can be sent thru level translators to a 3.3v system.

3.3v to the 5 volt system should work without translating.

Other than this, don't understand what you are asking.



No technical PMs.
If you are asked a question, please respond with an answer.
If you are asked for more information, please supply it.
If you need clarification, ask for help.

DrDiettrich

How many power supplies do you use?

Usually earth gives a common ground potential (reference voltage). But without a separate neutral wire in mains outlets, or with 2 pole connectors, a voltage can exist between floating grounds of multiple devices.

To be sure you can measure the voltage between two ground wires before connecting them. Or connect the grounds using a 100R resistor and measure the voltage on that resistor, giving the compensation current between both devices. Apart from high voltages or currents, which indicate wrong choice of a ground wire, low currents can result from ground loops. Ground loops are very nasty, because even if broken the voltage between device grounds can vary, often depending on some device load.

Wired communication between devices usually requires a common ground. Use the resistor trick before connecting grounds. Use common power supplies when possible.

super7800

it is common in DC circuits to tie all the grounds of the design together regardless of the voltages (within reason), unless the circuits need to be isolated. on a pcb design oftentimes there is an entire layer dedicated to grounds.

to sum it up, it is 100% correct to connect all the grounds together, as long as there is only one main electrical feed in.


i.e. 12vdc adapter  => 5v regulator => device
                            => 3.3v regulator => device

but generally dont tie all the grounds together when:

i.e. 12vdc adapter #1  => 5v regulator => device
      12vdc adapter #2 => 3.3v regulator => device

hope this helps

junebug12851

#4
Aug 23, 2019, 07:41 pm Last Edit: Aug 23, 2019, 07:45 pm by junebug12851
I have a 3.3v microprocessor on the same board as a few other components all 3.3v and I'm wanting to connect the 3.3v systems GND to the GND of a 5v system with it's own 5v components.

@super7800

Thank you for your answer! My board is getting power from an external 5v system so I used the quick and cheap method of using a resistor pair to shift the level down from 5v to 3.3v. The external system providing the 5v power also provides a ground that I directly connect my 3.3v system GND to as-is (making path of least resistance, hence no level shifters or protection from a 5v system)

So there is only 1 positive point (level shifted) and one negative point (unprotected) in/out and I'm gathering from your post it's A-OK because despite the voltage differences between the two systems, connecting the two GND is fine as long as theirs not two different sources of power which, in my case, theirs not, only one and it's coming from the 5v system.

@DrDiettrich

just saw your post and missed it in my initial reply, sorry. I'll keep it in mind to try before assuming it's safe. Thank you.

Railroader

Without a common GND, how will the 5 resp. 3.3 volt system have any idea about the voltage of the other? GNDs should be connected in 1 unic point to awoid noice problems. GND is, like already said, the reference.
Use Your knowledge. If that's not enough, look for education.
Having knowledge, think outside the box to gain more of it. Only trains run like the train, on the rails. The rest run between the rails.

bluejets

I have a 3.3v microprocessor on the same board as a few other components all 3.3v and I'm wanting to connect the 3.3v systems GND to the GND of a 5v system with it's own 5v components.

@super7800

Thank you for your answer! My board is getting power from an external 5v system so I used the quick and cheap method of using a resistor pair to shift the level down from 5v to 3.3v. The external system providing the 5v power also provides a ground that I directly connect my 3.3v system GND to as-is (making path of least resistance, hence no level shifters or protection from a 5v system)

So there is only 1 positive point (level shifted) and one negative point (unprotected) in/out and I'm gathering from your post it's A-OK because despite the voltage differences between the two systems, connecting the two GND is fine as long as theirs not two different sources of power which, in my case, theirs not, only one and it's coming from the 5v system.

@DrDiettrich

just saw your post and missed it in my initial reply, sorry. I'll keep it in mind to try before assuming it's safe. Thank you.
Using resistors to divide supply voltage down is not good.

junebug12851

How would you recommend I lower the voltage then, I'm fairly new to electronics.

The issue is I have 32 pins I'm interfacing with and all are to a 5v system. The supply and RST is 5v and the data lines in and out are all 5v based TTL. The solution I need to use is also on a tiny PCB board that's somewhat cramped.

If resistors is a bad way to do this then how do I do I convert ~22 pins (excluding ground and output-only lines) to 3.3v on a tiny cramped pcb and keep it low cost.

I've seen some other solutions but I'm not sure what else to use, if you have better feedback or suggestions I would greatly appreciate feedback on this as it's been a huge problem I'm trying to figure out.

larryd

No technical PMs.
If you are asked a question, please respond with an answer.
If you are asked for more information, please supply it.
If you need clarification, ask for help.

Slumpert

how do I do I convert ~22 pins (excluding ground and output-only lines) to 3.3v on a tiny cramped pcb and keep it low cost.
22 converter lines or simply use a 5v MCU... 


Grumpy_Mike

Quote
Using resistors to divide supply voltage down is not good.
I think she wants to reduce the signal from a 5V system to feed into a 3V3 system. For that using resistors is fine. However keep the resistor values low, that is between 500R and 2K.

Connecting the grounds together is not only possible it is compulsory, if you want things to work.
See this http://www.thebox.myzen.co.uk/Tutorial/Power_Supplies.html

bluejets

#11
Aug 24, 2019, 09:47 am Last Edit: Aug 24, 2019, 09:48 am by bluejets
How would you recommend I lower the voltage then, I'm fairly new to electronics.

The issue is I have 32 pins I'm interfacing with and all are to a 5v system. The supply and RST is 5v and the data lines in and out are all 5v based TTL. The solution I need to use is also on a tiny PCB board that's somewhat cramped.

If resistors is a bad way to do this then how do I do I convert ~22 pins (excluding ground and output-only lines) to 3.3v on a tiny cramped pcb and keep it low cost.

I've seen some other solutions but I'm not sure what else to use, if you have better feedback or suggestions I would greatly appreciate feedback on this as it's been a huge problem I'm trying to figure out.
level shifter for one of your queries, how much space you have is beyond control other than from your design I would imagine.

https://www.youtube.com/watch?v=g3a0X30Le-8

junebug12851

@larryd

Thanks so much for the links, it's quite a lot of information to take in but I'm sifting through it. While I can see a number of solutions I'm still not understanding of why a resistor pair is bad. I'll definitely bookmark the links, thanks for them.

@Slumpert

That's what I've already laid out in the schematic and board design, 22 converter lines using resistor pairs. I drew that out that days or weeks before I posted this question on here.

@Grumpy_Mike

Currently my resistor pairs are 1K and 2K. Should I do something else? In simulations any resistor value I entered 1K/2K or 5K/10K or higher or lower as long as the 2nd number was double the first number it all produced a 3.3v signal from a 5v signal so I didn't think there was any difference what the resistor pair was as long as it produced 3.3v based on simulations.

@bluejets

I know of a device that can linear shift a 5v or higher signal to a 3.3v and that it's not the most efficient method but it's cheap and it works. I also know the arduino series uses it. Is that what your talking about? I understand linear voltage converters only work for supply voltage and not for data lines.

--------------

Thanks for all the helpful answers

Grumpy_Mike

Quote
Currently my resistor pairs are 1K and 2K. Should I do something else?
No that is fine.
Myself I tend to use 510R and 1K but what you use is fine. Just don't use things like 1M and 2M. Or 10K and 20K as they produce more degradation of the rise and fall time as the signals due to stray capacitance.

Slumpert

I drew that out that days or weeks before I posted this question.
So in "all" that time, did you actually research the TTL signal your dealing with?
TTL is almost never 5 volts and may infact only be 3.3 volt as its "HIGH" output.

If it's as low and the minimum required (> 2.7v ), your voltage divider might drop it too far.


Go Up