74HC4053 serial multiplexer issue

Hi everyone,

i am trying to switch the Serial TTL lines from an FTDI 232 chip to multiple arduinos using a 74HC4053 multiplexer

because the multiplexer has 3 switches, i can join the switches together to get 4 different outputs.

problem is it seems to be interfering with the serial transmission.

if i have an arduino connected directly to the FTDI lines, the computer cannot talk to the arduino. if i disconnect the 74HC4053 chips from the transmit and receive lines, then the computer can communicate with the arduino.

so it seems like the 74HC4053 chips hold the line low or high in some way, but they shouldn't do because its supposed to be an open analog switch.

the 74HC4053 have power and the inhibit pin is low, the 3 inputs are high so there should be nothing restricting the serial lines.

does anyone have any ideas?

thanks

Post your circuit if you want the bug identified?

here is the schematic

test.png

That is one seriously confused attempt at a schematic! :astonished:

Your description is not helping either. Just what are you attempting to do, and why?

You have a lone USB-to-serial interface chip. To what sort of Arduinox are you trying to connect it? The UNO and Nano already have a USB-to-serial interface connected to the serial lines, potentially isolated by 1k resistors so an external device with a suitably low impedance can override the RX line. If you are using a 74HC4053 this should occur while the switches are set in the right manner and not if set otherwise.

But frankly, your diagram is just too confusing. :roll_eyes:

I can't see decoupling caps, and there seem to be random "+5V" labels scattered around the diagram - are
you connecting Vee to +5V - that will possibly fry the chips as well as jamming all the signals HIGH. Vee needs to be grounded for single-supply operation, its the analog negative supply.

The switching data paths seem sensible - I think you've just got the supply connections wrong - this has
possibly damaged something - fix this, add decoupling caps for every chip and see if it works.

[ Extra bit of advice - before you incorporate several chips into a complicated circuit like this, test each
one indivdually to make sure you understand it right - and you only endanger one device at a time if
you make a mistake - one small step at a time is often the fastest method to the result. ]

Hi everyone thank you very much for the replies.

firstly sorry about the wording of the original question and the schematic, i was in a situation which i am sure you are all familiar with, getting stuck in a problem for hours until all the datasheets blur together and you forget to eat and your brain stops working.

after a short break and some food the issue was obvious, as Mark T pointed out, i had supplied Vee with a positive voltage. i blindly read the datasheet that referred to Vee as a supply line (which is is) and just gave it a positive supply instead of a negative one. disconnecting the pin and grounding it solved the problem.

the schematic is hard to read but i usually keep my schematics hard to read for everyone but me, it makes it harder to have things stolen, but obviously in this case it didn't help at all.

for a better explanation of what this does:
the FTDI is connected to tri switch 74HC4053 chips which allow the Tx and Rx lines to be switched to different arduinos. that means one usb connector for 4 arduinos. but the actual switching is done by something which is only shown as headers in the schematic, and that is a MicroView. there is a single button which changes which arduino is being programmed and its name is displayed on the microview's screen.

thanks again for the help guys, i should have had a break before i gave up and asked for help.

Would it not have been easier to use a single 74HC4052 for a four-way interface?

Generally it is not necessary - or meaningful - to use a multiplexer to switch outputs as those not switched will be floating. I suspect - but your diagram is just too convoluted for me - that you would be using those two-way switches to switch each Arduino RX either to your serial chip or ground. Maybe not!

In this case, since the RX pin of each Arduino is already tied by a 1k resistor to its own serial chip which is held at "mark", it does make sense simply to switch the common output to the desired device to override that particular on-board serial interface. :grinning:

NeX:
the schematic is hard to read but i usually keep my schematics hard to read for everyone but me, it makes it harder to have things stolen, but obviously in this case it didn't help at all.

[soapbox]
Draw you diagrams so they can be used for trouble shooting, just making them confusing will not deter anyone seriously wanting to copy your circuit.
What you have presented is what I call a "search a word" diagram.
For example S1, S2, and S3, if I wanted to trouble shoot your "schematic" how do I know it is connected in "three" places, without wasting my time scanning your schematic.
The only way is to actually connect them with lines.

Also clean up the text, make it readable and not mixed in with the symbols.

Don't be so lazy, which is what this form of "schematic" promotes, draw a schematic that has supplies along the top of the page or supply symbols, gnd along the bottom or gnd symbols.
Also try to have some sort of signal flow from one side to the other.
If you want a job in the industry, learn to draw readable schematics.
[\soapbox]
Tom... :slight_smile:

Paul__B:
Would it not have been easier to use a single 74HC4052 for a four-way interface?

Generally it is not necessary - or meaningful - to use a multiplexer to switch outputs as those not switched will be floating. I suspect - but your diagram is just too convoluted for me - that you would be using those two-way switches to switch each Arduino RX either to your serial chip or ground. Maybe not!

In this case, since the RX pin of each Arduino is already tied by a 1k resistor to its own serial chip which is held at "mark", it does make sense simply to switch the common output to the desired device to override that particular on-board serial interface. :grinning:

my primary reason for using the 74HC4053 was that i had a whole bunch of them, but you are right, the 4052 looks like a much more compact solution. space was not an issue,

in the end it will go in a device with multiple "cores" that have their own independent supplies etc, this is a way to program them without accidently programming the wrong core.

TomGeorge:
[soapbox]
Draw you diagrams so they can be used for trouble shooting, just making them confusing will not deter anyone seriously wanting to copy your circuit.
What you have presented is what I call a "search a word" diagram.
For example S1, S2, and S3, if I wanted to trouble shoot your "schematic" how do I know it is connected in "three" places, without wasting my time scanning your schematic.
The only way is to actually connect them with lines.

Also clean up the text, make it readable and not mixed in with the symbols.

Don't be so lazy, which is what this form of "schematic" promotes, draw a schematic that has supplies along the top of the page or supply symbols, gnd along the bottom or gnd symbols.
Also try to have some sort of signal flow from one side to the other.
If you want a job in the industry, learn to draw readable schematics.
[\soapbox]
Tom... :slight_smile:

thanks for the tips, in Eagle, all connected paths can be highlighted using the highlighting tool which you probably know, that is how i normally deal with this. i am not sure if i would ever want to do this on an industrial level but you are right its always good practice to make it readable, even if its just me coming back to it after a few years. i don't suppose you have any examples of what you would consider a well drawn schematic?

Hi,
This is about what you have, but can't find where D1,D2,D3 AND D4 connect.

Tom... :slight_smile:
PS, Took about an hour , first draft, could still need some tarting up.

TomGeorge:
Hi,
This is about what you have, but can't find where D1,D2,D3 AND D4 connect.

Tom... :slight_smile:
PS, Took about an hour , first draft, could still need some tarting up.

ok wow that is beautiful thanks!

so i am seeing it is ok to use the 5v symbols for supply where they are needed, they don't need to all go to a 5v rail at the top of the schematic,

what did you draw that in?

the D outputs are diode isolated 5v supplies so when using the programmer all the arduinos are powered, but they themselves can be powered individually without back feeding the others. really those outputs should go to headers along with the Tx and Rx and Gnd (and reset) to the individual arduinos.

my one concern with connecting everything like you have in the drawing is that as soon as you get into a complicated design you end up with a lot of traces all over the place. this design was very basic, not even a micro processor in it so its not a big deal, but if you had something like a computer motherboard, it would be 90% traces with components having to be spread out to make space for the traces in between.

So the proper way is a hybrid of the two. Here is a schematic in which the important data lines are all shown wired, while the various details of the power supply and indicator LEDs require you to "connect the dots". Can't embed it in the post as it is a PDF.

Here is another schematic:

Expand!

which is more like yours but a lot more legible, detailing the more functional circuit elements. It is more usable as it does not require as much "connecting the dots" to determine the function. :grinning:

Paul__B:
So the proper way is a hybrid of the two. Here is a schematic in which the important data lines are all shown wired, while the various details of the power supply and indicator LEDs require you to “connect the dots”. Can’t embed it in the post as it is a PDF.

Here is another schematic:

Expand!

which is more like yours but a lot more legible, detailing the more functional circuit elements. It is more usable as it does not require as much “connecting the dots” to determine the function. :grinning:

thanks for the explanation they definitely are a lot easier to read. so the basic idea is to group the schematic into various sections, like all the components associated with USB in one area, all the components associated with the microprocessor in another area etc

attached is a a new version of the schematic, how does that look?

Hi,
Looking good, just I prefer white background.

Tom.. :slight_smile:

TomGeorge:
Looking good, just I prefer white background.

Note that back when we used the MDA and EGA, we had green or white text on black, but Windoze used black text on white.

Guess what? :roll_eyes:

Still not putting decoupling caps on each supply pin, you seem to have a single 100nF on the whole circuit.

TomGeorge:
Hi,
Looking good, just I prefer white background.

Tom.. :slight_smile:

agreed but when working on schematics white backgrounds can be hard on the eyes!

MarkT:
Still not putting decoupling caps on each supply pin, you seem to have a single 100nF on the whole circuit.

yep well spotted, the microview itself has internal caps so it shouldn't need any on its supply lines, the FTDI has one, but the multiplexers really should have some too.

Every logic IC needs decoupling, whether the datasheet mentions it or not, else unreliable behaviour is
possible.