74hc4052 multiplexer - weird crosstalk?

Status
Not open for further replies.
P

ppd83

New Member
Hi

I have a little problem with a 4052 MUX.
This is the scenario: I have 2 pressure sensors (piezo elements). One is connected to pin 1Y0 of the MUX and the other to pin 1Y1. If I hit the piezo connected to 1Y0 (thus generating a signal) I get a voltage also on pin 1Y1. The same is true the other way around (generate signal on 1Y1 and get signal on 1Y0 as well). This weird crosstalk doesn't show up with the 74HC4053 MUX (same setting).

Any ideas why this happens?

Thanks,
Paul.
 
You can always double scechk that those sensor really isn't conected together some way.

But, post the schematic, otherwise it's impossible to tell what might be wrong. Maybe your 4053 is defect. Have you tried more than one of same type?
 
The schematic is really simple. I tried several 4052, they all behave the same...
 

Attachments

  • sch.jpg
    61 KB · Views: 1,448
Unfortunately I don't have an oscilloscope, but in theory (and in practice) it looks something like in the figure I attached.
 

Attachments

  • wave.JPG
    7 KB · Views: 357
ppd83 said:
The schematic is really simple. I tried several 4052, they all behave the same...
Click to expand...

hi,
You are operating the 4052 outside its specified range, see image.
You will find that some CMOS devices will work just outside of their specification, others dont.

EDIT:
the amplitude and sense of the analog signal has to be within the range Vdd and 0V
 

Attachments

  • AAesp02.gif
    29 KB · Views: 423
Last edited:
No I'm not. I'm using 74HC4052 which has an operation range starting at 2V. What you described refers to the HCT version of the MUX.
Thanks.
 
ppd83 said:
No I'm not. I'm using 74HC4052 which has an operation range starting at 2V. What you described refers to the HCT version of the MUX.
Thanks.
Click to expand...
OK,
But I suspect the analog signal that you are applying is swinging +/- about 0V.???
If thats the case you should have a -2.4V on the VEE.

In the past I have had unexpected results in using these analog gates, they do vary in performance, even when using the same device types.
 
The thing is, I'm not interested in the negative value of the signal, just the positive. As I'm looking at the 4052 datasheet (from NXP btw.) I see that the input is limited to [Vee,Vcc] by 2 diodes so I don't think swinging around 0V should be a problem. Plus, I actually use a bridge rectifier after the piezo so the signal will always be positive. I didn't mention this before because I don't think it's important.
 

hi ppd,
If you are confident that the 4052 that you are testing is OK, you must be using it in some way which does not meet the devices specification.
Having checked the the 74HC device datasheet, which as you say will operate down to 2V. The devices I have the bench are all HEF or CD, so I cannot repeat your test.

How are you measuring the 'leak thru' signal at the output.??
Have you tried adding a resistive load to the output of both gates, say 10k.??
 
I bought 10 pieces of 4052. I tried 2 or 3 of them and all behave similar. I'll try some more when I get home but I really doubt it that all of them are malfunctioning.

I measured the 'leak thru' in 2 ways: with a simple voltmeter and thru the microcontroller's ADC.
With the voltmeter I can only make a rough measurement (i.e detect some voltage but not exactly how much).

With the ADC I can see when the 'leak thru' appears and with what amplitude. I attached 2 graphics of what it looks like (one showing a hard hit of the piezo and the other a softer hit).
The red signal belongs to the piezo that has been hit and the green signal to the other piezo.

I haven't tried adding a resistive load. I'll try that.
 

Attachments

  • hard_hit.JPG
    7.6 KB · Views: 276
  • soft_hit.JPG
    6.9 KB · Views: 273

Your schematic doesn't show the bridge rectifier - how are we meant to guess you were using one. You do also realise that the bridge rectifier will drop around 0.7v of signal dont you ?

You may also want to consider what happens to where the input voltage/current goes when you get a negative peak - yes the diodes are there to limit the voltage but it has to go somewhere and gets dumped into the power rails. If you don't have a big enough load/decoupling to absorb these then you can get some very unexpected results.
 
If you connect a 1N4148 or similar in parallel across the piezo (for testing purposes), you'll get the +ve amplitude with only around 0.4-0.6v of the negative which should be slightly more palateable to the 4052 input.

To be honest with you without looking at what is happening at the inputs/outputs/power rails with a proper scope its going to be a tricky one.

Make sure you have a 100nF between the 4052 power pins for decoupling, try something like a 10k resistor between each input and ground (and possibly output and ground) and see if that helps at all.
 
Last edited:
Ok. I've been doing some testing. It seems that if the amplitude of the input signal is lower than Vcc then the MUX functions properly. So I'll have to use a pair of limiting diodes myself...

Thank you for all your help!
 
Status
Not open for further replies.

Similar threads

Weird switching transformer noise, Audio attached
Replies
11
Views
3K
MacIntoshCZ
Charger for weird battery pack
Replies
5
Views
1K
Nigel Goodwin
double 1:2 switch/multiplexer
Replies
13
Views
2K
kubeek
K
P
4051 Analog Multiplexer Concept Question
Replies
4
Views
2K
Palbert
P
H
Help needed with CD4052 multiplexer input buffers. Something's wrong here...
Replies
8
Views
4K
unclejed613
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…