Analog multiplexor help

Not connecting E and Vee, for starters. Read the datasheet.

John

Hi, Jan. I'd guess you're talking about the 74HC4051 mentioned in an earlier post.

As Mr. Larkin said, the datasheet usually tells all. At least, for those with a basic electronics background who know where to look.

Since you're a newbie, though, you might want to check the website of the place where you purchase your ICs. Many have links to the IC datasheets available on the page where you see the price and availability.

One source for the HC4051 is On Semi -- here's a link to their datasheet:

The 4051 is a curious chip, with an analog section and a digital section. It's made so you can switch ground-referenced AC signals that go below GND. To do that, you have to provide a positive and negative analog voltage which is higher and lower than the maximum and minimum amplitude of your signal at Vcc and Vee (i.e. +/-5V analog supply).

However, and I don't believe it directly says this on this datasheet, if all your input signals are between the digital supply rails (like your LM35), you can tie Vee to GND and get away without a dual supply.

One thing it does say pretty clearly is that the "E" is an enable, and it has to be tied low to get it to work. This can be neat if you want to parallel the single multiplexed pins and, say, make a 24-to-1 analog multiplexer from three 4061s. But if you're just using one chip, and you don't want to turn the output off, just tie it to GND and be done with it.

The curious thing is, all datasheets aren't the same. If you look at the Fairchild datasheet:

the information about Vee is specifically mentioned on the first page. Go figure. In days of yore, I always kept the National Semi Logic databooks on the shelf for reference. I generally found them to be the most descriptive. You can now find that set of logic datasheets at the Fairchild site.

Good luck with your project

Cheers Chris

Chris skrev:

Yes

For a test the 8in/out puts are connected to a LED and the input/output is connected to VDC+.

So by setting the 3 signal pins I should be able to control which LED is on, right ?

The signal to the pins is from a picaxe, so its probaly around 4.7V and the supply is about 4.7V too.

I have tied E and Vee to GND now, but it didnt help. When I power the chip, it still sends out power to 2 of the LEDS. I thought that 1) all outputs is off till I select one, 2) only one output in on at a time.

I can disconnect the pin3 that holds the power I want to switch and they keep lit, so its using its own power to lit these 2 leds.

Its Y2 and Y6, the second top pin on both sides.

/Jan

Sounds like you might have a problem on the channel C input. What does it connect to? You cannot disconnect it - it has to be connected to either a high or a low. Same is true of the other pins.

Can you post a schematic?

Ed

ehsjr skrev: > Sounds like you might have a problem on the channel C

Right, it cant be disconnected since it will fluctate ? The 3 pins are connected to digital output pins of the picaxe, so will be 0V og 4.7V when off/on.

I have assembled it on breadboard, but I can draw up a schematic, will post it in an hour or 2.

/Jan

You must post to alt.binaries.schematics.electronic , of course, not sci.electronics.basics. s.e.b. doesn't allow images to be posted. Just in case you weren't aware...

FYI,

Here is my attempt in pad2pad, its a board design instead of a schmatic, but hopefully it gives you the information about connections.

The layers is just there for clarity, and I have omitted the LED connections.

/Jan

It will be "floating" at an indeterminate level, making the result unpredictable.

Good. Temporarily remove the connections from the channel select inputs to the PICAXE, and connect them to either low or high. You can manually test the chip that way by tieing each channel high or low per the datasheet to make sure you get the correct result. If it works properly that way, then the PICAXE is pulsing the C input, or putting it at an invalid level. Don't overlook the possibility of a flaky connection in the breadboard.

Ed

ehsjr skrev:

With a new chip (multiplexor) and the signal wired to GND I get a +vdc on a6 and a1.

Actually nomatter where I test the signal cables, those 2 and only those are active.

Could I have the wrong datasheet? "4051 - 8 CHANNEL ANALOG MULTIPLEXER" is all the data I have on it, and no link to datasheet from the supplier.

the chip reads hef4027bp which seems to be something different ?

/Jan

Of course it's something different, that's why they put part numbers on them. ;-) According to its datasheet that chip is a flip-flop:

Hopefully you haven't killed any output port pins on the PICAXE. From now on, you should probably use 1K resistors between the PIC i/o pins and the logic device being driven. That way if you accidentaly try to drive a pin high and it is shorted to ground, it won't burn out the pin since the resistor would limit the current to 5mA. The resistor will have virtually no effect on things when an output pin is connected to a true input pin.

Anthony Fremont skrev:

My supplier is just one of the electronics shops in denmark, I will try to get the right one.

I dont think the pic is damaged, but even if, the small picaxe is only 3$ :)

/Jan

A 4027 is a dual JK flip-flop - a totally different part. The pins you are calling A6 and A1 are the Q output from one flipflop, and the /Q output from the other.

You need a chip that is marked 4051 (or has "4051" somewhere in its part number).

Do a google search for "4051 datasheet" (without the quotes) - the first link I see is to Fairchild's product folder on the CD4051. All the details on that part are available there.

Note that the letter prefix and suffix around the number will vary between makers, but the number indicates the function of the part.