Maker Pro
Maker Pro

Need Multiplexer Solution

Hello,

For a project, I need to switch 5V DC through one of eight different
resistors using the digital outputs of a small USB data acuisition
device. I'm thinking of using a 74LS138 3 to 8 line decoder. Will this
work? I need the resistance of the chip to be 50 Ohms or less, as
anything higher will affect the project. I've come up with a design
using a relay network, but if there is a better way to do it with a
multiplexer chip, I really wnat to go that route. The resistances are
50-300 Ohms, and the 5 VDC will be coming from USB power. Thanx.

gm
 
L

Lord Garth

Jan 1, 1970
0
Hello,

For a project, I need to switch 5V DC through one of eight different
resistors using the digital outputs of a small USB data acuisition
device. I'm thinking of using a 74LS138 3 to 8 line decoder. Will this
work? I need the resistance of the chip to be 50 Ohms or less, as
anything higher will affect the project. I've come up with a design
using a relay network, but if there is a better way to do it with a
multiplexer chip, I really wnat to go that route. The resistances are
50-300 Ohms, and the 5 VDC will be coming from USB power. Thanx.

gm

No, the 138 is not going to work for you. The 138 has active low outputs
so the output is one of eight low also, TTL is a current sinking logic
family so the 74LS138 will sink current to ground far more effectively
than it can source 'something near' 5 volts as an output. You could
use the 4022 counter since it has decoded outputs that are active high.
It sounds to me like you really want a switch like the 4066. This old
part has an on resistance of about 80 ohms. Worse would be the 4067
which has a 125 ohm on resistance. There are much better analog
switches but these are in surface mount packages and only one switch per:

http://www.fairchildsemi.com/ds/FS/FSA1156.pdf
or
http://www.fairchildsemi.com/ds/FS/FSA4157.pdf

Watch the power supply on these to insure they will operate on 5 volts,
Have the type of switch configuration you need and have TTL compatible
inputs. The two links above are very low resistance devices... .9 ohms
and 1 ohm respectively. The first is a SPST N.O. switch while the second
is a SPDT. Either will operate on a 5 volts supply. Be aware of the
current
limitations through each switch.
 
Lord said:
No, the 138 is not going to work for you. The 138 has active low outputs
so the output is one of eight low also, TTL is a current sinking logic
family so the 74LS138 will sink current to ground far more effectively
than it can source 'something near' 5 volts as an output. You could
use the 4022 counter since it has decoded outputs that are active high.
It sounds to me like you really want a switch like the 4066. This old
part has an on resistance of about 80 ohms. Worse would be the 4067
which has a 125 ohm on resistance. There are much better analog
switches but these are in surface mount packages and only one switch per:

http://www.fairchildsemi.com/ds/FS/FSA1156.pdf
or
http://www.fairchildsemi.com/ds/FS/FSA4157.pdf

Watch the power supply on these to insure they will operate on 5 volts,
Have the type of switch configuration you need and have TTL compatible
inputs. The two links above are very low resistance devices... .9 ohms
and 1 ohm respectively. The first is a SPST N.O. switch while the second
is a SPDT. Either will operate on a 5 volts supply. Be aware of the
current
limitations through each switch.

Thanks for the help.

If I put my device between the 138 and ground, and the eight resistors
between the "outputs" and the 5V, will this destroy the device? Just
curious. But I think I may have a new solution, which is to use an
octal line driver with the voltage source going from the outputs to
ground. I can put my device between the ground pin and ground. I'll
need eight lines to drive it, but thats OK.
 
J

John Fields

Jan 1, 1970
0
Hello,

For a project, I need to switch 5V DC through one of eight different
resistors using the digital outputs of a small USB data acuisition
device. I'm thinking of using a 74LS138 3 to 8 line decoder. Will this
work? I need the resistance of the chip to be 50 Ohms or less, as
anything higher will affect the project. I've come up with a design
using a relay network, but if there is a better way to do it with a
multiplexer chip, I really wnat to go that route. The resistances are
50-300 Ohms, and the 5 VDC will be coming from USB power. Thanx.

gm

---

You could use eight P-channel logic-level MOSFETS, and an HC138, like
this:

+5V>---+-----------------------------+--[R8]----S D--+---->OUT
| | G |
| +-----------+ | | |
+-|G1 Y7-|O--------------|------------+ |
A2>----|A2 Y6-|O-----------+ | |
A1>----|A1 | | +--[R7]----S D--+
A0>----|A0 | | | G |
GND--+--O|G2A- | | | | |
+--0|G2B- Y0-|O--+ +--|------------+ |
+-----------+ | | |
HC138 | | |
. . .
. . .
. . .
| | |
| +--[R1]----S D--+
| G
| |
+------------------------+
 
Excellent suggestion, thank you very much. Another question: Is there
a digital potentiometer available that will carry a wiper current up to
25 ma continuously? If so, I could replace all those discreet
components with the pot, and get 256 resistor values instead of 8,
which would be nice for this project. The output of my circuit will
draw up to 20 ma of current at the lowest resistor value. I looked at
the AD8400 and it will only allow a max current of 5 ma when the wiper
is set at the lowest (50 ohm) setting. Is there another device similar
to the AD8400 that is designed to carry more current? Thanks for your
help.

gm
 
J

John Fields

Jan 1, 1970
0
Excellent suggestion, thank you very much. Another question: Is there
a digital potentiometer available that will carry a wiper current up to
25 ma continuously?
---
Dunno.
---

If so, I could replace all those discreet
components with the pot, and get 256 resistor values instead of 8,
which would be nice for this project.

---
What would _really_ have been nice would have been if that's what you
said you wanted in the first place. However, if you can't find a pot
you can replace the '138 with an HC594/595 and accomplish the same
thing.
---
 
L

Lord Garth

Jan 1, 1970
0
Excellent suggestion, thank you very much. Another question: Is there
a digital potentiometer available that will carry a wiper current up to
25 ma continuously? If so, I could replace all those discreet
components with the pot, and get 256 resistor values instead of 8,
which would be nice for this project. The output of my circuit will
draw up to 20 ma of current at the lowest resistor value. I looked at
the AD8400 and it will only allow a max current of 5 ma when the wiper
is set at the lowest (50 ohm) setting. Is there another device similar
to the AD8400 that is designed to carry more current? Thanks for your
help.

gm

Here is some reading re digital pots:
http://pdfserv.maxim-ic.com/en/an/AN1956.pdf

Looks like the current is always going to be a problem with analog switches.
Think about the voltage drop across a digital pot...that would be the
current
through the pot times the resistance of the pot. You might have to figure a
way to buffer the signal. You should not allow the signal to exceed the
power supply limits unless you want the magic smoke to appear.
 
R

Rich Grise

Jan 1, 1970
0
Excellent suggestion, thank you very much. Another question: Is there
a digital potentiometer available that will carry a wiper current up to
25 ma continuously? If so, I could replace all those discreet
components with the pot, and get 256 resistor values instead of 8,
which would be nice for this project. The output of my circuit will
draw up to 20 ma of current at the lowest resistor value. I looked at
the AD8400 and it will only allow a max current of 5 ma when the wiper
is set at the lowest (50 ohm) setting. Is there another device similar
to the AD8400 that is designed to carry more current? Thanks for your
help.

This makes it actually sound like you're looking for a multiplying DAC
and buffer.

Is it a programmable _resistance_ you're actually looking for? Or are
you making a divider or something, to give a programmable voltage or
current?

Good Luck!
Rich
 
Top