DPDT Solid-State Relays - Why so hard to find?

[DerStrom8]

I don't understand that logic. Why is analog (like the CD4016) better than digital when dealing with digital signals? Please expand on that reasoning.

John

Quality-wise it may not be. There may be no difference at all. I simply said that I might go with an analog switch because I already found one.

View attachment 93287
One chip: 40 cents

Hmm, I was thinking the NLAS44599 was 30 cents, but I must have been thinking of a different device. The 74ls153 would be a cheaper option, but after starting the design I realized I need two DPDT "switches". Now the question is whether to use one 16-pin device with analog switches on a board with little space, or use two 16-pin devices with digital switches on a board with little space? I forgot to mention, by the way, that I only have one I/O pin available to control both DPDT switches.

 

[Tony Stewart]

None of these will work until you define;

Speed, logic levels, source type; load , including cable impedance & length, EMI levels, need for isolation, crosstalk, Supply source& sequence on powerup

 

[Wade Hassler]

None of these will work until you define;

Speed, logic levels, source type; load , including cable impedance & length, EMI levels, need for isolation, crosstalk, Supply source& sequence on powerup

You and your basics!

.. but right as rain.

 

[dougy83]

The name of the device you seem to be after is called a data selector, or a multiplexer. Here are some 4000-series ICs that allow 4 outputs and 4 pairs of inputs and a control line to choose which input is switched to the output:
- 40257 Quad 2-line to 1-line data selector/multiplexer (tri-state)
- 4551 Quad 2-channel analog multiplexer/demultiplexer
- 4519 Quad 2-input multiplexer (data selector) -- requires additional inverter

 

[BobW]

One point that has not been addressed is whether the data flow is unidirectional or bidirectional. If it's bidirectional, then an analog switch such as a CD4016, or analog multiplexer such as CD4053 is needed. If it's unidirectional then standard gate logic can be used.
I think the 4053 would likely address all of the requirements with a single chip.

 

[dougy83]

I think the 4053 would likely address all of the requirements with a single chip.

Except that it's only three channels, and four channels are required (see post #21)

 

[BobW]

Right. What I should have stated was that a single 4053 will act as a DPDT switch. For two DPDT switches, you'll need two 4053's.

 

[crutschow]

How about a CD4551 (quad 2-channel switch).

 

[DerStrom8]

How about a CD4551 (quad 2-channel switch).

I'm thinking I would need a couple of 4551s, plus a couple of inverters in order to make that work as two DPDT switches. I may have to do something of that sort anyway, but still hoping for a one-chip solution. Then again, the one I already found doesn't require any inverters and it's got both DPDT switches in a single 16-pin package.

 

[dougy83]

I'm thinking I would need a couple of 4551s, plus a couple of inverters in order to make that work as two DPDT switches. I may have to do something of that sort anyway, but still hoping for a one-chip solution.

The 4551 is a 4PDT analogue switch, so no inverters or extra chips needed. Here's a datasheet http://www.cmos4000.com/media/cmos/ic-cmos-4551.pdf
Likewise with the 40257, only a single IC is required. Datasheet here http://www.ti.com/lit/ds/symlink/cd40257b.pdf

 

[DerStrom8]

The 4551 is a 4PDT analogue switch, so no inverters or extra chips needed. Here's a datasheet http://www.cmos4000.com/media/cmos/ic-cmos-4551.pdf

Ahh, that's the 14551. I was just looking at the 4551. Even so, I don't think I would need two chips, but in order to get the control on the 4551 to work properly (to swap the "positions" of two data lines), I would still need an inverter on one of the switches. The 14551 looks more promising (unless I'm completely misreading this, which is a distinct possibility. Since Friday I've been a bit under the weather and my brain hasn't been working very well )

Likewise with the 40257, only a single IC is required. Datasheet here http://www.ti.com/lit/ds/symlink/cd40257b.pdf

That's a thought. Going to have to look into that one....

 

[DerStrom8]

Likewise with the 40257, only a single IC is required. Datasheet here https://www.ti.com/lit/ds/symlink/cd40257b.pdf

Hmm, was looking up this part, in its cheapest form it's a $0.75 device. Hoping for closer to $0.40

Even so, it still looks like a good option. I'll have to speak with the purchasing department and my supervisor when I return to work tomorrow.

 

[dougy83]

Ahh, that's the 14551. I was just looking at the 4551. Even so, I don't think I would need two chips, but in order to get the control on the 4551 to work properly (to swap the "positions" of two data lines), I would still need an inverter on one of the switches. The 14551 looks more promising (unless I'm completely misreading this, which is a distinct possibility. Since Friday I've been a bit under the weather and my brain hasn't been working very well ).

The 4551 and 14451 are the same thing (just from different manufacturers). I linked to the more readable datasheet.

Here's the data swapping circuit:

Get well soon.