Creating DPDT Solid State Relay with FETs

Question

I feel like this should be a very simple question but so far I have not found a solution. My goal is to make the following DPDT relay circuit using MOSFETs:

The goal is to connect INX to OUTX or OUTY and INY to OUTY or OUTX depending on the state of an input signal (I have two data lines available for the switch control). Normally I would use a mechanical relay but due to the nature of my project this really won't work, I need a solid-state option. DPDT solid-state relays are very difficult to come by, and any I have found cannot be connected to work in the way shown in the image.

I was looking at using two enhancement mode and two depletion mode N-channel MOSFETs, but since a FET needs a G-S voltage to switch on I would not be able to simply pass the data signals straight through the FETs as they need to be isolated from ground.

So how might one construct a DPDT solid state relay entirely out of MOSFETs that would allow them to swap the two outputs (connecting INX to OUTY and INY to OUTX, or vice versa)?

EDIT: I forgot to mention, I had been looking at using bilateral switches but most of them can only handle about 30mA. These particular data lines can carry up to 200mA, so I'm looking for a ~500mA solution (hence using discrete FETs).

EDIT 2: I was looking at these: http://download.siliconexpert.com/pdfs/2014/7/28/1/10/31/351/aro_/manual/semi_eng_ge2a_aqw21_e.pdf

My thought was to connect them as follows (please pardon the crudity of the drawing, it had to be done in MS Paint, and I didn't want to put detail of the FETs and all in it. Just know that each pair of wires on the output are effectively connected to a switch:

Answer 1

You could combine multiple SSRs to do this. Two form B and two form A will give you a DPDT SSR. Cost would be fairly high (about $20 in singles) but this is really a weird set of requirements. Switching will be slow.

For example, two LCB710 and two CLA230 (IXYS) would handle 700mA.

You could also buy PV optoisolators and use them to drive back-to-back MOSFETs for each switch. Since the former come in pairs you'd end up with 8 parts (plus maybe 4 resistors). Again, slow switching.

Edit: Since you imply in the comment below that there is no requirement for power-off normally closed, the simplest method is probably to use two dual form A SSRs such as TLP222A-2.

Again you could use dual MOSFETs and PV drivers, but I don't see much point in that since the current requirements (and presumably your voltage requirements) are met by common parts.

Answer 2

As the OP said, it is very difficult to find DPDT SSR (solid state relays). But there are a some ways around this; use two SPDT relays, like this one, or use one DPST-NO and one DPST-NC as proposed in another answer, or finally, use two DPST, with an inverter in front of the control for the second relay.

For this particular application, the need to handle relatively high currents (at least 500 mA) with a low on-resistance weere the critical specifications.

I found this DPST C347S SSR which is rated for 1A. 150 mΩ. So you would need two, and use an inverter in front of the second IC's control lead. $8.09 @ in singles, $4.72 in thousands.

Answer 3

I think I found the simplest option: Instead of a "Solid-State Relay" like I kept asking for, what I think I was actually looking for was a DPDT analog switch like the NLAS44599 (Datasheet: http://www.onsemi.com/pub_link/Collateral/NLAS44599-D.PDF). The functional diagram looks like this:

I hope this might help someone in the future who's looking for the same thing I was. Sure, it's a 16-pin device, but it does exactly what I need it to.

Answer 4

this is the digital equiv of an h circuit, fwd rev motor drivers. try the tlp222 optical isolated fet driver, and if the current demands are really high follow the design and use opto isolators to keep one side that is on from inverting the side that is off and reverse biasing the fets. some diodes might come in handy during development

Answer 5

I would approach this as follows:

Use a pair of photovoltaic optos to generate an isolated voltage source.

Connect enough N-channel MOSFETs (back-to-back in pairs) as you need. That would be 4 pairs.

Finally, use opto-isolators to drive the gates of the MOSFETs. You want to use the optos in a tptem-pole configuration so as to minimize the current from the photovoltaic source.

I'll draw up a quick schematic when I get in front of a computer.

This gives you the ability to switch very quickly and handle as much current as you need.

Answer 6

I've done it for a bms using discrete FETs

I've looked at analog switches as well, but the issues I seem to run into is the difficultly in managing the supply and common mode voltages