0
\$\begingroup\$

First of all, I'm a real noob at this, I was driven by a need I'll explain afterwards, but I'd like to start with the general question. This is exactly what was done in previous question (MOSFET with *actual pin* for body), but the answers don't really address my issue (or I cannot understand them).

I seem to understand that a MOSFET is actually a four-terminal device, where the potential between gate and bulk establishes the field that enables current to flow in the channel between source and drain, and it could be used like that (here my confirmation) but... it is (or rather has become) almost impossible to find a MOSFET that has been manufactured without linking two terminals (say gate and source) making four independent pins available (Why are discrete four-terminal MOSFETs so hard to find?)

My idea is that, if I could find such a MOSFET I could use it, without further tricks, as a relay. I don't want source nor drain to be connected to the gnd, and I need to have no power consumption if I keep the relay on (MOSFET: When can we not assume that the gate current is 0?). Gate and bulk act like a charged capacitor with no further current flow.

How to use MOSFETs as relays has already been asked (How to replace a relay with a MOSFET? and How can I replace a relay with a MOSFET Transistor?), but again I couldn't find an answer satisfying my requirements.

So, to sum up my questions:

  1. would my naive idea just work fine? In addition to the confirmation in the link above, further and more specific opinions would not harm!
  2. is it really impossible to find a four-terminal MOSFET (independently from the answer to the first question: it would be anyway an interesting device)
  3. is there a different way to realize what I need? I've looked for "MOSFET relays" and "solid-state-relays", but I seem to understand that they usually do absorb power for just being activated

Thank you!


Ok, I understand that even if I had my 4-terminal relay it woudn't be able to mimic completely the behavior of a mechanical relay with the benefit of close-to-zero absorption when only activated (dissipation while conducting was expected and not an issue). In particular it won't be like a simple switch being open or closed insensitive to applied voltages and thus able to let any kind of current pass. So let me add a couple of details of my purpose, because I still think that in my simple case they could do the trick in the easiest way.

A certain number of MOSFETs would be linked in series (source of the first going into the drain of the second, and so on). Each of them can or not be activated by the same voltage levels (say "vcc" and "gnd"), and they may be waiting relatively long times (thus the no-absorption requirement). Sometimes vcc will be applied at the beginning of the chain and gnd to the end, and if all the MOSFETs are activated current will flow. If I had the source of an intermediate MOSFET shorted to gnd, that would provide a path for the current to flow making all the subsequent MOSFETs useless.

I've read that 4-pin MOSFETs are becoming increasingly hard to find, but maybe it's not yet impossible! I'd like to know if the are still available in any case, but for this particular purpose they'd provide me an easy solution without the need to look for alternatives ("AND" gates?).

\$\endgroup\$
4
  • \$\begingroup\$ Is your application for simply switching DC signals or is it for AC signals (that is both positive and negative swinging voltages) ? This will have a major impact on the solutions that are available .... Also, does there need to be isolation between the relay's active terminals and its switching terminals ? \$\endgroup\$
    – citizen
    Oct 27 '20 at 12:29
  • \$\begingroup\$ Don't forget that even a simple mechanical relay will have some degree of power consumption, either in its NO or in its NC position. A solid state relay will most likely consume 100 - 1000 times less power ... But NO power consumption is most likely not going to be possible, rather, very small power consuption is more realistic. \$\endgroup\$
    – citizen
    Oct 27 '20 at 13:01
  • 1
    \$\begingroup\$ You might want to say something about the voltages and currents you are interested in. 4 terminal MOSFETs are available, but usually at low power. \$\endgroup\$ Oct 27 '20 at 13:50
  • \$\begingroup\$ @BrianDrummond You're right, I should have told: vcc would come from a battery (details to be decided), so it may be between 3 and 12V. Once it's decided, it will always be the same and with the same polarity \$\endgroup\$
    – lesath82
    Oct 27 '20 at 13:57
1
\$\begingroup\$

I think what you are suggesting is that you use a MOSFET as the contacts of the relay, allowing current to flow in either direction. That is commonly done in VLSI design using MOSFETs. The key terms include "pass transistor" and "pass gate" or "transmission gate". You might also find some optoisolated MOSFETs that will do what you want.

The problem with doing what you want to do is controlling the gate voltage so that the relay is "open" or "closed" regardless of what voltages are applied at the source and drain. That requires circuitry, and circuits consume power. If you use one of the optoisolators then you need to provide power for the LED. It is also true that every MOSFET will consume some power itself when passing current, because the resistance from source to drain is not zero.

\$\endgroup\$
1
  • \$\begingroup\$ Thank you, this is useful information. Hoping in further insight, I'm adding details to my question! \$\endgroup\$
    – lesath82
    Oct 27 '20 at 13:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.