Can someone help me find a relay the works this way or a workaround? (EE design proj)

General Questions
#1

I am doing a design project for a EE class and am having a hell of a time finding relays to do what I need. I am pretty sure relays are the only way to do what I need. Here is what I am trying to do:

In a nutshell I have 4 pins on a circuit. I need to be able to put the 4 pins into 2 different states, all at the same time. I need one state to short all 4 pins together, to make them act as one giant node essentially. I need the other state to send each pin into another circuit independently of the other 3 pins. So all 4 pins are connected to another circuit at the same time, but are not connected to each other.

There will be as high as 1.5 or so amps traveling thru the nodes when they are shorted together. I had thought about using some FETs to short the pins together, but with the high current going thru them I don’t want to risk smoking the entire circuit. I see relays as the only option here.

I think that I need 4pin double throw relays. 4 pin out on both sides and one 4 pin input. I have looked all over for something like this but cannot find any. Power useage is not a concern here but price is. I have a $400 budget for this project and there are many other things involved in the project that are eating up money. I am pretty sure this is the most expensive part of the design though because I need 16 of these relays, unless there are 8 pin double throw which I doubt would be affordabel even if there were.

Hopefully I am clear in my question. If I am not please ask me if I left out critical info. If there is another way to do this without relays I am all ears. I just cant think of another way.

#2

1.5 A isn’t much. You could use BUZ71A or MTP3055E, which have Id[sub]max[/sub] of 12 A, and are pretty common and cheap.

#3

OTOH, I can get 3 A rated 4PDT relays for less than US$5 each retail, so even 16 of them won’t kill the budget.

#4

Hmm. The BUZ71A looks like it could work. Can you answer another question for me though.

Another concern about using any sort of FET or BJT to short the 4 pins out is that I need to make sure that, when the pins are shorted, that anything happening in the circuit connected to the 4 pin supernode wont create biasing problems with the shorting FET/BJT.

Is there a way for me to use a FET/BJT to short the 4 pins together and not have to worry about changing the biasing of the shorting FET/BJT under any circumstance? Basically I need whatever is happening on the circuit that the supernode is connected to to not change the biasing of the shorting FET under any circumstance.

Does that sound doable with a FET or do I have to go with a relay?

#5

What’s the voltage?

#6

Do you have a link to that by chance? I am looking around on google now and I have found some, but so far they are all about 8-10 a piece for less than 50.

The maximum voltage that could be at the supernode at any time is +/- 15 volts.

#7

Not really, unless you use a parallel arrangement of NMOS and PMOS FETs.

I’m in Australia, but here you go:

Jaycar Electronics

#8

Thanks. I found some at allelectronics as well. I was able to find some for $4 a piece new, or $2.25 a piece used.

How do you feel about used relays? I am thinking I could get 20 of the used ones for 45 bucks and that would give me 4 to screw around with or smoke. This is only for a school project so longevity of the relay is not too important.

Also, using relays, another concern, though small, is driving all 16 relays at once using digital logic. Right now it is looking like I am going to use a 4 input NAND gate to take 4 states from elsewhere in a circuit, and when they all match switch states on all 16 relays at once.

I am thinking that a single NAND gate may struggle with that so I am going to try to send the output of the NAND gate into a SPT relay and have that drive the 16 relays. Does that sound neccessary or do you think I can do it straight from the logic?

FTR the relays I found are 12VDC 4PDT cube type. 160ohm coil. Thats about 75mA per relay when the coil is charged.

#9

Get the used relays. They either work or they don’t.

There’s just no way you can drive relays from CMOS logic (and you’d struggle even using HTTL logic).

Use the logic to drive one power MOSFET, e.g. a MTP3055E, and use the MOSFET to drive the relays.

Make sure you connect a reverse biased diode (like a 1N404) across the relay coils.

#10

Typo: Should have been 1N4004.

#11

I found a 5V 36ohm DPST normally open relay that can handle 5 amps at the contacts. I figure that the 16 relays will need 1.2 amps total to drive them all. I am thinking that if I send all 4 mux states to an AND gate, maybe a power FET, which will output a high to the 5V relay mentioned above which will send 12V to the 16 relays at once, engaging them all at one time. The 5V relay is rated up to 5 amps so that should be good.

Does this sound workable? I am trying to make sure of this stuff before I go and order these parts.

Thanks for your help!

#12

You don’t need to interpose the 5 V relay. Like I said: Drive the MOSFET from the logic, and use the MOSFET to drive the 12 V relays.

The two MOSFETs I mentioned earlier require a (worst case) V[sub]GS[/sub] of 4 V, so if you’re using 5 V logic, there’s no problem.

#13

And on second thought, the 1N4004 should be something bigger, like a 1N5404.

#14

Helpful hint: If money is a problem many manufacturers will send samples. I’ve often called a manufacturer asking for distributors for their product and when I mention only needing a few they’ve sent “samples” at no cost.
[/hijack]