Solid state alternative to relay for switching AC current?

abenn

Senior Member
I'm familiar with the standard transistor and MOSFET interfaces for switching DC current, but I'd like to know if there's a simple equivalent -- other than a mechanical relay -- that can switch DC current. I'm looking at a model railway application, and would need to switch approx 16v AC at up to 5 amps or so, using an 08M2.
 

BESQUEUT

Senior Member
I'm familiar with the standard transistor and MOSFET interfaces for switching DC current, but I'd like to know if there's a simple equivalent -- other than a mechanical relay -- that can switch DC current. I'm looking at a model railway application, and would need to switch approx 16v AC at up to 5 amps or so, using an 08M2.
Statics relays are featured for 24V AC minimum...
 
Last edited:

johnlong

Senior Member
Hi
I have being using sharps S216S02F for switching the AC voltage through a auto transformer (140,170,190,230) for 2 years with an 18m2
They are very simple to use just conect picaxe pin through a ressistor (I used 220ohms) directly to the + pin and - to ground I put a 270v MOV across the AC pins to supress spikes job done. You could use these and drop the AC down to your 16v through a 240 in 16 ACV out transformer. They are zero crossing with built in optocoupling so a simple circuit just Plug and Play and like I say 2 years with out a clitch
sharp relay.jpg

regards john
 

abenn

Senior Member
Thanks John, that's a new type of device to me, but it sounds like just the job. My AC is already at 16v from a transformer, so no further drop is required. I'll look around and see if there's a lower-voltage version at, hopefully, a lower unit price!
 

Jeremy Harris

Senior Member
Search for "Solid State Relay". There a thousands around and all that I've ever seen have had a Picaxe-compatible, opto-isolated, 3 to 5 V connection.

Finding one that works down to 16V AC may be a challenge, but I suggest having a hunt around all those listed by the common suppliers in your area (Farnell, RS Components, Mouser, Digikey etc), as they usually have good parametric component search capabilities in their online stores.
 

abenn

Senior Member
Thank you Jeremy. I'm puzzled by your statement about finding one that works down to 16v AC -- this one, for instance, quotes a maximum of 2.5A per MOSFET (i.e. 5.0A if you parallel the outputs for use with DC), but doesn't quote any minimum voltage so far as I can see.

Anyway, as you say, there's plenty to browse through now that I know what search terms to use.
 

Circuit

Senior Member
Thank you Jeremy. I'm puzzled by your statement about finding one that works down to 16v AC -- this one, for instance, quotes a maximum of 2.5A per MOSFET (i.e. 5.0A if you parallel the outputs for use with DC), but doesn't quote any minimum voltage so far as I can see.

Anyway, as you say, there's plenty to browse through now that I know what search terms to use.
The ASSR 1611 is excellent for model railway purposes - I have loads of them doing just that; would you really need 5 amps? I think that the 2.5amp limit continuous run is usually quite sufficient for model railway purposes.
 

abenn

Senior Member
MPep, I get "The requested page could not be found" when I click on your link. But I'll browse the "All about circuits" forum later.

Circuit, I agree that 2.5A is probably enough, especially as my layout is N-gauge with a maximum of, say, four trains running at once -- I put 5A in my opening post just to get things going. At the moment I've got 08M2s driving servos to switch my points, and I'm pondering the idea (and cost!) of using solid state to switch the frogs, instead of the Peco micro switches I'm currently using.
 

Jeremy Harris

Senior Member
Thank you Jeremy. I'm puzzled by your statement about finding one that works down to 16v AC -- this one, for instance, quotes a maximum of 2.5A per MOSFET (i.e. 5.0A if you parallel the outputs for use with DC), but doesn't quote any minimum voltage so far as I can see.

Anyway, as you say, there's plenty to browse through now that I know what search terms to use.
Some have a threshold voltage for the zero crossing point that can be a few volts, as they are intended for use at high voltages where this doesn't matter, that's all. I've only ever used them for switching mains from a Picaxe, where they excel.
 

hippy

Technical Support
Staff member
Thanks for spotting the error. I've corrected that in the earlier post and in the above. It can be tricky copying shortened URL's around in the forum.
 

abenn

Senior Member
Thanks for that correction, and to MPep for the link. The article is very interesting, and confirms what I suspected, that maybe two MOSFETs can do the job.

The ASSR 1611 looks like it's two MOSFETs in a single package, so should be more convenient.
 

MPep

Senior Member
MPep, I get "The requested page could not be found" when I click on your link. But I'll browse the "All about circuits" forum later.
Strange, loads fine for me using Chrome.

NB Never mind, sorry about the ], hadn't realised it was there. :(


A solid state unit is MUCH more convenient, able to be driven from DC directly.
 
Last edited:

Circuit

Senior Member
Circuit, I agree that 2.5A is probably enough, especially as my layout is N-gauge with a maximum of, say, four trains running at once -- I put 5A in my opening post just to get things going. At the moment I've got 08M2s driving servos to switch my points, and I'm pondering the idea (and cost!) of using solid state to switch the frogs, instead of the Peco micro switches I'm currently using.
The ASSR1611 is excellent for frog switching - the only issue is that you need two of them and now you are running into cost considerations compared with a microswitch. I bought a tube of fifty from Farnell to gain an optimum price point - the next useful bulk reduction occurs at 500 units (http://uk.farnell.com/broadcom-limited/assr-1611-001e/ssr-2-5a-60v-dip6/dp/1708441?iscrfnonsku=true). Nevertheless, if you have a lot of turnouts to wire up the cost is not minor. What I really like about these devices is that they can be driven quite safely directly from a PICAXE and because they are optocoupled they offer a high level of immunity from all the spikey noise of a model railway track.
And as for 2.5A being "probably enough"; what brand of n-gauge loco are you going to drive across the turnouts? :)
Seriously though, I think the ASSR1611 is the best such opto-relay for model railways - the choice is actually quite limited.
 

abenn

Senior Member
Thanks Circuit. I have 20+ sets of points and, as you say, I would need two ASSR1611 for each, so it gets quite expensive. What I'm thinking of, to avoid shorts which happen if the micro switches are slightly out of sync with the point blades, is a program sequence which would disconnect the frog, actuate the servo to move the points, then connect the frog to the appropriate rail. But that will still cause a short if there's a mechanical disconnect between the servo and the points. But I'm going to experiment anyway :)

I run GF class 66, GF class 37, a Peco HST, a Peco 0-6-0, and a Peco Hall. I have a ring-main with feed at every track joint, but nevertheless there's the potential for the total current used by four locos to be passing through any particular point in the track, isn't there?

I was looking at RS, but I've just checked Farnell and they're not too bad at £2.64 for 25+, compared with about £1.70 for a single Peco microswitch.
 

abenn

Senior Member
Update, and thanks.

Thanks to everyone for their input, and especially to Circuit for confirming that the ASSR1611 is ideal for my intended use. I've now got a prototype 08M2 circuit up and running on my layout, driving one points servo, switching the AC supply to the frog, and switching the LEDs on my panel. I think the convenience of solid-state outweighs its cost disadvantage compared with mechanical switches.
 

rossko57

Senior Member
But that will still cause a short if there's a mechanical disconnect between the servo and the points.
I would think on a model railway you'd find it essential to use supplies/controllers that deal sensibly and safely with track short circuits anyway. So a short under fault conditions, due to points or logic failure, would be acceptable?
 

abenn

Senior Member
Yes rossko57, my controller automatically shuts down the power in case of any kind of short, such as derailment, points set wrong when loco tries to cross them, etc. I can't remember what the response time is, but it's nanoseconds.
 

westaust55

Moderator
Another device that yo0u may wish to consider is the Si4500BDY - a complementary MOSFET Half-Bridge (with N- and P-Channel transistors)
http://www.vishay.com/docs/72281/72281.pdf

Tie the two gate pins together and you will then drive the output to the positive of negative/0V rail.
I have used some (in pairs) directly driven by the PICAXE 08M2 as a H bridge to control small DC motors.
The N channel side can handle around 6 Amps continuous and the P channel side just over 3 amps.
 

westaust55

Moderator
I would think on a model railway you'd find it essential to use supplies/controllers that deal sensibly and safely with track short circuits anyway. So a short under fault conditions, due to points or logic failure, would be acceptable?
For my own Model railway, I use auto style festoon lamps that limit the current to 1 Amp in the event of any "short" arising from derailments or entering points set the wrong way.
Only exception is a diamond crossing where there are "Frog Juicers" that auto switch the polarity of the two frogs in the event of a short for which the festoon type lamps limit the current to around 1.75 Amps to give the electronic frog switching circuit a change to operate.
 

abenn

Senior Member
Another device that yo0u may wish to consider is the Si4500BDY - a complementary MOSFET Half-Bridge (with N- and P-Channel transistors)
http://www.vishay.com/docs/72281/72281.pdf

Tie the two gate pins together and you will then drive the output to the positive of negative/0V rail.
I have used some (in pairs) directly driven by the PICAXE 08M2 as a H bridge to control small DC motors.
The N channel side can handle around 6 Amps continuous and the P channel side just over 3 amps.
Thanks westaust55. I've just checked my regular supplier's web site (Farnell) and they show it as discontinued. Anyway, as was pointed out by Circuit, 2.5A capability should be sufficient for my N-gauge layout . . . . and I've just finished designing my PCB to suit the 1611, and there's more pins on the 4500!
 

The bear

Senior Member
@ abenn,
RS,(UK) stock an alternative. Si4532CDY, N- and P-Channel 30V (Drain-Source) MOSFET Data Sheet.
787-9020
Vishay
SI4532CDY-T1-GE3
Brilliant service.
 

abenn

Senior Member
Thank you The bear. The price difference between them and the 1611 makes them well worth looking at.

Please bear with my limited understanding of electronics theory, but am I right in thinking that they'll operate as an AC relay if I connect the outputs D1 and S2, and the outputs D2 and S1 each in parallel? But G1 and G2 need + and - voltages between 1v and 3v?
 

westaust55

Moderator
@ abenn,

In retrospect I think that the Si4500BDY and/or Si4532CDY will not achieve your goal.
Had the power source been just DC and a square wave relative to a 0 Volt line then yes, either device would work to enable you to switch the outputs (Drain) connections for the point frogs between the positive and 0 volt rails.
However as an A.C. signal where the polarity is reversed for the second half of the square wave they will not function as the transistors will only conduct in one direction.

Apologies for the misguidance.
 

abenn

Senior Member
No problem :) I'm learning a little every time I come onto this forum. My thought was that, since each package effectively contains the equivalent of a PNP and NPN transistor (or so I thought) they could somehow be configured to pass current in either direction.

The 1611 is a simple enough solution, so I'll stick with that for the moment.
 
Top