create a shore-gen-invertor-off switch using solid state parts

I have been wondering about that selector switch to choose between various power sources.
Thing is they have to break before make or you might accidently join 2 power systems together.

They would have to do 4 poles to keep twin 30 amp neutrals and hots separate.
2 black wires and 2 white wires.

Currently I have a 2 pole gen-off-shore switch rated at 60 amps.
I have twin 30 amp power and a polarity warning light and Buzzer.
So I am switching the hots and ganging the neutral together which is not ideal although it works I am sure it does not meet ABCY standards.
If the shore power receptacle on the dock was wired reversed, then plugging in would energize the neutrals, light the light and buzzer would scream. It might also blow the breakers on shore if ground was touching neutral somewhere.

Still that is not truly good enough standard wise.
I would ultimately like to have a 4 pole switch to also add in the invertor to power the breaker panel. It currently powers only the outlets by way of a DPDT switch to chose between invertor and ship-shore power.
It all works well enough.

You can of course pay hundreds for one of those switches.
So can you use triacs to choose AC power sources?

http://www.scribd.com/doc/78872555/29/The-Triac-Used-as-an-AC-Switch

 

Yes, you could. There are different type of Triac’s, what I remember, some have some electronics build in which could be used nicely for your application, like the one with an IR sensor. It means that with a micro or a triple 555 some delays between switching could be achieved. Let me think about it for a while and see whether I can cough up a simple circuit for you. I would switch during zero crossing, wait a few cycles and then switch the next one off or on. I am under the impression that you like to switch the hot and neutral simultaneously. Thus with 4 Triac’s.

 

Have a look whether you can purchase 2 x NTE 56024. or 56026 or 8 Forget for the time being to have the neutral switched also. Get also a 47 Ohm resistor.
Place a 1 Kw element heater between life and MT1 and connect the stud to neutral. (preferable a type of Ohmic resistance as a load, later on we will place a capacitor and a resistor, if you want to use an inductive load. But for the first circuit to start understanding triacs, it is better to have an Ohmic resistance)
Connect the 47 Ohm from MT1 to Gate and your heater will be on.
Connect the 47 Ohm from Gate to stud and your heater will be off.

As soon you have done that we will take you to the next stage. I am going back to watch the Olympics on the TV.

 

If you designed that switch, then it would certainly be worth doing considering that a manual source selector switch of 4 poles and multiple sources is ridiculously pricey.
It might not be easy to control the triacs seeing really, what power source would you use to turn them on and off?
Mechanical switch does not have that issue. Every boat is likely to have 12 V power available all the time. So using battery power to trigger the triac?

This Newmar switch has a cheaper price than the Blue Sea switch.
http://www.newmarpower.com/AC_Source_Switches_Panels/AC_Source_Switches_Panels.html

Model SS-15* is 126 amp with 2 poles
2 + "OFF"
GEN-OFF-SHORE
and may be configured as a 63 amp, 4 pole switch




Such as here, notice that price! And looks to be only 3 poles too.
http://www.fisheriessupply.com/productgroupdetail.aspx?cid=216543&did=6979

 

Lovely switches. I would not easy use Triacs, as you need a number of components to make it work.
4 x TRIACS up to 60 - 100 Ampere, which aren't cheap
4 x IR sensing TRIACS 2,5 Amp like the V23100-S032-B402 or replacement as this item is obsolite
2 x 0,1 UF capacitors
2 x 150Ohm 2 watt resistors
1 x toggle switch with zero point
2 x electrolytic caps 10uF 16 Volt
2 x 270 Ohm resistors 1/4 watt
2 x 2 KOhm resistors 1/4 watt
1 printed circuit board to mount components
1 x box to protect you from hitting the hotwires.

and a 5 to 12 Volt DC source.

It can be done, but is it worth it?
The working.
A DC source charge via the 270 Ohm a 10 uF capacitor when the switch is changed from shore to 110 AC generator
Because the switch will be a double throw switch, at the same time the other capacitor will be shortened and therefore it takes a little while to charge the other 10 uF capacitor before the infra red diode of the triac makes the triac switching the big triac on.
As soon the toggle switch is thrown over the 10 uf capacitor is shortened and now the other 10 uF capacitor takes a little while to charge up and makes the other IR triac switching the 2nd big TRIAC on. Because the 10 uF is shortened, the IR triac is no longer conducting and the first big traic is switched off before the second big triac is switched on.

Is it worth it?? I don't think so.
Bert

 

The short way to achieve what you want is to use SSR's (solid state relays). They have isolated 5-32V inputs, with or without zero voltage switching circuitry and can handle load currents up to 150 Amps. The control current is just 20 mA so you can use any small panel switch with 3 positions (1-0-1).

 

The problem what is worrying him is, the break before make situation. I haven't seen solid state relays which can handle 2 change over positions. Maybe somebody knows of a model number or manufacturer. If he uses 2 solid state relays and does not want to get himself into a make before break situation, one has to add electronics to avoid such situation. Those hand turning switches are still a beautifull solution, but if he is worried about explosion danger, he need to go for solid state. He wants to break mains, both life and neutral and switch this over to a running generator life and neutral also. One needs at least 60 Hz = 8 millisecond time delay. He is worried by manual switching when he switches fast, the spark will create a path for a short circuit, using a manual switch of which he has no technical data on.

 

If you use SS relays, then the break before make will be handled by a type of

on-off-on-off-on type switch.
You need to make 3 different sources of AC power. Shore, Gen, Invertor
Perhaps some type of rotary switch would still be needed.
This is shorting, but need to be non shorting.
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=023-662&utm_source=googleps



Where the switch is located, explosion proof is not critical, meaning it is not in the bilge, it is up on the helm station.

How many different relays would be required to switch 4 poles (wires) between 3 different AC sources? Would each source require 4 relays = 12 relays total? This could get pricey.
relays needed? I am sleepy this AM and have not planned out any scheme yet. Absolutely must not have more than one source connected at a time.

I have read the reason to keep neutrals not ganged together is if you have twin 30 amp power and your drawing significant power from both hots, if the neutrals are joined, it is possible for one of the neutrals to carry more current than 30 amps and this can burn socket plugs. This would be due to one of the returns having a higher resistance than the other. This however so far has never been a problem for me.

I use 8 gauge wire to connect from my shore socket to my panel instead of 10 gauge. My shore cords to the dock are the standard 10 gauge yellow jacket cords.

Alternately, perhaps I could retain the current shore-off-gen switch and add relays to the neutrals only.
I wonder if there is some type of code about switching sources that says the switch must be self contained and not multipart, made of various components.

 

here is a triac 35 amp for $8.95
http://www.jameco.com/webapp/wcs/st...a=658275&productId=658275&keyCode=WSF&CID=GMC

What would be nice is a multipole triac type of device. Still would have to be able to control the triac from a DC battery so would have to somehow create some AC signal?

If you could get triac price under $5 this might be worth playing with.

$7.79 40 amps but have to buy a lot like 25 items
http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_869777_-1

 

Another less elegant way to do this would be to use 3 double pole breakers backfeeding into a standard breaker panel. You would want some kind of interlock handle so only one source breaker could be on at a time. Or just be very careful turning them on and off.

I like my current square D breaker panel but would have enough room to take it out and put in another bigger panel. Likely wont do any such thing.

 

Unfortunately, you may have to forget to do it with Triacs or contacters. Only with a microprocessor can one do it maybe with lots of clever software writing. Sometimes one has to accept that a switch like at your thread No 4 (at the right) is better than electronically with all sorts of tricks. Not even your 3 double pole breakers would be a solution. Maybe except if somebody can come up with a method to prevent a big bang. Should any of the contacters engage at the same time.

I am using my development computer in such a way that I don't have virus protection software nor that it has seen Internet for 3 years since new. I only use new CD disks to transfer something from this computer to the computer linked to the Internet. Never I use any other disc's and connections on this development computer. Only my Internet computer sees the Internet.

I will, when I have time, transfer a circuit diagram and you can see that just for 2 position it is already clumsy. Now to do it for 3 postions, i.e., 6 powerfull Triacs, 6 IR small triacs to drive those powefull big triacs or to do it with 6 IR power Triacs Relays like CDK proposed. The software will take quite some time to make such system work and at the end it will be too expensive and very clumsy. I am so sorry to give you this bad news.

Maybe somebody else is able to help you here.
Bert

 

a) If you can find a power triac with currents of up to 100 Ampere and it has already zero crossing and IR sensing build in, use them. And combine the small triac with the power triac.
b) If you like to have also the 3 neutrals switched, place the IR inputs to those triacs parallel or in series to the output of the micro PICF675 and use the internal oscillator instead of the external one Otherwise you don't have enough pins.
c) I tried NOT to do it with a micro, but with a rotating switch. The cross means it does not work and nobody can get it to work in that way.
d) But with a micro, you can loop and have 50 Milliseconds or whatever time delay you like and every time you change a toggle switch, there is a time-delay and the ports must be switch off first before the correct triac or two triacs are switched on. If the first switch is on and also the last one, the loop does not allow you to get to the last switch, so visa versa.

Cost for everything, is in my view more than that beautiful manual rotating switch >off> shore>>inverter>>generator.
I haven't time in the next three weeks to help you to make and program a micro, but maybe thereafter.

 

Given your load requirements, the solid-state-relay idea is nice, but will run into £100s each. More realistic is a triac. There are loads of triac circuits on the web, including this PDF, which you might find useful...

http://www.google.com/url?sa=t&rct=...sg=AFQjCNHZqfLJgdRR0di18J7Hgv60sHAzoQ&cad=rja

Using a microprocessor for this application seems sensible, as you can then determine break-times etc. You could potentially use the uP for voltage and current measurement (and display). If I were undertaking this project, a microProcessor would certainly be at the centre of it.

Tim B.

 

Excellent information Tim. The problem he likes to have resolved, is to have 3 x the life and 3 x neutral switched with a triac. I have never done it and probably it may work if you have one in the life and one at the other end in neutral. Most likely, one needs a "leak resistor" to kick the conducting in. But 6 x power triac’s with IR sensing, will, in my view cost more, than that beautiful switch as per #4. Why going second class when you can do something first class? He cannot write for a microprocessor in assembler language, nor has a development platform. I tried it with a rotating switch, forget it , it will blow the triac’s up during the short circuit. Only with a micro, one can create time delays between switching or with a couple of logic IC’s.

 

That is correct.
I use a very old fashioned rotary switch to change between grid and inverter in my home. It has 3 positions with 'OFF' in the middle and some ball/spring mechanism to reach each position.
The switch itself carries only the coil current of two industrial AC relays. Because the knob must be turned 180 degrees to go from one source to the other it takes far more than 8 milliseconds to switch. It is in place for more than 12 years now and the inverter suffered no damage.
With SSR's you can make lots of circuits to create a delay, but I prefer the simplicity of a mechanical switch. Even the rotary type you show can be used if you take a 7 position type and wire it 1-0-0-1-0-0-1.