GE low voltage lighting

[jimandy]

Just bought a house as we begin downsizing in our old age. One feature was an extensively remodeled (for a disabled person) two-room basement in which they used GE remote control (wired) lighting. The recessed fluorescents are controlled by 24v relays actuated by rocker switches. The firm that did the work has since gone out of business. I found info on this type of system on-line and have finally identified the relays used (GE part RR8) of which there are 17 in all. The RR8’s are discontinued but apparently the RR9 will sub.

The previous owner left no documentation on the wiring, none of the cables are marked and some of the relays appear inoperative with at least one stuck in the “onâ€

[gerty]

Haven't seen one of these in years. I thought they were awful proud of them. http://www.lightcontrol.net/a3RR9.htm
The problem with going solid state ( using existing number of wires ) is going to be the latching. One alternative would be changing out the switches from momentary to maintained (ie 3way) and replacing the relays with solid state relays. Of course that would mean changing from ac to dc, and that would eliminate existing 3 way configurations. You cannot be the only one in this siuation, maybe there is a commercial 'fix'.
Are you familliar with this system, we learned how to install these in school back in the '60s.

[Viking]

These are magnetically latching 2 position switches, so replacing them with solid state versions is not going to be a simple matter. You could replace the momentry control switches with latching switches and use opto-isolated solid state switches, but I'm not familiar enough with this system to know if that would have an effect elswhere in the system.
Regards
Rob

[gerty]

The major problem would be if he has a 3 way switch arrangement. Turning the light on with a maintained would prohibit turning off from another location.

[MrAl]

Hi there,


I agree that replacing the existing relays with solid state devices wont
be easy unless you can find someone that makes such a device.
What makes it harder is that if the existing swtiches are momentary
contact (for use with a latching relay) then power does not normally
get supplied to the relay unit.

What might work (i dont really have a schematic of the system in
front of me so im guessing a bit here on how the system is set up)
is to jumper the switches with resistors. This would turn all the
switches into two level (analog) devices that either provide a low
resistance to 24v (switch pressed) or a high resistance to 24v
(switch not pressed). The high resistance mode would keep the
solid state replacement active when the switch is not pressed.
When the switch is pressed, the solid state replacement would
sense that the voltage went up (with the right load of course)
and that would tell it to do turn on or off whatever side of the
(presumably) two or three state solid state replacement relay.

Using this sort of system it shouldnt be too hard to design
a simple solid state replacement, using probably triacs to
do the actual 120vac switching.

This assumes the system works similar to this:

1. Press switch one way, light turns on.
2. Press switch the other way, light turns off.

or

1. Press switch one way, light turns on.
2. Press another switch, light turns off.

The only trick is to jump ALL the switches with resistors (or possibly
just some of the switches) so anything used as a solid state replacement
for the latching type relay would be able to get power all the time.

The other catch is that depending on how the system is set up
after a power failure ends all the lighting would come on, or just
some of it.

What else i dont know is how much room there is for replacing
one of the existing relays with a circuit board. What size circuit
board would be possible, 2x4 inches, 4x5 inches, etc.

There are other possibilities too though, because it appears the system
is 24vac and not 24vdc, such as to jump the switches with simple
diodes, where pressing the switch sends a rectified negative part
of the ac signal, but not pressing sends the positive part of the
ac signal. This would tell the solid state device what to do.
In fact, you could get three states this way: one would be to
pass the positive half cycle, another to pass the negative half,
and the third to pass the entire wave (both pos and neg half cycles).

[jimandy]

I am looking at the RR9 spec from the link provided by gerty (20A tungsten filament 125 VA) and thinking $119.00 is a lot of money. From Jameco I see a solid state relay rated at 240 VAC, 20 Amp for under $18.00 in quantity. That plus a simple TTL flip-flop to handle the latching (and some key bounce filtering) and it seems like active parts for the equivalent function is under $20.00. Of course there’s the problem of sending back a 24 v signal to light the pilot bulb on those GE switches, but I think I can live without that.
So does this make sense?

[reloadron]

Outch, I saw the price in the link gerty was kind enough to provide. About $120 EA. for those mechanical latching relays. Those relays rated at 30 Amps have to be substantial overkill for a home lighting application.

I see you having a few viable and lower cost options.

You could replace the existing relays with for example Magnecraft Class W250AML relays using an 11 pin octal base. A Magnecraft W250AML2CPX-8 (24 VAC Coil) relay runs about $16.00 and has a pair of contacts rated at 10 Amps giving 20 Amps rating if you parallel them. 11 Pin Din Rail mount sockets would run you a little over $2.00 EA.

You mention there are 17 Circuits. That is quite a few circuits for lighting 2 rooms? You may be able to get that down to 10 or less.

Another option is to eliminate all the latching and simply remove the relays and replace with wall switches centrally located where the existing relays are?

You have several options. Pretty much a matter of cost and how much work and change you want to incorporate.

Ron

[jimandy]

Yes 17 circuits are a lot. There are two finished rooms, each with 4 fluorescent bays (and the 3 tubes in several are apparently individually controlled) plus some incandescents here and there. The relay box is in an unfinished area and would not be suitable for rewiring with standard switches.

I am still toying with the idea of a SSR (Omron G3NA series) and a TTL latch. Would only use it for selected circuits and as a replacement for those expensive GE relays as they fail. Thought by now someone would have told me what a dumb idea that is.

BTW, discovered they left a TASH Ultra4 receiver/switched AC receptacle that, after some research, found is an ultrasonic controlled switch generally used by those with impaired motor functions to turn on/off a light or other appliance. Problem is, they did not leave the transmitter that I found online in the range of $90.00. Guess it will go in the junk box for now.

[reloadron]

That could work. Using for example an Omron G3NA-210BDC5-24 which when heat sinked can handle 10 Amps. They cost about $13 EA. and then you are looking at heatsinks.

The only potential problem I see is that line of SSR has an inptu voltage of 5 to 24 VDC which might be a problem going directly from a TTL latch. I would look for maybe a 3 volt input SSR?

Ron

[jimandy]

The only potential problem I see is that line of SSR has an inptu voltage of 5 to 24 VDC which might be a problem going directly from a TTL latch. I would look for maybe a 3 volt input SSR?

Maybe the latch could drive an open collector buffer and I could push, say 12 V, to the SSR. More chips, of course, but I would be using multi gate packages anyway like a 7407.

[MrAl]

Hi again,

One thing to remember is if you replace the relays with more relays
then you end up with roughly the same reliability. If you replace them
with triac circuits (controlled with a flip flop perhaps) you end up with
a fully solid state solution. The triacs work in on/off mode only so they
dont bother the fluorescents.

[reloadron]

The only potential problem I see is that line of SSR has an inptu voltage of 5 to 24 VDC which might be a problem going directly from a TTL latch. I would look for maybe a 3 volt input SSR?

Maybe the latch could drive an open collector buffer and I could push, say 12 V, to the SSR. More chips, of course, but I would be using multi gate packages anyway like a 7407.

I think I would just look to use a type D flip flop like a 7474 and work the inputs around the switch bounce issues. Then I would come off the outputs with a handful of simple 2N2222 transistors. However yeah, I can see going with a hex buffer package. Long as you can get the needed input voltage for the SSR. Maybe start with a proto of a single circuit or two and see how it works.

I still can't get over those existing latching relays.

Best
Ron

[jimandy]

The relay was the clue to my discovering what this system was all about. At some point they had replaced a relay and left the empty box for the replacement on top of the relay cabinet. It had GE and the number RR8 with a schiz for the connections. I don’t understand why the RR8 was discontinued. It was a version that fed signal back to the pilot on the rocker switch, but apparently it was not isolated and the RR9 corrected that.

Using these terms I Googled and finally found stuff archived on the GE web site. I’m keeping the box as a collectible for the museum of lighting history!

[Dean Huster]

Gotta love the "innovative" 1960s. Friend in Gulfport MS custom built a house and installed the GE LV system himself (electrician was required by the city to do the 120v part). His was the little momentary rockers by each door and a small panel with two twin rotary switches (one for "OFF" the other for "ON", turn to select the light and push to activate) at the front door, another at the back door and a third in the master bedroom. Trouble is, the LV wiring for these is not as robust as Romex and deterioriates rapidly in attic temps as do the relays. I've seen two or three of these houses lately and all are plague with electrical problems and all have had "rigging" done to them in some weird way.

With standard latching relays, the house remains "green". With solid state relays, you have to have everything powered up all the time and the green goes away. I suppose that CFLs will work OK with most SSRs, but don't know that for sure. The 30a rating on those relays was necessary because the systems were usually installed in upscale houses and a single circuit might turn on 10 or 15 spots in the ceiling at one time. I'm not convinced that the doubled wiring requirements (LV control and HV power) were worth the effort. For sure, now is not the time to be poking around in the attic -- we're running 104°F outside right now in SE Missouri which makes the attic an inferno. Climbing into our smallish above ground pool is same as taking a warm bath.

Dean

[jimandy]

With solid state relays, you have to have everything powered up all the time and the green goes away.

Good point. However, the honking big 24 V xformer in the GE-LV system is on all the time so it is not a squeaky green system. In my proposed replacement plan there would be only a minimal low voltage power supply to power the 7474 and buffer chips (maybe a 74C74?) and I would think that the SSR would draw almost no current when not turned on. Course if I hybridize, part solid state and part GE, I will have lost a little ground in that regard.

Biggest problem is the rats nest of wires in a rather smallish enclosure with all those LV and HV runs coming in and out. They must have spent a fortune on wire nuts alone.

Is it just me or isn't it weird no one makes a SS latching relay? Seems like it would be useful either in a two wire or pulse to toggle application.