Laser emergency stop over ethernet

[sugeek]

It looks good, I can't wait to see your first prototype!

-Adam

[dave]

I hate to throw another spanner in the works, but you cannot get to the spare pins of the cat5 socket in the qm2000.net

[loopee]

Not only that, but how really wants to hack there several thousand dollar QM.net device!

I’d be happy to do a cheap switch but I’d rather not damage the QM box, even if it would only be the network adapter part of it.

Personally I’m not that sold on this design (IMHO)

On one hand you’re trying to minimise cabling for the Kill/key switch but then one the other hand you’ve made the power connectivity more complex.

Here are some suggestions

1. Remove the Power Outlets with relays box, just put an IEC socket in the projector to power the QM.net off but controlled from the same relay as the lasers.
This means you don’t need to run power cables back to a specific power board or cabling to multiple controlled power boxes. (Unfortunately your design doesn’t show how you’re controlling more than one projector)

I don’t know how you’ve built your projectors but I’d look at building the QM.net box inside, this way you can break out the data pairs and feed them on to the QM box while the Kill/Key switch pairs can go off to the relays.

2. Put the power supply for the kill/Key switch in the same box; don’t use the power supply for the switch because it probably can’t handle the additional load of the relays with the volt drop. This also allows you to wire all switch ports the same and not have to worry about where you’re plugging things in. It also removes having to add an additional power supply per projector.

3. I think the point to a separate Key switch and kill switch is mute. If you hit the kill switch it will turn off all lasers and close the shutter on all projectors yet if you turn the Key switch off you kill all the lasers and the QM boxes on all projector, but you still don’t have control over individual QM boxes therefore what’s the point.

Additionally if you do reset the QM boxes there is a possibility that due to the fact that you’re not using the shutter control of the QM box that there maybe same exposure emitted as the shutter would be open before the QM box has booted up.
You’re initial issue around stray emissions when powering down the QM boxes doesn’t exist anymore as you now have a shutter on each projector.
Also, by removing the separate circuits for the Kill switch and Key switch means you can parallel up the blue and brown pair to:
1. Minimise damage to a laptop if accidentally connected to a power Cat5 lead and
2. Reduce the volt drop across the runs between the Kill switch, network switch and Projectors

4. If you really want the separate the kill switch and Key switch I would suggest you look at using shielded cat5 cable and using the shield as a ground, this will also provide the additional benefit of increasing the durability on the cat5 cables

[mccarrot]

I hate to throw another spanner in the works, but you cannot get to the spare pins of the cat5 socket in the qm2000.net

I have build my QM board + .net OEM board in a box and wired the network connector to a Neutrik RJ45. I have done this becouse the LAPTOP QM.net boxed version Pangolin sells is really expensive. a OEM.net and QM is way cheaper, and now I have access to the RJ45 pins.

[mccarrot]

[quote=loopee;87404]Not only that, but how really wants to hack there several thousand dollar QM.net device!

I’d be happy to do a cheap switch but I’d rather not damage the QM box, even if it would only be the network adapter part of it.

As you can read in the previous post I do have access to the pins on the QM.net, and if you have original boxed qm.net you can always use the breakout cable plugs dave suggested.

Personally I’m not that sold on this design (IMHO)

On one hand you’re trying to minimise cabling for the Kill/key switch but then one the other hand you’ve made the power connectivity more complex.

When you are at a rave or other event, the biggest distance is always between your operator area and the stage, this is the reason I only want 1 cable from laptop to stage

Here are some suggestions

1. Remove the Power Outlets with relays box, just put an IEC socket in the projector to power the QM.net off but controlled from the same relay as the lasers.
This means you don’t need to run power cables back to a specific power board or cabling to multiple controlled power boxes. (Unfortunately your design doesn’t show how you’re controlling more than one projector)

Great idea, this will safe us another box!


I don’t know how you’ve built your projectors but I’d look at building the QM.net box inside, this way you can break out the data pairs and feed them on to the QM box while the Kill/Key switch pairs can go off to the relays.

I know this is the best way, but at this moment I dont have the money to put a QM.net in every laser I have....

But I still want to have the possibility to turn off the QM.net (and laser) seperate from the laser. this is becouse QM.net can lock up.
Hugo (LaserImage) is using the same method (only not over cat5) and he told me its a must have to have the possibility to power down the QM's remotely. (Like when 10 QM.net disspeared during a big event(Qlimax)




2. Put the power supply for the kill/Key switch in the same box; don’t use the power supply for the switch because it probably can’t handle the additional load of the relays with the volt drop. This also allows you to wire all switch ports the same and not have to worry about where you’re plugging things in. It also removes having to add an additional power supply per projector.

I dont want to have another power supply, but I'm planning to use solid state relais, SS relais only take up 5mA.

3. I think the point to a separate Key switch and kill switch is mute. If you hit the kill switch it will turn off all lasers and close the shutter on all projectors yet if you turn the Key switch off you kill all the lasers and the QM boxes on all projector, but you still don’t have control over individual QM boxes therefore what’s the point.

I dont see the problem if you have 1 failing qm.net you need to power down all QM.net to get the QM's up and running again.

Additionally if you do reset the QM boxes there is a possibility that due to the fact that you’re not using the shutter control of the QM box that there maybe same exposure emitted as the shutter would be open before the QM box has booted up.
You’re initial issue around stray emissions when powering down the QM boxes doesn’t exist anymore as you now have a shutter on each projector.
Also, by removing the separate circuits for the Kill switch and Key switch means you can parallel up the blue and brown pair to:
1. Minimise damage to a laptop if accidentally connected to a power Cat5 lead and
2. Reduce the volt drop across the runs between the Kill switch, network switch and Projectors .

If you first power up the equipment by turning the key switch to ON, and after all QM's are discovered in the software you release the shutter with the STOP switch you should not have a problem with stray emmisions




4. If you really want the separate the kill switch and Key switch I would suggest you look at using shielded cat5 cable and using the shield as a ground, this will also provide the additional benefit of increasing the durability on the cat5 cables

Good point about shielded cable, but with SS relais the wire distance is not a problem any more.

[mccarrot]

I found another problem, putting 12V or 24V over the ILDA cable is not that safe. When the laser is on you can by excident short pin 4 with pin 5 and causes 12V or 24V at the input of the green laser input. This can kill the laser.

So I think the interlock signal should not exceed 5V to be safe.

I think I will go for a all optocoupler design working at 5V with a repeater optocoupler in the switch

[mccarrot]

I have made a new system diagram with optocouplers and only one 5V power supply.

The max distance between shutter and PSU is also reduced to maximum 100 meters instead of 200 meter with this design.

I intentally left out the resistors becouse this is just a system diagram.

Input please

[mccarrot]

wat do you think about this design:

[mccarrot]

I now have a final design i'm putting togheter at this moment.

I have allready one modified qm.net, and 1 modiefied network switch with a bypass for pair 3 and 4

[andyf97]

Interesting project there but unless i missed something it sounds like over complex methods being described.


I have a thought here:

You have a master device that has the features:
You keep this near the laptop.
STOP
Keyswitch
Function switches.
Transmits Data to the LAN in IP format every x milliseconds.



You have little slave boxes:
You have one of these per projector.
Allows power to the laser system and qmnet.
Has an output to toggle the keyswitches on the projector.
Has other outputs to toggle features on the projector.
Receives data from the Master in IP format.



Failure states
If no specific data is being received we have or we have a LAN failure and the projector is kept off, in other words specific data must exist to power up the projectors and qmnet. You could also with a bit of code have two outputs, one 220v for the qmnet and another for the Projector allowing selective re-starting.

If data is being received, then take actions according to the data, remote features etc..

Simple and effective. Takes power and delivers it to the projector.

Want to reboot the qmnets, just hit the reboot option and power can be cycled to all remote devices..

Sounds almost like a dmx dimmer but with some intelligence to only work if very specific data exists.

no point using a system that has no power, if no power then the projectors dont work either.

A wiznet, some power control and little bit of code.

Something like this:
http://www.pdu.at/pdu_power_distribu...wer_switch.htm