emergency shut down

[thecleancompany]

Greetings from California everyone.
As most of you know I have converted a powerchair into a work horse which I use nearly daily in my business. I have the RoboteQ 2550 controller and a DX spectrum radio. Currently I have a simple fail safe of a switch on the unit in case it gets away and a cord that disconnects the 24 volts from the motors with a pull cord.

Looking for a better way to do emergency shut down in case of loss of control.

Any ideas?

Thanks,
Tony

[woodygb]

In my hobby of robots ( Battlebot / Robotwars ) we use an Anderson connector with a loop of wire linking both connections... this plugs into another unlinked Anderson on one of the battery lines....simply unplug the link to kill everything.



You could purchase a battery immoboliser switch ( manual kill ) or a solid state relay for use with a R/C kill switch http://www.jaycarelectronics.co.uk/prod ... &SUBCATID=

[Burgerman]

Are you talking about via RC or do you sit in this? The latest Roboteq firmware, and the matching new instructions, show how to make it fail safe in the event of a shorted mosfet etc. Using in built checking and a contactor.

[thecleancompany]

Hey Woody,

Great suggestions I will get out tomorrow and purchase a the large kill switch like the one pictured. Thanks for the link too.

Tony

[thecleancompany]

Via r/c. Not sure what you are suggesting. Is it software I need to download or something already in the RoboteQ 2550

[Burgerman]

Have a re-read of the new manual. Its updated, as is the firmware. It may help you.

[woodygb]

Via r/c. Not sure what you are suggesting. Is it software I need to download or something already in the RoboteQ 2550

You'd need to read up on the Roboteq manual to find out if it has any outputs suitable to drive an ON/OFF relay inserted into a battery power supply lead.

I'd suggest the 100 Amp SSR that I linked to in combination with a R/C Switch ... e.g. http://www.hobbyking.com/hobbyking/stor ... itch_.html

Simply use a spare channel with a toggle switch on your transmitter to 1st operate the small R/C switch ...use the output of this to turn ON/OFF the larger Amperage SSR ( Solid State Relay )... you should also be able to set the small R/C relay to "failsafe" i.e. Turn OFF the SSR and thus your machine if the R/C signal from your transmitter is lost or corrupted ... this is unlikely with the new DSM R/C systems ..BUT...

[Burgerman]

You'd need to read up on the Roboteq manual to find out if it has any outputs suitable to drive an ON/OFF relay inserted into a battery power supply lead.

The HDC 2450 that I have here does this. In the event of the output stage doing anything funny, an internal fault or error, or say a fused or shorted output mosfet, it turns off the power via a relay. This is new in the firmware (needs you to update it), and is detailed in the new updated manual. Your controller may do it too.

The same auto stop relay, can ALSO be activated via a failsafe Reciever and RC switch, (not all spectrum recievers have true failsafe, but the 9 channel one does for EG) and so signal loss, or via a switch on the transmitter, the vehicle will also stop.

Thats 3 ways. Automatically in case of a roboteq fault, via a transmitter switch, and on signal loss after 1 second.

[LROBBINS]

If you use an ssr such as Woody suggested do note that it will be carrying full battery amps and will get hot - it MUST have a substantial heat sink (see the pdf spec sheet). That one is 100 Amps. Does anyone know of one that's big enough for the Roboteq that John's using? An old-fashioned mechanical relay (contactor) will only have the heat generated by having its coil constantly live. It won't live too long if it's repeatedly opened under high load, but except for its emergency function it will probably not have to make or break significant current, just pass it after already closed. Lastly, even the SSR sheet says that diode protection is needed for inductive loads (e.g. our motors), but that's a bit of a problem with regeneration - the diode would be a short for regenerative current even if the relay opens. Maybe one could use a smallish sacrificial diode that will fry itself (and have to be replaced) soon after the contacts open. Ciao, Lenny

[Burgerman]

Roboteq is 150 per channel at up to 50 volts. With 250 for short periods peak per channel. So it COULD briefly want 500 amps... If everything lines up, at half speed. But that wont happen with powerchair motors. At least I hope not!

[woodygb]

http://www.albrightinternational.com/lang/en/index.html

[LROBBINS]

Woody's link is to a wide range of electromechanical emergency cutoffs that could handle even the Roboteq at it's most extreme. An example, maybe a little small, is the SD150AL line that is both a relay and manual mushroom switch. It can handle continuous 125A and break 800A. Tradeoff for using something like this though is the current draw of the continuously active coil to close the normally open contacts - 7 to 13 watts, which at 24 Volts translates to 0.3 to 0.5 Amps; not insignificant. Could one achieve an adequate level of safety using a normally closed relay whose coil is activated only when there's a problem? Do normally-closed relays exist with high enough carry and break current ratings? Ciao, Lenny

[woodygb]

Lenny ..I've never investigated this myself ..
But I believe that the following is true.
With a solenoid ...the initial pull in amps are high ..but once pulled in the coil amps could / can be significantly reduced with some electrickery.

[LROBBINS]

Woody, Given that the drop out voltage is substantially lower than the pull-in voltage, I think you are right. A time-delay relay or solid-state switch that put a resistor in series with the coil would do, but it does add to the parts count. At least it would be safe-fail; if the resistor stayed in circuit the chair wouldn't go, if it stayed out of circuit, the chair would go but current drain would increase. I think I should go look at the latest Roboteq downloads to see what they actually recommend. Ciao, Lenny

[ex-Gooserider]

Depends on the solenoid to some extent... If one is trying to get high speed / power on closing, then it is common for there to be two coils (or one with a tap partway into it) with one coil wired for high current and generating a lot of flux, and the other much lower current and flux. They are wired so that there is an initial pulse going to the "pull-in" section, and then it drops to just "holding current" through the second section. It is sort of like the "starter windings" you see on larger AC motors....

I used to work on pin-ball machines, and they would use this style on the "flippers" and other coils where the coil could get held on for a while. Bumpers just used a high power draw coil that fired and released... Ditto other coils that only fired momentarily.. For the ones that didn't need fast response (like drop target resets) they used a lower power relay that didn't draw as much to start with.

The only other way to do it is to have some way to drop the voltage going to the coil so that it doesn't pull as many amps.

ex-Gooserider

[woodygb]

Silly me ...voltage not amps.

PWM the solenoid and reduce the duty cycle.

[Burgerman]

Use a cheap 5 amp forwards only electric flight speed controller.

[LROBBINS]

You guys seem to want to complicate things. Here's a circuit using one cheap relay and some passives to drop voltage to 1/3 (i.e. 8 V) after 50 ms. This assumed 1/2 current draw at 24 V and hence ca. 50 ohm coil and 20 msec pull-in time at full voltage. The summary for the disconnect relay+switch that Woody linked gives ranges, and coil resistance surely differs for different voltages, so the values here might have to be changed for the actual 24V unit. Total cost, except for the contactor, less than E10.

contactor time delay.jpg (17.35 KiB) Viewed 9850 times

I've drawn this as though using a relay without the mechanical mushroom, so there's a small emergency switch as recommended by Roboteq. For Rachele, who could not herself hit the emergency switch anyway, I think I'd just use a small phone plug with a lanyard.

On the other hand, the current draw problem goes away if one finds an SSR that can handle the current. They exist:http://www.goldssr.com/DC_SSR_12_57.html, but how easily one can get a hold of just 1 and how much it would cost I don't know. I'll bet the Albright electromechanical contactors are pretty expensive though.

Ciao,
Lenny

[Burgerman]

I saved the drawing!

[LROBBINS]

Ooops, the RC time delay has a couple problems: (1) with just 100 ohms, the current flow while charging the cap would require a high wattage resistor, (2) the time may be a bit short as the relay may pull-in before reaching 66% of 24V (= RC time). This revised drawing is a better. A 1 watt resistor will do and the time constant is increased to 150 msec.

contactor time delay.jpg (17.61 KiB) Viewed 9793 times

Ciao, Lenny

[Burgerman]

Thanks. Re saved!