Current rating of Faston / Lucar connectors

[Adrian Brentnall]

It all depends on how far you want to go with the paranoia! <grin>

My background is in industrial control - so I'm plenty paranoid..

In 'dumb' control systems you tended to have independent monitoring/safety systems.

With more advanced, micro-based systems - you either rely on the controller to look after itself, and operate 'safety' relays - or go you for full-blown belt&braces and have a second system (ideally with as small a component count as possible) looking over the micro's shoulder - and with the ability to stop proceedings if it thinks fit.

In this system I have an additional system (based on a Raspberry Pi) which feeds real-time info to a remote pc. It wouldn't be at all difficult for that system to also have a hand in the safety system - just an output from the RPi and a relay in series with the main controller's relay. That way, if the controller should lose its marbles, then the RPi would stand half a chance of saving the day..

Crowbar protection! That takes me back. I was contracted to work on a bulk petroleum tanker loading system back in the 80's. The USA-based designers of the computer part decided to add a simple backup PSU, based on their own design, with a big Dryfit lead-acid battery. To protect the micro, they fitted an over-voltage crowbar system. Unfortunately the crowbar was prone to being tripped by transients, and was connected directly across the Dryfit... Hmmmm!

Thanks

[Adrian Brentnall]

Never come across that - but it makes sense... I think we can be a little more sophisticated than that - and have you seen the price of gold recently <grin>

Thanks

[Adrian Brentnall]

Yes - I wasn't sure whether 2 poles or 2 relays was the way to go...

As I say, the plan is for these relays to be 'service replacement' items

- so it's really about making sure they run happily for the 3 years or so between replacement...

Thanks

[Adrian Brentnall]

Well - if you can find a thermal fuse that trips at 820c then I'm all ears <grin>

I'm sure something could be done with 'the right kind of wire' - but could also be done just as elegantly with the watchdog chip..

[Adrian Brentnall]

Yes - that's how it works. At frequent intervals the micro looks at the 'travelling setpoint'

- which is where it thinks the temperature _ought_ to be, and tweaks things up or down to compensate. There's a 'learning' process running all the time...

This one is fairly easy on the controller as the kiln is lined with ceramic fibre

- so there is very little thermal mass to consider - unlike a brick-based kiln where a lot of the heat goes into warming the bricks up.

Yes - can get exciting.. As with all these situations - it's about doing the "what if's" - and having a contingency plan!

[Fredxx]

I use an SSR for a 20A kiln. It's fitted in a die-cast box that also acts as it's heatsink. To be honest I wasn't aware of much temperature increase.

[The Other Mike]

A 'selected' silver solder or brazing alloy would get you something in that region.

House it in a slug of steel, with a hollow chamber, a ceramic plug at the top holding some rod shaped contacts maybe made of stainless with deadspace below. Some positive means of retention of the ceramic cap would also be required, as would some means of venting the chamber to ensure it didn't become pressurised both during operation and during the manufacturing / resetting process.

Ensuring the chamber walls were non-wettable and the contacts were the opposite would be advantageous for reuse.

To manufacture the fuse material is melted with a propane torch to bridge the contacts At the selected temperature the fuse material melts and then gravity moves the means of conduction into the chamber below and away from the contacts, breaking the circuit.

Reset and reuse would then simply require inversion, application of a brazing torch to the housing until gravity ensures the contacts are bridged again.

Once the contacts remake, turn off the heat and let it cool.

--

[Steve Walker]

Use two force-guided relays in series and use the NC contacts for monitoring? If one relay welds shut, the other should still cut the power and you'll get a warning that one has failed.

SteveW

[Steve Walker]

P gives proportional control, giving higher control effort the further the measured value is away from the setpoint. Moving rapidly towards the setpoint but ending up with some overshoot and final offset.

I gives integral control, higher control effort the longer the measured value is away from the setpoint. Closing the offset to give a zero error.

D gives derivative control, reacting to rate of change of the error to allow maximum rate of change, while reducing it at the last moment to reduce overshoot.

Switching on and measuring the rate of change, overshoot and offset allows a simple calculation of the figures to enter into the controller for the best response. Autotune just does the calculation in the controller and stores the figures itself, cutting out the human part of it.

I've walked around a bottle making plant, seeing slugs of semi-molten glass drop through holes in the ceiling, be deflected by strategically located channels and fire off through other holes to the next part of the process.

SteveW

[Adrian Brentnall]

Having worked on industrial control systems - i can't watch those sort of factories on tv without thinking to myself 'what happens when it all goes wrong...?'

It's bad enough in the food factories, where all they're making is biscuits or chocolates, but when you've got molten glass plying about the place, it could all get much more exciting much more quickly..

They never show you that bit!

[Adrian Brentnall]

Hi Steve Not sure about 'force-guided' - not a term I know?

The idea of the safety relay is that is is in series with the 'main' relay - but the safety relay is only switched once per firing schedule, while the 'main' relay can be cycled several thousand times per schedule.

On that basis, the 'safety' relay should wear out less fast, and should be reliable enough that it can be expected to interrupt the power when/if asked to do so.

If I was designing the controller from scratch I would want to see a pair of 'signal' contacts on each relay - so I could do the usual

- ask for the relay to close

- wait a bit (in computer terms... to allow the contacts to close and stop bouncing)

- check the signal contacts to confirm that it's closed

- throw an error / shut things down if it hasn't

and vice-versa when asking the relay to open.

But, in this case, if I can allow the controller to drop out the 'safety' relay when it feels uneasy about what's occuring, and, possibly, use the remote monitoring Raspberry Pi as a second sanity check (by watching the kiln temperature with a second thermocouple and breaking the 'safety' chain if it goes above expected limits), then I reckon it'll be a vast improvement over what's there already.

Thanks

[T i m]

<snip>

Thanks for that.

I think I'm running an older version of Marlin in my Mendlemax 3D printer but loath to change it because it works (and often like the 'tool' I always hoped it to be, after the project / hobby it initially was).

If you could pull the existing firmware out of an Arduino Mega I might be more keen to give the latest one a try but I don't think you can.

I think I have a complete set of spare electronics so might set that up with the latest firmware and swap the current one out as a test.

That sounds like a good idea, as long as it works etc.

I think I can trump that round a steelworks (whilst looking at 'Heavy Industry' with a college group for a week) and that looks and *is* very dangerous (or maybe was more dangerous then).

Second to that was probably a Lead Acid battery factory and a chroming plant. ;-)

Cheers, T i m

[jrwalliker]

How about a copper refinery where I had to walk along an overhead steel gantry with rather feeble looking railings over a huge open crucible of molten copper.

John

[Steve Walker]

Force-guided relays have NO and NC contacts that are physically connected, so if a NO contact welds together, the NC contact cannot close, thus allowing allowing failure to be detected.

Perfectly reasonable.

That's where the force-guided relay comes in - allowing the NC contacts to be running at control voltages rather than power voltages, but showing when there is a fault.

Sounds like a reasonable plan.

SteveW