Control Panel: PIDs and Switches, or one large touch LCD?

passedpawn said:

I had a crimped faston burn up after a couple of years in service inside my panel. Not sure how it got loosened up. Maybe hot/cold cycles fatigued the metal.

My new panel will have very little wire at all. I've got to mind PCB trace widths / copper weight for 30A though

Click to expand...

doug293cz said:

So, what kind of trace cross sectional area are you using for 30A? I know 2A thru a 240 sq um (0.37 sq mil) is too much (but I can't tell you why I know.)

Brew on

Click to expand...

passedpawn said:

My design criteria is to keep the trace temperature (due to I^2*R heating) below 20C above ambient.

If I go with 5 ounce (oz/sqft) copper plating, I only need trace widths of about 0.120" (that's for 23A). 5oz copper is not good for digital stuff because the min trace width is relatively large, and you couldn't route traces to processors and such, so I'm going to put the high-volt and high-amp stuff on a separate board that will plug into the main digital board. That gives me the added advantage of just changing this module to support, say, a 50A system if I ever run the wire for it.

If I did't mention it before, the SSRs are PCB version that solder right into the PCB. No wires. Datasheet here: Crydom LR600240D40

Click to expand...

doug293cz said:

5 oz Cu at LSL thickness of 0.0012"/oz is about 6 mi thick. Times 120 mil width is 720 mil^2 or 0.46 mm^2. This is somewhere between 21AWG and 20AWG. And yet, based on actual PWB current carrying experiments (that I can't provide details on) your design seems conservative. Curious. Be interesting to see what your actual temp rise is in operation. Have you got enough extra area on your board to put in about a 0.5" x 0.5" serpentine @ 0.120" lines and 0.120" spaces. This could be a good target for an IR thermometer, which would allow measuring actual temp rise.

Brew on

Click to expand...

kal said:

Heat is created when current flows through an area that is too small to handle the amount of current. This can happen:

(a) with a loose connection where only a few strands are making contact because the wire is not properly tightened/fastened, and/or
(b) a wire that is too small to handle the current is used, and/or
(c) when a properly sized wire is used but some of the copper strands were cut/trimmed to get it to fit (this should never be done), or
(d) if an electrical socket or blade is dirty or charred or if the plug is not pushed and locked in properly with the receptacle.

All of these cases effectively reduce the contact area meaning that more current flows through a smaller area which in turns creates more heat which can melt the insulation on wires and damage nearby components.

So it's not the amount of current being pulled through the entire cord and both plugs exactly, it's because the area through the problematic plug had been reduced somehow to the point where the current was too much for the connection point to hanel. So it's completely normal that the other SwitchCraft end was fine since the area through which the current was flowing wasn't affected.

This is a fairly common problem for those that go the DIY route. Make sure all high current connection points are tight, make sure to use *ALL* of the wire strands. I like to screw the wire tight and then twist/turn/pull it to confirm, then tighten again. Use good quality industrial nylon plugs and receptacles too. I like Leviton (USA made). Some of the Chinese junk is well, exactly that: Junk. It's much easier to have a poor connection or a connection go loose over the years.

Kal

Click to expand...

augiedoggy said:

Reason I thought it was ironic was that it was the 20a nema connectors and not the switchcraft ones because a lot of folks were against using the switchcraft ones stating that proper nema connectors should be used and the switchcraft would not likely stand up to the amperage well. (even though they are rated at 30A)

Click to expand...

I have yet to check the element since I swapped it with the old chinese 4500w watt element from my boil kettle that I bought from your site.
it was an apcom brand I believe and an american made element.. it looks fine but I was in the middle of a brew and didnt want to take chances.

Click to expand...

jcav said:

Love this thread! You electronic geniuses on this forum are amazing and your blowing my mind with your lingo, logic and capabilities! All I can say is Andrew doesn't do anything half a$$ and that is going to be one awesome control panel when it is finished. Photos of the progress would be good too... hint hint!

John

Click to expand...

BrunDog said:

Congrats on twins!!

Click to expand...

deprecated said:

Darn you for making me second-guess myself even more, passedpawn.

My last HERMs system was PID based, albeit cobbled together in a plastic tool box. It was about as simple (and probably unsafe) as you could get. A single PID, single relay, one outlet. I'd just shut off spa panel, unplug HLT, plug in BK, turn on spa panel and set PID to manual. It made beer.

While planning my new HERMs and brewing space I went back and forth between control panel PID or software based (Doug's SBE or CraftBeerPi). I figured I'd stick with PID based because it had been reliable for me previously and being in IT for a career, I didn't really need more bits and bytes to tend to.

Lately I've been second-guessing that approach and thinking that if I can convince myself of SBE or CBP reliability I might go that route, mainly for flexibility and it would just fit my brew space better than a wall-mounted or table-top control panel.

Maybe I should try both and alternate using until I settle on one. I've started a new panel already based on the PJ diagram below.

Click to expand...