burned out element detection

Add a back up or secondary controller that kick in when more heat is needed due to an element failure. And sound an alarm or indicator.

-or-

Can you crimp a durable heat tolerant tap to the resistive element as a voltage divider circuit and drive a voltage sense circuit. There should be a way. They had to attach a lead wire or connector to the beginning and end.

[ tap element length / total element length ] times applied voltage = sample voltage

If the tap opens you see applied volts. If the element opens any were else you would see zero volts.

Paul P.

The mention of using a playback head to sense current reminded me of a similar solution at

This approach uses a common inductor as the sensor, which should be considerably cheaper than a tape head if you need to buy it. You could make the detection circuit a bit simpler by using a quad comparator IC such as the LM339. Sensitivity could be made adjustable to handle different current levels.

David masondg44 at comcast dot net

On a sunny day (Sat, 07 Nov 2009 08:38:49 -0500) it happened Dave M wrote in :

Yes, that method should also work. The playback-heads are so small that you can just glue those against the wire. After all the talk here about glue, finding a suitable glue should be easy :-). Maybe if you have more then one, as each head has about 10 cm of shielded wire, connect those to one big shielded multi core cable, and do the electronics is a different more convenient place. After all the talk here about soldering that should be easy too :-). As to the cost, I think if you shop around, less then a dollar a piece, for the playback-heads, but that may well go up over time as that article is getting scarce... Strip an old walkman perhaps.

Somebody here in this group mentioned he had a whole lot of those in a box, so maybe he wants to sell some? I use a 1/4 LM324 on 5 V as preamp, and 1/4 as peak detector, so 1 chip can do 2 PB heads. Since the output is an analog voltage, I can actually measure current, and use it to monitor the electric heater here, the PC calculates the kWh and electricity cost from that: ftp://panteltje.com/pub/xhcs.jpg System has been working OK for many years...

I've a few if needed.

--
Best Regards:
                     Baron.

r

For that kind of money, you'd get something that has a good output waveform and takes calibration. Neither quality is required. You could put ten turns of wire around a soft steel washer for simple detection.

It's bothers me when only the COTS parts are considered. Heck, a common coathanger is a perfectly good 10W resistor, you don't always have to buy something with "10W resistor" printed next to it in a catalog.

.

the

to=20

Oh yeah, Speff is onto something alright. Use current transformers, set the current ratio to provide about 5 mA into an antiparrallel pair of LEDs. Light =3D=3D current to load. Cheap simple and works nice.

.

transformer

If you do not know what you are doing (with current transformers) you can get skinned real easily. So i will give some clues:

The core only needs to be sized to support the output load power (usually milliwatts). For an example take apart a GFCI outlet, see just how big those cores are, yet they detect and trip on 5 mA out of a possible 15 A load.

A transformer is a transformer, the windings ratings are properties of the copper wire size, the insulation withstanding voltage, and the expected temperature rise. To a large extent the core as well. The canonical load of a current transformer is nearly a short circuit.

Depending on volume is may be cheaper to just hand wind a few (dozen or hundred).

Provided that the core and windings can take the currents and heat surprisingly small audio transformers can be used. Much the same with hand wound, see GFCI above.

Pulse systems have very different requirements than steady state line power systems.

That is about as much as i can say without getting into the actual application.