Led limiting resistor

[HankMcSpank]

Thank you Demon,
I am testing now a ACS712 and in the scope it looks like a sine wave - 60 HZ, swinging between +1 volts to + 3.5 volts
I will get tomorrow an LTC1966 from Linear Technology , which is an RMS to DC converter and start working from there
Regards
Ion

a liitle more meat on how you anticipate handling the 'trapping' of an element going open cicuit might be useful..... ie the whole signal chain, because I'm left wondering why you're using an RMS to dc converter (and a comparator) for a simple task that a pic can readily do for you.

[igeorge]

The pic is supposed to monitor 16 heaters. Can be done with a multiplexer, but my boss wants a visual LED for each burned heater, so in this case i decided to have 16 separated inputs. He wants the leds to represent the status of the inputs. He does not believe in multiplexing .... and he sign my check also

[HankMcSpank]

The pic is supposed to monitor 16 heaters. Can be done with a multiplexer, but my boss wants a visual LED for each burned heater, so in this case i decided to have 16 separated inputs. He wants the leds to represent the status of the inputs. He does not believe in multiplexing .... and he sign my check also

No, that bit is clear...what I'm getting at is how are you going to have the PIC 'monitor' (ie what's the PIC gonna be looking for ....a DC level? - therefore ADC, comparator, a small AC signal? therefore ADC to extract a non zero level)....you mentioned comparators....you mentioned RMS to DC converters .....a bit more clarity on your anticipated approach at the PIC itself will help everyone help you towards your end goal.

[igeorge]

Some how i have to bring to pic a level 1 for each good heater and zero for a burned one. They are not enough A/D on the pic to handle 16 channels.
So, what i need is to convert before pic the signal in zero and one . Zero for any current lees than one amp and one for anything above that

[HankMcSpank]

Well, in your shoes, I'd be inclined to use two PICs and just step through the ADC channels to pick out out a 'non AC' signal (ie a dead element), else you're looking at a fair bit of 'gumph' just to put in between your 16 'monitored' signals & the PIC. (ok, so I know your boss said no multiplexing....but would he know that the ADC pins are multiplexed internally anyway?!!)

Also you make no mention of how quick you need to be alerted....is there a 'failed heater response unit' on hot standby looking at the leds, is a minute or two ok, must this be notified within a second etc.....such requirements all help steer the final deployed solution.

[igeorge]

1-2 seconds will be fine

[HankMcSpank]

In which case, I'd not be going dicking about with hw comparators (that said when you think about it, ADC'ing a signal and then comparing against a preset level in software is a comparator...a sw comparator), RMS to DC converters etc....just tap off some reduced AC from each element feed (obviously 5V 'peak to peak' max, but at least here when your element is only drawing 1A there'll still be 1V AC peak to peak - more than enough to work with) ...squirt the 16 tapoffs into a PIC's ADC pins (use two PICs to get around the 'not enough ADC channels' problem). Then step through each pin all sequentially, staying on each pin for long enough to ensure you capture a 'non zero' reading (note: 'zero' here being half your VCC, so for 8 bits, and a VCC of say 5V, 2.5V = 127.....so say any ADC reading less than 110 or more than 144 means there's AC present....if so move on, if not then latch a red led, then move on etc)