Need Help Wiring a load Fault Indicator

[cowboybob]

Actually i went ahead and wired the entire thing with both fans and the relay and the first led when both fans are running is out but the second one is on then when the 210f stat is triggered then the led comes on. So wjhen both are running pin one voltage is at 1.9vdc

At half speed run in series led one runs trough 2 shunts.... maybe making them 3feet long? Also what would happen if you increased the resistor rating from the led to a higher ohm? Wouldnt that drop the voltage? i can post pin voltages but when both fans are running at low speed the voltsges are really close to what you said

Thanks for the updates,Trent.

The resistor(s) in series with the LED(s) are "current limiting" devices meant to protect the LED(s) from damage. While it is true they dump some voltage it's so little that the overall voltage level across the devices is not significantly altered.

Very interesting how the circuit is responding. Might be able to salvage this after all. I'll play some more with the existing breadboard circuit, taking into your account of how it's reacting and see if I can get where we need to be, i.e., LEDs OUT when the fans are functioning properly.

 

[alec_t]

Don't know which verion of the TLo82 you have, but the spec for the TL082C gives the max input offset as 15mV. Could that be causing a problem, given that your design input reference is only 7mV? Perhaps upping the ref to, say, 20mV would help?

 

[cowboybob]

Don't know which verion of the TLo82 you have, but the spec for the TL082C gives the max input offset as 15mV. Could that be causing a problem, given that your design input reference is only 7mV? Perhaps upping the ref to, say, 20mV would help?

Thanks, alec.

As it stands, the sig off the shunt is generating 25 - 45mV +, so that's not being a problem. It is a problem, though, in how I've expected the TL082 to perform/react.

This project looked straight forward but is, as it stands, not something that one can rely of a SIM for a final, working circuit. I'm confident that it can be made to work - just more there than met the eye...

 

[Trent W]

Ill post the reading i was getting, also i was going to try and make my shunt longer and see what happens. If worse ill shorten it. Maybe its just the length that is causing an issue. Thst is the only thing that is diffrent from the original design.

 

[cowboybob]

The shortened shunt will just produce less voltage which won't make much difference since the original ground leads are also producing some of the voltage.

KISS provided a possible workaround (adding a negative 5Vdc to the TL082) that I'm going to try this weekend.

I'll post the results.

 

[cowboybob]

The negative supply is the ticket.

 

[Trent W]

Ya you were right the shuts made it worse... burnt out an led as well. So how exactly would you add a negitive supply? Would this supply be going on the inverting input?

 

[cowboybob]

It's called a MAX1044. Readily available from Digikey, Mouser (https://www.mouser.com/ProductDetail/Maxim-Integrated/MAX1044CPA /?qs=sGAEpiMZZMtitjHzVIkrqUa/3DrsydEwOqnTvjSnloI= ), etc..

It supplies a -5Vdc at 20mA (from a + 5Vdc source) but used as the negative supply voltage (Pin 4) for the TL082, it only has to suppy <10mA, which makes sense since the +5Vdc supply is carrying the LEDs' load. The - 5Vdc supply is providing the negative rail for the circuit and little else. Being able to get the comparators outputs to drop below 0 (zero) volts makes the circuit work a advertised.

I'm ordering a couple tomorrow. When I get them I'll test it on the TL082 comparator circuit I have breadboarded, and let you know how it goes.

If it works, and I fully expect it to,, this will be a simple fix with minimal alteration (you may have to replace the TL082) to the PCB circuit you have already built.

 

[KeepItSimpleStupid]

Would this supply be going on the inverting input?

Early OP amp has a plus and minus symmetric supply, and those are V- and V+ on the OP amp itself. There was no ground power connection the the OP amp.
So, a negative supply, if used, would connect to the V- pin of the OP amp. Certain OP amps are designated as "single supply", but not without issues namely operating close to the power supply rails. In your case 0V and 5V. Voltages have to be above a certain value: e.g. 200 mV and the output can only got to 1.2 V of the supply for one particular single supply OP amp. I picked an early single supply OP amp. Now, you can operate closer to the rails, but not close enough. So, a negative supply allows one to have signal inputs and outputs nearly 0V. In some cases, a "rail splitter" is used, but then the OP amp in your case would have to operate at +-2.5V as power.

In general, what I said is true, but as usual, there are exceptions.

Note to CBB: You might also need a diode, resistor and transistor like the 2n3904 to ground reference the comparator output.

 

[cowboybob]

Trent,

Here's a revised (preliminary) schematic for the TL082 comparator circuit:

No significant change to the current board you've made.
---->> <EDIT below> PIN 2, NOT Pin 3!
In the SIM, by adding a negative 5 Vdc source to pin 4 and a 45mV sig to either pin 3 or 6, the outputs of the TL082 swung from about + 3.75 to - 3.75 Vdc, fully turning off the LEDs, one at a time. And the load on the negative supply was < 11mA, well within the sourcing ability of the MAX1044.

I've got 10 MAX1044s coming from Mouser (I can see lots of uses for the extras, if this works) to confirm all this. Should be here (USPS Priority Mail) by the end of the week.

...Note to CBB: You might also need a diode, resistor and transistor like the 2n3904 to ground reference the comparator output.

KISS, Thanks. I tried something like that (although without the diode) but it didn't help. Shouldn't need that with the neg supply to the TL082, right?

I can post "snips" of the SIM's (on and off conditions with values) if you like.

 

[KeepItSimpleStupid]

Yea, kinda forgot your driving an LED and not a 12 V relay. You might add a IN5404 (reversed biased across the 12 V power) and a 18 V Bidirectional or Unidirectional Transorb or TVS diode just to protect the circuit in general from the automotive electrical system.

 

[cowboybob]

Yea, kinda forgot your driving an LED and not a 12 V relay. You might add a IN5404 (reversed biased across the 12 V power) and a 18 V Bidirectional or Unidirectional Transorb or TVS diode just to protect the circuit in general from the automotive electrical system.

Sound advice. Will do. Keep forgetting how noisy (and nasty) automotive power can be...
Thanks.

 

[cowboybob]

Just an update.

MAX1044s have shipped from CA.

 

[Trent W]

Just an update.

MAX1044s have shipped from CA.

Ya i ordered 2 of these since ive built two pcbs of these im positive i can find a use for my spare haha

 

[cowboybob]

I must say that;s a very postive attitude, Trent.

You might want to hold off actually constructing the circuit until I've had a chance to refine it (even though I'm pretty confident that it'll work).

Your call.

 

[cowboybob]

Ok. Got the MAX1044s today.

Used one of them in a breadboarded TL082 comparator circuit, as configured in post #70, and the whole thing worked as advertised cool.

I did not check for ripple but should you get some noise (from your radio, for instance) add the lead from pin 1 to pin 8 (as shown) and that should eliminate it.

I used both electrolytic AND tantalum 10uf caps in the MAX1044 circuit and both worked well.

Interestingly though, the tantalum caps provided a very slightly lower output from the +5Vdc input, i.e., -4.80Vdc with the tantalum vs. -4.78 Vdc with the electrolytics. Don't know why. Maybe different ESRs. Of course, that difference has absolutley no effect on the circuit.

 

[KeepItSimpleStupid]

Leakage is a parallel path. I'd look at the actual capacitor datasheets.

Concrats! You passed OP Amps 101, Real World Addition

 

[cowboybob]

Leakage is a parallel path. I'd look at the actual capacitor datasheets.

Concrats! You passed OP Amps 101, Real World Addition

I am so proud...!

(don't remind me how long I've been in the class)

 

[Trent W]

Ok. Got the MAX1044s today.

Used one of them in a breadboarded TL082 comparator circuit, as configured in post #70, and the whole thing worked as advertised cool.

I did not check for ripple but should you get some noise (from your radio, for instance) add the lead from pin 1 to pin 8 (as shown) and that should eliminate it.

I used both electrolytic AND tantalum 10uf caps in the MAX1044 circuit and both worked well.

Interestingly though, the tantalum caps provided a very slightly lower output from the +5Vdc input, i.e., -4.80Vdc with the tantalum vs. -4.78 Vdc with the electrolytics. Don't know why. Maybe different ESRs. Of course, that difference has absolutley no effect on the circuit.

As soon as i get the MAX1044 i'll post some pics of the finished product its turning out really nice the led's and switch mounted it the dash look like it came stock from the factory like that.

Also i'd thanks for your guys help on this one I really appreciate all the time you put in on this one. I would have gave up long ago and wouldn't have even tried to put led's in haha.

 

[Trent W]

wow must need a little more sleep and stop staying up soldering pcb's and maybe i can write coherent sentences.... what I meant to say is

THANKS!!!