Question about sinking in op-amps

[ds18s20]

Hi

I have been testing a little sensor project and just to get things straight I built a little test circuit and discovered some strange stuff. So if someone could explain what I am seeing, I will be very happy:

Basically in example #2 all works exactly as I expect - when the output is "positive" the LED is totally dead with about 0.12V across and when the output is "negative" (sinking) the LED is fully ON with about 2.5V across

However, in example #1 - the LED is fully ON when the output is positive (sourcing) but when the output goes nagative, the LED is not completely OFF. There is about 2.3V across and the LED is 50% ON and my question is why is that?

I suspect this has to do with the internal structure, and if yes, then my follow-up question is - why is it that the TTL world had adopted sinking as the predominant principle instead of sourcing? Why is it that most TTL ICs are built with sinking in mind?


Thanks
~B**broken link removed**

 

[rat_race]

Re: Op Amp question

why dont you take a look at Gray/Meyer's book ? i suggestbyou to do so.

 

[ds18s20]

Re: Op Amp question

why dont you take a look at Gray/Meyer's book ? i suggestbyou to do so.

Hi

I can't find anything meanigful on Google - is this a paper book or is it available as PDF?

THanks
~B

 

[fantusox]

Re: Op Amp question

Your Led is a model from manufactuor? if so, check the reverse breakdown voltage of the device.

 

[ds18s20]

Re: Op Amp question

Thanks for everyone's reply, but does anyone see what the difference between the two schematics is?

Yes - there is a difference.

~B

 

[turttle]

Re: Op Amp question

It seems LED2 is drawn with wrong polarity. If LED1 lights even at output low is because minimum output over negative rail for that OA is high enough to do that (you can't get a lower output from that device because of its output stage design). For that OA, maximum output is closer to possitive rail and the difference is low enough to prevent light to be emitted from LED. With another OA model results can be just the opposite or work fine for both circuits (rail-2-rail output).

Cheers ;-)

 

[Kral]

Re: Op Amp question

ds18s20,
The LEDs are connected with reverse polarity. A LED wil emit light if its reverse voltage breakdoen is exceeded, but the results are unpredictable.
Regards,
Kral

 

[ds18s20]

Re: Op Amp question

It seems LED2 is drawn with wrong polarity. If LED1 lights even at output low is because minimum output over negative rail for that OA is high enough to do that (you can't get a lower output from that device because of its output stage design). For that OA, maximum output is closer to possitive rail and the difference is low enough to prevent light to be emitted from LED. With another OA model results can be just the opposite or work fine for both circuits (rail-2-rail output).

Turttle

You have precisely understood my question – thank you. If you ignore for a moment my retarded drawing the LEDs with incorrect polarity my issue is that the results in example 1 are NOT just the opposite in example 2 as you suggested they might be.

A low OA output measured over ground rail is around 2.1V while a high output measured over the positive rail is only about 0.15V – and we have arrived to my question – why?

I would expect that a low OA output measured over ground rail should too be some negligible voltage, far insufficient to light an LED but it is NOT – the 2.1V provides sufficient energy for the LED to be “somewhat” on.

~B

 

[Sal]

Op Amp question

Hi

regardless the drawing of the led, I would be good if you posted the IC's code number, so we can look for the internal circuit and have a broader idea (wheather your IC is sinking or sourcing)

From the circuit in example #1, I assume that your IC is a sourcing one. Now, you didnt post the value of the resistors, so we dont know how negative becomes the (+) input of the OpAmp with respect to the (-) input. Because of the precense of R5, the most negative value has a limit equal to 5*R5/(R5+R1) . assuming that the upper lead of R1 is connected to 5V. One thing you can do is make R5 very small respect ot R1, or simply eliminate it, and try again.

Now, I dont know exactly why TTL designers chose to use sinking but 2 strong reasons I would consider are: 1)With sinking you can control voltages different from TTL Vcc, e.g. 20 V, this means making an interface to a different Vcc circuit. 2)The first CI doesnt need to spend energy in supplying current to the next CI, there is the extra reason which make the sinking type robust against short circuits. Definitely, it is better than sourcing.

cheers

Sal

 

[ds18s20]

Re: Op Amp question

Thanks Sal,

You are right - I skipped the numbers. The OA is LM741CN. RN1A and RN1B are 2M each. R5 is 10K and the pot R1 is 20K so that adjusting R1 can put the OA in equlibrium or not, depending on the sensor which is connected at the other end of R1

R3 is something like 120Ohm, just per the LED spec sheet.

Thanks
~B

 

[turttle]

Re: Op Amp question

@ds18s20

LM741 is not thought to operate with single supply that's the cause of such a disappointing behaviour about negative rail. Use a LM358 instead or another OA designed to work with a single supply: they manage properly negative rail approaches, either at input as at output.

Cheers ;-)