sensor opamp with LM741-did I cook the chip?

[case-sensitive]

Hello,

just build an signal amplifier which is meant to amplify by 3. Now all it does is having about 4.5 V at the output and the inverting input, whatever happens at the input. Did I kill the op amp?

case

 

[mneary]

The LM741 is difficult to damage.

It's not a single-supply op amp. Your circuit requires a separate negative supply (a minimum of 5V) on pin 4.

 

[ericgibbs]

Hello,

just build an signal amplifier which is meant to amplify by 3. Now all it does is having about 4.5 V at the output and the inverting input, whatever happens at the input. Did I kill the op amp?

case

hi,
What is J1-2 connected to.? [ end of the 100k] and what is the voltage on J1-2.??

 

[case-sensitive]

I see. I put in +6V and 0 (battery) and its going to run on +9/0 V, just trying it first on a veroboard before I do 8 of them. So it wants to have a balanced voltage?
Shall I rather go for another chip, LM358 or so?

edit: r100k goes to an LM35, had 180mV. But I pulled the resistor out, just to check, same. Put 1.5V onto the input of the LM741, still the same. Shortened the 20k resistor, voltage went down (same both ends of the 20k, thats what its meant to do). But still, seems to go into saturation? (suppose 4.5 is max when you put a nominal 6V accupack on)

 

[ericgibbs]

I see. I put in +6V and 0 (battery) and its going to run on +9/0 V, just trying it first on a veroboard before I do 8 of them. So it wants to have a balanced voltage?
Shall I rather go for another chip, LM358 or so?

edit: r100k goes to an LM35, had 180mV. But I pulled the resistor out, just to check, same. Put 1.5V onto the input of the LM741, still the same. Shortened the 20k resistor, voltage went down (same both ends of the 20k, thats what its meant to do). But still, seems to go into saturation? (suppose 4.5 is max when you put a nominal 6V accupack on)

The 741 has a very poor spec,.
The output voltage will never go higher than about [Vsupply-1.5V] and never lower than +1.5V above 0V.

A LM358 has the same high output limit but will go close to about 20mV above 0v.

These figures are for a single supply.

 

[case-sensitive]

and it needs symmetric voltage? Yes, did Multisim with 1V at the input, was ok. Only today I checked with less than 1V, all of sudden it has more than amplification 3 (in my case with R1 10k, R2 20k).
For upper limit: will be 3 V max, 9 V supply, all good.

 

[AllVol]

and it needs symmetric voltage? Yes, did Multisim with 1V at the input, was ok. Only today I checked with less than 1V, all of sudden it has more than amplification 3 (in my case with R1 10k, R2 20k).
For upper limit: will be 3 V max, 9 V supply, all good.

While the 741 is an old chip, it is very suitable for what you are doing. However, it does reguire a dual power supply to operate properly. It can be used single supply, but with an entirely different circuit than what you are showing.

For a simple dual supply, simply hook up two 9V batteries head to tail, with the ground drawn from between them and your circuit should be fine.

Edit: see attachment

 

[case-sensitive]

Ok, full explanation: I put in LM 35s on 3 wire cables up to 10 or so m long. Their voltage can go down to 20mV/2°C. I have issues with noise, so I decided to put preamps in. At present there is +9V, 0V and LM 35 output. Now I would have to replace the wiring with 4 wired cables, +9V, 0V, -9v and LM 35 output.

So I'd prefer the LM 358 if it runs on single voltage and can cope with 20mV upwards (actually it looks better in Multisim)

 

[audioguru]

The inputs on the LM741 do not work when their voltage is less than 3V above the negative supply which is 0v in your circuit.
Some 741 opamps do not work when their supply is less than 10V.
It seems that Multisim does not know these limitations.

The LM741 and the LM358 are very noisy. You might see the noise when the input voltage is low.

An TLE2141 opamp is low noise, has low input offset voltage, has wide bandwidth, has a minimum supply voltage of 4V and has inputs that work down to 0V in a single-polarity supply circuit. The output goes down very close to 0V in a single-polarity supply circuit.

 

[case-sensitive]

Thanks, plugged it into Multisim, starts at 80mV, so I am missing out at the first 8°C?
LM358 does better, starts at 30mV, only 3°missing. Thats better, will have 3°C or less in here. At present the logomatic drops out/has offset of about 5°/50mV or even more on a lousy day.

 

[case-sensitive]

ok, bought the TLE2141s, put one in and what happens it pulls up the input/sensors/LM35s. With open input it would see 0V but goes to full blast, 4.5V on the output. Bought LM358s as well. Wired them up. (sensors/op amps). Held icecubes against them. LM358 goes down to 13° (130mV), TLE2141 to 15°C(150mV).

 

[audioguru]

The LM358 and TLE2141 need an output load resistor to ground so that their output voltage can go down lower.
But their output voltage will never go down to as low a voltage as a modern "rail-to-rail" opamp.

 

[case-sensitive]

replaced the LM35 by 100k/470 ohm to get 24mV and it works like a charm, get 72mV at the output.

Load resistor: shall I try as in the LM35 spec, 75ohm/1µF from signal to ground?

just remembered, have lighter refill gas, can use it as ice spray (to cool the LM35, ices the lighter quite well.

edit: the spray works perfect, got it down next to 0V, amplification is still 3! Just a bit smelly. The TL2141 is easier to work with (single amp, needs less current I suppose plus the board layout is simpler.
edit 2: it works fine with 10cm cable, but not with 10m. So I am where I was before.... What to do now, twisted pair?

 

[case-sensitive]

how big does the load resistor have to be?
first column is the amplifier, the others LM35 unamplified, RC filter
with RC filter as in the LM35 datasheet:
0 24 19 28 39 64 33 22
0 25 19 25 41 60 33 24
132 24 19 26 39 58 34 24
734 25 19 25 36 59 34 24
6 25 18 26 39 59 34 23
0 25 20 26 28 56 33 22
756 24 20 24 51 58 33 24
68 26 20 26 31 55 32 24
60 25 19 25 41 58 33 20
72 25 20 25 38 59 34 24
614 25 20 25 43 59 34 23
28 26 20 25 37 58 34 24

without RC in the amp
145 25 17 26 37 55 33 23
186 24 17 24 33 60 33 23
194 25 18 25 36 58 33 22
162 25 17 25 35 59 34 24
196 25 17 24 38 60 34 23
185 25 16 25 37 59 33 23
178 25 17 27 39 60 34 21
212 24 17 24 41 60 35 22
218 25 20 25 37 56 34 24
154 25 17 25 36 58 34 24
159 25 18 24 37 59 33 24
164 25 17 24 35 59 34 24
143 25 16 26 37 62 34 23
144 25 17 26 37 54 34 26

without RC and 2.2k resistor in the output

192 22 14 21 32 56 31 20
192 22 14 21 32 56 31 19
190 24 13 20 27 52 30 22
191 22 14 22 34 56 28 20
191 22 14 22 34 59 29 21
193 22 13 21 32 59 31 19
188 22 14 21 32 60 32 21
190 22 14 20 39 53 30 22
190 22 14 21 36 58 29 20
188 22 14 22 35 57 30 20
190 22 14 21 35 57 30 20
189 22 16 20 37 56 30 21
188 22 15 19 40 60 31 21
189 22 15 20 26 57 32 21
190 23 14 23 33 54 30 21
188 23 13 23 41 58 30 22
(cooling down a bit, but way less noise)

edit: getting worse with 10k. So guess I should do it by the LM35 datasheet, 2kB. Should not use the RC filter, but divided by 3 its still quite good. (the sensors that are in the house already have the RCs in, and I just going to wire the amps into the existing cable)

 

[case-sensitive]

so, for the future reader:
LM35 gets 75 Ohm and 1µF in series from signal output to ground. That fits on a 4 x 4 hole piece of veroboard, still a quite small sensor bit. Then comes 1-2m wire, then I put another piece of veroboard, 1K Ohm into the TL2141, then 10k from Opamp input to ground and 20k from opamp output to input (to achieve 3x amplification). Then 2k at the opamp output and then so ans so much wire to the data logger. Works fine, plus put averaging into Excel for 3 consecutive values.
Another idea seems to be: if possible put all the sensors next to each other, make a few readings, say at least 100 values. It turned out that there is different offsets in the different ADC channels of the logomatic, remained the same over several days, so I included the offset into my calculations in Excel.