1
\$\begingroup\$

I am trying to measure the output from a Hall effect sensor (LEM-LF 310-S) for a welding application.

The maximum output of the sensor is 0-0.25A when sensing 0-500A. This range I need to convert to 0-1V for my measurement equipment (Oscilloscope for testing, later RedPitaya).

So I am using a Shunt with 100mOhm which should give me an output voltage of U=R*I=0.1Ohm*0.25A =0.025V.

This voltage I wanted to amplify with a non-inverted OpAmp (MCP6292) to get 1V as Output-Voltage -> Amplify with a ratio of 40 (a=32.04dB). According to an online calculator 100Ohm & 3.9kOhm are sufficient for the required amplification. The used OpAmp is a dual OpAmp where I just use one of the two.

The supply Voltage of my Sensor is +-15V DC and the supply voltage of my OP-AMP is +5V DC.

I tried to read 0-1V as output with the following circuit but I am receiving nothing but noise:

enter image description here

What I tried already:

  • Measuring the Hall-Sensor output directly with a 4Ohm measuring resistance -> as expected
  • Checked wiring with multimeter
  • Measured output of DC-DC Converter (15V to 5V) = Supply Voltage for OpAmp -> Output voltage is 5.6V (Should be ok according to OpAmps datasheet?)
  • Measured resistance of Resistors and Shunt -> as expected

I also thought about adding a capacitor of 100nF to the OpAmps PowerSupply, but I think this is not the problem here?

Hope someone can help me figure out the problem.

\$\endgroup\$
0
\$\begingroup\$

According to your suggestions I modified the circuit ...

Looks like you didn't :)

You have placed decoupling capacitors in series to supply pins. Since capacitors don't allow the DC to pass through, your opamp cannot get supplied and thus can't work.

What I just found out is that my Converter is an isolated one.

That explains everything. Isolated converters have isolated references (i.e. GND). So, in your first schematic, the sensor outputs wrt AC-DC converter's GND, but the opamp output wrt DC-DC converter's GND. Since they are isolated, the opamp outputs zero.

Here's a schematic for you:

schematic

simulate this circuit – Schematic created using CircuitLab

There's no need to use a DC-DC converter, a 78L05 is completely enough. C1, C2 and C3 are decoupling capacitors (look how they are placed).

NOTE: 4-pin shunt's measurement pins seem unconnected but it's just a drawing.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ Thanks for the correction - As I am not an electrical engineer I am quite unexperienced in this field. But your schematic and explanations helped a lot! Thank you very much. \$\endgroup\$ – timosmd Jun 4 at 10:53
  • \$\begingroup\$ @timosmd You're welcome. Please see my edited schematic. I've done some modifications regarding stability. \$\endgroup\$ – Rohat Kılıç Jun 5 at 15:25
  • \$\begingroup\$ Thank you! I will try it out as soon as possible and let you know if it works as expected :) \$\endgroup\$ – timosmd Jun 6 at 9:20
  • \$\begingroup\$ I tried the suggested circuit today but unfortunately it doesn't work. The signal when measuring directly at the shunt comes out as expected. So I guess the problem is somewhere around the Op-Amp. Checked the soldered board a million times but can't find the problem... \$\endgroup\$ – timosmd Jun 15 at 14:25
  • \$\begingroup\$ @timosmd are you sure that the grounds of sensor, supply and amplifier are all connected? \$\endgroup\$ – Rohat Kılıç Jun 15 at 14:37
0
\$\begingroup\$

0R1 is too low for the secondary side shunt (RM). If you read the datasheet then you'll notice that all the performance characteristics are given for RM = 10R. You may worry about the dissipation for RM = 10R compared to that with 0R1. So you can use RM = 1R.

Besides, with RM = 0R1 the full-scale voltage will be 25mV for IL = 500A. The offset voltage of the opamp can be as high as 3mV or as low as -3mV, so you may have a measurement error of (±3) x 40 = ±120mV which translates into ±12%. But for RM = 1R the full-scale voltage will be 0.25V, for for 0-1V range a non-inv. amplifier with a gain of 4 (R2 = 3k6 1%, R1 = 1k2 1%) is enough and the error will be significantly reduced.

I also thought about adding a capacitor of 100nF to the OpAmps PowerSupply, but I think this is not the problem here?

It's always a good practice to place decoupling capacitors right next to the supply pins.

| improve this answer | |
\$\endgroup\$
  • 1
    \$\begingroup\$ I'm currently working with a differential opamp and one problem I saw was the power supply of the opamp and the inputs of the opamps should be referenced to the same ground. In your circuit, the power supply is referenced wrt Vout of DC-DC converter while inputs are referenced to GND of DC-DC converter. The DC-DC converter I used dint have the same ground on the input and the output. If that's the case in your circuit, the opamp would not work. This happened tome practically and when I simulated the circuit on Tina. I just want to point out. \$\endgroup\$ – varun Jun 3 at 17:15
  • \$\begingroup\$ Thanks for your quick responses! Rohat Kilic: Yes you are right about the error and the 10Ohms as Rm in datasheet. But do you think this might be the reason why I can't see any useful output on my Oszilloscope? Anyways - will change Rm to 1Ohm if possible. @varun: I also thought that it might be a problem with some GND but I didn't find what I did wrong. In case DC-DC doesnt share GND it should be sufficient to wire VSS to COM of AC-DC converter instead of wiring with GND of DC-DC correct? \$\endgroup\$ – timosmd Jun 3 at 17:58
  • 1
    \$\begingroup\$ @timosmd could be. I don't know what your test current is, but you may try to disconnect the shunt from the amplifier and measure the voltage across the shunt while the test current is high enough (e.g. 200A) to make sure the correct voltage is induced. And also, I would tie the shunt's U1 pin to opamp's VSS for better referencing. \$\endgroup\$ – Rohat Kılıç Jun 3 at 18:13
  • \$\begingroup\$ That depends on the opamp. The opamp (lm358) that I used could take +32 to -32V. If your opamp can take 15V, you don't need that 5V DC-DC converter. Just connect +15V to VCC of opamp and COM to the ground of the opamp. You will have one less component and a common ground for inputs and power supply of the opamp. You will also have more room before the output of the opamp saturates(More wiggle room to chose your opamp gain ). \$\endgroup\$ – varun Jun 3 at 18:51
  • \$\begingroup\$ My test-current is arround 150-200A. I tried to measure the induced voltage but couldn't see anything useful. Maybe the output voltage of ~0,015V is too low to see it. \$\endgroup\$ – timosmd Jun 4 at 7:29
0
\$\begingroup\$

According to your suggestions I modified the circuit and included Capacitors for the Amps Power supply, changed the measuring resistance to 1Ohm (R1&R2 accordingly) and connected the DC-DC-Converters GND to AC-DC-Converters GND so all OpAmp-Inputs do have the same GND:

enter image description here

Please correct me if I'm wrong?

Edit: The DC-DC-Converter is of the Type "TMA 1505S" (TRACOPOWER)

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ We don't know the characteristics of your DC-DC converter (maybe a model number would help), in most cases it's okay to short its 0V inputs but still there's a problem with decoupling capacitors. Please see my answer and make their connections correct. \$\endgroup\$ – Rohat Kılıç Jun 4 at 10:34
  • \$\begingroup\$ Edited my answer. \$\endgroup\$ – timosmd Jun 4 at 11:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.