# current sensing and problem with amplification by AD623

#### patpin

Joined Sep 15, 2012
401
Hello folks,
I have been making this clamp on current probe with Hall sensor and it gives me an output of 1mV/Amp.

The problem I have is the amplification of the signal. The purpose is to measure currents in the order of 1-50 amps changing at a speed of about 1000Hz.
I have connected a AD623 like that.
With Rg(ain)=1 K (for gain 100) between pin 1 and 8; The output of IC1a (=input of AD623) remains 0 when I measure a current when I connect to the AD623.
With disconnected inputs the IC1a changes from 0 to 1mV when I measure 1 amp.
Then I diconnected the resistance to ground goiing to IN- of AD623. ... it was getting better: The input of AD623 also changed form 0 to 1 mV when I measured a current of 1 Amp, which is OK. But the output (between Pin 6 and REF pin 5) remains at 0,225 mV even if input (pin 2-3) changes from 0 to 1 mV.
Other Rg's (10K;100K;1M) give same result, exept for Rg 10M: then output changes with 2mV (from 0,299mV to 0.301) for input change of 1mV...
Could somebody help me?

#### AnalogKid

Joined Aug 1, 2013
11,055
A copy of the 623 datasheet.
A real schematic that relates the external components to a more conventional opamp or inamp symbol so we can see the actual circuit.
Why is there a 4.5% attenuator in each input, if those are inputs?

ak

#### patpin

Joined Sep 15, 2012
401
A copy of the 623 datasheet.
A real schematic that relates the external components to a more conventional opamp or inamp symbol so we can see the actual circuit.
Why is there a 4.5% attenuator in each input, if those are inputs?

ak
Dont have the tools to produce a complete diagram...
I removed the attenuator and results remain as described above.
After the Clamp on diagram reproduced at the top of the tread the AD623 connection diagram is now

#### patpin

Joined Sep 15, 2012
401
Sorry cannot add own circuit to that one

#### ronv

Joined Nov 12, 2008
3,770
View attachment 101337 Sorry cannot add own circuit to that one
Try disconnecting the Ref pin and taking the reading from the output to ground.

#### patpin

Joined Sep 15, 2012
401
Try disconnecting the Ref pin and taking the reading from the output to ground.
I choosed a Rg of 2.2K for gain of about 48. Dont have put any condensersexcept one of 100n between + and Grnd.
I disconnected the ref pin and let it float and measured between GRND and pin 6 (out): It measures 1.917 V and does not change if INPUTS changes from 2.5 to 3.5 mV.
As a second test I have put REF to GRND; Output is 1.615 V; It does not change as input changes BUT input only changes 0.2mV now instead of the 1mV before (and as whished).

Last edited:

#### crutschow

Joined Mar 14, 2008
34,464
What is the value of Rg?

#### patpin

Joined Sep 15, 2012
401

#### ronv

Joined Nov 12, 2008
3,770
I choosed a Rg of 2.2K for gain of about 48. Dont have put any condensersexcept one of 100n between + and Grnd.
I disconnected the ref pin and let it float and measured between GRND and pin 6 (out): It measures 1.917 V and does not change if INPUTS changes from 2.5 to 3.5 mV.
As a second test I have put REF to GRND; Output is 1.615 V; It does not change as input changes BUT input only changes 0.2mV now instead of the 1mV before (and as whished).
Try measuring between Vout and Ref. Nothing else tied to Ref.

#### ronv

Joined Nov 12, 2008
3,770
Hmm, I guess I don't understand this IC very well. But I think if you want the output around ground you will need 2 supplies +/- for the op amp.
Or bias up Vref.

#### patpin

Joined Sep 15, 2012
401
This is what I tried with putting 2.5V on Vref. But no effect. Do you think it would be better that the input of the AD625 is in the mid of VCC eg the hall EMF (2,50V+ 1mV/amp then I could put that on IN+ and 2,5V on IN- as realised in the current-probe diagram at treat start on pin 3 of IC1a ?

#### ericgibbs

Joined Jan 29, 2010
18,865
hi pat,
Using a single 5V supply the Vout of the AD623 will not go below approx +0.15V.
This is with the Vref pin connected to 0V.

What is the common mode input voltage relative to 0V, on the input pins of the AD623.?

Eric

#### Attachments

• 20.5 KB Views: 12

#### patpin

Joined Sep 15, 2012
401
Try measuring between Vout and Ref. Nothing else tied to Ref.
I disconnected the ref pin and let it float and measured between GRND and pin 6 (out): It measures 1.917 V and does not change when INPUTS changes from 2.5 to 3.5 mV..

#### ericgibbs

Joined Jan 29, 2010
18,865
I disconnected the ref pin and let it float
Hi,
You must not let the Vref pin float, else the AD623 will give incorrect outputs
The two 50K's on the neg input of the output OPA form a 2:1 potential divider, for the signal coming from the lower input OPA.

#### patpin

Joined Sep 15, 2012
401
hi pat,
Using a single 5V supply the Vout of the AD623 will not go below approx +0.15V.
This is with the Vref pin connected to 0V.

What is the common mode input voltage relative to 0V, on the input pins of the AD623.?

Eric
The input is 2.46V relative to GRND at both IN's when I do not measure any current with the probe and when I do measure 1 amp.the IN- is 1mV higher (2.47V) and IN+ 2.46 V

#### patpin

Joined Sep 15, 2012
401
Hi,
You must not let the Vref pin float, else the AD623 will give incorrect outputs
The two 50K's on the neg input of the output OPA form a 2:1 potential divider, for the signal coming from the lower input OPA.
Thats why I put REF at 2.5V with a voltage divider myself. Let it float was not an option for me.

#### ericgibbs

Joined Jan 29, 2010
18,865
hi,
OK.
What is the resistance of the current meter.?

#### Attachments

• 10.4 KB Views: 1

#### patpin

Joined Sep 15, 2012
401
hi,
OK.
What is the resistance of the current meter.?
I guess you mean at this output unconnected: 5.7Mohm

#### ericgibbs

Joined Jan 29, 2010
18,865
hi [atpin,
I meant the resistance of the meter that you are using to measure the Vout of the AD623, just in case your meter has a low resistance value and is loading the Vout.
Its unlikely that it is low, but otherwise the AD623 should be working as you are expecting.

I use the AD623 in a number of products/projects and they work just fine.

Is it connected as shown in this Sim.?

E

#### Attachments

• 28.9 KB Views: 16

#### patpin

Joined Sep 15, 2012
401
hi [atpin,
I meant the resistance of the meter that you are using to measure the Vout of the AD623, just in case your meter has a low resistance value and is loading the Vout.
Its unlikely that it is low, but otherwise the AD623 should be working as you are expecting.

I use the AD623 in a number of products/projects and they work just fine.

Is it connected as shown in this Sim.?

E
My DVM is 10Mohm; Thanks for diagram. i"ll give it a try. Should I include the 2 R's at the input as in yr Sim. Wont it suppresse the input coming from IC1a to much?

And second question: What will happen if I measure a negative current. Will I be able to display it on the Bitscope BS05 which I want to hang on the AD623 output.

Last edited: