circuit idea for automatic gain controlled amplifier(AGC)

Hello,

Could you provide more information on your transimpedance application, such as transimpedance gain, bandwidth, photodiode characteristics (capacitance that will be placed at the input of the amplifier), etc...?

What gain range are you looking for?  What other constraints in your system do you have (power supply available, etc...)?

Thanks,

Hello,

           Thanks for reply Mr.Xavier Ramus.I want to give  my application details. I am using discrete IR emitter & PIN photodiode @ wavelength of 850-950nm.The emitter is driven @ 20mA  of continous constant current source(not pulsed current).The emitter & detector is placed inline @ distance of 2cm.Actually that  emitter & detector is used as transmissive type sensor for sensing the yarn.Please consider the yarn is in Y-axis & beam path is in X-axis.Kindly note that the yarn diameter is 0.25mm. My requirement is to findout  wheather yarn is present or not.And also have to find the yarn is in static or moving position.So The photo diode current has to be processed  as per those physical information & have to provide the logical signal.In my circuit design, I am using LM324 Quad opamp.

          My problem is,The sensor(i.e sensor itself & sensor to sensor) has wide dynamic range of value in photo current  over one decade @ without object.So i deceided to include the AGC or VGA in my circuit.

         Transimpedance gain is 20KOhm, required bandwidth 10K-20KHz(for moving yarn detection),Photodiode capacitance is 11pF,single power supply from 5- 15V.

Thanks & Regards

Vasanthi Ranganthan

Hello Vasanthi,

As with many designs, there are multiple approaches to accomplish the general design requirements.  Based on the description above, it sounds as though your present focus is to convert the output of your sensor to indicate a condition where there is no interruption in the IR path to the sensor by the yarn indicating that the yarn is off of its desired path.  You mentioned that there is a wide dynamic range from the sensor when the yarn is not in its proper position.  If a typical transimpedance approach using a single resistor value does not accommodate the full output range of the sensor based on the variance in signal due to yarn position, a variable type of current conversion / transimpedance gain may be useful. 

The level of complexity may change depending upon the level of variance from the sensor.  A couple of approaches could be to use a typical transimpedance approach using multiple feedback resistors in parallel that are switched in and out based on whether the output of the amplifier is saturated or not.  This one may be the simplest approach to designing to an input signal that is just beyond what is capable of achieving with a single gain resistor.  Another approach would be to use a diode and resistor in series as an additional feedback path to essentially switch in an additional feedback path to reduce the feedback impedance value at a certain point.  This essentially creates multiple gains based on the output voltage.  This approach will not have the same circuit-to-circuit control that the switched in resistors and will require the diode conduction point be evaluated and determined based on the application characteristics.

There are likely additional methods that would also work but these are a couple that should be able to get you started down the path of the variable gain capability for your circuit.

Mr.Scott Hill,

                    Thank you Mr.Scott Hill.I want to give the explanation regarding my existing circuit design concept.Due to interruption of yarn in beam path, some photon loss will occur,It is very samll amount(because the yarn diameter is 0.25mm).This is the information to findout the yarn  present or not interms of electrical signal (i,e photocurrent).That means consider if photo current is 100uA without yarn, While interrupting  yarn the photocurrent  will reduce to 95uA.I  amplify that  window level of samll change in  signal from 85uA-100uA level as of 0-15uA by discarding(or subracting) upto 85uA from 100uA .From there it is converted as logical signal.These process is done after converting the photocurrent into voltage by transimpedance amplifier.

                My problem is, That photocurrent is not @ level of 100uA (without yarn). That means this level is not same for all sensors( it will vary  sensor to sensor  as wide dynamic range).And also the sensor itself that level will vary as per the ambient condition( that change in current  around 5uA)  without yarn. I want to know how to impement in my circuit, the idea which you have given above?  And how to process the signal as per my application with those sensor difficulties?I need some other circuit idea  also for my application.

Thanks & Regards

Vasanthi Ranganathan

Hello Vasanthi,

Thanks for the added application information.  It does give me some additional ideas as to how you could possibly go about this.  What will be a tricky issue is how to adjust the circuit to account for the circuit-to-circuit variances with the light source and sensor as well as how the sensor responds over temperature.

Bottom line here is that if the inherent errors in the sensing due to circuit and temperature variances is unacceptable, the circuit has to be calibrated in order to remove these error sources.  Many sensors have fairly consistent temperature characteristics from sensor to sensor.  If your sensors exhibit similar behaviour over temperature, you could characterize a sensor over temperature to create a look-up table.  Using a temperature sensor to measure the sensor temperature, you could then correct for the sensor variance due to the temperature shift.

The variance in the no-yarn condition ~100uA signal would likely have to be calibrated on a sensor by sensor basis in order to calibrate the offset(shift from 100uA) for each circuit.  If the deviation from no-yarn to yarn current was always 5uA or some other constant value, simply knowing the deviation from the no-yarn sensor current from 100uA, the offset could be removed from the output signal through software correction.

Another approach you may want to look at is something similar to the LOG114.  This could allow you to build in a reference current which would then allow for the subtraction of the difference from 100uA, convert to voltage, and then amplify the desired signal to a usable voltage output range.

Mr.Scott Hill

          Thanks for reply.The current withstanding capability of emitter is upto  maximum 100mA of continuous current.I am driving the emitter by less than 20mA. The emitter does not dissipate heat energy in its body and   I tested with temperature sensor in emitter side .So  I identified that behaviour is not mainly over temperature.I want to list here reasons for my sensor difficulties in my point of view, I think generally the variencies will  occur in intensity of light beam from source itself(that means emitter to emitter), due to skew in assembling the sensor in PCB board and the light beam will process with space(i,e air and ambient condition) while travelling from emitter to detector.I think these are all self aspects of optical signal.

         I tried to design a circuit by using log amplifier.Log amplifier will not suitable for my application.Because it compress the signal, so the samll change in photocurrent due to interruption of yarn will not reflect in electrical signal.

          My constraint is, sensor should have the capable of self calibrating or auto calibrating option and I should use positive power supply only.

          Actually in my application the input is physical action(i,e yarn present or not),it has to be converted interms of electrically logical signal.My tricky issue is, change in photo current due to interruption of yarn is very less  compared with the change in photocurrent due to the self aspects of optical signal.I should extract my input information from unwanted information in electrical signal.So my circuit should capable to tackle this issue.

           I have idea about some discrete blocks of circuit conceptual as per my application, as follows. I  request to suggest  best circuit idea with refence of my explanations and suggest ICs from TI to fullfill the circuit for my application.

1. Variable threshold voltage  circuit concept
2. Automatioc gain contol amplifer concept
3. Voltage controlled gain amplifier concept
4. By using transconductance amplifier concept( for VCA,AGC etc)
5. By using Analog switch ICs 
6. By using Photocurrent monitoring ICs

Thanks & Regards

Vasanthi.R

Hello Mr.Scott Hill,

           Thank you for suggestion.The application requirement can be fullfilled by combining the aspects of pysical parameter & electrical parameter(interms of electronic circuits) , by the way the cabability of system can be defined inorder to address the tricky issues.

My existing circuit concept is same as what you said.I unable to understand the explanation of calibration circuitry which you mentioned.Please give the detailed explanation.

Regards

Vasanthi.R

hi

can u  give details of your IR transmitter     ( CW or Pulsed and PRF, power etc)

rgd

yash