LT1789-10 Maximum Gain + Input Limitations

Hi foetusmachine 

Congratulations on your interest and enthusiasm in the field of electronics, I wish you well, and keep up the good work! 

Regarding your inquiries, I have listed below my responses:

1st Question response: The LT1789-10 Spice was not appropriately modeled, as the Datasheet only guarantees that the Gain will work from 10 to 1,000. We have an online tool (In-Amp Diamond Plot Tool) to aid you during your simulation:

Diamond Plot online Tool: https://tools.analog.com/en/diamond/#difL=-1.45&difR=1.45&difSl=-1.45&gain=10&l=-13&pr=LT1789-10&r=12&sl=-13&tab=1&ty=2&vn=-15&vp=15&vr=0

2nd Question response: Upon checking your schematics, the two passive components were used to protect the LT1879 when for NO Dut state. The two have different functions;  the Zener was used to limit the input voltage and the Resistor was used to limit the current.

Thanks,

Thank you!

If I drive the LT1789 outside of the diamond, while it seems to simulate well at G=10000, are we saying the results may be unpredictable or is it perhaps that it will lock-up or damage the unit?

Do you have a recommendation for a replacement for the LT1789-10 that can deliver the 10/10000 gains i'm looking for?

Using the Diamond Plot Online Tool leads further into the second question.

While the DUT is not in place, I am not concerned with the output result from the LT1789. In this case, i'm only concerned about not sinking the 100mA current from the current source (to save energy) and to not exceed maximum limits on the LT1789. My real questions here are -

1. can I simply limit the current and not worry that i'm applying rail voltage to the input of the amp?

2. If I should be worried about inputting rail voltage to the inpuit, is the current scheme to drop the input voltage via the diode by about 1.1v acceptable?

3. Or should I be getting the input voltage down to around 0.23v to bring it into the Diamaon for my particular setup?

Thank you!

If I drive the LT1789 outside of the diamond, while it seems to simulate well at G=10000, are we saying the results may be unpredictable or is it perhaps that it will lock-up or damage the unit?

Do you have a recommendation for a replacement for the LT1789-10 that can deliver the 10/10000 gains i'm looking for?

Using the Diamond Plot Online Tool leads further into the second question.

While the DUT is not in place, I am not concerned with the output result from the LT1789. In this case, i'm only concerned about not sinking the 100mA current from the current source (to save energy) and to not exceed maximum limits on the LT1789. My real questions here are -

1. can I simply limit the current and not worry that i'm applying rail voltage to the input of the amp?

2. If I should be worried about inputting rail voltage to the inpuit, is the current scheme to drop the input voltage via the diode by about 1.1v acceptable?

3. Or should I be getting the input voltage down to around 0.23v to bring it into the Diamaon for my particular setup?

Hi foetusmachine 

Thank you for sending these follow up questions and please see my responses below:

Do you have a recommendation for a replacement for the LT1789-10 that can deliver the 10/10000 gains i'm looking for?

ArSol: A good alternatives are AD8421 and AD8222 if you are still using a similar Supply Voltage of 2.5V.

1. can I simply limit the current and not worry that i'm applying rail voltage to the input of the amp?

ArSol: The LT1789-10, AD8421, and AD8222 have a protection mechanism features internally, which can withstand beyond the Max/Abs ratings but it still recommended to include a resistor and diode to protect for any input overvoltage/ current events.

2. If I should be worried about inputting rail voltage to the input, is the current scheme to drop the input voltage via the diode by about 1.1v acceptable?

ArSol: If the applications still use a supply of 2.5V then, this is still within the recommended Input voltage Range for LT1789-10, and AD8222 and as the Datasheet Suggests.

3. Or should I be getting the input voltage down to around 0.23v to bring it into the Diamaon for my particular setup?

ArSol: From your applications; this will use a Gain =10,000 and Supply= 2.5V. I suggest bringing down your input voltage to <250uV.

To be able to give the proper recommendations, I would like to ask the following design requirements in your end:

1. Intended function for LT1789-10 or alternative In-Amp

2. Target Input voltage

3. Supply Voltage, single or dual configuration

Thanks,

Thank you.

The intended circuit is a micro and milliohm meter. The In-Amp at x10 will be for the milliohm range, the x10000 will be for the microohm range. The LTspice model I shared is a general concept.

When there is a DUT in place, the intended input range will be 0-200uV for the microohm range and 0-200mV for the milliohm range. When there is no DUT in place (kevin clamps sitting empty), the voltage at the inputs will be rail voltage minus whatever the "protection" diode can drop.

This is a battery operated single supply project. I'm currently sitting at 5v supply but am starting to move towards 9v.

I've had a go at modelling the LT1167 today in place of the LT1789, but i'm having a bit of trouble getting it to work.

I've been playing with the LT1167 a bit more, but i'm struggling a bit to get reliable results with it.

Firstly i'm struggling to get it to work with a single supply and a virtual ground, so i'm modelling it with a split supply for now.

I'm also seeing the output of the LT1167 being particularly sensitive to load - I started putting a 20k load on the output... to high. 5k load... too low. 10k load... seems to work, but that seems too sensitive to me.

Hi foetusmachine 

 Good alternatives are AD8421 and AD8222 if you are still using a similar Supply Voltage of 2.5V and above.

Hi Arsol Alright, so i'm slowly getting the message.

I've modified the circuit and worked in the AD8421. I believe I've set everything up as it should be but i'm getting some unusual results on the InAmp output. I'm not quite sure what i'm missing here.

See the attached LTspice model. Any idea where i'm going wrong?

Milliohm Meter13.5.asc

Hi foetusmachine 

Thank you for your feedback and for using AD8421 as the In-Amp alternative part; I reviewed your applications circuit and came up how to isolate where are the faults. please follow these recommendations carefully:

1. please isolate other circuitries connected after the DUT. disconnect the connection going to the AD8421.

2. insert a ground connection on the low side of the DUT resistor (100mOhm) or after R15.

3. Re run again the sim and the current flowing through DUT Res should be approximately 100mA.

4. The IR drop on DUT is 100mA x 100mOhm = 10mV. This Voltage should be then measured by the AD8421 depending on the Gain set. x10 or x100 and final Voutput must <9V Supply.

I hope this suggestion will help you.

Thanks

I think the issue is going to be that a virtual ground or basic rail splitter circuit is not going to be able to handle the 100mA to ground when the DUT is in place.

For the DUT to be tied to ground, the supply would need to be a true split supply - two batteries or a proper split rail regulator that can handle more than 100mA on the +ve rail.

Be benefit of my very original design using the LT1789 was that the DUT didn't need to be tied to ground.

Hi foetusmachine 

No need for the DUT to be tied to ground; Your overall circuit is not grounded properly and you may connect it to a chasis ground instead.

Thanks,