Part Number: PGA281
Other Parts Discussed in Thread: PGA855, PGA854
Hi We are using PGA281 in our design, attached the tina simulation and schematics, we want to know why traingular and square wave response has distortion at peak.PGA281_afb23_a1.TSC
Hello,
This behavior is occurring due to the bandwidth limitations at lower gains. Please see figure 16 in the datasheet as shown below.
I made a few adjustments to the simulation. One is the VCVS blocks were changed to 500mV so that the input signals are just changed differentially but not amplified on the input.
When I increased the gain and leaving the input signal the same, the output signal improved. Another way is the modify the input signal. The PGA281 is optimized for low frequency, slow moving input signals.
Gain = 1
Gain = 2
Please let me know if you have further questions.
Thank you!
Best regards,
Ashley
Hello,
I only have the PGA855 data right now, I will provide the PGA281 data by end of day Tuesday.
For all these charts, test conditions are as follows: VS = ±12V, LVDD = 5V, LVSS = GND, VOCM = 2.5V, Gain is 0.125V/V.
For a precise 7Vpp input the expected output is 875mV, I am not using a precision signal generator - therefore assume error in PGA855 is minor (equipment used is at fault for inaccuracies). I see around 950mV in my imprecise equipment.
I also tested a larger 10Vpp input, this is the max the signal generator can provide, expected output is 1.25V. I see around 1.34V in my imprecise equipment.
The input is the CH1 (yellow), OUT+ is CH2 (blue), OUT- is CH3 (pink), the Math channel (red) is CH2-CH3 (difference between OUT+ and OUT-).
For 7V input with a 25us rise time, the PGA855 provides an expected output:
For 10V input with a 25us rise time, the PGA855 provides an expected output:
For 10V input with a 5us rise time, the PGA855 provides an expected output:
For 10V input with a 1us rise time, the PGA855 provides an output that is being slightly reduced:
All the best,
Carolina
Hello,
As a follow-up, here is the PGA281 performance in the same conditions.
7V, 25us rise time
10V, 25us rise time
10V, 5us rise time
10V, 1us rise time
My recommendation is to go forward with the PGA855 instead of PGA281.
All the best,
Carolina
Hi Thanks for sharing Triangular wave results appreciated the efforts been spend on explaining the issue with PGA281. We need same graphs for Sinusoidal and Square wave too to compare PFA281 and PGA855. Also could you check in provided schematics(Deleted IP) what we doing, any thing wrong with respect to capture ADC data.
No problem:
Problem 1:
Problem 2:
Hi Carolina,
Currently what we need to validate is as below only,
our interest is with gain of 0.125/0.25 given input at PGA281 and what is the output at PGA281 and What we get output at LTC1566-1IS8#PBF(Antialiasing filter).
Here is the circuit,
The DAC results what is shared you can Ignore those are coming after FPGA, they are for our internal purpose to validate the input signals.
ADC --> FPGA --> DAC
Rgds
Chandra
Hi,
Thank you for your patience, here is the data for the PGA281 (with the conditions outlined above).
| Input signal (2Hz sine) | Input common mode | Input differential | PGA281 |
| 5Vpp | 2.5V | 2.5V |  |
| 7.4Vpp | 3.7V | 3.7V |  |
| 8Vpp | 4V | 4V |  |
As you can see, I am not able to replicate the test results on the EVM. I think the problem is the voltage on VSOP (pin 9) and VSON (pin 12) on the PGA281. Please double check this voltage in the testing.
All the best,
Carolina
Hi Carolina,
Thanks for your scope shots,
It seems we identified the issue who us causing the problem but still why this is problem we are not sure, Here below experiment we did,
This is the input of PGA and Output of the PGA with gain of 8. PGA seems working proper.
Any thing above 5V will starts clipping as expected VSOP is 5V in our case.
Now real problem after Antialiasing filter,
This is the capture at input of PGA and output of PGA for 40mV input.
This is the capture at input of PGA and output of Antialiasing filter(AAF). We have gain 4V/4 that is why do we have amplified signal at AAF output. But why it clips for >1.65V at AAF output?
If you see below anything more than 70mV, we start clipping at AAF output, refer below.
We see here it started clipping >1.65V output. Are we missing anything here.
My Oscilloscope having less resolution so ignore the CH1 reading measurements for Lower voltages(40mV, 70mV).
our application is DC coupled input.
Thanks Carolina,
This helped us a lot,
Need few clarifications of faster raise time response of PG855, Why do we see such behavior? Is it raise time is too fast for PG855? Can me adjust the raise time and test with 8V/10V, We wuld like to know for 1Mhz signal what could be raise time acceptable by PG855. Any input capacitance limiting our fast raise time?
Rgds
Chandra
Hi Chandra,
I think the limitation is internal to the device. If we consider Figure 7-41 from the datasheet, we see the full power bandwidth response from the output of the PGA855, by definition this is done with a sine wave.
If we look at the block diagram of the PGA855, the output stage is in a gain of 1 (see blue), and the input stage is in a non-inverting gain, whose gain equation is 1 + RF/RIN (see orange).
Although we cannot see what is in the gain network, we do know the minimum gain on the input stage is 1. The attenuation therefore occurs somewhere in the middle box (labeled gain network).
Assuming everything is in a gain of 1, the 8V input signal is bandwidth limited around 700kHz. As I mentioned earlier, this graph is for a sine wave, depending on process variation (the bandwidth can shift 30%), input signal, balanced inputs - I think it most apparent in the square wave but to any extent it is a device internal limitation.
All the best,
Carolina
