THS4521: THS4521

Hi Rohit,

“Do you want to convert a single-ended bipolar signal of +/-2V to differential signal with unipolar output of 0 to 3V? If yes, then I would recommend supplying the op-amp - THS4521 with dual supplies because it helps in processing the negative portion of the input signal swing during the single-ended to differential conversion.”

No, I want to covert a unity gain buffer of a differential input signal of +/- 2V to differential output. Will the attached schematic work?

Mini DL-1 Page 2.pdf

Hi Neguse,

Although the attached schematic does perform the required function, I don't think it will work primarily because of the following reasons.

1. The supply voltage of +/-3V violates the absolute maximum rating of +/-2.75V (or +5.5V) for both the LMV794 and THS4521. I would recommend to bring the supply voltage down to +/-2.5V in-order to use either of these devices. 

2. The LMV794 is a de-compensated amplifier which is not stable for unity gain buffer application. Instead, I would recommend to use the OPA354 which can be configured as unity gain buffer offering improved performance.

3. The THS4521 Rf value of 12.1k will probably result in peaked-up frequency response. You could use an Rf of 1.2k and Rg of 1k instead in-order to get rid of the peaking in the frequency response.

Also, is there a reason for the R10 and R12 to be 10M-ohms? The LMV794 or OPA354 non-inverting input impedance is on the order of 1T-ohms which might make the R10/R12 of 10M-ohms as redundant.

Best Regards,

Rohit

Mini DL-1 with OPA354.TSCMini DL-1 with OPA354 and THS4521.TSC6278.Mini DL-1 Page 2.pdfHi Rohit,

When I simulated the unity gain buffer using OPA354 it seem to work, I am getting the correct output for +/- 1V DC, see attached file name (Mini DL-1 with OPA354) and Excel sheet file. However, I am having the following issues when I assembled the circuit on breadboard.

• Using TL7660 voltage converter , when I connect the -2.5V input power  to just one of the OPA354 IC in the unity gain buffer circuit, the voltage drop to about -2.2V and when connected to both IC the voltage drop to about -1.7V, why?

• Out of the unity gain buffer circuit, when input is +/- 1V the output is about+/-15mv, why? The output should equal to the input correct?  See attached schematic.

Regards,

Neguse G.

OPA354.xlsx

Hi Neguse,

1. It does not look like the TL7660 can provide the required load current to the amplifiers (OPA354/THS4521) when its input and required output voltage are equal. If you look in the TL7660 datasheet page 5, it shows the output voltage vs load current profile for the device. For symmetrically negative voltage conversion where Vout = -Vin, the TL7660 will source minimal load current and this happens to be your case with Vin = 2.5V and Vout = -2.5V. As a result, the OPA354 supplies will be current limited causing the supply voltage to reduce from -2.5V to -2.2V. Connecting another IC supply will further current limit the TL7660 output and cause the supply voltage to reduce even further to -1.7V. You can fix this by increasing the input voltage of the TL7660 to provide Vout = -2.5V depending upon the load current profile such that the amplifiers are not supply current limited. 

2. I think the second issue of the OPA354 output = +/-15mV for input = +/-1V is tied to the first issue where there is insufficient supply current for the op-amp. If you fix the 1st issue, then I believe you should be able to fix the reduced output swing issue of the OPA354.

The other thing is that if you think the OPA354 supply current is high for your application (4.9mA), then you might want to take a look at the OPA837 with 0.6mA supply current and is unity gain stable.

Best Regards,

Rohit

Hi Neguse,

You could use the below THS4521 circuit for converting a +/- 2V DC signal to 0-3V differential output. In the below schematic, I have used the OPA837 as input buffer amplifier instead of the OPA354 to save power. It is important to note that each of the THS4521 outputs swing 1.5Vpp giving a total of 3Vpp(diff) required by the ADC. Here is TINA-TI circuit for the same: THS4521_Diff-2-Diff.TSC

Best Regards,

Rohit

Hi Rohit,

Thank you, I simulated the circuit you suggested and looking at VG1 input  and VM1 output, the output is below 0, which it should not if you follow this formula.

Vout = 0.75Vin + 1.5V

Why is the output not shifted, another word when the input is-2V the output should be 3V and when the input is -2V the output should be 0V, what am I missing?

Regards,

Neguse G.THS4521_Diff-2-Diff (10).TSC

Hi Neguse,

The VM1 in the TINA circuit is a differential probe which cancels out any DC voltage between the two individual outputs - Vop and Vom. Since, the Vop and Vom are set to the same DC voltage of 1.5V, the VM1 output DC voltage will be centered around 0V and any AC swing will happen with respect to this. If you want to make sure that both the THS4521 outputs are centered at 1.5V, then I would recommend to look at each of the THS4521 outputs individually. As you can see in the above diagram, the Vop and Vom are both centered at 1.5V and the AC output swing happens with respect to this DC voltage.

The output swing in the THS4521 currently is 1.5Vpp single-ended at each outputs or 3Vpp differential centered at 1.5V DC. I have plotted below the 3Vpp differential output swing by not cancelling the 1.5V DC in-between the two outputs. Do you mean 3V for differential output or each of the THS4521 single-ended outputs?

Best Regards,

Rohit