OPA835

Eli,

±2.7V is a total supply of 5.4V. The absolute max rating is 5.5V, so this is pushing close to that boundary.

As far as THD is concerned, it is given in the datasheet in multiple places. In the 5V electrical characteristics section on page 6, it is given directly as -130dBc at 1kHz for 1Vrms (or 2.828Vpp).

THD can also be approximated from the first two harmonics with the following formula.

(Note: The more harmonics you include, the more accurate your THD number will be.)

Using this equation and estimations from figure 35 (which shows harmonic distortion vs output voltage), you can come up with the following results.

Granted these results are for 1MHz which may or may not be the frequency of interest. However, because the amplifier is linear, you can combine these results with figure 33 to approximate what the THD should be at lower frequencies. The only important thing to note is that around 2Vpp HD3 starts to behave poorly most likely due to slew rate (which isn't going to track linearly over frequency) so this method should really be limited to <2Vpp to get good approximations. So for your purposes, the numbers on figure 35 are probably going to be your best bet.

Always the best method is to build up a board and test it. This removes the uncertainty altogether and gives you real results for your configuration. There is a module you can order from the e-store, OPA835IDBVEVM, that comes fully populated so you can do quick testing or system level prototyping with your exact configuration/power levels.

Let us know if we can be of further help.

Regards,
Luke Lapointe
High Speed Amplifiers

Hi Luke,

See below customer feedback:

1)

We should take the graph related to Vs=5V, page 19, Figure 32.

Situation is better, considering at 4V_PP the change is not dramatic. (Figure 34) 

Note that we have 2 amps in each stage, meaning that available V_PP is double

2)

A. can we have from TI more detailed HD data in the form of function HD_2,3 = f (V_O, Fr., R_LOAD) or a table? 

B. I Checked OPA835 Spice model TI provided me in the past – I guess it is not supposed to model HD.

        I simulated it in typical conditions and saw only odd harmonies. (3^rd, 5^th, etc.) 

does TI have a dedicated Spice model for HD simulation?

 Eli

Hi Luke

Can we get your e-maill address for additional questions?

B.regards

Eli

Hello Eli,

Let me quickly address your follow up questions.

eli said:

1)

We should take the graph related to Vs=5V, page 19, Figure 32.

Situation is better, considering at 4V_PP the change is not dramatic. (Figure 34) 

Note that we have 2 amps in each stage, meaning that available V_PP is double

I'm not exactly sure what you mean, figure 32 is output overdrive recovery..  I also don't know what you mean by the voltage being double with 2 amps. Do you mean they are in series? Can you post a schematic?

eli said:

2)

A. can we have from TI more detailed HD data in the form of function HD_2,3 = f (V_O, Fr., R_LOAD) or a table? 

There are plots already in the datasheet with these comparisons.

Figure 33: HD_2,3 vs Fr
Figure 34: HD_2,3 vs R_LOAD
Figure 35: HD_2,3 vs V_O

Using all 3 you can extrapolate in their linear regions. For example, assume you want to know the HD₂ at 100kHz with 4Vpp.

From figure 33:

HD₂ @ 1MHz = -70dBc
HD₂ @ 100kHz = -105dBc

35dBc improvement

From figure 35: HD₂ @ 4Vpp = -66dBc

Add the 35dBc from before and you should get roughly -101dBc for HD₂. Granted this isn't a perfect method, but it gets you close. 

eli said:

B. I Checked OPA835 Spice model TI provided me in the past – I guess it is not supposed to model HD.

        I simulated it in typical conditions and saw only odd harmonies. (3^rd, 5^th, etc.) 

does TI have a dedicated Spice model for HD simulation?

We typically don't include HD simulations in SPICE due to the simplistic nature of SPICE. The best approach to getting results in your config is to either build a prototype and test or pickup one of the OPA835IDBVEVM's from the e-store (https://estore.ti.com/OPA835IDBVEVM-OPA835-in-DBV-package-EVM-P2432.aspx). These will typically ship within 1-2 business days and come fully populated so you can test it right out of the box or modify it to match your config. I would recommend this method to get the most reliable data.

Let me know if there are any other questions.

Regards,
Luke Lapointe
High Speed Amplifiers