OPA2835: Crosstalk vs frequency

Morning, 

It is hard to tell specifics on this particular plot and for a part that old, the much new OPA2837 shows kind of an incredible plot - essentially, these are network analyzer measurement driving one channel but measuring what comes through on the other, variables include - 

1. gain setting for both channels, 

2. loading

3. layout (using EVM here probably)

4. supply decoupling, 

The details might be more accessible for the OPA2837 plot - here, 

Hello Steffes-san,

I'm sorry, there are some points I could not understand.

>It is hard to tell specifics on this particular plot and for a part that old, the much new OPA2837 shows kind of an incredible plot 

Do you mean that measurement environment is different b/w OPA2835 and OPA2837, and it is hard to explain environment of OPA2835 ?

>essentially, these are network analyzer measurement driving one channel but measuring what comes through on the other, variables include - 

>1. gain setting for both channels, 

>2. loading

>3. layout (using EVM here probably)

>4. supply decoupling,

I could not understand what "variables" mean for. I also can not understand input signal and how you configure gain setting(According to note of figure, it seems that gain is set as unity...) and so on.

I do not require to provide exact circuit such as including constant value. But could you please tell me more detail about above ?

Best Regards,

What I was trying to say was that the more recent OPA2837 work might still have someone in the the product group that could answer these specific questions, the OPA2835 was probably similar, but done so long ago no one that might have done that work to ask at this point, 

This type of question can only be answered definitively by the product group, continue you discussion at this link, 

hsamps_apps@list.ti.com

Hello,

Measuring crosstalk for the OPA2835 is exactly as Micheal mentioned mentioned. Where measurements were done through a network analyzer, with one channel being driven and the other being measured. In this case data was captured using two analyzers for different frequency ranges.

For data from 20Hz to 115kHz the following parameters were used: 1Vrms input, G = 1, Vs = 5V, RL = 100k

For data from 130k to 50MHz  the following parameters were used: 2Vpp Input, G=1, Vs=5V, RL=60

Best,

Hasan Babiker

Hello Hasan-san and Steffes-san,

Thank you for your reply.

I understood that OPA2835 is little old device.

And I also checked OPAx83x series(OPA2837, OPA838, OPA2836), however OPA2835 is lowest Iq product. So I selected.

* Are test condition of both "1Vrms input, G = 1, Vs = 5V, RL = 100k" and "2Vpp Input, G=1, Vs=5V, RL=60" for driven channel ?

  Does measured channel not connect any load and input (just connect network analyzer to output )?

* Is above test condition for OPA2835 not OPA2837 ?

*  "In this case data was captured using two analyzers for different frequency ranges."

Did you use two network analyzer ? Did above not mean you used two frequency range with one network analyzer ?

Sorry for asking several questions, however please check about above questions.

Or, should I use mailing list which Steffes-san attached to continue this topic, please let me know.

Best Regards,

Hello Machida,

1) Yes the information provided was for the driven channel. The channel that is being measured can be put in unity gain and set to midsupply so that it is within the linear range for both input and output. 

2) The information I provided was for the OPA2835.

3) No, two different network analyzers were used for the test. 

No problem, I hope this clarified things. The discussion can be continued on this thread.

Best,

Hasan Babiker

Hello Hasan-san,

Thank you for your reply.

I understood that following configuration was used to confirm crosstalk vs frequency from your reply.

Is my understanding correct ?

Thanks in advance,

hello Ryuuichi, 

Some quick notes, 

That OPA2835 crosstalk is pretty odd changing loads across frequency

you might want to look at the brand new OPA2834, TINA model here, 

OPA2834.tsc