This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LMH6702 Inverting Gain Circuit Unexpected Peek Frequency Response

Other Parts Discussed in Thread: LMH6702, LMH730227

Dear Sir,

Using Figure 25. Recommendation Inverting Gain Circuit in datasheet, a customer tested and found unexpected peek frequency response at 1.5GHz (+10dB) when Rin 25ohm is changed to 0ohm (pin3 is connected to ground). There is no peek observed when 25ohm is existed. May I ask why?

The resistance network values are;

Rg=Rf=240 ohm and Rt=62 ohm. (Gain=-1)  

This phenomenon was not seen on TINA Simulation but actual measurement. 

If you gave me your email address I can send actual measurement data. 

Best regards,

Masa

  • Hi Masa,

    Are you using the LMH6702 EVAL board, or the customer's own board? With this device, the board / routing could be a factor in the results.

    Also, most of the datasheet curves were done with RL= 100ohm. Is your device loaded with 100ohm? For best results, you could use a 50ohm series output resistance which would then feed the scope terminated in 50ohm (for RL=100ohm).

    I assume that you have short output traces / leads and minimal capacitive loading.

    You may be able to reduce the peaking / bandwidth by raising the value of RF (since the LMH6702 is a Current Feedback architecture device), and RG (to keep the gain constant).

    • As far as the termination of the non-inverting input affecting the frequency response:

    With some of the ultra-fast devices, this is not unusual. Figure 25 shows the configuration which was used in gathering the Typical Performance plots of the datasheet.

    Regards,

    Hooman

  • Hi Hooman,

    Thank you very much for your support.

    The test is performed using SOIC type LMH730227 EVM for LMH6702.  50ohm series output resistances (one is for scope terminated in 50ohm and total RL=100ohm) is used.

    In the figure 25, what is the purpose of 25 ohm resistance?

    I think that Frequency Response will not be changed by resistance on non-inverting input, correct? I just want to understand the theory/idea. 

    Is there any specific reason to LMH6702?

    Best regards,

    Masa

  • Hi Masa,

    The non-inverting input of the LMH6702 (and also other current feedback devices) is a gain of x1 buffer (similar to an emitter follower). The closest analog I can think of to explain the impact of 25ohm to ground in Figure 25, is how an emitter follower can peak / oscillate unless some finite base resistance is inserted in series, as shown here:

    https://classes.soe.ucsc.edu/ee171/.../Lab3_Bipolar%20Transistors.pdf

    Hope this answers your question.

    Regards,

    Hooman

  • Hi Hooman,

    Thank you very much for your explanation.  

    It is known that adding a small value resistance to base of emitter follower can prevent from an unexpected Colpitts Oscillator caused by large collector current  that increases Tr gm as well. I could realized that current feedback OPA has a similar situation as you taught.

    Best regards,

    Masa