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.

THS4541: Limits of Feedback Resistor

Part Number: THS4541
Other Parts Discussed in Thread: TINA-TI, THS4551

We intend to use the THS4541 in a differential-to-differential configuration between a mixer which has 200 Ohms output impedance and an ADC which has a high impedance input termination (>= 1 kOhms). The mixer should be terminated in 200 Ohms. The amplifier should operate at 3.3 V, have a a voltage gain of 16, and a large signal bandwidth of 10 MHz. Since the amplifier specified large signal bandwidth is 300 MHz at G = 2 and Vout = 2 Vpp (THS4541 datasheet page 8), the amplifier gain-bandwidth product should be sufficient for our application.

To match the amplifier input to the mixer output, Rg of 100 Ohms are needed. To obtain the voltage gain of 16, the resulting Rf value is 1600 Ohms. In the datasheet, however, the largest mentioned Rf is only 499 Ohms (p. 43, 47).

Our question: Can we expect the design to work to our needs (G = 16, B = 10 MHz), or are there any practical limits for Rf in terms of maintaining the gain-bandwidth product of this amplifier, e.g. stability issues? We are aware of the elevated output noise.

  • Hi,

    why not running a TINA-TI simulation?

    THS4541 TINA-TI Reference Design (Rev. B)

    Kai

  • Well a lot of that datasheet was single to differential oriented where the calculations include the active input impedance effect kind of unique to an FDA - that was where Table 6 came from where I generated that with Rf fixed at 402ohms. 

    Your application with a matching 100ohms on each side from 100ohm on the output of each mixer with Rf =1600ohm will be operating at a noise gain of 9. Since the GBP is around 800MHz, you will be getting <100MHz closed loop BW. It will work, and exceed your 10Mhz target. 

  • Also, you may need to think about the mixer output impedance through very high frequencies - if it starts to increase you may run into stability issues - then, use that compensation approach using two feedback caps and then a cap across the inputs to shape the noise gain up a bit at higher frequencies independent of the mixer output Zo. I show that as part of the attenuator discussion in the THS4551 datasheet figure 71

  • Hello Michael

    Thanks for your replies. A simulation of the circuit showed that at a voltage gain of 16, both the small and large signal bandwidths are still close to 100 MHz (much more than we need), and the circuit is stable.

    Regarding the mixer output impedance: Between the mixer and the amplifier we have an LC lowpass filter ending in a shunt capacitor. So the source impedance seen by the amplifier converges to zero at very high frequencies.

    websurfer

  • You bet, that filter Z going to zero at higher F is ideal, you should be fine,