LMH5401: NG calculation

Matias

Part Number: LMH5401
Other Parts Discussed in Thread: THS3217

Hello,

I'm working with an LMH5401 for a 500MHz application where I have a gain of 7V/V. I was reading the datasheet where the stability considerations are described (section 9.1.1) and it's said that the NG must be smaller than 3.

Where NG = (Rf/Rin) +1.

In section 8.3.3.1 it's said that Rm combines in parallel to the sum of the two Rg resistors in order to obtain the differential input impedance. If I follow those guidelines to calculate the NG of the example given in section 9.1.1 next to figure 63 I don't get a NG of 5 for that 2 gain circuit as it's pointed out in the datasheet.

In a nutshell I'd like to calculate the NG for the circuit in the figure below to make sure it'll be stable or if a parallel resistor will be needed. I ran a PSpice simulation of that and it seems to work fine, but I want to make sure it works once I implement it on a PCB.

Thank you,

Matias

over 2 years ago

0 Sima Jalaleddine over 2 years ago

TI__Mastermind 25665 points

Hello Matias,

That is correct, this amplifier is slightly decompensated and recommended to use gain equal to or greater than 2V/V. Lower gain configurations can be used by compensating the amplifier as described in the same section. The parallel resistance won't be needed in your case because the above configuration is greater than a nose gain of 2V/V. There isn't an issue to use noise gains greater than 2 or 3V/V other than limiting your bandwidth. The noise gain curve will intersect the AOL at a lower frequency the higher the noise gain.

To determine the stability of the circuit, you would need to do an open loop analysis. This video walks through an example with FDAs. I can help with this circuit; just a few more questions. What is the loading to the circuit, and is the source impedance 100 ohms differential input?

Thank you,

Sima

0 Matias over 2 years ago in reply to Sima Jalaleddine

TI__Prodigy 20 points

Hello Sima,

Thank you for your reply.

The loading of the circuit is first another LMH5401 (I'm actually using two equal stages) and then it goes into a simple envelope detector with a Schottky diode. The source impedance for the first stage is purely capacitive.

The second LMH5401 stage goes into the following circuit:

Where the THS3217 has an equivalent input impedance in differential mode of 50 || 2.4 k [Ω || pF].

Thank you,

Matias

0 kai klaas69 over 2 years ago in reply to Matias

Guru 140200 points

Hi Matias,

the source impedance is purely capacitive? Is it symmetrically operating, being connected to the two 50R resistors? Or are there also common mode capacitances present, being referenced to signal ground?

Kai

0 Matias over 2 years ago in reply to kai klaas69

TI__Prodigy 20 points

Hi Kai,

Yes, it's purely capacitive (or at least mainly),at least that's what I was told, sorry for not being able to provide further details. Yes it's symmetrically operating.

Thanks,

Matias

0 kai klaas69 over 2 years ago in reply to Matias

Guru 140200 points

Hi Matias,

like that?

matias_lmh5401.TSC

The envelope follower cap looks a bit small:

Kai

0 Matias over 2 years ago in reply to kai klaas69

TI__Prodigy 20 points

Hi Kai,

I'm using PSpice for TI and for the diode I imported the model for a BAT15-03W (by infineon). The signal I get at the output of the envelope detector looks like this: