LMH6702QML-SP: Highspeed comparator circuit

This device does not have rail-to-rail I/Os, so I'm not sure if it can handle your input and desired output signals.

Please specify the supply, reference, and input voltages.

There are fast comparators like the TLV3801, but they have LVDS output. What are you going to do with the output signal?

The input signal is going to be a triangular waveform @ 500MHz LVDS 1.8V.

The output signal should be a square waveform @ 500MHz and it is going to be applied to a 50ohm load, we can amplify the level using RF amplifier in order to achive a certain level.

The idea is to use the output signal (square) as a ON/OFF singal @ 500MHz.

Thanks.

Pablo.

Dear Clemens,

Thinking a bit more, first we need to convert the LVDS signal into a single-end in order to compare with a reference level (below you can see the block diagram proposed)

1) Convert LVDS 1.8Vpp triangular waveform @ 500MHz into a single-end signal --> Use the LVDS Receiver SN55LVCP22-SP (>1000Mbps), using only 1 of its outputs.

2) Convert from triangular to square waveform -->Use the Fully Differential Amplifier LMH5401-SP (slew rate = 17500V/us & GBW = 6500MHz), so a 1.8V/17500V = 0.1ns (500MHz period is 2ns)

Our application requires space grade components, that is the reason why we have chosen those references.

Thank you very much in advanced.

Best.

Pablo

LVDS is a binary protocol; a triangular LVDS waveform does not make sense. Do you mean that the FPGA outputs a differential signal?

An LVDS receiver works like a comparator (and does not output a triangular waveform); it is not necessary to have two devices.

Is your load a 50 Ω resistor, or do you have a transmission line with a characteristic impedance of 50 Ω? In the first case, what are the required voltages at the receiver? And what supply voltages do you have available?

Dear Clemens,

Thank you for quick reply.

Yes, FPGA outputs a differential signal @ 500MHz that we need to convert in a square single-ended signal @ 500MHz, so we need to convert the differential signal to single-ended (maybe with a balun).

The output will feed a transmission line with an impedance of 50ohm (SMA connector)

Regarding voltages,

Supply voltages : positive & negative up to 15V.

Reference voltages: up to 15V

Input voltages: up to 1.8V

Maybe using a balun +  LMH6702QML-SP or LMH5401-SP could be enough.

Thank you very much in advanced.

Best.

Pablo.

Hello Pablo,

   Current-feedback amplifiers have a different input structure than voltage feedback amplifiers. Their input impedance at the inverting input of this architecture is low, and shouldn't be used in a comparator configuration.

    In the last post, you have the triangular to square wave conversion using just the LMH5401-SP instead. Is this the solution you are looking at now? If so, I will move this over to RFAMPs for further support on this device.

   I was looking at our available space rated comparators, and our fastest dedicated comparator is around 25ns if you are considering going down to this level: https://www.ti.com/product/LM119QML-SP. 

   Also, looking at our VFB space-rated high speed amplifiers, THS4304-SP is another choice that can be considered which is right around what you need, but with little to no headroom unless you can attenuate the signal prior. Also, when using op-amps as comparators, you have to watch out for differential voltage rating in abs max, if this is present this does suggest that there are back-to-back clamping diodes between the amplifier's inputs which would make it more suspectable to damage when used in comparator configuration. The THS4304-SP does have this rating which is specified at abs max differential input voltage of +/-2V. Here is an in-depth video of this topic on using op amps as comparators: video link. Also, here is a quick blog on this topic, but for some reason the official link is not working, here is a cached version. 

Thank you,
Sima 

Dear Sima,

Thank you for your reply.
The idea of use th LMH5401-SP was due its high slew rate (17500V/us), it could be a possible solution yes.

Best.

Hello Pablo,

  I was looking into past support on this type of application, and it has been recommended that LMH5401 not be used in open-loop configuration, but another solution would be maybe trying out saturating the amplifier. It will be in non-linear operation in the period of saturation, but it will be more predictable than using no negative feedback.

https://e2e.ti.com/support/amplifiers-group/amplifiers/f/amplifiers-forum/390432/lmh5401-as-a-comparator

5633.LMH5401_Comparator.TSC

Thank you,
Sima