I built the one on the top of the list as it covers 3.5 to 5MHz which is the exact frequency of the VFO coverage, although being A LPF it's going to filter everything above so may not make that much difference, but I noticed a difference.
I built?NT7S 5-8 MHZ filter and I don't hear any real difference. But the signal into T2 is much prettier?than the square wave. I'll do some more testing later, not much can be done at 5:30 in the morning.
Sorry about the ugly square wave, you are right, we can't see it properly on a regular scope, but I was looking for a sine wave which I got after the LPF.
I don't know if we need a sine wave or a square wave either, but after experimentation I have found the receiver has a lower noise floor and generally sounds better using a LPF on the output of the Si5351. I'm not a Engineer just a amateur:-) and I am stumbling along as everyone else. In NT7S's antenna analyzer project, he found the Si5351 putting out harmonics, (and yes, he does have the proper equipment to see the output), and thus recommended using a LPF to filter them out. I would think the same principal applies to VFO's, we only want the VFO frequency mixing and not the harmonics. NT7S is one, if not the first person to test these chips out to see if they would be suitable in Ham radio use, and has written the library's that most of us are using in our Arduino Sketches to control the Si5351, so I listen to him very closely.
I'm not sure if you need a sinewave for the BITX40 - perhaps Farhan could comment on that. Many mixers work better with a squarewave drive. But I am not sure about the BITX40.?
QRP Labs does also produce a 7-element LPF kit see ?which does plug into the VFO kit too. So that could be used to get a sinewave output if necessary.?
I'm not sure why you would think the Si5351A output is an ugly squarewave Joel, it should be very clean! RF squarewaves are VERY difficult (nearly impossible) to view accurately on an oscilloscope.?
The output from the Si5351 is a square wave and a rather ugly one at that, so yes it's a good idea to put a LPF between the VFO and the transceiver which will convert it to a sine wave. I did it and noticed a much lower noise floor on receive, no doubt because of removing harmonics.