The main point here is what kind of sample do You have. And what is the origin of the T1 noise, is it the solvent?

T1 noise is at best reduced in echo-antiecho sequences with gradient filters. It is also recommended to use ns=2 and turn on the trim pulse to clear out the unwanted magnetization. Of course you will get T1 noise if you use phase-cycled versions instead of gradient-filtered.

I, myself do not have any T1 noise in HSQCs on cryoprobes (see attch.).

Adiabatic pulses are used to get the uniform phase of all peaks and to invert all chemical shifts even at the edges of the spectrum in 13C dimension.

Phase sensitive - is not a problem it is a necessary thing in HSQC sequence. All good pulse sequences are phase sensitive - it is the method of detection. 

I use hsqcedetgpsisp2.4 for low-MW compounds. It allows the maximal sensitivity improvement and the best phases.

For protein aliphatics I use hsqcetgp with some modifications that include the decoupling of CO and aromatic CG. Or constant-time HSQC (hsqcctetgpsp)

For aromatic HSQC of proteins use aromatic TROSY.

I am having a very good experience using the hmsIST reconstruction method (http://gwagner.med.harvard.edu/intranet/hmsIST/). The method enables the shuffling of the acquisition of the indirect dimension.  Shuffling compensates of the hamonic temperature oscillation. Sinusoidal temperature oscillation are frequently the source of the T1 noise. Even if the temperature control of the room is within the specified range (+-0.5oC), if it is harmonic it is problematic.

Hello Konstantin and Fabio,

Thanks for your feedback. I will give the hmsIST method a try when I find the time. It seems like it has a great deal of potential.

However, I am very interested in the hsqcedetgpsisp2.4 pulse sequence as well. I have run one of these pulses and it looks like a fantastic reduction of noise in the spectra. I believe the T1 noise is coming from the solvent (or rather the water taken in by the solvent), as we are running in DMSO. Despite my best attempts at keeping everything dry, baked, and under argon, it seems like the moisture is creeping in and causing a problem in the 2D experiments.

The reason I ask about phase sensitive variants is because there is the fear that, especially in mixtures or extracts, overlapping peaks with close to the same shifts in the carbon dimension may cause a (+/-) cancelation and reduce signal. However, this is just a caution, as I have not yet encountered this as a problem worth stressing over.

Thanks for the answers, I appreciate the insight.

- Joe

Experimentally, if you have even a small amount of water in the sample, there will be some T1 noise. There are a couple of ways to reduce it through processing.

1) On Topspin, you can use 'qfil' in the direct dimension BCMOD, the kills the water T1 noise. 

2) MestreNova and ACDLabs Spectrus have a "Reduce T1 noise" option. But all these methods will mess with your actual peaks a little bit.

What Fabio above suggested sounds like NUS sampling, and could be very useful, but you will need a few trials to get things right and ake sure all the  signals are there.

In our lab we modify some of the sequences to add a water suppression step. As for enhancing sensitivity, there's a fairly new sequence that really cleans up the HSQC spectrum by mirroring all CH2 signals about the O2P and putting them in an otherwise blank region of the spectrum. In effect, there is complete removal of overlap between CH3/CH and CH2 signal, without the risk of signal cancellation as in traditional edited-HSQC, without the need for increasing digital resolution. We have found it to be quite useful for making more accurate assignments. http://www.sciencedirect.com/science/article/pii/S1090780715001585