We have a HP 1050 with a 4.6 X 250 cm C18 column. All chromatograms have baseline noise a the begining we have tried a back pressure regulator and using the mobile as our solvent to no avail.
Any sugestions?

Can you describe the "noise" ? also, is there any sign of pressure fluctuations ? I have seen this before on Beckman instruments with water to acetonitrile gradients and have found that microscopic bubbles in the flow cell are the cause. I would also try cleaning the check valves if you're seeing pressure changes.

If using TFA in the buffers and the disturbance is at the beginning of the gradient and appears as wavy baseline it depends on interactions between TFA and the gel. I have seen this on many chromatograms. This problem is mostly a visual disturbance and is not effecting the separation very much. Test to run without THF in buffer A. If the wavy baseline disappear I probably have guess right. You then have to try this:

1.lower TFA level in buffer A (to 0.060%, because you should not run without it)
2.test with 3% or 5% ACN in buffer A
3.Use wave length to 215 nm instead of 214 if possible (I do not no what you using today)

The chromatogram has a negitive peak then a positive spike. The pump module says the pressure is stable. We are runing ACN:Water 70:30 to 99:1 in 15 min.

What wave length are you monitoring at? Its not unusual to see baseline disturbance at the start of a gradient run if your monitoring at low (say 220nm) wavelength. Its a refractive index thing.Also try equilibrating the column longer before starting your next run. You could be seeing residual mixing of your end mp with beginning mp.

I hate to ask the dumb question but....What time is this happening at? could it be the peak from void volume? Also, maybe a posttime is needed based on what B.Buglio said (above) if there is a residual mixing problem.