You don't say whether or not the image is washing out, but since everyone else seems to assume you're underexposing when using the vellum, why not try it a double your usual exposure? If that won't wash out, drop it down to 4-1/2 minutes. If it all washes out, go to 12 minutes. The idea of doubling or halving times is one used in photography and brings an element of consistency to testing, rather than just picking a number at random, and when that doesn't work, pull another one out of thin air.
You should really invest in an exposure testing sheet. There are some free ones, but a good one is the one made by Chromaline. It will get you in the ballpark with one exposure test. You'll need to run exposure tests every time you change emulsions anyway, so the test sheet just makes the job easier. Technically, you should also do it for each mesh count, but what I've done with success is do my tests on a 195 or 230 mesh, which is in the middle range of mesh counts I use, and then extrapolate what should be a good exposure time for mesh counts above and below that. I then fine-tune them base on the amount of unexposed emulsion left on the squeegie side of the mesh at washout time.