DIY fan/heatsink combos

A specific point I'd like to bring up, if you're venting X CFM you don't want your lamps total CFM going over that amount. It might result in hot air circulating back down below the canopy. I've been playing around with some heatsinkusa sinks and it's something that entered my mind and thought I would share.

I was originally going to put a 100CFM fan on each lamp (4 emitters per lamp) for 25 CFM per heatsink, but this got me to thinking. I'm venting my 2x4 with a 150CFM duct booster. Works great with the passive cooled lamp I'm currently using. Throwing 200 CFM from fans with high static pressure ratings up into the air would likely result in a down draft. For this reason I've ordered several fans for testing purposes. I'm going to test both the 4.6" and 3.945" profile heat sinks from heatsinkusa to determine which by cost is better with a fan, and then try to determine how low I can go with the CFMs and perhaps add an inch or two on the sinks if needed. It will be a week or two but I will be happy to post my results here.

Any other advice on custom building fan/sink combos?

Interesting thought and I don't really have an answer.
I'm just going to think out loud because you have my brain going now...

-I would think that the volume of your room would be very important and be the biggest factor. I have much more air being moved around by circulating fans and led fans than I am exchanging through the exhaust. And I use passive intake.
-if a fans mounted to the heatsink...it should be exhausting/blowing down the channels to the sides more than down, if not bouncing off and up slightly without a fin guard on. The heatsink is blocking down draft.
-the fans should not be pulling additional air into the environment and changing the negative pressure as long as the room volume is greater than the air movement inside from the fans.

That's a good point. I assumed with fans sitting on heatsinks they would pull from the sides but pushing air out the sides would cause some recirculation rather than upward pressure and the overall flow of air would remain similar.

Problem solved.

Anyway, the experiment goes like this: I have a 4 inch piece of the 4.6" profile heatsink and a 3 inch piece of the 3.945" heatsink. The 3.6" was a total of $9 and the 3.945" was $9.81, so similar price and similar weight. I will mount a CXB3070 to each and power it at 1.4A (49 watts). A duct will be made of cardboard for each heatsink. A 25CFM fan will blow air into the duct at a right angle and force it through the heatsinks. Tj at the test point will be measure at 30 minutes for each sink.

Once I've determined which heatsink works best in that scenario I will attempt to determine how low the fan can go and maintain reasonable temps. This should give me a rough idea of what type of fan power will be required per emitter.

The CXB3070 arrived today. It's tiny compared to the Vero29. I have the sinks and a couple fans to play with. I'm still waiting on the holder and a driver. I have a 100CFM fan to play with as well, so am seriously considering ordering 3 more cobs to see if 100CFM will cool them all @1.4A. I'm not sure if 1 being cooled by a 25CFM fan will indicate anything other than what it is. There's also the issue of enclosure design. 4 sinks in a straight duct would cause more pressure (single vent) but 4 in an enclosure with 4 vents would split the airflow. I have no idea how the math regarding these issues would work.