I just wanted to bring backup this topic because it kind of died off when it was getting interesting.
When I read, The Reef Aquarium, Vol. 3 by Julian Sprung, J. Charles Delbeek, in regards to RDSB they stated that a V shaped settling chamber could be used before the water enters the DSB chamber. You have to have the chamber large enough to allow the water to flow slow enough that detritus will settle to the bottom where it can be easily cleaned during weekly or bi-weekly water changes. Also, they stated that there was much success when the per hour flow was between three and five times the volume of the tank. They also stated that a size of approx 1/3 the volume of the tank seemed to be the standard that had measurable performance increases and still easily achievable from an engineering standpoint. The also seemed to only mention systems that had 10" of sand, but perhaps that is all I remember. They did bring up the hydrogen sulfide issue and never really resolved it saying that some people had great success with their 10" beds.
For my 90GAL:
This is the plan I have with the addition of a separate chaeto chamber. The first chamber is a V shape near a 40 degree angle and approx 10" wide and 18" deep. It will have a very small number of bio-balls at the top to break up detritus as it enters the settling chamber. The next 16" x 18" will be a 10" deep sand bed. The final 10" chamber will be the return and chaeto chamber. This will be lit 18 on and 6 off. I may make room for a small skimmer too, I do not know yet. I currently have an HOB I will probably run half the week. I plan to use a 40gal breeder for this. The flow rate will be approx 450gph, I am opting for the larger flow rate because I am over the 1/3 mark.
I believe this method solves the problems of having a need to clean the DSB for any reason and still removing excess detrius before it even reaches the DSB without using screens, filters etc which need constant attention even when you go on vacation, cost money, and are traps for bacteria. I could even ditch the bio-balls probably and have the same results. (This was again another recommendation by Sprung, Delbeeck.)
Any comments on this idea would be great. I thought it seemed to be the most sensible solution I have read.
Regards,
Brandon
PS: Great thread