Issues/concerns with AsicMiner Block Erupter Blade v2 and P2Pool

Hi all,

Wasn't sure whether to post this here or under mining, but will try here first I guess:

I have an AsicMiner Block Erupter Blade v2 in a backplane, which I've been running with P2Pool and BFGMiner as a stratum proxy. I've noticed a few issues with running the blade on P2Pool and was wondering if anyone else has run into these issues and has any suggestions/fixes:

1) To start, I tend to get an abnormally high number of rejects (>50-75%), as well as a few hardware errors. I used to GPU mine with multiple GPU cards using P2Pool and was not seeing anywhere near the same number of rejects. I've decreased the queue to 0, but once in awhile I end up getting a work underrun and BFGMiner automatically increases the queue size, which appears to increase the number of rejects as well. I've also tried tweaking the scan time and expiry (I haven't actually found a good explanation or suggestion as to what these two values do and should be set to either) to no avail. According to the blade output, it looks like I average 7-8 GH/sec (although P2Pool shows ~10-11 GH/sec local rate). If I mine on Eligius on the other hand, I have barely any rejects and I maintain ~10-11 GH/sec according to the blade output itself. I do tend to have a much better turnaround on P2Pool than Eligius, but I think it could be improved further if I could sort out these other issues.

2) Is there any way to increase the difficulty of the shares I receive with P2Pool, generally I see "Diff xx/2". I've tried the --request-diff option in BFGMiner, but it doesn't seem to help at all. However, again, mining on Eligius, I routing see "Diff xx/4", "Diff xx/8", "Diff xx/16", etc. Is there any way I can tweak this in P2Pool or is this a limitation of the P2Pool program itself?

3) As one other less critical question, how has everybody else mounted the Block Erupter Blades & Backplane?

Thanks in advance.