I agree, the solder jumper will surely cause problems.

I had a similar issue ... somehow if I use load balanced mining .... with 40-42GH on 3 pools im getting 13GH at BTCGuild ... 13GH .... at eligius ... and 12 GH at Bitminter ...

if I change to mine 100% on BTCGuild it shows me 20-30 GH only .... i tryed with 16,32 & 64 diff

You could be having a different issue, but I noticed the stratum proxy is behaving strangely with BTC Guild. If my minimum worker difficulty is set to 1, the proxy correctly picks up when the difficulty quickly increases, and doesn't try to send shares which don't meet the requirement. However, if I set the minimum difficulty so high that it never has to be increased by the pool, the proxy keeps on thinking that the target is diff 1, and sends every share to the pool. Obviously most of them get rejected, and this might even overload a router.

I didn't have any actual problems with this, though. BTC Guild always reported my full hashrate, even when I was flooding it with tons of rejected shares. Of course flooding the server isn't exactly nice for the pool OP

I had this issue as well... I had my difficulty set at 8 in btcguild, I only have 18GH.  Stratum proxy would send everything back at diff 1, and none of them were accepted.  I set my min difficulty to 1 in btcguild, and now it negotiates its way up to 8.

There's a protocol implementation error in the official Stratum proxy.  The original spec always stated the first message a pool should be receiving is a mining.subscribe method.  However, if you try to make the Stratum Proxy do the authorization (-cu -cp args), it sends those *before* the subscribe method.  As a result, you're getting your difficulty set to 2/4/8/etc before the "default subscribe response" is sent.  A workaround is being added to the BTC Guild stratum code today to hopefully prevent this mixup.

Hmmm. Another issue. I have disabled Autotune for all chips. I run miner with console output. And suddenly, after processing 800-1000 jobs, miner prints "SPI update chip xxx", and my Noncerate starts dropping drastically... If I checkrun/shm/.stat.log chips have aIFDSo parameters.
How is it possible?

Hey spiccioli - go get your hashing card fixed and stop giving away secrets like that....  

Word will get around that people make lots of coins on HHTT and it will get too big...

The problem is, fixing the error in the proxy would require slush to actually still care about mining.  He hasn't been seen on bitcointalk in months.  Additionally, it'd require every user to upgrade to a forked proxy that fixes the problem.  It's much easier to just make the pool workaround the broken proxy rather than hope everybody else upgrades.

The error is only related to minimum difficulty though.  If you have your minimum difficulty set to 1, it will work just fine, and vardiff will adjust your difficulty as needed.

BTC Guild has now deployed a workaround for the reverse auth/subscribe order in slush's Stratum Mining Proxy to prevent it from spamming low difficulty shares when a worker has a pre-set starting difficulty.