Miner acting up on certain pools.

[Travis726]

I have an old antminer s5 that loses half its hash every once in a while. When on btc.com, viabtc, or slushes it happens very rarely. But if I mine on suprnova, or blockfactory it almost instantly drops to around 600GH/s.  I have googled around and a lot of people have had one board drop out, but everywhere I read when they have it happen to them, one board totally loses power. That isnt the case with me. Both boards still have power, and im pretty sure ive have it happen with both boards at times which im hoping means my hash boards are fine. Maybe the "circuit board"?  That would be a great answer as Ive seen them pretty cheap.

Does anyone have any guesses as to why it would happen WAY more often on blockfactory, suprnova and zpool. And VERY rarely on viabtc, slush's and btc.com? And when it does happen on the big pools, it fixes itself within a couple minutes. At blockfactory ect, it will spike up to its normal hash for a minute or two, but drop back down like clockwork.

Any help would be appreciated.  In a way Id like the option to mine stuff like digibyte in hopes it will rise in price more than btc since im playing with very little hash, but right now im stuck with btc or bch unless I want to give up half my hash (i dont)

[wavelengthsf]

Where are you measuring the hash rate? On the miner, or on the pool?

One theory is that those pools are setting the difficulty to something your miner can't handle, and its not able to submit a share until the difficulty adjusts. The bigger pools have a lower starting difficulty.

Another option could be their stratum servers location to you. If their infrastructure doesn't have nodes near you, your connection might be timing out and having to reconnect.

[Travis726]

Both the pool and antminers screen show the hash rate drop, and I can feel that one board is not heating up. One board isnt hashing without a doubt, but it doesnt lose power (red light) and happens WAY more frequently on some pools.  I know on blockfactory you can specify a difficulty. It seems to use 4096 when I use stratum on that pool and I have set it to that via their instructions (adding something to worker name) but it didnt seem to make a difference as I was wondering if it had to do with difficulty also.

Thanks for the response, any more from you or others is appreciated.  It also seems to work well mini solo on ck's pool.

[Travis726]

also, overclocking/underslocking seems to make no difference. I have a second fan added (powered via wall plug, not the miner) and temps stay in the low 50's at 375 setting. Ive tried to clock all the way down to under 300 and makes no diference at all so Im as high as it goes without HW errors getting up over .005%.

[wavelengthsf]

What's the status page look like when you lose power?

Its possible you just have some bad chips, and its coincidence its when you switch to one of those pools.

[Travis726]

What's the status page look like when you lose power?

Its possible you just have some bad chips, and its coincidence its when you switch to one of those pools.

It looks normal, sometimes if I reboot the miner rather than wait it out, one of the chains will show ----- at first, but a few seconds later updates and looks/acts normal. It most definitely isnt just a coincidence though (even if I do have hardware problems) Something about those pools is effecting my problems in that case. I can make it happen at will by switching to one of those pools. Its night and day as far as frequency of occurrence.

[Travis726]

I have gone as far as taking the sides of the miner off to take a look with my ignorant eyes at the back side of the chips and nothing looks burnt (like some pictures Ive seen)

edit: Im not 100% sure it has happened to both (although I kinda think so). recently it has been effecting the side/board that the circuit board is NOT powered through (if that helps at all).

The psu I am using is a 1000 watt server psu. It came with the miner used.It only has two wires coming off of it, with two connectors each. I do have a 750 watt evga psu with 4 separate pci-e cables somewhere if anyone thinks that may help.... I have just been using it on the pools it works on since i got it. (mid November) but wouldnt mind mining something with a little more risk (and possible reward). Ive gotten enough of my money back from it that why not.... hence the solo mining sometimes

[Travis726]

ok now Im having a different issue that may be easier to diagnose.  It is showing 2 of the 3 pools as "dead" most the time. Sometimes it is "pool 2" that is alive, sometimes 3 ect.  It happens when I assign the same pool to all three so I know the pools arent actually down. Last night it dropped the 3rd pool and started beeping at me.  Also, I have noticed when I boot it, it sometimes takes longer to connect than normal. During this time on the "system" page it says the hardware version is "x.x.x.x" . 

Im guessing its one of the two boards (other than the hashing boards)... any help?