I had two A9 with 30% rejects, but found a solution. If you guys still have problems with rejects - drop me a pm. I can possibly help
This great BBS won't let me send a pm, so here it is...
Greetings,
I have a few of these A9 devices and literally HALF of them are crap (31 to 41kSol/s avg hash at the pool.)
What kind of interesting solution did you use?
Kind regards,
Lee
---------------------
Wie geht's,
Ich habe dieser A9-Mineren und die Haelfte von scheiss davon sind Muell (31 bis 41 kSol Durchschn).
Was fuer eine interessante Loesung hast du benutzt? Danke.
Tschuss,
Lee
Did you contacted support to get new firmware dated 6/30? Which pools are you using?
I got stable 46 kH/s for any pools now, with 1-2% rejects max.
Thank you for your comment.
I did indeed update the firmware on those laggard A9 devices and indeed the reject rate went way down but the hash rate is still the same on two different pools.
Seems like I got a bunch of duds.
My friend bought 2 ZMaster and i bought one, and his run everytime more hashes.
The only difference was, that he had bigger windows and wide open, i only had 1 windows open and not fully only tilt.
And he had also put ventilator directional to the miners, which blowed the air directed to the window.
His miner showed also two cards at around at around 64°-65° Celsius, and only one card at 71° Celsius.
Mine ZMaster had two cards running at 72-73° and only one below ~67-68° Celsius.
So i had everytime in my backhad that it might be caused by temperature 'underclocking'.
We brought now the 3 ZMaster to our "Mining Hub" where we suck out the air with 3 axial fans, and with a pipe construction directly from the miners.
Now the miner of him runs at 63° - 71° - 64° and mine runs at 67° - 72° - 65° and now i got steady average slightly over 50 k/Sol right now hourly 51 k/Sol, the hash numbers are from the antpool panel.
Only 7,07% rejects so i am at 92,93% accepted rate, also from antpool panel.
(Before i got 41-48 k/Sols, and higher temperature like mentioned before)
So guys what i figured out is what you might have to optimise or what might cause the problems:
1. Pool :
Some pools have to optimise for the ASICs, since ASICs are quite new for Equihash. So look out for one which says our seems to be already adapted the ASICs.
2. Cooling:
Look at your Innosilicon Panel, it seems to be that the middle card is getting hottest, but the 2 on the side should be below 70°. Otherwise i guess there are underclocking or get to much physical electric resistance caused by heat. So watch out for your temperature and cooling.
3. Network:
I personally had network issues, some switches and some network controller of the asics, can have problems. The Symptoms are that the panels kinda loops in "tunning" but when you look in the logs, you see that he loops at connecting to the pools.
I am still in to solving that problem i ordered a new Switch and maybe some cat5 cables, will see and report how that works out, and if it proves the theory that i might be caused by automatic 100 full duplex.
For now the one originally faulty device, was put in a other network at home at a friends place. The secound one got similar problem after bringing all 3 together, but it could be fixed by just pressing "Update" other Network.
To note, the first days i run mine at home and he run his two at home, and we had no problems at all, beside my heating "problem".
