Quote from: Gaim on July 02, 2013, 02:52:59 PM

Dear ZTEX,
I need some help.. Sad

Today, my new 1.15y2 arrived from German-group-buy.

Some strange things are happening:

Code:

008-0: ztex_ufm1_15y1-04A39E1212-1: f=220.00MHz, errorRate=0.51%, maxErrorRate=1.10%, hashRate=218.9MH/s, T=46.4°C, submitted 15 new nonces, luckFactor=1.05
008-0: ztex_ufm1_15y1-04A39E1212-2: f=216.00MHz, errorRate=1.21%, maxErrorRate=1.81%, hashRate=213.4MH/s, T=45.8°C, submitted 15 new nonces, luckFactor=1.06
008-0: ztex_ufm1_15y1-04A39E1212-3: f=224.00MHz, errorRate=0.00%, maxErrorRate=0.00%, hashRate=224.0MH/s, T=52.2°C, submitted 12 new nonces, luckFactor=0.87
008-0: ztex_ufm1_15y1-04A39E1212-4: f=220.00MHz, errorRate=0.00%, maxErrorRate=0.00%, hashRate=220.0MH/s, T=45.1°C, submitted 11 new nonces, luckFactor=0.93

With your miner the values are relative ok, because the client is able to change the frequence of all 4 independly.
I reflashed the firmware once again, too.

I removed the heatsink from 1,2 & 3 removed paste (a bit too much) and attached them again, but I can't change it:

1. Chip 3 is running with 8 degrees more then chip 4 - something to worry about?
2. Chip2 is producing more errors then the other three, which results to a lower frequency and lower hashrate (of course).

On Chip too, the paste at the heatsink looked like it has only connection to the edges and not at the middle of the chip. Is this possible? Any advices? Is this within the normal range?

Thanky in advance,
Gaim

edit: In fact, after I reworked the paste on the chips, It behaves much better on bfgminer too... maybe that was the reason why.
edit2: The reason for posting is/was.. on my really first try the 3rd chip ( I assume no. 3 is c-suffix in bfgminer ) was running super fast, while the 2nd dropped much. I'm currently testing the values after the above written rework of the past on the spartan chips (while the no. 2 had over 300 errors)

edit3: Relaunched bfgminer to make a long-run test:

Device	Enabled	Status	MHS av	MHS 30s	Accepted	Rejected	Hardware Errors	Frequency
ZTX1a	Y	Alive	213.988	215.727	102	0	10	216.000
ZTX1b	Y	Alive	214.320	215.715	115	1	259	216.000
ZTX1c	Y	Alive	210.527	211.729	115	1	2	216.000
ZTX1d	Y	Alive	222.402	223.756	144	0	71	216.000

that is what I wanted to show you. .. it is for off with the hardware error rate.

1. The Tem. displayed is not of the chip (inside). Its somewhere on the PCB, probably under the chip. So read this Value with a grain of salt!.

2. Not every chip behaves the same way. I have boards that do 228 minimum on all four chips and others that wont make 204 or more than 199...

3. Error rate is calculated every the bitstream clocks up and down. I have a lot of HW errors with my Ztex on CGminer in the beginning (when it clocks up) and then it normalizes.