Quote from: macgyver007 on July 31, 2014, 07:26:56 PM

Quote from: ZiG on July 26, 2014, 03:02:45 PM

Quote from: finlof on July 25, 2014, 03:58:24 AM

Quote from: philipma1957 on July 25, 2014, 03:36:57 AM

Quote from: dlasher on July 23, 2014, 10:10:22 PM

generated a table to work with, based on the S1/S2 tables:

Quote

...package 'cgminer'

config 'asic-freq' 'default'

   #option 'freq_value' '1486' #262.5M
   #option 'chip_freq' '262.5'
   #option 'timeout' '15'

   #option 'freq_value' '1386' #250M
   #option 'chip_freq' '250'
   #option 'timeout' '16'

   #option 'freq_value' '1306' #243.75M
   #option 'chip_freq' '243.75'
   #option 'timeout' '17'

   #option 'freq_value' '1286' #237.5M
   #option 'chip_freq' '237.5'
   #option 'timeout' '17'

   #option 'freq_value' '1206' #231.25M
   #option 'chip_freq' '231.5'
   #option 'timeout' '18'

   #option 'freq_value' '1185' #225M
   #option 'chip_freq' '225'
   #option 'timeout' '18'

   option 'freq_value' '1106' #218.75M
   option 'chip_freq' '218.75'
   option 'timeout' '18'

   #option 'freq_value' '1086' #212.5M
   #option 'chip_freq' '212.5'
   #option 'timeout' '18'

   #option 'freq_value' '0782' #200M
   #option 'chip_freq' '200'
   #option 'timeout' '20'

first off thanks to you.

second can you find a number between 200 and 212.50 ( 206.25 ) comes to mind

third can you find something for 215.625

my guess for 206.25 would be (use at your own risk):
   option 'freq_value' '1006' #206.25M
   option 'chip_freq' '206.25'
   option 'timeout' '19'

With all my due respect for your time and effort, but this table is NOT very applicable to S3...

You can NOT apply the same S1 math/calculations... to S3, because these are using different chips with different PFD (PLL Frequency Divider...I guess), subject to different conditions...due to chips redesign...I guess... Wink

BM frequency table - from BM1382 datasheet ...:

Work Input timing
Clock Freq.(M)   Hash rate(G)   Reg.value   Ideal delay(ms)   Recommenddelay(ms)
100    6.30     0783    42.6     38
125    7.88     0983    34.0     31
150    9.45     0b83    28.4     26
175    11.03     0d83    24.3     22
200    12.60     0782    21.3     19
225    14.18     0882    18.9    17
250    15.75     0982    17.0     15
275    17.33     0a82    15.5     14
300    18.90     0b82    14.2     13
325    20.48    0c82    13.1     12
350    22.05    0d82    12.2     11
375    23.63    0e82    11.3     10
400    25.20     08f2    10.6     10

These are the same in S3's "asic-freq" file...

Conditions for S3 valid frequency divider's values now are different...check them here...:

https://bitmaintech.com/files/download/BM1382_Datasheet_v3.0.pdf ...

Pay attention to ..." For normal operation the following constrains MUST be satisfied..."

I did NOT check Your value(s) on my S3's, but seems like they are NOT the same "animals"... Grin

Cheers

ZiG

Agree Bitmain chart is much different for the same values...and they follow the equation..I wouldn't use the above chart values only the ones from the BITMAIN specsheet

I am trying to get the pattern of the smaller steps...like 12.5MHz...6.75Mhz ...etc... Wink