Short solution making it brief: Simply SSH into your miner raspberry pi. Change the clock speed to lower than 1000, then mine again check status on pool and miner see if any errors, repeat the process until you achieve the speed advertised on the pool which should be around 1TH/s.
A little bit more detail can be found here https://bitcointalk.org/index.php?topic=545642.msg7120969#msg7120969
Good reference also here https://bitcointa.lk/threads/dragon-ore-mining-machine-five-modules-t-class-overclocking-img-installation.296670/
I know that some 5 blades 1TH/s miner shipped with the English Raspberry pi version, then this should have you a lot of time following direction below. If not then you would need to flash your SD card to the English version or the SSH being reset so you can login and tweak
(see post here https://bitcointalk.org/index.php?topic=409701.msg7117387#msg7117387 or see below)
i recommend reimaging with the dragon firmware. then if you would like to further tweak each individual board to get your best possible results from each one you can ssh into the raspberry pi and edit the run.sh file to increase/reduce the speed of each board. I had to do that with the 3 units i ordered from them because there were way too many hardware errors once i imaged them to the dragon firmware because it runs all boards at 1000. before i tweaked the settings none of the units were hashing above 850 at the pool level due to hardware errors. so after tweaking the 3 units i had them hashing at 970, 950, and 920 at the pool level (internally shows higher but also has some hardware errors still).
1. image over to dragon firmware (default IP 192.168.1.134 i believe)
2. set your IP and pool settings
3. let it mine for 15 – 30 min so you can see where your hardware errors are per board in the miner.php page stats
4. ssh into raspberry pi (username pi, password raspberry) – you can also use winscp if you are less command-line savvy
5. cd /var/www (dont leave out the space after cd)
6. sudo chmod 777 run.sh (makes the file editable)
7. edit run.sh (nano, vi, etc.) scroll over through the line that runs cgminer and change the boards from 1000 to a lower number to reduce hardware errors. you will notice there are settings in the cgminer line for 5 boards. only the 1st 4 make a difference (unless of course you have a 5-module unit). save your file
8. sudo chmod 555 run.sh (makes the file read-only executable. if you skip this step any changes to the pool settings OR if you stop/restart the miner through the web interface then it will overwrite this file. on the flip side if you make the file read-only then any changes to the pool you make in the web interface will not save. you can still stop/restart the miner from the web interface)
9. sudo pkill cgminer
10. sudo ./run.sh
11. return to step 3 and repeat until you’ve achieved desired results (personally less than 1% hardware errors is the goal)