Billions of Bytes Programmed at JET Speed
High demands made of electronic storage media and media production:
exploding production quantities and ever increasing storage capacities keep the
electronics manufacturing industry on their toes. The new SmartJET high-speed technology
from ertec is the ideal solution, achieving high programming rates in mass production
and taking automated device programming to new quality and speed dimensions.
One billion has nine zeros, one gigabyte has even more bytes, and if we go down
as far as bits, then 1 gigabyte contains just short of 10 billion bits. But what
use is a gigabyte to us tomorrow when individual chips with a Flash capacity of
16 gigabytes are already on the market today? Today, more and more electronic products
have such huge memory capacities, which in many cases have to be programmed "bit
for bit" during the production process. Conventional processes are simply not up
to the job, as can be shown in a simple example:
Even if just 100 nanoseconds per byte are additionally needed for Flash programming,
for an 8 GB module this additional time amounts to as much as a quarter of an hour
per module. This is because conventional microcontroller processing does not allow
for shorter processing times.
In device programming the time needed is broken down as follows: the time for the
PC to access the programming data (hard drive, network, external data storage device...),
program running times inside the operating system (computing speeds), time needed
to transfer the programming data from the storage location to the Flash module (data
transmission), the software overhead inside the programming system (processing times
for data transfer, retrieval, evaluation..) and finally the time needed to actually
program the Flash memory.
The new P67805 JET-based programming modules from ertec feature smart, high-speed
JET technology based on the optimum combination of a fast microcontroller master
and a super fast FPGA module as a slave. In addition, the transmission time for
the programming data is zero and PC computing time next to nothing, since the data
is taken directly from pre-loaded image storage cards and programmed directly into
the Flash in the fastest possible time. Furthermore, thanks to 4-fold parallelization
per programming module and up to 64-fold parallelization in the overall system,
a total programming rate of more than 5 billion bits per second is achieved.
In sum, the programming robot BEAVER from ertec features 64 programming sockets
to allow for a fully automated programming rate of around 650 devices per hour,
even if the theoretical individual module programming time is more than 5 minutes.
With BEAVER programming, the modules are taken from the input tape, programmed,
laser marked, undergo a visual check for marking errors and mechanical pin check
(2D or 3D check, as required) and re-packed in output tape in a single complete
process. Tight multi-level safety checks are, of course, also performed during the
entire process to ensure compliance with the relevant quality standards for data
consistency, transparency and traceability.
Author
© ertec GmbH, 91058 Erlangen/Germany
Author: Dipl.-Ing. Peter Nickel ((Managing Director - Technical Operations))