Re: [NTLK] Knowledgenavigator

On Sun, Sep 7, 2008 at 12:21 AM, Sonny Hung <sonnyhung@gmail.com> wrote:
> What I don't get is that there on his site he has photos of the speeds.
> Boost ON for the eMate 300 it only reflects 27.0722222222222 mHz - Where is
> the 30 mHz Ferdi advertised?
> Boost OFF it reflects 19.9555555555556 mHz - Well at least is about 20 mHz
> These number do reflect if a 3.686 mHz crystal and 5.000 mHz are used for
> the eMate 300 hack.
>

Interesting :-?

So I track down the implant.pkg (Implant Status PKG) and first find out the
version I got was to test if the Implant was Active or not...
Duh, this only works with the Implant 2000 installed and maybe the Implant
130.
http://www.unna.org/view.php?/utilites/ImplantStatus

I tracked down another copy of the Implant Speedtest over at Tek-Ed's site.
http://tek-ed.com/pkgdldr.cfm?type=1&index=42
* Can this be put up on UNNA please :D

I got a copy from Pixsolution way back direct when I contact Oliver Harms
(forgive spelling if I err'd)

Ran my test and see that it reflects close to Ferdi's numbers.
Mine came out to 19.9833333333333 mHz - Hm, now my brain ticks a few and
recalls that this program was specifically written for Implant 2000 and in
NSBasic so when testing this on an eMate this could be why we are getting
this lower number which is significantly slower than the 25 mHz reported by
Apple in their spec sheet for the eMate 300.

So with nothing running I slide the switch to the Turbo setting and re-run
the test again. I would assume that my number should be in line with Ferdi's
number again.

Yes, 27.0833333333333 mHz in Turbo (overclocked mode).

So in conclusion at least the numbers depicted in his photos are valid.
The question still bugs me as to how come the stock speed of 25 mHz is not
reflected in the test. Is it as I suspect that the application written in
NSBasic is not calibrated to reflect correctly?

20 mHz - CPU-Clock 20 MHz, all baud rate on external Serialport are support
27 mHz - Implant-130 has variable frequency, no baudrates onexternal
Serialpoprt was supported
So I double checked and confirmed that MP110,120,130's all run at 20 mHz.

Pulled from Apple site.
- 25-MHz ARM 710a RISC processor (eMate 300)
- ARM 610 RISC processor at 20 MHz (MP130)

Hm, here's a thought -
PCBman stated the following in reference to the MP2x00:
- Simply put, each transition of the base clock yields 44 transitions of the
master clock.

This number is what is used to calculate the speed of your MP or eMate.
It would be probably that the Implant Speedtest package like was programmed
with the MP1x0 OR MP2x00 family of Newtons (5.4295 OR 44) and not take into
consideration the eMate 300 which uses a variant of the same ARM processor
ARM 710a which runs at 25 mHz versus the ARM 610 which runs at 20 mHz. The
multiplier in this case would then have been 6.7824 which would have
resulted with a correct speed of 25 mHz/27 mHz respectively to the
stock/turbo (overclocked) crystal speeds.

So there is an error in the advertisement Ferdi has on his website,
hopefully he will correct said error shortly :-)

And everyone else will understand also that some of the discrepancies are
possibly due to the software not being properly coded to accommodate the
eMate 300. My 2 cents/theory on the matter.

-- 
God bless,
Sonny Hung
the Hung Family
==================================================================== 
The NewtonTalk Mailing List - http://www.newtontalk.net/
The Official Newton FAQ     - http://www.splorp.com/newton/faq/
The Newton Glossary         - http://www.splorp.com/newton/glossary/
WikiWikiNewt                - http://tools.unna.org/wikiwikinewt/
====================================================================