> Finally, I'd like to address one inaccuracy in Dennis' post here...
> Dennis' assertion that Edelbrock is about the same price is, well,
> maybe undocumented would be a good way to put it. Edelbrock shows
> their product at $995, for the entry level version, on their website.
Edelbrock, and for that matter many of the "big guys" like Moroso, Holley, etc.
publish MSRPs in ads and company literature that are _nowhere near_ the street
prices their products actually sell for in the real world.
I found a website that was selling the Basic QwikData for $750. I wanted to give
my business to an autocross supporter, so I called Sam Strano and asked if he
could meet that price. He could, and I bought the system.
Will Sam continue to sell QwikDatas for that price? I dunno. But if I'm able to
find 2 places in very short order who will sell them for $750-ish, then they can
be had for that price. If I understand how Edelbrock's wholesale price policy
works (bigger dealers get better prices) one might be able to get the same
system from a mail-order place like Summit for even less.
> Our Racer's Bundle is a complete system at $395. That's a lot of
> difference.
> Even with the Palm, as Dennis likes to make the comparison, expect to
> spend about $555 for the full GEEZ, G-Cube, and Palm III package.
Well, that means you got your Palm for $150. My visit to www.palm.com showed
that the cheapest Palm III was $250-ish, and one could drop on the order of
$500-$600 for the Palm V.
Not to mention that you need a cradle, and sync software, etc. etc. etc.
Now to be fair, I'm playing the same MSRP game here in reverse - I'm sure one
could find discounted prices on Palms the same way I got discounted prices on
the QwikData. My research (done quickly, to get a ballpark feel) showed me that
to get a new, base model Palm with all the accessories required would run
$200-$250. To get a more featured Palm (hey, if you gotta buy a Palm, you might
as well get one to do stuff outside of data aquisition) was in the $300-$350
arena.
So look at it this way - the best case price of entry for GEEZ!+palm (used?) is
$550. The "reasonable" case is about $650-$750, and the worst case is as much as
you can spend - but then you're looking for more than just GEEZ! features, so
it's hard to use the cost of the titanium and carbon fibre Palm against GEEZ!.
;)
For the QwikData, the best case (from Summit) is undetermined. I got mine for
$750, and I had 2 sources at that price. The worst case is MSRP, at $995.
That makes the QwikData anywhere from $0, to $100, to $200, to $450 more
expensive, depending on whose numbers you choose to use. If we use each "side"'s
"best case" numbers, ($550 vs $750) you get $200.
Given that you want accelerometers no matter what, that $200 gets you 14 more
data channels. That's $14 a channel. :)
For me, that's a better deal. For others, that may not be. Edelbrock doesn't pay
me anything, and I don't own Edelbrock stock. YMMV.
> It's okay with me that Dennis prefers the Edelbrock.
It's also worth pointing out that I have NOTHING against Byron or his product. I
made my evaluation straight up, and I have posted my reasons for and against
both products. I prefer the Edelbrock unit, but people are free to make their
own decisions.
> I would also like to point out that Dennis is not a customer,
This is correct...
> and has never called or e-mailed for help in using the system. He did >
e-mail and ask about whether or not I would give him the source code to > our
mapping algorythms. I declined.
Well, that's not entirely true. When GEEZ! first appeared, I contacted Byron
about the possibility of doing a Linux port. He declined. When I had an
opportunity to use GEEZ! (this was before the Palm version was out) I both
e-mailed and called to discuss my observations about the product. When I
discovered the Edelbrock unit, I e-mailed again to make Byron aware of the
unit's existence, and to discuss the possibility of a version of GEEZ! that
worked with the Edelbrock data files. He wasn't interested. I did ask for the
mapping algorythms, but I wasn't seriously expecting to get them.
In short, I gave Byron an opportunity to meet my requirements and take my money
at every stage of the game, and he declined.
Incidently, that's FINE. If Byron feels that my requirements or suggestions do
not meet those of his target audience, more power to him. I am probably not
representative of the average GEEZ! customer, and so it's not terribly suprising
or bad that he wasn't interested in doing what I wanted. It's hard to get tone
out of e-mail, but there really are no sour grapes here.
But it's equally true then that there's nothing terribly suprising or bad about
my choosing a different product that does meet my requirements - or that at
least gives me a warm fuzzy that they will be met sometime in the future.
> Another quick note, this one about Dennis' supposed lateral and braking
> G readings. The indicated error which Dennis attributes to roll would
> require a HUGE amount of roll to get a good usable number, something on
> the order of 20 degrees.
On advice from some other posters, plus a hunch of my own, I went out and got
readings from the unit at 1G values (by flipping it on the appropriate sides)
and discovered that the unit is not correctly calibrated in 3 of the 4 axis. It
was, however, repeatable (flip it from 0 to 1G 20 times, and you always got the
same number)
So I called Edelbrock, and they're going to add a calibration routine to their
software (it currently only has a "zero acceleromters" feature)
So I'm satisfied that I cannot trust the raw values of the accelerometers, but
that I can trust them to consistantly read the same wrong numbers in the same
conditions, and that they are consistant relative to each other - which is fine
for the kinds of analysis I'm currently doing.
I'm taking all my analysis directly from the book "Data Power" (aka "Data Scrape
to Win") and 95% of the analysis there using the accelerometers is based on
trace shape and time/distance between peaks. Raw values don't matter in the
general case when you have a measured speed line.
Now that's not to say that I wouldn't prefer to have the raw values correct, and
when the software is updated I will, but I don't _depend_ on the raw values for
anything save bragging rights, so none of my analysis to date has been
invalidated - with the exception that no, my Talon really doesn't pull more
lateral G than an Atlantic. ;)
> We attempt to help the user learn what led to speed.
That's no different from what I do with the Edelbrock unit. The speed line is
the results line - if I see a .3G higher peak in a certain place between 2
overlaid runs, but the speed lines are the same, then that extra .3G did
nothing.
The speed line tells you what happened. The G traces, the throttle position
trace, the brake trace, and the RPM trace tell you WHY that happened.
Anyway, this got way longer than I intended. The gist of all this is that for me
at least, the Edelbrock was a better value, and that I'm finding it easier to
use. I have nothing against Byron or his product, nor his customers. I think
data analysis is going to revolutionize my performance, and quite possibly
autocross in general, and I think discussions about data analysis and the
conclusions one came to, and how one came to those conclusions, are far more
interesting than squabbling over who makes the best widget.
Fair enough? We all friends again?
DG
|