[Gghc-discuss] Proposed initial purchase - too much?

Wed Mar 30 18:23:46 CDT 2011

Cool on speccing this out! Thanks for looking at this. I have some concerns
though. Generally, SparkFun and RobotShop are kind of expensive for some
thins. Some more specific comments:

* There are better prices on xbee and xbee arduino shields out there.
(closer to $40 for the combo than $48). Even more affordably -- is there
really that much work into breadboarding an xbee to an arduino?

* Getting motor controllers and just breadboarding them to be arduino
controlled will probably cost $10 (or less) per prototype robot, and only
take an extra few minutes to set up.

* Are there robot platforms that are not machined for the arduino, but
perhaps just motors and wheels, or even better, motors, motor controllers
and wheels, which we can just drive with digital logic? Making a mount hold
an arduino is one of those really easy things we can do "good enough for
prototyping" and maybe spend a lot less money.

* The price of each of those upconverting power supplies is greater than the
cost of a second battery in series. We can take care of the extra voltage
easily with a voltage divider or voltage regulator for like $.50 .

* Paying $4 per mini-usb cable is insane. You can get the same cable from
monoprice.com for $1.

My overall point here is that I know we need to get this going now, but
taking a few minutes to shop around for good prices, and taking a few
minutes to look at alternatives and how much actual effort is involved in
them (vs cost) seems prudent.

I guess my thinking is we should try to not be too wasteful with that money,
since while it is R&D money for whatever we need to do the projects, we can
also use some of that money towards general makerspace improvement and
equipment, not just project specific stuff -- since that could be part of
the R&D effort too.

Further delving into the hardware/software in parallel thing: there is
enough software that can be written with a bare arduino + xbee  (no robot
immediately) that having a week of putting hte robot together time is not
going to block anyone. Further, the hardware design will be finished enough
for software development/testing as soon as we have breadboard circuits
working -- no need to wait on PCB and soldering. Basically the choice
between fully formed starting points and diy everything starting points is a
false dichotomy. I am convinced there is a way that isn't as expensive,
involves parallel development and modification of other parts, but is
completely doable.

I appologize for not articulating this as well last night, but I needed to
actually see the parts and prices before I realized that basically a lot of
the pre-fab circuits are just a couple components.

Regards,
Erich

On Wed, Mar 30, 2011 at 5:01 PM, Jonathan Manton <jmanton at illinois.edu>wrote:

> I put together a Google Docs spreadsheet with my proposed initial purchase
> of stuff.
>
> https://spreadsheets.google.com/ccc?key=0Atq-9wMS4xTIdGVpbnhuMmpOVmo2cGFHWEliYlBsbEE&hl=en&authkey=CNuInqUO(read-only link - send me your email and I'll share it with you read/write -
> didn't want to open it up to public edits).
>
> As an explanation, the "base station" is the part that hooks up to the
> laptop, the "proto robot" is a fully-functional robot with motors and such,
> the "pseudo-robot" is a robot without a base or motor (but with
> battery/charger) that we can move around by hand, and the "training station"
> is an Arduino + XBee (no battery), for use in learning about the Arduino.
>  The "pseudo-robot" and "training station" can be turned into full robots
> later with the addition of a base, motors, motor controller, and battery
> (for the training station).
>
> It comes out to $451.93, which is a distressingly-high percentage of our
> overall budget of $900.  And that is before shipping, which will probably be
> $30-$40 if we get 2-day shipping (the order comes from two different
> vendors, $15-$20 each, 2-day so we can start this weekend or Monday at
> latest).
>
> Two things that have kind of high costs are the Arduino board ($30 ea.) and
> the XBee module + shield + headers ($50).
>
> We need to decide as a group how we want to end up spending the $900 (at a
> high level).
>
> (option 1)
> On one extreme, if we went with the configuration I'm proposing for
> everything (we don't build custom hardware at all), then each base station
> costs $51.85, each robot costs $191.64, which means we would have enough for
> the two base stations and 4 robots.
>
> (option 2)
> At the other extreme, the base stations cost the same, but we design the
> robots from scratch (no prototype platform), and can probably cut the robot
> cost down to about $100 each.  We'd end up with two base stations and 8
> robots.  But then we have no Arduino boards for people to learn with, and
> probably 2 weeks lead time (minimum) before we'd have the first prototype
> robot up and running.
>
> (option 3)
> The middle route is to go ahead with this initial order.  We'd have about
> $420 left (depending on how much shipping is).  That leaves enough for the
> other base station ($51), enough to turn the "training station" and "pseudo
> robot" into real robots (~$70, since they both already have an Arduino and
> XBee), and build 3 robots from scratch at $100 each.  At the end we'd end up
> with 6 robots and the base stations.
>
> (many more options possible - this is just what I came up with)
>
> My personal opinion on this is I don't want to buy everything (option 1)
> because I want to help design the custom robot and make the circuit boards
> and stuff (this is what Erich articulated yesterday and I strongly agree
> with him).  But I don't think we should try so hard to save money by going
> with the cheapest route (option 2) that we run out of time (and have half
> the people on the project waiting around with nothing to do until the robots
> are built, and put pressure on the people doing the hardware design... in
> parallel with getting ready for Mini Maker Faire U-C).  Even though this
> consumes more than half our budget up-front, I'd advocate for the middle
> road (option 3), and go ahead and make this purchase now.
>
> One thing we *could* do to save money is not get the LiPo battery packs,
> charger, or switching regulator, and instead use 4 AA batteries for each
> robot.  This would save about $80 from the initial order (which is then
> compensated by having to buy batteries).  That might be OK for development,
> but for the final product we want batteries that are light and can go
> through lots of charge/discharge cycles, rather than disposable AA
> batteries.  Even rechargeable batteries will be kind of a pain to deal with
> if they have to be removed to be recharged - 6 robots x 4 batteries each is
> a lot of rechargeable AAs to manage.
>
> Personally I don't think getting AA batteries rather than rechargeable
> LiPoly batteries is a good idea, because we'd then end up spending money on
> something that is consumable (batteries).  In the stuff I'm recommending,
> everything is used for the end product - there are no parts that are used
> *just* for development (with the exception of a couple of USB cables maybe).
>
>
> _______________________________________________
> Gghc-discuss mailing list
> Gghc-discuss at lists.chambana.net
> http://lists.chambana.net/mailman/listinfo/gghc-discuss
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.chambana.net/pipermail/gghc-discuss/attachments/20110330/6fd79867/attachment-0001.html>