pingswept.org - Latest Comments

Re: pingswept.org: blogging since the end of time

Thaps — Wed, 09 May 2012 13:26:31 -0000

How would one write the programme without the modules , numpy and scipy?

Re: pingswept.org: blogging since the end of time

Nathan Crock — Fri, 20 Apr 2012 15:09:53 -0000

I like what you've done here. Thank you for taking the time to make this post!

Re: pingswept.org: blogging since the end of time

Padma Bala1 — Mon, 06 Feb 2012 04:38:59 -0000

Please send us a simple program of least square fitting in python

Re: Designing embedded systems with web frameworks

Access Control Systems — Thu, 19 Jan 2012 07:10:16 -0000

his post really conveyed the part which i was looking for. Waiting for your next entry.

Re: pingswept.org: blogging since the end of time

nattapon — Wed, 18 Jan 2012 12:34:01 -0000

Thank you for posting this. It helps me very much.
btw, to create Vandermonde matrix, I think you should use A = np.vander(t, degree+1)

Re: Building a 150 foot zipline

pingswept — Sat, 13 Aug 2011 12:40:45 -0000

Hi Bob,

Not 100% sure, but I believe the pulleys are these: http://www.rei.com/product/471...

They were definitely from REI, and the pictures look right. But we found the pulleys were getting pretty hot when a lot of people were riding the zipline, so you might try Petzl "speed pulleys" which are designed for ziplines.

One other suggestion-- use bare wire, not wire with a rubbery coating. The coating definitely slows you down.

Re: Building a 150 foot zipline

Bob — Fri, 29 Jul 2011 21:48:35 -0000

yo what is the exact name of the pulley and where can i get it? How much does it cost?
Thanks

Re: pingswept.org: blogging since the end of time

KevinInConcord — Wed, 09 Mar 2011 14:42:00 -0000

One way to reduce your water usage further is to install a Bricor showerhead. I did a lot of research into water efficiency before buying our house last year and stumbled upon their website. I bought two 1.0gpm showerheads, and not only do they give quite good showers, but they actually clock in at only 0.9gpm or so (depends on the local water pressure). They're 2x as expensive as other showerheads, but I think the payback would be fairly quick (not just heat, but here in Concord, MA, our combined water+sewer rate is roughly 1.5c/gal, which adds up quick).
Our houses sound similar, and I actually almost went with a Viessmann when replacing our HWH (went with an indirect stainless from SuperStor, but want to get a Viessmann dual-coil indirect some day for a solar hot water system to connect to the SuperStor as a backup). Still have our gas boiler w/ steam pipes though.

Re: pingswept.org: blogging since the end of time

Ardarga — Fri, 08 Oct 2010 15:54:22 -0000

Very nice and useful program. Thank you for shearing

Re: pingswept.org: blogging since the end of time

pingswept — Mon, 20 Sep 2010 15:19:41 -0000

That's a decent question, Chris. I don't know the answer off the top of my head, but I'm sure it's a solved problem. I'd guess that you'd look at the average residual, normalized by the average output value, but that probably falls apart when your data spans many orders of magnitude. To Google!

Re: pingswept.org: blogging since the end of time

Chris — Mon, 20 Sep 2010 15:10:14 -0000

A very intuitive post. Your code worked perfectly for me. However, how do you determine the overall "fitness" of the estimated curve? Is there some metric, like linear regression's r^2, but for a logistic model?

Re: http://pingswept.org/2010/04/29/liquid-metal-batteries/

pingswept — Wed, 18 Aug 2010 16:23:17 -0000

I think molten salt batteries are a subset of liquid metal batteries, but you'd have to ask a chemist to be certain. Both have electrodes that are metals, or at least good conductors. Molten salt batteries have electrolytes made of salt, but there may be liquid metal batteries that have electrolytes that are not salts.

Re: http://pingswept.org/2010/04/29/liquid-metal-batteries/

reemrevnivek — Wed, 18 Aug 2010 16:10:21 -0000

What's the difference between a liquid metal battery, and a molten salt battery?
http://en.wikipedia.org/wiki/M...

Re: Building Blender 2.5 from source on Ubuntu 9.04

Mzsade — Wed, 30 Jun 2010 23:56:13 -0000

sade@sade ~ $ svn checkout https://svn.blender.org/svnroo... /home/brandon/blender2.5
svn: URL 'https://svn.blender.org/svnroo... doesn't exist
sade@sade ~ $

Re: http://pingswept.org/2010/04/29/liquid-metal-batteries/

Charley Robinson — Tue, 29 Jun 2010 23:02:25 -0000

Seems like the size of such a device would be pretty substantial in comparison to what we consumer-grade individuals accept as "battery sized". Even so, I'm curious about the energy density of such a clever device -- is it comparable to something we layfolk may be more familiar with? Say, more energy per unit volume of a lithium-ion process, but less than a can of gasoline?

Re: http://pingswept.org/2010/04/29/liquid-metal-batteries/

pingswept — Mon, 07 Jun 2010 20:44:50 -0000

Glad you're interested, but I'm afraid I don't have any particularly new insight right now. The battery team at MIT is currently building battery prototypes and testing different electrodes; there's a decent chance that I'll be building some larger test system hardware in December and early 2011.

Still seems like a decent idea, but there's a long way to go to commercialization.

Re: http://pingswept.org/2010/04/29/liquid-metal-batteries/

John F. — Mon, 07 Jun 2010 19:00:39 -0000

"creating carbon dioxide, which is then vented to the atmosphere to help keep the planet warm"

:-D

"... for the next month or so."

So you might be winding down your work for them right about now? When you do, I'd love to hear your thoughts again.

Re: Drain heat recovery

malcolm — Sun, 30 May 2010 06:58:51 -0000

I think your savings calc has an overly optimistic starting point because you're ignoring the substantial heat loss between the showerhead and the drain. All those little droplets of 40C water are a pretty good water-air heat exchanger - you can feel the big difference between the temperature of the water at the heat and the drain next time you're in the shower.

cheers

Malcolm

Re: http://pingswept.org/2010/04/12/designing-embedded-systems-with-web-frameworks/

pingswept — Wed, 14 Apr 2010 13:31:17 -0000

Forget buffer overflows-- I'll probably leave the root password blank, or something stupid like that.

If I could somehow involve medical records or credit cards, we'd really be in business. . .

Re: http://pingswept.org/2010/04/12/designing-embedded-systems-with-web-frameworks/

Jerry — Wed, 14 Apr 2010 13:07:17 -0000

The reason it's a Good Thing is that, by exploiting a buffer overflow, some kind soul can now adjust your thermometer for you from the interwebs.

Re: http://pingswept.org/2010/04/12/designing-embedded-systems-with-web-frameworks/

pingswept — Tue, 13 Apr 2010 17:10:51 -0000

Hi Henry! Glad to hear you're lurking. I think you just signed up to be a beta tester as soon as I have a prototype working.

$4k after an educational discount is a lot of money, especially when you still have to buy the cameras. But, I suspect that the National Instruments image processing algorithms will work well. Seems like a little bit of a missed learning opportunity (who doesn't love a Hough transform every now and again?), but I'm sure there's plenty of other fun stuff to play with, like the lawnmower's blade.

By the way, that dice-rolling machine makes me seem reasonable!

Re: http://pingswept.org/2010/04/12/designing-embedded-systems-with-web-frameworks/

Henry Snow — Tue, 13 Apr 2010 00:41:17 -0000

I find myself popping back to your website every few weeks and always find these fun articles. At Case Western Reserve University, I've been working on a autonomous lawnmower project and we recently purchased NI's Embedded Vision System (NI EVS-1464RT) to run our obstacle and localization image processing algorithms. I'll just say, if you like LabView and need to run some computationally intensive algorithms (at 9-30VDC), this will be your friend. At the same time though, an educational discount brings the price down to around $4k + camera costs. For this application (prototyping out real-time, multi-camera image processing) it is worth every dime. From what I can tell, the features available on this system are far beyond what most applications would require. One of my favorite low-cost vision systems was the Dice-O-Matic (http://bit.ly/yElUW) that showed up the webs about a year ago.

Anyhow, soapboxes aside, I completely agree with you about these embedded systems becoming web platforms. A friends company is about to start work on making gym exercise equipment web-enabled for automated workout statistics. I'm sure over the next few years, more and more every day devices will be fitted with low-cost wifi-enabled microprocessors. I just can't wait to be able to use google analytics on my coffee maker...

Re: http://pingswept.org/2010/04/12/designing-embedded-systems-with-web-frameworks/

pingswept — Mon, 12 Apr 2010 21:05:57 -0000

As a quick answer, I don't think it's a particularly good thing so much as an incremental advance. It's kind of like the displacement of carburetors by electronic fuel injection in car engines of the late 70s and early 80s. Fuel injection is cheaper and more reliable, but most drivers don't know or care about the difference.

The reason I care about the change in embedded systems is that I work with the old style of systems on a regular basis, and they're maddening. Last week, for example, a friend of mine was saying that he had gotten a quote for a machine vision system for $30k. The task was to look a a solar cell on a conveyor belt with a camera, and then tell the machine controller whether the cell was lined up right or not. He thought there was a good chance we could literally ziptie an iPhone into the machine and use its camera and wifi connection to do the job, which would cost substantially less than $30k.

Anyway, that's the quick version. (The long version is kind of boring.)

Re: http://pingswept.org/2010/04/12/designing-embedded-systems-with-web-frameworks/

Justin Reich — Mon, 12 Apr 2010 20:45:18 -0000

The next post in the series should explain why this is a good thing... This is at the "sheer genius assuming everyone knows why it's sheer genius" level. Take a step back for us little people...

Re: http://pingswept.org/2010/03/31/picking-a-cortex-a8-cpu/

Mike Goelzer — Thu, 01 Apr 2010 23:41:30 -0000

I've looked at the idea of building a ~50-100 keV xray source with something like a Marx generator and an old rect. tube, but I'm not at all convinced I could do it without electrocuting myself (which I see as a bigger risk than the xray exposure, assuming reasonable sheilding). Overall, though, I'm more excited about single photon detection experiments. Let me know if you go through with it, though, because I've got some ISO 32000 (!) polaroid film that I'd like to test before it goes bad later this year.