Comments on: Electric (motor)cycle extra

By: michael(HOLLYWOOD)raines

michael(HOLLYWOOD)raines — Mon, 14 May 2007 22:14:29 +0000

zachs bike is cool wish he had a video like me on you tube I'd love to see it run. soo.....go to my websiteand see my electric bike do a burnout.
HWOOD.
http://mysite.verizon.net/res83zys/

http://www.youtube.com/holywdtughguy

By: Sean

Sean — Wed, 09 May 2007 08:01:59 +0000

ISTR the PC parallel port will give you an additional five bits if you take a look at some of the control lines, and don’t mind one or two of them being inverted, and are willing to wiggle port 0×379 and 0x37A, in addition to port 0×378.

Though, the Wikipedia article on the Parallel port implies you can get an additional nine (!) bits when using the control lines, for a total of seventeen bits. Interesting.

By: andrew

andrew — Wed, 09 May 2007 05:09:47 +0000

i tend to agree with #8 on the source of the jitter. here are a few clarifications: first, the temperature probes are at the end of two ~15-foot cables, meaning “thermal conductance” really has nothing to do with it. threepointone: temperature sensitivity on the part of the 555 is also not an issue. the circuit itself is kept indoors in a constant-temperature environment and is not subject to the same temperature changes as the sensors. but i agree about the poor quality of the 555 as a clock source, especially when looking at its waveform on the oscope :) however, the frequency is likely too low for the 555′s irregularities to make a big difference. plus, at least one other design I’ve seen using the ADC0808 chip uses a 555 for the clock signal without problems. wolfmankurd: I’ll likely implement that into the logging program, shouldn’t be too hard.

all in all, the circuit’s doing its job fine for what i’m using it for, but if i had to do it again, i would a) use a different clock and b) use a 10- or 12-bit ADC and send the extra bits through the Status line on the parallel port.

By: ...

... — Wed, 09 May 2007 02:41:31 +0000

The noise in the thermometer comes from not using enough bits on the A/D, not the clock.

The clock only serves to tell the a/d to update its value at 10khz, you could double the frequency and it would not affect the output.

The noise he has is normal for 8bit a/d, the chips are generally speced at +/-2 counts or so, so when you only have 250 counts that is a +/- 1% drift.

If he wanted to clean it up he could use more bits in his a/d (although then you run into the problem that a serial port only has 8 lines, so you have to start finding unique ways to get more data through) or had his program average every 10 samples or so before making a point (but then you are using 10x as many resources...).

that is why most people just use microprocessors ;)

By: ElectEngr

ElectEngr — Wed, 09 May 2007 02:07:53 +0000

I’m amazed that the parallel port A/D converter was built by a sophomore in high school (at least that’s what his website says). I know I couldn’t have built that or done any of that programming as a sophomore in high school.

By: threepointone

threepointone — Tue, 08 May 2007 23:09:04 +0000

I get this feeling that some of the inaccuracy comes from using the 555 chip for a clock. It's fine for simple timing apps, but don't even think about using it as a clock, especially for measuring temperature. the frequency accuracy is only 10% or so (some of that jumping around might be from jitter when sampling the data or something funky like that), 10ns of jitter (pretty bad for timing), and it's has pretty terrible temperature stability (when you're trying to measure temperature!)

At this accuracy, I'm honestly not all that certain if the 8-bit adc or the 555 timer is the main source of the jumping data, but if you're going for something more accurate, you'd better dump that 555 for a real clock.

app note on clock oscillators: http://www.maxim-ic.com/appnotes.cfm/an_pk/569

By: Bob7k

Bob7k — Tue, 08 May 2007 19:47:23 +0000

This thing has got to sound of pure awesome going down the street. I agree, Li-Po batteries are very nice, i have one powering my archos av500, thing lasts longer than an ipod. however, i dont see many companies using them, at least not in person, it itsnt a common battery whare i live. however, i have seen the chocolate, that phone, i think its from verizon, that one uses a Li-Po battery. i was just wondering, do you think you could efficently power a high end metal detector with a home made battery array thingy if it was designed to run in a bunch of AAs? also, is it me or do these things have a rapid recharge rate?

By: Wolfmankurd

Wolfmankurd — Tue, 08 May 2007 18:27:54 +0000

The parallel port A/D converter would be much more accurate if he took a moving average of the temps, or just averaged every 1sec to reduce it’s jumping around. Or he could have wrapped the probe in something with quite low thermal conductance to get the same effect.

By: electricmotorcycles.net

electricmotorcycles.net — Tue, 08 May 2007 16:47:15 +0000

"I (Michael Raines) have an incredible mechanical ability, along with the intriguing, off-the-wall personality."

At least he doesn't take himself too serously.

By: PKM

PKM — Tue, 08 May 2007 15:10:24 +0000

Hmm.. I will admit the electric chopper built by michael raines does look pretty sweet, but if you don’t like things that make you cringe, the rest of the site may not be for you…

“I (Michael Raines) have an incredible mechanical ability, along with the intriguing, off-the-wall personality.”

By: ...

... — Tue, 08 May 2007 11:02:05 +0000

this one is pretty cheap
http://cgi.ebay.com/RC-Li-po-Lithium-Polymer-3500mAh-7-4v-lipo-Battery-UP_W0QQitemZ270117482843QQihZ017QQcategoryZ34056QQrdZ1QQcmdZViewItem
admititadly it will bid up to about $20ish, but when you condiser $/w/hr that is only about $1/w/hr...

By: Bob7k

Bob7k — Tue, 08 May 2007 08:46:27 +0000

Those Li-Po batteries arent cheep…