New Products Flash Sale
Jack Crossfire's blog View Details
Posted by Jack Crossfire | May 30, 2015 @ 11:45 PM | 2,778 Views



The 1st video card. It was PCI. None of this AGP, PCI-E nonsense. It's not about the performance as much as the memories. It was upgraded to some ghastly amount of memory for its time. The day it was upgraded, there was a geeky old guy in the same room, slobbering over a dual Pentium II he just received. It was the seed for the idea of a dual CPU. In the meantime between then & the dual CPU, this card went from 16 to 24 bit color. What an improvement.


The 2nd video card was for the dual CPU & had a lot of issues. After this came the era of freebie's which lasted until 2006. The next video card will be the 1st one bought since 1999.

...Continue Reading
Posted by Jack Crossfire | May 27, 2015 @ 11:17 PM | 2,352 Views



Recycled the last of the through hole components.


After recycling most of 2 old power supplies, it became clear that more through hole components had been reclaimed than would ever be used in the rest of my life, so many parts were left.


The 20 year old Soundblaster had some through hole. A lot of memories came from that card. It was a miracle for a computer to play any amount of PCM audio, in those days. This card still had a muddy sound, compared to the best. It was a rare one which could do full duplex, with enough coaxing.

It only did 44.1khz. Had to upgrade to a new one when 48khz DVD's came along. The 48khz couldn't record higher than -3dB. It gave up a 24khz crystal, which makes the 44.1khz limitation seem like a driver issue. It would never have been replaced if that was known.

...Continue Reading
Posted by Jack Crossfire | May 25, 2015 @ 07:18 PM | 1,834 Views





Decided to revive the XV-11 LIDAR, after failing to find a convenient way to wear a sonar transmitter. It wasn't expected to work in daylight. The XV-11 remanes to this day the only 2D ranging device for a comprehensible cost. 4 years of kickstarter projects could do no better. 2 of the LIDAR modules were probably salvageable, with the 3rd taking some doing.

After a year since the XV-11 arrived, it was finally rebuilt & plotting 2D scans as designed. The error rate bottomed out at 250rpm. Below & above, it increased. It depends on stable motor movement as well as speed. The scan resolution is limited to 360 points. You'd get higher resolution scans by sending a fake photo interruptor pattern down the TX line to the photo interruptor pad. Clock it as fast as possible & spin it with a hobby motor being driven in stepper mode.

After building up the software to assemble images & the dual power supply, it was completely useless in sunlight. If the camera got pointed at the sun, it shut down until power cycled. It couldn't detect any surface lit by the sun. In shade, things got more interesting. It could resolve a concrete path, grass on 1 side, & a curb on the other side.


...Continue Reading
Posted by Jack Crossfire | May 17, 2015 @ 10:09 PM | 3,601 Views




Fired up a segment of the last sonar circuit, built 3 years ago. Stepped up the voltage to 41V with a 43k 100k parallel resistor network & after a period of fits & starts at 500mA, it ran off 2.5V 100mA just fine. It drove a pulse every 50ms. The voltage hitting the transducer would have been slightly less than 41V.

The next steps are building up a new receiver & recording the waveform on an actual drive. Skepticism keeps progress slow.
Posted by Jack Crossfire | May 16, 2015 @ 02:48 AM | 3,385 Views
It's quite clear that there's not enough time to do any projects at home, anymore. After 3 hours of commuting, the amount of stretching & high vegetable eating required to do all the things this body does consume about all the time.

There's an idea for a monocopter based on a single chip for the radio & computer, with a hacked ESC for silent operation. There's an idea for path following.

Tracking the athlete once again presented a better solution than tracking the path. Sonar presented a better solution than vision, because of the problem of consistent lighting of a target. It would depend on GPS to get the direction of the path. It would use the athlete's position to get the sideways position on the path.

For all the times sonar appeared, it has never worked. The ambient noise on the receiver side is too high. It's too directional. There are too many echos. There aren't any echos outside. The athlete would have to stay in a cone of directionality.

But there are times when staring at a schematic for the day job when you remember staring at schematics as a young boy. They were neat looking drawings of which you knew nothing of the function, only that the collection of lines did something important & better than any human could.

Now, sometimes, you do get paid to stare at the schematics of complete consumer products & know the functions of most of the parts. A lot of livelihoods depend on you to make the once mysterious drawings work, sometimes rework them, sometimes design them. There are still a few mysteries in the analog stages & it's not long term at the rate prices are rising, but for now it's what you want to do, you run along the waterfront every day, you get free lunches, & your boss isn't completely crazy.
Posted by Jack Crossfire | May 14, 2015 @ 11:25 PM | 8,210 Views


So do you feel like an idiot for spending $1400 on a 30" 2560x1600 monitor, 7 years ago? Of course not. The 30" provided $1400 of content. It's going to continue doing so until a system capable of producing 4k video becomes affordable.

Making software look right on a 4k monitor is another matter. For now, it amounts to doubling the font size which looks right on a 2k monitor & hoping for the best. There's also examining the software in the Apple store. These monitors are 3840x2160, only 1.5x the resolution. There's no interest in slowing a program down with dynamically scalable assets.


No-one wonders anymore what technical challenges had to be overcome to make a 4k monitor. They just magically appear from South Korea.
Posted by Jack Crossfire | May 13, 2015 @ 11:08 PM | 2,371 Views

Photos of the world wars get more fascinating as they drift farther into the past. Germans of WWI look absolutely alien, nowadays. Part of it might be the alien gas field they're fighting in.

The ruins of WWII still surround us every day. They are a reminder that a world war in what was perceived as a modern age was fought with nothing more than crude sticks & stones. That was the best they had.

Why hasn't something like that happened again in almost 100 years & when will we have to fight something like that again? As long as 1 country held a dramatic advantage no-one else could win against, there was no world war, but the field is leveling again. The happy time won't last forever. It mathematically can't. All armies must regress towards the mean & have a chance against each other, eventually.
Posted by Jack Crossfire | May 12, 2015 @ 01:29 AM | 3,789 Views
Another day of testing path footage revealed the mane problem is detecting the horizon. The RALPH algorithm requires the camera to be pointed down from high enough to fill the frame with some distance down the path. The Tamiya is too low. Even with GPS aid, it was still unsolvable without spherincal LIDAR. 10 years on, spherincal LIDAR is still more than the cost of a new car.

It turns out all the LIDAR modules GOOG uses are made by 1 company, invented by 1 guy in search of a use for a real estate investment that went bust in 2001. There was a nifty video about it.

Velodyne: On Sound, LiDar, and Marine Technologies (25 min 7 sec)


A good reminder that if a real estate investment had to go bust for the innovation to happen, how much innovation can happen in a world without recessions, where there is no interest on credit, no limit to borrowing, any invention no matter how small has infinte valuatioin.

Also a good reminder that 10 years on, no-one at the Goog has a clue how to make an audio limiter. They just know how to recite 50 path following algorithms in interviews.
Posted by Jack Crossfire | May 11, 2015 @ 12:35 AM | 2,710 Views
Goog produced many videos vaguely describing how their self driving car worked, in 2013. Then, they stopped discussing how it worked & focused more on philosophy. No matter who the employee is, whether it's a marketing guy, middle manager, or an engineer, they all show the same slides, show the same videos, read the same lines. It's a perfect corporate machine.

Embedded Linux Conference 2013 - KEYNOTE Google's Self Driving Cars (24 min 27 sec)


They reveal hardly anything about how it works. The mane nuggets are it uses GPS to get a rough position fix. Then it matches the current LIDAR image with a 3D map of the world to get the exact position. LIDAR detects landmarks & pedestrians. Radar detects metal objects like cars. They have to manually drive every route the computer drives, to make a 3D map of the world, manually entering all the traffic rules, all the lanes, all the traffic light positions.

It doesn't use vision for anything besides traffic light color. They made no mention of vision detecting lane markings, unlike lane keeping systems. Lane keeping systems must have proven too unreliable for a fully autonomous car.

Infiniti Q50 Active Lane Control - Selfdriving Car (2 min 6 sec)
...Continue Reading
Posted by Jack Crossfire | May 10, 2015 @ 06:33 PM | 2,023 Views



Next to go was the dual 1.4Ghz Athlon. It was the 1st affordable SMP motherboard for an AMD chip. It was quite a breakthrough, for its time, but it only did 12Mbit USB & had an extremely slow bus. Contention between disk I/O & PCI cards abounded. It couldn't ingest video from a PCI card that a dual 933Mhz Pentium had no problem with.

Tyan offered a server equivalent simultaneously. In those days, the Thunder K7, Tiger MP, & Tiger MPX were the duck's guts. It was the last golden age for a 32 bit CPU.



...Continue Reading
Posted by Jack Crossfire | May 09, 2015 @ 06:54 PM | 1,807 Views
There is somewhat of a desire to build a fully tricked out Pixhawk based copter from scratch, make the electronics from scratch, make a custom radio module for it out of an 8 year old XBee, & make a brushless gimbal for it. It would be just like the original plan to build a copter around Rotomotion's code fragments, but it would work.

The problem is the severe limitations of hobby grade copters are still around, no matter how many billions every quad copter company is now worth. There were a few issues in a recent video:

Flying the DJI Inspire 1 Quadcopter with Adam Savage (10 min 38 sec)


Modern radios continue to require line of sight. The radio & limited battery life still limit them to flying in a cone directly overhead. All quad copter videos look the same. They fly 1 mile horizontally, then have to turn around.

Navigation is a lot better, but it's still not possible to make a timelapse of a quad copter by having it stand perfectly still. They still look out of control in windy environments. They still can't be flown reliably enough near people & property, eliminating most of their use. They're still far too dangerous to be the household objects they're now marketed as.

Another video showed how useless optical flow continues to be, despite being the latest thing. It still required constant attention to keep from drifting away. Low altitude continues to require constant attention, as the barometers still drift.

Hands-On with DJI&#
...Continue Reading
Posted by Jack Crossfire | May 07, 2015 @ 10:17 PM | 1,518 Views





It was finally time to tear down this 15 year old beast. It was intended for verifying NTSC compliance, but NTSC was short lived. It was used for maybe an hour upon arrival, in an unsuccessful attempt to receive local TV stations, then was briefly valuable in burning a DVD after 7 years, then never again for 8 years.







The innards are revealed for the 1st time.


...Continue Reading
Posted by Jack Crossfire | May 03, 2015 @ 05:15 PM | 2,418 Views



It has become quite clear that not being able to take it all with you is part of daily life, even for stuff that used to be considered valuable like electronics. It's especially hard to get rid of something that was $5000, but is now worthless. The problem is a large mass of antique computers has accumulated over 20 years, which is unaffordable to keep in an age of $10,000 rent.




The dual 2.4Ghz Opteron was $5,000 when it was new & served from 2006 - Summer 2010. If it wasn't for a misdiagnosed power supply issue, it would still be in use. Now, it's long been outdone by the laptops.

The dual 2.5Ghz Opteron was unreliable & slower than the 2.4Ghz, because of a limited memory bus, but it served for a very long time.

The dual 1.4Ghz Athlon was a quick decision which ended up very low performing.

The dual 933Mhz Pentium III can be outdone by a $30 phone, but it was a fortune when it was new. It was the standard office computer in 2000.

There were still bits from a dual 450Mhz Celeron. It was a true miracle when it was new, the 1st SMP machine. The moment of inspiration for building it is still a strong memory of waking up 1 sunny day thinking it was time to go SMP.

There was a 166Mhz Cyrix which was slated to become a router, but now you can solder something like that from loose parts.

For a long time, the plan was either to boot them up again for embedded applications or recycle parts as needed. It...Continue Reading
Posted by Jack Crossfire | May 03, 2015 @ 12:35 AM | 2,568 Views
After a frenzy of activity lasting 2 days, the old dump which was home for 14 years & 6 months was no more. It was like everything that happened there never was, yet if I woke up there again & everything was in its old place, it would feel as at home as if the move never happened. Perhaps in a parallel universe, a fake test pilot still lives there & life continues as it always did. Just 4 months ago, everything seemed as if it was going to continue as it always did, forever. There would be another summer there, another XMas, another year of waking up in the same place, but it suddenly ended.



Posted by Jack Crossfire | Apr 18, 2015 @ 12:00 AM | 3,606 Views
It certainly wasn't the biggest government bungle, but the giant welding machine began to tilt immediately after it was built. There were rumors that it was shedding bearings.




3 months after it was built, it was leaning .06 degrees or 1/4" out of alignment at its highest point. The whole thing was torn down with plans to rebuild it, someday. In typical government contractor fashion, the swedish contractor was supposed to reinforce the foundation & just simply didn't. Tough beans.


http://spacenews.com/fix-in-works-fo...lding-machine/







Posted by Jack Crossfire | Apr 16, 2015 @ 11:09 PM | 2,769 Views
https://vid.me/i6o5


In a touching finale, the mangled engines & landing gear that someone worked his ass off to get working are revealed before the $60 million piece explodes.



There was a rumor that the mane engine can't throttle low enough to descend. They have to start the engine close enough to the landing pad to reach 0m/s at 0 altitude.
Posted by Jack Crossfire | Apr 15, 2015 @ 10:07 PM | 11,302 Views
CRS-6 First Stage Landing (0 min 23 sec)


After many ice obstructed videos from onboard the rocket, lost viewfinder videos from a chase plane, & partially visible gopro videos from the barge, it was the 1st watchable footage of a landing ever captured. It was most impressive by how hard it was to make the video & how far they continued with no footage of a previous landing to go by.

The final landing attempt was the most aggressive out of control fall, following by last minute suicide burn. The legs deployed 7 seconds before touchdown, while in previous videos they deployed 10 & 9 seconds before touchdown. If they used the more conservative approach of the past landings, it would have made it.

They obviously tried to stretch the fuel farthest with the least stable approach they could get away with, but it's not stable enough to recover from. They need a customer with an even lighter payload than NASA, but so far, NASA is the only customer willing to throw away enough upmass to get this far.

There are 2 more landing attempts this year, in the form of 2 more NASA missions.
Posted by Jack Crossfire | Apr 12, 2015 @ 11:22 PM | 2,769 Views
The blurry, shaky cellphone cams reveal a world of dead ends, buzzwords, & few practical demos. It's like a frozen moment in time when Goog glass just came out & everyone wanted to be bought out by doing a copy of that. 3 years after these concepts hit the kickstarter circuit, none ever became mass produced & all attention shifted to virtual reality goggles.



WEARABLE COMPUTER WITH HAND HELD KEYBOARD AND MOUSE (6 min 20 sec)


That guy loves joysticks & patents.

A reversion to the trackball, but with a pad instead of a ball, might get more mileage than either a joystick or a full pad.


ETAOI - a five key Android keyboard for phones and wearable computers (15 min 25 sec)
...Continue Reading
Posted by Jack Crossfire | Apr 12, 2015 @ 12:11 AM | 2,217 Views
So the idea came up of a simple commuter setup that would allow someone to type while standing up in a crowded train. The system would be compact & light enough to run down 1 mile of city streets. How could it be done as simply & cheaply as possible?

It amounts to 2 inventions that went absolutely nowhere: the virtual reality goggle & the wearable keyboard. Google glass would have been the ideal form factor, but it only did 640x360. No virtual reality goggle has ever been mass produced. Google cardboard might work, but would need a camera projecting outside video in a window. Maybe a new kind of goggle with no sides can be invented. The cheap phone is only 480x800.

The most effective keyboard leans toward a rubber thing. The rubber thing contains a full keyboard & touchpad. A different full keyboard & touchpad is worn on each side of the abdomen. If not a full keyboard, each side overlaps the other by a few keys. Maybe it could be a single rubber keyboard on the abdomen.

This amounts to a lot of money. Maybe there's an incremental step from the bog standard phone.

The average commuter stares at a phone screen, desperately trying to be entertained by what meager, meaningless news bites it can download between tunnels. Don't know what they did 30 years ago.

The mane limitation with this scheme is 1 handed or 1 fingered typing being extremely slow. What you need is a 3rd arm to hold the phone so both hands can type. If an arm can be designed, it conceivably simplifies the problem quite a bit. A simple arm to hold the phone combined with the abdomen keyboard might do the job.

Searches for wearable computing show a wasteland of vaporware stretching 5 years into the past. There's much more vaporware than the old days, showing how much longer vaporware can survive without making money than it used to.
Posted by Jack Crossfire | Apr 09, 2015 @ 10:57 PM | 2,241 Views
Free nuggets about lane following are few & far between. There are many nuggets about line following, but not lane following. This story was fascinating for 2 reasons

http://www.roboticstrends.com/articl...riving_in_1995

It was work my generation did when it was in college, with the tools available during our time. It was so primitive to modern eyes, yet it was the bleeding edge for someone living in that time. There was no GPS. Capturing video on a computer was nearly impossible.

The full text of their path following algorithm costs money, but there's a rough description

http://www.cs.cmu.edu/afs/cs/user/tj...haa/ralph.html

The key to their algorithm was a scanline intensity profile. It was done on a 30x32 greyscale image, on a 486. The scanline intensity profile was derived from a test of lane curvature. The lane curvature can be neglected, since the current issue is a straight path.

They captured the entire lane width, converted the trapezoid shaped path to a rectangle by widening & shifting farther rows horizontally. When making the rectangle shape, they made several images with the farther rows shifted left or right by different amounts. Then they summed each column in the rectangle image. The adjacent columns had the maximum differences when the shifting of rows matched the path's true position.


The sums of each column made up the scanline intensity profile. The scanline intensity profile when the vehicle was centered in the lane could be compared to the current scanline intensity profile to give its lateral offset. The current scanline intensity profile was iteratively shifted left or right until it matched the centered one.


They claimed better results this way than with edge detection. The key advantage was immunity to shadows, relying only on visual features running parallel to the road. This method does require training the algorithm with known scanline intensity profiles for a centered vehicle on different sections of road. There was another issue of cropping the image to where the path should be.