In an effort to help Tracy Allen in his extensive work in providing functional code for Mark in NH and the Rocketeers,
I've been trying to run his latest Vinculum code.
The port connection assignments for the Vinculum are identical to those shown in the Parallax Demo code.
Tracy wrote the code for a BS2 Stamp, but the Board of Education I use to verify the code has a BS2p Stamp.
I made the obvious change in baud rate to enable the BS2p to communicate with the Vinculum. However,
I still found the Vinculum was not initializing reliably.
An external AC powered wall-wart was used to power the BOE.
Fixing the baud rate constant for use with a BS2p, Tracy's made code run a little better. Immediately after the code
was downloaded from the PC to the Stamp, the sync characters were recognized when the sync count reached
anywhere from 2 to 8.
Running from a cold start, however, it would fail every time, with the sync count reaching 9.
Another odd symptom running from a cold start was that the Vinculum red and green LEDs would begin flashing
and the first DEBUG statement in the code would NEVER EXECUTE. It was as if the Vinculum was somehow
interfering with the BS2p code and preventing it from running.
Very occasionally they would turn off after 15 to 30 seconds and Tracy's Vinculum initialization sequence
would complete before the sync count reached its limit of 9. Nearly all other times they would continue in the blinking state
for several minutes or until I manually turned off the power.
There are two LEDs on the Viniculm Datalogger assembly, a red on and a green one. When power is first applied to
that unit via the Stamp 5v regulator, they blink at about twice a second, alternating green and red for 3 cycles followed
by a short 1 second pause when the both turn off. Usually this sequence stops after 3-5 seconds with the red LED
on.
After the green-red sequence finally stops, Tracy's code manages to successfully send "E" and receive the correct
response and the red LED on the Vinculum stays on.
As an aid in debugging, I stripped out all the other code except for the Vinculum related initialization sections. To ensure
there weren't any baud related issues, I used the Basic Stamp conditional directive and changed the baud rate constant
from vBaud to T9600.
See the attached file(s). The BS2 code runs reliably. The BS2p code doesn't.
During a phone call to Parallax Tech Support today, I spoke with Chris Savage and he said that those blinking LEDs
meant that the Vinculum was trying to establish communication with the flash drive plugged into the Datalogger.
They just keep on flashing until it finds a flash drive attached.
Chris said that the only difference between the BOE and Basic Stamp states when code is being downloaded vs a cold
start is that the Vinculum has had enough time to establish the flash drive communications link prior to download, whereas
on a cold start, it has to first create that link before any transfers between the Stamp and Vinculum can occur.
Replacing the BS2p stamp with a BS2px made no difference with respect to the cold start hang up.
However, when I installed a BS2, the cold start issue disappeared.
Another difference observed when using a BS2 is that the first DEBUG statement appears almost immediately after
power on, once the code has been downloaded
Four questions to the forum:
1)What is different about using a BS2p or BS2px with a Vinculum instead of a BS2 that causes the Vinculum never being
able to establish communications with the flash drive from a cold start?
2) Is it a hardware issue, a code issue or both?
3) After waiting over a minute following a cold start using a BS2p, why does pressing the RESET button on the BOE cause
the flashing red and green LEDs on the Vinculum to quit cycling and enter the ready state?
4) Why does the Parallax demo code run with any flavor of Stamp (BS2, BS2p,BS2px), but my stripped down version of
Tracy's doesn't?
I might also add that all the symptoms described above occurred using two separate Memory Stick Dataloggers,
as well as a BS2p and BS2px Stamp. They weren't caused by a bad Datalogger.
phil
Post Edited (phil kenny) : 3/11/2008 1:57:15 AM GMT
·· So what kind of gizmo did you have in your rocket that printed out the graph that·you posted? It doesn't look like it''s from a Black Sky AltAcc (since the AltAcc doesn't read temperature.) Can you post a picture? How big was the rocket? Are you L2 or L3 cert?
I won't tell the Washington County 4H guys or the Madison West team a thing·(they always qualify for TARC anyway...!) Have they made their qualification launches yet? What are they getting for time and altitude? We·made our first test flights last Thursday and we were waaaaaaaaaaaay off on altitude (513', 535', 569'). We had to take off all of our 'add-on' weights and even then we were way under altitude. Our rocket is·obviously too heavy and so we had to re-build·a·new·booster (from cardboard, not phenolic.) We'll test launch again tomorrow at around 990 grams on a G80-7.
·· Boy, that's a bundle of questions! I'm eager to find out what the problem ultimately is. For what it's worth (*and consider the inexperienced source),
recall that when we·had a similar problem·with the Datalogger initializing, we·made the·PAUSE after the initialization command longer. At first we set it
to 5 seconds; but ultimately we found out it could be set as low as 1 second. When Tracy posted the latest version of ROCKET_ROBOT, a similar thing
happened. I added a second "GOSUB init" command after the PAUSE (see comments several posts ago) and that fixed it. Why, I cannot say for sure.
I wish I had more to offer but I don't. As previously demonstrated, my 'expertise'······ (and I use that word very loosely) is limited. Keep·me posted!
If you look closely at my stripped down version of Tracy's code, you'll notice that
I changed the two PAUSE 512 statements at the very start to PAUSE 2000. Close
to the change you made. I think they helped.
Phil, Thanks for testing the Vinculum code. I wonder if it could be a power supply issue, noise or something, that shows up only with the more power-hungry BS2p? I will contact you by PM so that we can try to resolve that there outside this thread. However, I would like to hear from anyone else who has tips on using the Vinculum.
·· Thanks for your post about the wiring scheme. We'll keep them in mind as we (finally, or finely) begin to wire the robot for the last time. And I'm eager to scoop some Goop and try it out (in a well ventilated arae as you suggested.
You asked several posts ago (and I thought Andrew had already posted an answer): "What does the debug screen display for "sync=? and disk=?" The numbers·are "sync=2" and "disk=1" virtually every time. It looks like the Datalogger is initializing on the first try, every time (knock of wood!) Bravo!
So what do you make of Phil's conundrum? Strange that a different stamp runs the same program differently, isn't it...?
I just tried a BS2pe and it has the same cold start issue. The current draw for the BS2pe
is speced at 15 ma and the BS2 is 3 ma. Hard to see that 12 ma more load would
make the difference.
·· So what kind of gizmo did you have in your rocket that printed out the graph that·you posted? It doesn't look like it''s from a Black Sky AltAcc (since the AltAcc doesn't read temperature.) Can you post a picture? How big was the rocket? Are you L2 or L3 cert?
I won't tell the Washington County 4H guys or the Madison West team a thing·(they always qualify for TARC anyway...!) Have they made their qualification launches yet? What are they getting for time and altitude? We·made our first test flights last Thursday and we were waaaaaaaaaaaay off on altitude (513', 535', 569'). We had to take off all of our 'add-on' weights and even then we were way under altitude. Our rocket is·obviously too heavy and so we had to re-build·a·new·booster (from cardboard, not phenolic.) We'll test launch again tomorrow at around 990 grams on a G80-7.
Regards,
Mark
Mark -
The data were collected using a Vernier Labpro (www.vernier.com), with a 25G accelerometer and a fast-response temperature probe, whose tip came out through a small hole in a hatch. The rocket is 5.5" diameter, about 8' 4" tall. I'm NAR L2.
The Labpro was pretty easy to use, but large, heavy, and difficult to mount. I probably won't ever fly it again, and instead build something using a Stamp module. I've got an adapter so I can use the Vernier sensors with a Basic Stamp. The big issue is one you've already faced: interfacing a real-time clock. The acceleration data aren't worth much if I don't have an accurate time scale. I expect to spend a lot of time scouring your thread to pick out the great advice and help you're getting here.
I don't know what the current TARC flights·look like here - I do know that those folks flew some qualification flights at Bong two weeks ago, when I was flying a small rocket with one of my students for my RDF grant.
The mystery behind the peculiar operation of the Datalogger when using anything other than a BS2 Stamp with Tracy's latest code has been cleared up.
Today I powered just the Datalogger and thumb drive from a completely separate regulated 5v supply. Now, the Viniculum
is able to complete its internal handshaking with the thumb drive and Tracy's program runs fine, with minor adjustments to
the initial PAUSE statements, independent of the Stamp model.
Using a Fluke 87 DVM to measure the 5v current pulled by the Viniculum and thumb drive showed 155 - 160 ma using a 'real'
256 Mb thumb drive. I also measured the current using a USB card reader with a thumb drive form factor plus a 32Mb SD card.
That combination pulled 48 - 55 ma. These were just average readings and not peak.
Even pulling 100 ma less current, the USB card reader with an SD card wouldn't let the Viniculum finish its internal sequence of
establishing communication with the flash drive.
All of the peculiar Stamp behavior that I saw when running Tracy's code can be attributed to short term overloading of the BOE onboard power supply regulator.
However, I still can't explain why the Parallax Demo code doesn't exhibit the same cold start problems when using a BS2p,
BS2px, or BS2pe.
I wasn't able to find any info regarding current drain of the Memory Stick Datalogger on the Parallax website. Perhaps that's
because it is heavily dependent on the flash drive attached to it.
phil
Post Edited (phil kenny) : 3/12/2008 1:01:22 AM GMT
·· I'm glad to hear that you solved the problem! So you were essentially drawing too much power (amperage?) which preveted the program from completing the·initialization sequence? How does that work? Why does it slow down initialization?
The "Junior Rockteers" (the younger rocketry team) did the second round of their test launching today and it was like "Wide World of Sports" (I'm aging myself now.) It was "the agonoy of defeat", followed by the "thrill of victory." On their second test launch they forgot to attach the payload bay (a cargo module with their egg 'astronauts' and altimeter inside) and it went straight in from 800' feet up! Luckily it went into deep snow! Deep. Real deep. The eggs were very scrambled and we destroyed an altimeter. If it was a qualifying·flight (to qualify for nationals)·they would have been DQ'd because their 'astronauts' didn's survive. Luckily their sturdy payload bay survived.
On the very next flight their rocket came within one foot of the target altitude of 750', and 1.2 seconds with the target flight time of 45 seconds (749' in 46.2 seconds.) If this had been a qualifying flight they would be on their way to nationals. A score like that at nationals could well have won nationals! It certainly would have put them in the top 10 for scholarship money, and the opportunity to make a proposal to NASA for yet another SLI project like one you've been helping with.
Good call. I see from the Parallax data sheet (memoryStickDataloggerV1.0.pdf ) that it draws 25 mA (running) and 2mA (standby). So the rest of the ~160 mA would be due to the USB memory device. I can understand a surge of current for a few milliseconds when it writes data. I have no idea what it is doing when it first starts up. Maybe it fires up all its subsystems and those draw the current. It is indeed perplexing that it behaves differently with the different Stamps and with different demo code. Maybe a fat filter capacitor on the power supply would eliminate the glitch. Say 1000 µF, preferably installed close to the datalogger power connections.
Mark and rocketeers,
Is it still working on your system? No glitches? All the data being collected looks OK?
I'm attaching an uncluttered copy of the planning and layout template for the BOE breadboard area. I made this diagram for the Earth Measurements text.
Our two TARC teams·did their second round of test launching today and it was like "Wide World of Sports". It was "the agonoy of defeat", followed by the "thrill of victory." On their second test launch they forgot to attach the quick-link to the payload bay and it 'lawn darted' from·750' up.····· Luckily it went into deep snow! Real deep snow. It was a snow shovel recovery.·The eggs were very scrambled and we also destroyed an altimeter.····
·
On the very next flight their rocket came within one foot of the target altitude of 750', and 1.2 seconds with the target flight time of 45 seconds (749' in 46.2 seconds.)
Tell Madison West team that the Rocketeers from New Hampshire are ready to give them a run for their money at Nationals! Thanks for your post.
··· "Daedalus"continues to sing like Pavarotti (and·babble like a child in Morse code!) No glitches at all (knock on wood), and the program initializes·at 'sync=2, disk=1' every time.
Thanks for the breadboard schematic. It will be really helpful, for posterity,·in documenting where the wires go, and for anyone else who is interested. We'll post it as soon as we get it sketched. Tyler and Andrew should already be working on it for Sunday.
I was glad to hear that Phil's problem seems resolved. Even though I only understood about half of the conversation, it was interesting to follow.
You'll have to read the post below about our very successful test launch of our 'other teams' rocket this afternoon...! Thanks and,
The 5v regulator on the BOE can only supply a limited current. When that limit is reached, and the
load tries to pull more, the output voltage begins to sag or drop.
The Basic Stamp module has an undervoltage detector (UV) circuit that is set to trip somewhere
around 4 volts. When that happens, the UV circuit generates a system RESET. This in turn causes
the Stamp to go into a low power state and all output pins are left floating.
The RESET state is the same one generated by the black RESET button on the BOE. However, this
RESET signal isn't seen by the Vinculum.
With the Stamp now asking for less current, the 5v regulator may let its output voltage rise. If it climbs
high enough to cross the UV threshold, the Stamp comes out of RESET, pulls more current, and this
cycle can start all over.
All it takes is a momentary spike in the load current for this RESET cycle to occur.
Using a separate power supply for the Vinculum prevented the Stamp from being reset by any extra
load pulled by the Datalogger.
The "Junior Rockteers" (the younger rocketry team) did the second round of their test launching today and it was like "Wide World of Sports" (I'm aging myself now.) It was "the agonoy of defeat", followed by the "thrill of victory." On their second test launch they forgot to attach the payload bay (a cargo module with their egg 'astronauts' and altimeter inside) and it went straight in from 800' feet up! Luckily it went into deep snow! Deep. Real deep. The eggs were very scrambled and we destroyed an altimeter. If it was a qualifying·flight (to qualify for nationals)·they would have been DQ'd because their 'astronauts' didn's survive. Luckily their sturdy payload bay survived.
On the very next flight their rocket came within one foot of the target altitude of 750', and 1.2 seconds with the target flight time of 45 seconds (749' in 46.2 seconds.) If this had been a qualifying flight they would be on their way to nationals. A score like that at nationals could well have won nationals! It certainly would have put them in the top 10 for scholarship money, and the opportunity to make a proposal to NASA for yet another SLI project like one you've been helping with.
Wow. I've witnessed quite a few TARC flights, but nothing close to that.
My project is a college-level separate grant with the Wisconsin Space Grant Consortium, which is a state level funding arm of NASA. No qualifying involved, other than writing a convincing proposal.
With the experience you've gotten in the last few months, you're an expert. No doubt about it. You're going to give Pavel and Brent (Mad West) a run for their money.
·· Thanks for the really clear explanation. That makes compete sense to me and now I understand what·was happening.·Basically it was just RESET-ing itself because there was a spike in power (voltage, or amperage?), yes?·So how's it running now?
.-.·· .··· ...·· .··· -···· !!
Mark
phil kenny said...
Mark,
The 5v regulator on the BOE can only supply a limited current. When that limit is reached, and the
load tries to pull more, the output voltage begins to sag or drop.
The Basic Stamp module has an undervoltage detector (UV) circuit that is set to trip somewhere
around 4 volts. When that happens, the UV circuit generates a system RESET. This in turn causes
the Stamp to go into a low power state and all output pins are left floating.
The RESET state is the same one generated by the black RESET button on the BOE. However, this
RESET signal isn't seen by the Vinculum.
With the Stamp now asking for less current, the 5v regulator may let its output voltage rise. If it climbs
high enough to cross the UV threshold, the Stamp comes out of RESET, pulls more current, and this
cycle can start all over.
All it takes is a momentary spike in the load current for this RESET cycle to occur.
Using a separate power supply for the Vinculum prevented the Stamp from being reset by any extra
load pulled by the Datalogger.
·· The "Jr. Rocketeers" (TARC team) test launched again today in exactly the·same weather conditions, at·the exact same launch weight. THIS time their rocket went:
802' - 47 seconds
790' - 46 seconds
We think a loose fin on the 'perfect' flight the other day (749' in 46.1 seconds) caused the rocket not to achieve it's maximum altitude.
Recalculate, reweigh, reload, relaunch... our qualification flights are 9 days away!
Mark and the Rocketeers,
'Daedalus' ( ....·· ..·)··and Bandit the Rocket Dog
Mr. Kibler and Andrew,
Here is a diagram of the breadboard before any new arrangement has been made. Do you want me to go ahead and experiment with my board to find the best arrangement before our meeting on Sunday?
·· Thanks for posting to the forum!· I'd leave it like it is for now, especially if it's wired like ours, and working. We'll 'tweak' it on Sunday. Has your RTC and cyrstal arrived yet? Did you get·them ordered like you said? We'll solder and attach them on Sunday like we did with Andrew's last Sunday. Thanks again for posting to the forum.
Mr. Kibler again. Have you included your diagram, with a label and an explanation,·with your section of the FRR report? It would go really well in the report. Please reply to the forum.
Funny you should ask. On 3/25/2008 I send Mark Kibler a message, asking the same question. I asked Mark if they were all on vacation. Here is the response I got then:
- - - - -
No vacation here (though the idea IS appealing!) The SLI team had last Sunday off for Easter as we regroup and recharge our batteries before we do our video teleconference with NASA next Monday.
"Meanwhile, back at the ranch" the two TARC (versus SLI) rocketry teams have been in a flurry of activity making test launches and getting ready for qualifying attempts. It's been straight out-- after school and on weekends--- getting their rockets dialed in.
Last week we test launched 13 times in 0-2 mph winds. Then on Saturday we went to the first qualifying launch weekend where there were 35 mph winds! Our first rocket ended up 60 feet up in a tree (we got it back) and all we did was test launch until 4:00 PM, when we made an ill-advised (and ill-fated) qualifying attempt. More of the same all week long, then back to the robot and rocket on Sunday. I think the kids needed and appreciated Easter Sunday 'off'! Class is starting; more later.
- - - - -
So, I guess we can expect them back when they ready to be here.
Regards,
Bruce Bates
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Involvement and committment can be best understood by looking at a plate of ham and eggs. The chicken was involved, but the pig was committed. ANON
For their entertainment (sorry, no Parallax content. I won't make a habit of this):
I was at a launch today where two of the Mad West TARC teams flew a couple of tests. They seem to know what they're doing - their flights were fine - but with fairly stiff wind, their scores were pretty mediocre (or would have been were they official launches).
The SLI team flew as well, also a test launch on a K motor with a dummy payload. The flight was successful: I didn't stick around to see what kind of altitude they got. I'm not sure if the payload is a secret - let's just say there will be sensors, and its an interesting project that doesn't duplicate anything I've seen before.
I flew a smaller rocket on an F and then a 3" diameter high power rocket to 2964 feet on a Cesaroni I285, using a Perfectflite HA-45 for deployment.
·· No, we were not abducted by aliens. And yes,·things got·very busy, VERY fast. We are back on-line (not that we ever left; we checked in periodically.) So where do I pick up? What a ride it's been....! WHEW!
First, we wrote, and passed, our Flight Readiness Review report and video teleconference with NASA and were deemed (as you can guess) "flight ready". That's a big hurdle crossed.
Next, after 25 test launches on high-power G80-7 engines (at $20 a launch) our Team America Rocketry Challenge (TARC) team qualified for the national finals... on our very last qualifying attempt and engine! Second hurdle crossed. We're officially in for the 2008 finals (PAUL - we nailed a score of 8.8. Flight time = 46.9; altitude = 755')
Then, per: NASA's mandate that I get qualified "Level 2" by the National Association of Rocketry (NAR) before we go to Marshall Space Flight Center (NEXT WEEK!), I had to design, build, and launch an (8 ft tall, 6" diameter, 15 lb) rocket and recover it intact. Of course I had to study for an pass a fairly rigorous written exam first (all while teaching and lead a Boy Scout troop.)
But enough of the non-robotic chat. The robot is DONE!· Andrew figured out that if we attach a piezzo buzzer to the p12 and p13 holes, it beeps in Morse code when the LED blinks! That was cool in and of itself. THEN we discovered that if we plug a second buzzer into one of the servomotor ports it buzzes continuosuly. This remedied the NASA requirement that our rocket and robot be equipped with a "locator device" (in this case a dual sonic locator.)
Next we wired the ASP (air sampling probe in the nose cone) exactly the same as the robot. It's function is simple to gather humidity and temperature data to compare against the robot's. Problem was, the ASP doesn't have servomotors and we only wanted it to take samples every second. Period. It kept running servomotor subroutine at 2:00 minutes (for 26 seconds).Then it took samples every 26 seconds (after the no-extant servomotors 'ran'.) We trouble-shot this problem... and actually figured it out ourselves! I'll post the debug solution later (but it simply involved deleting the 'OR' option after the 'RUN $01 minute' OR ... command. Now it collects samples every second for 120 minutes, 180 minutes, etc. We'll ground truth this next Sunday, our last meeting before we leave for Marshall Space Flight Center on April 23.
On Sunday a newspaper reporter came and interviewed the team for 1-1/2hours. She was enthralled by the team and their project and she promised us a front-page feature story later in the week. I made SURE to mention everyone's help on the forum by name, and I hope they include it! It's the Manchester (NH) Union-Leader and I'll post the link once the article appears.
OH! We ended up 'attaching' the robot's wires to the breadboard with clear·silicon sealant/'glue', and it works perfectly! The datalogger we anchored down with an electrical 'zip-tie', with a thin piece of G10 fiberglass underneath it to keep it from 'jiggling.' I'll post good quality, close-up·pictures of the wiring scheme and the final set-up. I'll also post the 'final' version of program for all to see and use.
So that's where we are today. Tomorrow morning at 7 AM we·make one of many fund-raising presentations to the rotary club, seeking funds to help the TARC team get to nationals (May 14-17.) What a strange and exciting journey it's been. But it's not over!
One of the kids' fathers bought me a book called '50 rocket project for the mad scientist' (or something like that.) In the back are about·a DOZEN parallax rocket project, and source code: accelerometers, ejection deployment timer system, dual axis roll sensor, etc. I'm in hog heaven (no offense to hogs intended, nor to those who believe in heaven.·It's a relative thing I suppose.)
I'd like to get framed team photos out to all who helped with (can you say, "Did") the programming with (er, 'for') us. That's you Tracy, and Brucem and Mike Greens. No promises it'll be tomorrow. But I certainly haven't forgotten you, nor your kindness. I will always remember how generous you were with sharing your expertise and time. I'm sure the kids learned a life-long lesson on kindness from you. More later.
Glad to see you're back on line actively. We'll all be here if you need us.
It's really great to hear all the good news! Our best to you and all the rocketeers.
Regards,
Bruce Bates
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Involvement and committment can be best understood by looking at a plate of ham and eggs. The chicken was involved, but the pig was committed. ANON
Post Edited (Bruce Bates) : 4/15/2008 4:23:09 AM GMT
Congratulations! Great flights - doing 25 test flights paid off, it seems.
As usual, all (three) of the Madison West teams here qualified - they were flying like mad a couple of weekends ago.
Did you get your L2, then? If so, congratulations on that as well. Building a six-incher for it? I flew mine on a four-inch diameter rocket on a J350, dual deployment.
I can't wait to hear how your contest flights go - I hope you'll have some video uploaded. It's going to be an interesting summer.
·· It's great to be back in touch and hear from everyone again. Life was empty without you! Good to hear from you, too.
Paul,
·· So all THREE Mad-West teams qualified for TARC? WOW! Congratulations! That's quite an accomplishment, even once.
Yes, I got my L2 cert on a clunky but functional 6" diameter behemoth of a rocket. Like yours, it was flown on a J350W-M and used dual deployment. Nice low and slow·flight profile, 58" 'drogue' at apogee (if you can call a 58" 'chute a drogue), and dual main 'chutes right on cue at 500' using a PerfectFlite altimeter. Touchdown was soft and gentle in the rain-dampened grass, with nary a scratch (now if I would just get around to upgrading this slooooooooooooow dial-up Internet service I could post some excellent photos of rocket, robot, and the team...!)
·· It's great to be back in touch and hear from everyone again. Life was empty without you! Good to hear from you, too.
Paul,
·· So all THREE Mad-West teams qualified for TARC? WOW! Congratulations! That's quite an accomplishment, even once.
Yes, I got my L2 cert on a clunky but functional 6" diameter behemoth of a rocket. Like yours, it was flown on a J350W-M and used dual deployment. Nice low and slow·flight profile, 58" 'drogue' at apogee (if you can call a 58" 'chute a drogue), and dual main 'chutes right on cue at 500' using a PerfectFlite altimeter. Touchdown was soft and gentle in the rain-dampened grass, with nary a scratch (now if I would just get around to upgrading this slooooooooooooow dial-up Internet service I could post some excellent photos of rocket, robot, and the team...!)
Regards,
Mark
Congratulations, and nice job. I'm glad to hear you used dual deployment - I'm kinda appalled that so many people L2 with a motor ejection kit. It hardly seems like a challenge. I used a Missileworks RRC2 (the older style) for mine, but I've got about a half dozen of the PerfectFlite MAWDs, and one of their HA-45s as well, and they've all worked every time I've used them.
Mad West has great mentors. If you don't know Brent and Pavel, introduce yourself when you're down there, and tell 'em you've been talking with Paul Smith. They're fun guys, though Pavel is a bit insane at times <grin>. Ask him about firefighting at Bong.
Paul
Comments
I've been trying to run his latest Vinculum code.
See http://forums.parallax.com/attachment.php?attachmentid=52629
The port connection assignments for the Vinculum are identical to those shown in the Parallax Demo code.
Tracy wrote the code for a BS2 Stamp, but the Board of Education I use to verify the code has a BS2p Stamp.
I made the obvious change in baud rate to enable the BS2p to communicate with the Vinculum. However,
I still found the Vinculum was not initializing reliably.
An external AC powered wall-wart was used to power the BOE.
Fixing the baud rate constant for use with a BS2p, Tracy's made code run a little better. Immediately after the code
was downloaded from the PC to the Stamp, the sync characters were recognized when the sync count reached
anywhere from 2 to 8.
Running from a cold start, however, it would fail every time, with the sync count reaching 9.
Another odd symptom running from a cold start was that the Vinculum red and green LEDs would begin flashing
and the first DEBUG statement in the code would NEVER EXECUTE. It was as if the Vinculum was somehow
interfering with the BS2p code and preventing it from running.
Very occasionally they would turn off after 15 to 30 seconds and Tracy's Vinculum initialization sequence
would complete before the sync count reached its limit of 9. Nearly all other times they would continue in the blinking state
for several minutes or until I manually turned off the power.
There are two LEDs on the Viniculm Datalogger assembly, a red on and a green one. When power is first applied to
that unit via the Stamp 5v regulator, they blink at about twice a second, alternating green and red for 3 cycles followed
by a short 1 second pause when the both turn off. Usually this sequence stops after 3-5 seconds with the red LED
on.
After the green-red sequence finally stops, Tracy's code manages to successfully send "E" and receive the correct
response and the red LED on the Vinculum stays on.
As an aid in debugging, I stripped out all the other code except for the Vinculum related initialization sections. To ensure
there weren't any baud related issues, I used the Basic Stamp conditional directive and changed the baud rate constant
from vBaud to T9600.
See the attached file(s). The BS2 code runs reliably. The BS2p code doesn't.
During a phone call to Parallax Tech Support today, I spoke with Chris Savage and he said that those blinking LEDs
meant that the Vinculum was trying to establish communication with the flash drive plugged into the Datalogger.
They just keep on flashing until it finds a flash drive attached.
Chris said that the only difference between the BOE and Basic Stamp states when code is being downloaded vs a cold
start is that the Vinculum has had enough time to establish the flash drive communications link prior to download, whereas
on a cold start, it has to first create that link before any transfers between the Stamp and Vinculum can occur.
Replacing the BS2p stamp with a BS2px made no difference with respect to the cold start hang up.
However, when I installed a BS2, the cold start issue disappeared.
Another difference observed when using a BS2 is that the first DEBUG statement appears almost immediately after
power on, once the code has been downloaded
Four questions to the forum:
1)What is different about using a BS2p or BS2px with a Vinculum instead of a BS2 that causes the Vinculum never being
able to establish communications with the flash drive from a cold start?
2) Is it a hardware issue, a code issue or both?
3) After waiting over a minute following a cold start using a BS2p, why does pressing the RESET button on the BOE cause
the flashing red and green LEDs on the Vinculum to quit cycling and enter the ready state?
4) Why does the Parallax demo code run with any flavor of Stamp (BS2, BS2p,BS2px), but my stripped down version of
Tracy's doesn't?
I might also add that all the symptoms described above occurred using two separate Memory Stick Dataloggers,
as well as a BS2p and BS2px Stamp. They weren't caused by a bad Datalogger.
phil
Post Edited (phil kenny) : 3/11/2008 1:57:15 AM GMT
·· So what kind of gizmo did you have in your rocket that printed out the graph that·you posted? It doesn't look like it''s from a Black Sky AltAcc (since the AltAcc doesn't read temperature.) Can you post a picture? How big was the rocket? Are you L2 or L3 cert?
I won't tell the Washington County 4H guys or the Madison West team a thing·(they always qualify for TARC anyway...!) Have they made their qualification launches yet? What are they getting for time and altitude? We·made our first test flights last Thursday and we were waaaaaaaaaaaay off on altitude (513', 535', 569'). We had to take off all of our 'add-on' weights and even then we were way under altitude. Our rocket is·obviously too heavy and so we had to re-build·a·new·booster (from cardboard, not phenolic.) We'll test launch again tomorrow at around 990 grams on a G80-7.
Regards,
Mark
·· Boy, that's a bundle of questions! I'm eager to find out what the problem ultimately is. For what it's worth (*and consider the inexperienced source),
recall that when we·had a similar problem·with the Datalogger initializing, we·made the·PAUSE after the initialization command longer. At first we set it
to 5 seconds; but ultimately we found out it could be set as low as 1 second. When Tracy posted the latest version of ROCKET_ROBOT, a similar thing
happened. I added a second "GOSUB init" command after the PAUSE (see comments several posts ago) and that fixed it. Why, I cannot say for sure.
I wish I had more to offer but I don't. As previously demonstrated, my 'expertise'···
Mark and the Rocketeers
If you look closely at my stripped down version of Tracy's code, you'll notice that
I changed the two PAUSE 512 statements at the very start to PAUSE 2000. Close
to the change you made. I think they helped.
phil
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com
·· Thanks for your post about the wiring scheme. We'll keep them in mind as we (finally, or finely) begin to wire the robot for the last time. And I'm eager to scoop some Goop and try it out (in a well ventilated arae as you suggested.
You asked several posts ago (and I thought Andrew had already posted an answer): "What does the debug screen display for "sync=? and disk=?" The numbers·are "sync=2" and "disk=1" virtually every time. It looks like the Datalogger is initializing on the first try, every time (knock of wood!) Bravo!
So what do you make of Phil's conundrum? Strange that a different stamp runs the same program differently, isn't it...?
Regards,
Mark
I just tried a BS2pe and it has the same cold start issue. The current draw for the BS2pe
is speced at 15 ma and the BS2 is 3 ma. Hard to see that 12 ma more load would
make the difference.
phil
The data were collected using a Vernier Labpro (www.vernier.com), with a 25G accelerometer and a fast-response temperature probe, whose tip came out through a small hole in a hatch. The rocket is 5.5" diameter, about 8' 4" tall. I'm NAR L2.
The Labpro was pretty easy to use, but large, heavy, and difficult to mount. I probably won't ever fly it again, and instead build something using a Stamp module. I've got an adapter so I can use the Vernier sensors with a Basic Stamp. The big issue is one you've already faced: interfacing a real-time clock. The acceleration data aren't worth much if I don't have an accurate time scale. I expect to spend a lot of time scouring your thread to pick out the great advice and help you're getting here.
I don't know what the current TARC flights·look like here - I do know that those folks flew some qualification flights at Bong two weeks ago, when I was flying a small rocket with one of my students for my RDF grant.
The mystery behind the peculiar operation of the Datalogger when using anything other than a BS2 Stamp with Tracy's latest code has been cleared up.
Today I powered just the Datalogger and thumb drive from a completely separate regulated 5v supply. Now, the Viniculum
is able to complete its internal handshaking with the thumb drive and Tracy's program runs fine, with minor adjustments to
the initial PAUSE statements, independent of the Stamp model.
Using a Fluke 87 DVM to measure the 5v current pulled by the Viniculum and thumb drive showed 155 - 160 ma using a 'real'
256 Mb thumb drive. I also measured the current using a USB card reader with a thumb drive form factor plus a 32Mb SD card.
That combination pulled 48 - 55 ma. These were just average readings and not peak.
Even pulling 100 ma less current, the USB card reader with an SD card wouldn't let the Viniculum finish its internal sequence of
establishing communication with the flash drive.
All of the peculiar Stamp behavior that I saw when running Tracy's code can be attributed to short term overloading of the BOE onboard power supply regulator.
However, I still can't explain why the Parallax Demo code doesn't exhibit the same cold start problems when using a BS2p,
BS2px, or BS2pe.
I wasn't able to find any info regarding current drain of the Memory Stick Datalogger on the Parallax website. Perhaps that's
because it is heavily dependent on the flash drive attached to it.
phil
Post Edited (phil kenny) : 3/12/2008 1:01:22 AM GMT
·· I'm glad to hear that you solved the problem! So you were essentially drawing too much power (amperage?) which preveted the program from completing the·initialization sequence? How does that work? Why does it slow down initialization?
The "Junior Rockteers" (the younger rocketry team) did the second round of their test launching today and it was like "Wide World of Sports" (I'm aging myself now.) It was "the agonoy of defeat", followed by the "thrill of victory." On their second test launch they forgot to attach the payload bay (a cargo module with their egg 'astronauts' and altimeter inside) and it went straight in from 800' feet up! Luckily it went into deep snow! Deep. Real deep. The eggs were very scrambled and we destroyed an altimeter. If it was a qualifying·flight (to qualify for nationals)·they would have been DQ'd because their 'astronauts' didn's survive. Luckily their sturdy payload bay survived.
On the very next flight their rocket came within one foot of the target altitude of 750', and 1.2 seconds with the target flight time of 45 seconds (749' in 46.2 seconds.) If this had been a qualifying flight they would be on their way to nationals. A score like that at nationals could well have won nationals! It certainly would have put them in the top 10 for scholarship money, and the opportunity to make a proposal to NASA for yet another SLI project like one you've been helping with.
Glad to hear that you'reup and running!
Mark
Good call. I see from the Parallax data sheet (memoryStickDataloggerV1.0.pdf ) that it draws 25 mA (running) and 2mA (standby). So the rest of the ~160 mA would be due to the USB memory device. I can understand a surge of current for a few milliseconds when it writes data. I have no idea what it is doing when it first starts up. Maybe it fires up all its subsystems and those draw the current. It is indeed perplexing that it behaves differently with the different Stamps and with different demo code. Maybe a fat filter capacitor on the power supply would eliminate the glitch. Say 1000 µF, preferably installed close to the datalogger power connections.
Mark and rocketeers,
Is it still working on your system? No glitches? All the data being collected looks OK?
I'm attaching an uncluttered copy of the planning and layout template for the BOE breadboard area. I made this diagram for the Earth Measurements text.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com
··· "Daedalus"continues to sing like Pavarotti (and·babble like a child in Morse code!) No glitches at all (knock on wood), and the program initializes·at 'sync=2, disk=1' every time.
Thanks for the breadboard schematic. It will be really helpful, for posterity,·in documenting where the wires go, and for anyone else who is interested. We'll post it as soon as we get it sketched. Tyler and Andrew should already be working on it for Sunday.
I was glad to hear that Phil's problem seems resolved. Even though I only understood about half of the conversation, it was interesting to follow.
You'll have to read the post below about our very successful test launch of our 'other teams' rocket this afternoon...! Thanks and,
Regards,
Mark and the Rocketeers
The 5v regulator on the BOE can only supply a limited current. When that limit is reached, and the
load tries to pull more, the output voltage begins to sag or drop.
The Basic Stamp module has an undervoltage detector (UV) circuit that is set to trip somewhere
around 4 volts. When that happens, the UV circuit generates a system RESET. This in turn causes
the Stamp to go into a low power state and all output pins are left floating.
The RESET state is the same one generated by the black RESET button on the BOE. However, this
RESET signal isn't seen by the Vinculum.
With the Stamp now asking for less current, the 5v regulator may let its output voltage rise. If it climbs
high enough to cross the UV threshold, the Stamp comes out of RESET, pulls more current, and this
cycle can start all over.
All it takes is a momentary spike in the load current for this RESET cycle to occur.
Using a separate power supply for the Vinculum prevented the Stamp from being reset by any extra
load pulled by the Datalogger.
phil
My project is a college-level separate grant with the Wisconsin Space Grant Consortium, which is a state level funding arm of NASA. No qualifying involved, other than writing a convincing proposal.
With the experience you've gotten in the last few months, you're an expert. No doubt about it. You're going to give Pavel and Brent (Mad West) a run for their money.
·· Thanks for the really clear explanation. That makes compete sense to me and now I understand what·was happening.·Basically it was just RESET-ing itself because there was a spike in power (voltage, or amperage?), yes?·So how's it running now?
.-.·· .··· ...·· .··· -···· !!
Mark
·· The "Jr. Rocketeers" (TARC team) test launched again today in exactly the·same weather conditions, at·the exact same launch weight. THIS time their rocket went:
802' - 47 seconds
790' - 46 seconds
We think a loose fin on the 'perfect' flight the other day (749' in 46.1 seconds) caused the rocket not to achieve it's maximum altitude.
Recalculate, reweigh, reload, relaunch... our qualification flights are 9 days away!
Mark and the Rocketeers,
'Daedalus' ( ....·· ..·)··and Bandit the Rocket Dog
Here is a diagram of the breadboard before any new arrangement has been made. Do you want me to go ahead and experiment with my board to find the best arrangement before our meeting on Sunday?
·· Thanks for posting to the forum!· I'd leave it like it is for now, especially if it's wired like ours, and working. We'll 'tweak' it on Sunday. Has your RTC and cyrstal arrived yet? Did you get·them ordered like you said? We'll solder and attach them on Sunday like we did with Andrew's last Sunday. Thanks again for posting to the forum.
Great diagram!····
Mr. Kibler
Mr. Kibler again. Have you included your diagram, with a label and an explanation,·with your section of the FRR report? It would go really well in the report. Please reply to the forum.
Mr. Kibler
PAR
Funny you should ask. On 3/25/2008 I send Mark Kibler a message, asking the same question. I asked Mark if they were all on vacation. Here is the response I got then:
- - - - -
No vacation here (though the idea IS appealing!) The SLI team had last Sunday off for Easter as we regroup and recharge our batteries before we do our video teleconference with NASA next Monday.
"Meanwhile, back at the ranch" the two TARC (versus SLI) rocketry teams have been in a flurry of activity making test launches and getting ready for qualifying attempts. It's been straight out-- after school and on weekends--- getting their rockets dialed in.
Last week we test launched 13 times in 0-2 mph winds. Then on Saturday we went to the first qualifying launch weekend where there were 35 mph winds! Our first rocket ended up 60 feet up in a tree (we got it back) and all we did was test launch until 4:00 PM, when we made an ill-advised (and ill-fated) qualifying attempt. More of the same all week long, then back to the robot and rocket on Sunday. I think the kids needed and appreciated Easter Sunday 'off'! Class is starting; more later.
- - - - -
So, I guess we can expect them back when they ready to be here.
Regards,
Bruce Bates
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Involvement and committment can be best understood by looking at a plate of ham and eggs. The chicken was involved, but the pig was committed. ANON
I was at a launch today where two of the Mad West TARC teams flew a couple of tests. They seem to know what they're doing - their flights were fine - but with fairly stiff wind, their scores were pretty mediocre (or would have been were they official launches).
The SLI team flew as well, also a test launch on a K motor with a dummy payload. The flight was successful: I didn't stick around to see what kind of altitude they got. I'm not sure if the payload is a secret - let's just say there will be sensors, and its an interesting project that doesn't duplicate anything I've seen before.
I flew a smaller rocket on an F and then a 3" diameter high power rocket to 2964 feet on a Cesaroni I285, using a Perfectflite HA-45 for deployment.
Paul
·· No, we were not abducted by aliens. And yes,·things got·very busy, VERY fast. We are back on-line (not that we ever left; we checked in periodically.) So where do I pick up? What a ride it's been....! WHEW!
First, we wrote, and passed, our Flight Readiness Review report and video teleconference with NASA and were deemed (as you can guess) "flight ready". That's a big hurdle crossed.
Next, after 25 test launches on high-power G80-7 engines (at $20 a launch) our Team America Rocketry Challenge (TARC) team qualified for the national finals... on our very last qualifying attempt and engine! Second hurdle crossed. We're officially in for the 2008 finals (PAUL - we nailed a score of 8.8. Flight time = 46.9; altitude = 755')
Then, per: NASA's mandate that I get qualified "Level 2" by the National Association of Rocketry (NAR) before we go to Marshall Space Flight Center (NEXT WEEK!), I had to design, build, and launch an (8 ft tall, 6" diameter, 15 lb) rocket and recover it intact. Of course I had to study for an pass a fairly rigorous written exam first (all while teaching and lead a Boy Scout troop.)
But enough of the non-robotic chat. The robot is DONE!·
Next we wired the ASP (air sampling probe in the nose cone) exactly the same as the robot. It's function is simple to gather humidity and temperature data to compare against the robot's. Problem was, the ASP doesn't have servomotors and we only wanted it to take samples every second. Period. It kept running servomotor subroutine at 2:00 minutes (for 26 seconds).Then it took samples every 26 seconds (after the no-extant servomotors 'ran'.) We trouble-shot this problem... and actually figured it out ourselves! I'll post the debug solution later (but it simply involved deleting the 'OR' option after the 'RUN $01 minute' OR ... command. Now it collects samples every second for 120 minutes, 180 minutes, etc. We'll ground truth this next Sunday, our last meeting before we leave for Marshall Space Flight Center on April 23.
On Sunday a newspaper reporter came and interviewed the team for 1-1/2hours. She was enthralled by the team and their project and she promised us a front-page feature story later in the week. I made SURE to mention everyone's help on the forum by name, and I hope they include it! It's the Manchester (NH) Union-Leader and I'll post the link once the article appears.
OH! We ended up 'attaching' the robot's wires to the breadboard with clear·silicon sealant/'glue', and it works perfectly! The datalogger we anchored down with an electrical 'zip-tie', with a thin piece of G10 fiberglass underneath it to keep it from 'jiggling.' I'll post good quality, close-up·pictures of the wiring scheme and the final set-up. I'll also post the 'final' version of program for all to see and use.
So that's where we are today. Tomorrow morning at 7 AM we·make one of many fund-raising presentations to the rotary club, seeking funds to help the TARC team get to nationals (May 14-17.) What a strange and exciting journey it's been. But it's not over!
One of the kids' fathers bought me a book called '50 rocket project for the mad scientist' (or something like that.) In the back are about·a DOZEN parallax rocket project, and source code: accelerometers, ejection deployment timer system, dual axis roll sensor, etc. I'm in hog heaven (no offense to hogs intended, nor to those who believe in heaven.·It's a relative thing I suppose.)
I'd like to get framed team photos out to all who helped with (can you say, "Did") the programming with (er, 'for') us. That's you Tracy, and Brucem and Mike Greens. No promises it'll be tomorrow. But I certainly haven't forgotten you, nor your kindness. I will always remember how generous you were with sharing your expertise and time. I'm sure the kids learned a life-long lesson on kindness from you. More later.
As ever, yours in robotic bliss,
Mark and the Rocketeers!·
·
Glad to see you're back on line actively. We'll all be here if you need us.
It's really great to hear all the good news! Our best to you and all the rocketeers.
Regards,
Bruce Bates
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Involvement and committment can be best understood by looking at a plate of ham and eggs. The chicken was involved, but the pig was committed. ANON
Post Edited (Bruce Bates) : 4/15/2008 4:23:09 AM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com
As usual, all (three) of the Madison West teams here qualified - they were flying like mad a couple of weekends ago.
Did you get your L2, then? If so, congratulations on that as well. Building a six-incher for it? I flew mine on a four-inch diameter rocket on a J350, dual deployment.
I can't wait to hear how your contest flights go - I hope you'll have some video uploaded. It's going to be an interesting summer.
·· It's great to be back in touch and hear from everyone again. Life was empty without you! Good to hear from you, too.
Paul,
·· So all THREE Mad-West teams qualified for TARC? WOW! Congratulations! That's quite an accomplishment, even once.
Yes, I got my L2 cert on a clunky but functional 6" diameter behemoth of a rocket. Like yours, it was flown on a J350W-M and used dual deployment. Nice low and slow·flight profile, 58" 'drogue' at apogee (if you can call a 58" 'chute a drogue), and dual main 'chutes right on cue at 500' using a PerfectFlite altimeter. Touchdown was soft and gentle in the rain-dampened grass, with nary a scratch (now if I would just get around to upgrading this slooooooooooooow dial-up Internet service I could post some excellent photos of rocket, robot, and the team...!)
Regards,
Mark
Mad West has great mentors. If you don't know Brent and Pavel, introduce yourself when you're down there, and tell 'em you've been talking with Paul Smith. They're fun guys, though Pavel is a bit insane at times <grin>. Ask him about firefighting at Bong.
Paul