TREX: HotFilm Messages: 40 Entries..

The power inverter lent by Oncley to power Ripple in the hot-film/stn0 electronics box was unable to operate due to low battery power of unknown origin. A brief effort to ascertain the genesis of the low power was made approximately 30 minutes prior to the end of T-REX (00UTC 1 May). We soon decided that the benefit was limited given the end of the project (we had originally intended to run until the Central tower was shut down).

The hot-films were gathered by Poulos at ~ 1900 UTC 30 Apr.

Thanks for all the hard work and the gathering of some 300 hours of data.

Greg

24Apr06, 17:40PDT, Central
Began hotfilm test..

End. 25 Apr 2006 ~230pm local.

Dual backup of data completed 26 Apr 2006

21Apr06
GPS coords under the 1m sonic
36.80169
118.16019
3776 

20Apr06,

17:20PDT began new hotfilm run with 'fresh' pocketec disk.

ntp on daisy was mis-configured, and as a result its
clock was  591.58 seconds ahead.

This was corrected at 09:19 this morning, and the
system rebooted, so data recorded after this time are OK.


I believe it has been mis-configured since it was
put on the 192.168.15 network early in the project.

Started hotfilm sampling at 19:30 PDT, Apr 16.


This run lasted until 20Apr06 ~12:30 when it was taken down for awhile to
allow the full pocketec to be off-loaded at the base

ripple (hotfilm Dell notebook) died this afternoon

It's internal battery was discharged. The inverter that provides
power was giving a little beep every 5 seconds or so and the green
led on the inverter would brighten and then dim.

Looks like the voltage at the 5m tower is not high enough to drive
the inverter.  The two batteries that were at the tower were moved
to the 30m tower this morning.  I moved one back over to the 5m
tower and then the inverter was happy.  However, then the main
power breaker on the enclosure would flip. ripple's internal
battery was discharged, so it was pulling it's highest load.

Installed a power 'Y' so that the power to the inverter is coming
straight from the battery, rather then via the aux plug on the
enclosure.   Seems to work. It is running right now.

Another hot-film data system test was started at ~ 1420pm (without hot-films themselves) during the latest Central visit.

Hotfilm data system is up - but no probes are installed.

The prometheus is running a modified version 5.91 of the DSCUD 
driver and library.

The old version was 5.9.  I doubt 5.91 has any bug fixes - most
likely it just includes support for a new board from Diamond.

I've made a slight modification to the 5.91 driver - declaring
as volatile two variables that are shared between the 
interrupt routine and the user read method.  The A2D stayed
up all last night and today but we didn't always have sonics
connected, and never had motors connected.

14Apr06 6:45PDT  Ripple halted
Gordon had to use a 4gb usbstick instead of the pocketec for this test.  Since the ingest was down due to full disk, I halted ripple to save power since it's cloudy and the batteries are down to 12.15

Begin 10Apr06,19:45 PDT Started hotfilm.
End   11Apr06 12:15 PDT

The fuel cell running ripple is looking good: 1200psi remaining, and the 2 batteries are at 13.6 on the f.c. before we started.  At the end, the batteries were still 13.6 and the f.c. 1100psi.
The main system batteries were at 12.55, and thus ~11.55 going into daisy at the end of the long cable when we began.

4/8/06 ~17:00PDT Begin
4/9/06 ~11:30PDT End of run


Fuel Cell began at 1400psi for ripple supply.
Fuel Cell end   at 1250psi, Vbatt for 2 running ripple=13.5 on F.C.
Central power system for Daisy, etc. was looking good also after 3 sunny days of solar charging.

Winds picking up today.
Begin ~22:30Z

No hotfilms installed and CTA's not running but otherwise daisy/motors/ripple are up so the Gordon can install some new scripts for improved data recovery over the periodic reboot method.

Note: Fuel cell was still running OK.  By that time the batteries were at 12.7 on the f.c. charging.   Swapped in 2 fully charged batteries.

Winds picking up today.
Stop Test: Apr 7th, 097: 18:35Z
Started: Apr 5th, 

Note: Fuel Cell was barely running when I arrived.  The regulator showed 0-psi on both gauges.  The battery voltages were at 12.2 at that point.  When I checked the valve it was barely cracked open.  When I openned it all the way the pressure went to 1600psi and etc.  At that point I had shutdown everything.

The hotfilm is up and running at ~18:00.

Hot Film:
	The hot film system consists of 4 major components: (1) power source, 
(2) sensors (hot film and sonics), (3) data collection (Prometheus and Ripple), 
and (4) the communication system. The system can run without the communication 
link going however you will not be able to monitor on a real time basis.

Instructions:
SETUP: 
     This assumes that power is off on all systems and all sensor cables are 
plugged in.

1: Power system
     The power system is designed to be running at all times. The power system 
consists of 2 parts; the batteries and the fuel cell. The fuel cell keeps the 
batteries charged.  Verify that the fuel cell system is running. If it is not 
operating then check the hydrogen tank. The tank pressure should be at least 
300 psi. If it is below 300 psi then it should be changed.  If you do change 
the tank remember what the setting of the regulator is  between 10 and 20 psi 
on the second stage. After the tank is attached then you can start the fuel 
cell. There is an on/off switch on the fuel cell located below the hydrogen 
line. Turn the switch on and the fuel cell should start. It will make a noise 
for about 15 seconds then start running.  If it stops you may need to replace 
the 9 volt battery located in the fuel cell. This requires removing the fuel 
cell from the enclosure. The battery is located on the side of the cell.
IF YOU ARE NOT SURE HOW TO OPERATE THE FUEL CELL THEN TALK TO AN EXPERIENCE 
PERSON.

2:  Startup Ripple
     Ripple is located inside the electronics box attached to the top of the 
lid. Open the box and plug the power cable into one of the power ports on the 
12 volt battery splitter cable. Attach the USB drive if it is not connected. 
This is necessary in order for the aster processes to get going. Undo the 
bungee cord so you can open the DELL. BE CAREFUL UNHOOKING THE BUNGEE CORD. 
Power up the DELL and login when requested. 
Do a check_aster command to verify that the aster process is running. 
Also do a ping aster to verify that the wireless communication link is up.

3: motor controllers
     The motor control systems should be plugged into serial ports 9 (1m), 
10 (2m), and 11 (3m). With the lid open verify that the internal power cord to 
the Prometheus stack is unplugged. Now turn the main power switch on. This 
should start the 3 motor systems. The zero position sequence starts with the 
motor shaft moving CW to about 135 degrees. It stops for a second then goes 
CCW to the zero position. Zero is parallel to the sonic boom direction. You 
want to make sure all three probe support arms follow the sequence above. If 
something does not go right then turn off the main power switch immediately.
NOTE: At this time NO probes should be mounted.

4: Tri-axial probes
	The tri-axial probes can be installed. Note the reference number on 
each probe box. The probe order is: #1 – 1m, #2 – 2m, #4 – 3m. When 
inserting the probes note the notch cut out at the back end of the probe. 
This should be aligned with a corresponding piece on the probe support. 
Be extra careful handling the probes!  Put the yellow protection cap, 
on the end of the support arms into the probe box.
	
5: Prometheus
	The Prometheus can now be plugged in. You can run minicom on ripple 
if you want to watch it boot up. After a minute or two you should see a cockpit 
come up on ripple. Verify everything is running as expected. Also verify data 
is being written to the USB drive. This can be done by either watching for a 
green light flash or do a DF command on ripple and verify that it is increasing.
At this time the hot film data on the analog channels should be around zero 
volts.

6: CTA power
	Now plug in the CTA power cable. It is plugged into on of the aux ports
on electronics box.  Move it over to the other aux port. You should be getting 
A/D on the cockpit.

SHUTDOWN:
	Shutdown is pretty much in the reverse order as startup

1: CTA power
	Unplug the CTA power cord and move it back to the aux port.

2: Prometheus
	Via minicom or ssh to daisy, do a halt command to stop the Prometheus.

3: Tri-axial probes
	Remove the hot film probes. Do them one at a time to ensure they get 
put back in their correct box. 
Remember to put back on the yellow protection cap.

4: Ripple
	Ripple can now be shutdown. Once it is shutdown remove the USB drive so
that the data can be downloaded back at the base. 
Remember to unplug the power cable and put it back inside the box.

Other items of interest:
Sonics: the sonics are routed to 2 locations: Prometheus and central. 
Only the transmit signal is routed to the Prometheus. If you ever have to 
disconnect the sonic cables from the Prometheus box make sure you note how 
they are connected.
Sonic 1m – port 5
Sonic 2m – port 8
Sonic 3m – port 7

Ripple: Ripple may be running. Tests have shown that the fuel cell is apply to 
keep ripple up all the time if its internal battery is fully charged. However, 
if the USB drive is not attached the aster processes may not be running!

The hot film system was up and running around 16:00 yesterday. The fuel cell
system was able to keep ripple going since it was powered up for Gordon on
Friday.
The deadband was set to 20 degrees.

The hotfilm was shutdown at about 10:45am local. The sonics were moved over
to central nids.
The 35 watt fuel cell kept things going and the batteries still are up
around 12;5 volts.

The hot film was offically back on the air at 16:35.
During the process the 1m, 2m, and 3m sonics were removed from nids central
and move back to daisy. An option to eliminate this cable changing is to
build special Y cables to route the sonic xmt line into daisy. Maybe in the
future.

The hot film system was left up overnight to collect sonics data.
Unfortunately the batteries could not keep things up and the system went down.
Earlier this morning I changed the batteries and brought the system back up.
I re-routed power for daisy to come from the nids power source. The fuel cell
system will now provide power for ripple only.

The sytem was taking down once we got the sonics interfaced to nids (refer to
messages 117 and 118). We will see if the fuel cell keeps the batteries
charged at hot film.

The hotfilm site was brought back on the air collecting sonci data only.
This was done vias channel_config. There are now wo channel_configs in
OPS. "channel_config.sonci" is used for only sonic data while
"channel_config.hf" is the hotfilms added. You have to copy either of
these files to "channel_config" before starting things up.

The battery voltage levels were down around 11.75 volts. The hydrogen
tank was at ~1800 psi. This will be a good test to see if the fuel cell
can recharge the batteries wihtout the hot film stuff going. I also
disconnected power to the CTA moduels.

The hot film ran through out the night until ~10:30 this morning. The data
are bieng downloaded to exthd7 in the base. Once the pocketec drive is
cleared off it will be takne back out to central where the system will be
turned on to collect sonic data only.
NOTE: the battery level was down to 12 volts. The fuel cell was running
fine so this down time will be good for the batteries.

Deployed all three hot film probes this afternoon for a night run.

Data started about 1500 pst.

Mods have been completed on the motor controller circuit. All 3 units are
now deployed at central. The plan is to do a run tonight with breakable
wooden sticks in place of the probes. If the motor controllers go beyond
the limit switches the wooden probes should break.
If things run goo over night then we are back in business to collect hotfilm
data.

Work was spent on the motor controller systems. SInce one of the optic units
was not acting correctly I decided to re-wire it to a different format. There
is now a pull down resistor on the emitter side of the detector. The
result was very positive. The signal levels pull all the way to 0 volts as
wanted. I will need to re-write the code because the trigger logic is now
reversed; ie, it triggers on a falling edge verses a rising edge. I should
get this done tonight so that we can deploy them tomorrow.

The hotfilm system was shut down yesterday. I brought the motor controller
systems back to the base. We are having a problem with the coupler between
the motor shaft and the vertical probe support shaft. Also problems with the
coupler and the optics units.

Based on the archive files on ripple, diasy shut down around 3:45 this morning.
I assume that the reset program "ndaqrestart" used in the crontab may have a
bug. I will talk to Gordon.

I plan to go out there in teh next 1/2 hour to shut the system down and bring
back the USB drive.

All shafts on probe position system have loosened up. I believe the problem
is the tolerences between the optics and the shaft coupler along with the pin
used to trigger the optic. I have seen two problems:
(1) The shaft coupler piece has had contact with an optic detector causing
the detector to move out of position.
(2) The roll pin has made contact with the optic detector as it passes through.
The contact point has been at the bottom of the optic detector.

Possible solution:
(1) reduce the diameter of the shaft coupler towiden the space between the
coupler and the optics.
(2) change the roll pin to a flat surface piece. This would provide more
clearence between the top and bottom of the optic piece.
 

Installed hot film probes and began running at ~1712 pst.

Had problems with orientation shafts slipping, so turned off orientation motors
and oriented probes ~ west.

1 m probe had problem with y bnc connector immediately after probe.  Steve
jiggered with it and got it connected for now. I believ this is the connector
that John had found to be bad.

The batteries were changed. The old olds were done to ~12 volts.The fuel
cell should be able to recharge the batteries during the daytime when
the system is off. 

Probe 1 at 1 m
Probe 2 at 2 m
Probe 4 at 3 m

   I believe the problem with ripple has been found. It may have been in the USB
port connection. The PCMCIA USB card is a USB2 connection while the USB
connection on the back of ripple is a USB1. I changed to the ripple USB at
about 11:30 this morning and it has been running fine every since.

  Prometheus is still having problems. It only wants to stay up for a few
minutes. The plan now is to install a forced reboot on Prometheus every xx
minutes where xx will be determined based on how it is running.

   The fuelcell is working like it was meant to be out there. It has used about
100 PSI over the last 24 hours.

During the night ripple shut down so no data was collected. The 35 watt
fuel cell kept runnin gok and the batteries are charged. I went out this
morning to restart ripple. I also remounted the 1m motor system. I had to
remove the 2m motor system becasue the motor shaft coupler screws were loose.

During the day the hotfilm was left running at central. The A/D did not stay
up all the time. Back at the base I got the 35 watt fuel cell running. It 
was taken out to central to serve as the main power source for the hotfim
stuff. Tis included ripple which is mounted inside the Prometheus box.
In addition Gordon made some changes so that the USB harddrive is now
interfaced to ripple instead of the Prometheus. The system was back on the
air at about 16:40 local time.

The system stayed up during the night. Based on the local storage file sizes
the A/D stayed up for about 7 hours. Ripple stayed up on an external battery
that is not on the charging grid. I will go out there later this morning to
check on the ripple battery and restart the A/D system.

The hotfilm was moved out to central this afternoon. There were some initial
crashes on the A/D. After some playing around it seemed to stay up. The 3m
sonic was not responding correctly. Who knows if this may have been having
some effect. It was disconnected. Ripple is running off an external battery
via the car charging unit that Steve Oncley had. There are no hotfilm probes
deployed. Also the motor controller at the 1m height was acting up. It was
not seeing the zero position switch. It appears to be a similar problem we
were having back in Boulder. What is real frustrating is that unit was in the
base during the tests over the last few days and it worked without any
problems.

The hotfilm stayed up last night. It was configured in the base with ripple
tied directly to the prometheus via a swapper ethernet cable. The plan is
to move it out to the site later today with ripple housed inside the
Prometheus box. The external hard drive on ripple will not be used since
the local USB drive files look good. We may try to get ripple hooked up to the
wireless link so we can monitor things back at the base.

The hot film ran for about 4 hours in the base then the A/D died. I will restart daisy for another run.

Due to the discovery of the daisychain on the sonics, any data looked at since this time should be done with great care. In fact any sonic data representing 2m and 3m should be ignored.

Steve S and Tom removed the hot film data system and 1 m sonic from central
site this afternoon to work on it in lab.

Steve discovered that the 2 and 3 m sonics were never cabled within the data 
system.  The 3-sonic replicator cable was still transmitting the 1 m sonic 
data to all three channels.It was also discovered that serial channel 201 was
not responding to a sonic input. So a change was made in channels.txt on daisy
(/home/isff). Channel 201 was moved from /dev/ttyS5 to /dev/ttyS7

Later that evening:
  Steve set the Hotfilm system up in the trailer. It has been running for
~1hour with one sonic and 9 analog at 2000 hz. It is now runing with 3
sonics and 9 analog. It will be left up over night to see what happens.

As yet another test, I've just changed the config to enable sampling of only
1 triple hotfilm probe at 1000 and all 3 sonics.  The usb disk didn't come up.
I had tried to set the sample rate to 500 on all hotfilms, but the prometheus
went into a continuous ndaq restarting cycle -- perhaps 500 sps is not a valid
rate?

I had to start archiving manually, which is now being done to ripple's disk.
I figured that we might as well be acquiring data from 3 sonics, since they
don't significantly add to the data rate.

P.S. Even this test was a failure!  It ran for only 20 minutes and died.  
Data rates in the file were ok (no dropouts -- frequency as expected).

cockpit shows that the A/D process on the hotfilm system died sometime between
John and I leaving at about 7 last night and us arriving about 8:45 this 
morning.  I'll leave it going, since the sonics (only the lowest 2 are enabled)
are still running.

Spent much of today at central (John/Greg/ sometimes Steve&Brad) installing
the hotfilm stuff.  Left it with middle (nominally 2m) wire installed, but
motors not working at about 3:10.

Also checked tensions -- somewhat low.  We need to retension sometime.