SHEBA: Thermal_Generators Messages: 10 Entries..

TEG: 
Propane Maintenance schedule
			Sled-1		Sled-2		Sled-3		Sled-4
			------------	--------	-----------	-------
Fill Dates		10/11, 2 ea.	10/14, 2ea.	10/12, 2 ea.	10/18,(#3,#5)
Next Fill drop-dead	31-Oct		3-Nov		1-Nov		7-Nov


Factory Ratings:
    VOC, element	3.99		3.90		3.94		3.87
    Gauge Pres. kPa	58		63		58		64

Current Settings:
	10/10/97	58/3.56		63/3.83		58/3.78		63/3.62
	10/14				64
	10/15		57				56

				Sled-1	Sled-2	Sled-3	Sled-4
DC Supply voltage (VOC)		15.0	14.5	14.5	15.0

Note: Increasing the DC supply voltage increased the power/current provided by
	the TEGS to the PAMs significantly over the factory settings of 14.1

Pressure Setting Note:
	John Hutchinson from the Global factory in Calgary said that setting 
	the pressures to factory levels seemed to be required for him at sea
	level despite what the manual says about reducing it by about 9kPa.
	He said that shouldn't damage the elements.
	Obviously periodic checking of the open circuit voltage to be less
	than the max rating of 4.1 is necessary.
 
Air/Fuel Mizture Note:
	The factory air/fuel mixture setting starting point is by
	turning full closed in the CCW direction, and then opening
	back up by 4.5 turns CW.

	On 10/10, all four TEGs were set at 1/4 turn CCW from what the
	factory setting was, making them slightly more fuel rich.

Looking at baro data for sleds 1,2,3,4
it appears that the insulation for the sleds is too good.
This morning the temps:

sled	trh	baro/electronics
  1	-17	50	full config, deployed
  2	-16	31	checkout config., no sonic 
  3	-17	50	full config., deployed
  4	n/a	25.4	checkout config, trh, 4-comp, no sonic
			This station had vent insulation removed.

MUST DO:
All three sleds 1,2,3 need to have their vent insulation removed.

The barometer temp rests on the electronics box 'bottom' plate
directly over the propane tanks.  It is probably slightly elevated
in temp. from the overall box because of the EVE heat generated.

  The TEG at Baltimore went down yesterday. At this time it is
unknown why it shut down. The battery voltage is dropping, approximately
1 volt in 48 hours. According to logbook entry 33, the station had
a full set of propane bottles on the 12th. We may be looking at a
maximum run of 12 days on propane. Station 2, Cleveland, should be monitored
closely since it is approaching this time window.

  It appears that the auotmatic sparker had something
to do with the spiking we were getting on the TEG/Battery
voltage/current plots. It maybe the sparker was inducing
noise on the dc signals that were sampled.

  Station 4 had TEG problems. It would not start period.
Even with the torch approach it would not work. I pulled
it out of the box and put it in a hut to warm up. After
supper, Jeff and I went back out to try to start it. Upon
inspection, we found a oily paste substance in the line from
the pressure regular/starter to the orifice to the burner.
The stuff was also in the air/fuel mixture space. After cleaning
it up and clearing out the orifice hole, we rebuilt it, put it
back in the box, nad started it up. It started right away. We
let it run for awhile to settle down. The pressure gauge was
jumping between 60 and 65 Kpa. The voltage output was around
3.5 volts. We decided to let it run the night and try to tune
it tomorrow.
  The size of the orifice hole was hard to determine. If fact
the jumping in the pressure may have something to do with
how much of the hole was cleaned out. I sent an email to
John to talk to TEG about the orifice, the oily stuff, and to 
get more orifices for backup. Also the butterfly valve from the
main line to the TEG was leaking when it was turned. John will
send up spares.
 

  I spent the morning working on adjusting the output
of the TEG at station 4. It seems to be running fine.
It will be monitored over the next few days.

On satureday and into sunday we have had a temperature problem at Florida 
evebox.
At 1300AST I removed the rad array reducing the electrical load by ~8.6 W
At 1900AST it began to snow
At 2200AST it was noted that the Tevebox had started to climb and was 
approaching 57 degrees.
I went to Florida, reconnected a 9 W load (Clevelands ventilator and heater,
and also opened the TEG box. I shoveled free the sled and cleaned off the flue.
The box temp dropped to -8.8 and then began to climb again. By 0500AST sunday
it levelled out at 45 degrees.

These are a variety of notes taken during the last couple of days
plopped here for safekeeping....

4/12 2:50Z Cleveland TEG air adjusted
	Turned air shutter down (CCW) 3/4 turn
	using the 'blue-flame' then back-off optimization method recommended
	by Global.  Ie. teg appears to have been running a bit air-rich.
	Disconnected PV to watch results.
	Very odd...both iload and icharge seem to have risen by an equivalent
	amount: perhaps .1 Amp.  After looking at Fla charge/load currents
	on 4/12 ~18Z I don't have complete faith that the readings are totally
	accurate.  Suspect that the charging did improve a bit and feedback
	into load circuit monitor.

	It occured to me however, that if the deposits Jeff has seen are formed
	by not enough air (as 1 hypothesis of Global) then this might be bass-
	akwards and it might be better to return to air rich and reduce fuel
	pressure.  With the addition of 2 - 2.5" air holes though the oxygen
	starvation question should addressed.  Increasing the fuel rich, could
	reduce the fuel endurance though.

	Preliminary Conclusion from noticing rise in current and further
	TEG observations to be noted below: TEG performance improved.

4/12 16:30Z Fla.
	- Ran overnight with TEG and 1-PV.  Load high with 2 30ohm heaters
	going on 4-comp SW.
	Batt dropped to 12.85 after disconnecting wind and dropped to 12.7 by
	morning. Iload=1.48,Icharge=1.77 on EVE.
	Ie battery not fully charged and charge monitor circuit is
	off.  Disconnected internal batt box ground, pin 16.  Iload=1.883,
	Icharge=1.728.  This strikes me as closer to truth
	but didn't have stuff to measure w/meter.
	Added wind gen back on to run station while inspecting TEG.  Plenty
	of wind and juice for job.  Still have noise/ground-loop problem
	on current monitors.

	- TEG Inspection:  Yukko.
	Shut down teg and removed from sled for working inside met hut.
	Removal of the internal filter is impossible with teg in sled and
	awkward with it out....not field procedure in arctic.
	Internal Filter looked clean.
	However golden yellow sticky goo mixed w/what looks like thread seal
	inside of trap.   Stuff will not flow out of the drain plug.
	Took sample to bring to Boulder.  Very smelly and messy.  Also
	saw it on teg rail and some fittings indicating it oozed out sometime.
	This goo also appears to be inside of the lines, but they are clear
	enough to keep working.
	Upstream Inline Filter inspected:
	Filter looked clean, but same residue was in the trap.  Jeff later said
	he hadn't seem that on last inspect 1-2 mos ago.
	Cleaned housing and Replaced Internal Filter (since we have new
	ones and just in case).
	Nozzle replaced:  However it looked pretty clean and Jeff said it was
	the best he'd seen of one being replaced.

	Conclusion 1: the inline filter helps and these need to be added to
	the other 3 stations.
	Conclusion 2: This crap	is in the fuel itself (need to confirm) or was
	from liquid propane condensing at some point. The residue on the nozzles
	is probably this stuff bubbling through and not being caused by lack
	of oxygen.

	23:03 Restarted TEG (after reinstallation in sled)
	23:33 VOC = 3.77 (Rated = 3.87), Pressure = 71 Kpa (Rated 64)
	00:30 VOC = 3.865.......This should be the stabilized output (good)
		but will keep open through night and recheck in am.
		May need to adjust regulator down a bit and air mixture
		like Cle.


4/12 23:05Z Cle TEG
	Shutdown of TEG, removal for inspection
	Same residue but more liquid and began slowly dripping from pres. reg
	intake.  This line was pretty fouled and even though the generator is
	running much more of this could clog it, or maybe it's bubbling up
	into the nozzle which was much dirtier than Fla.  
	Need to clean fittings at ship.

	Running overnight on battery and 1-PV
	So, is the battery good/fully charged?

4/13 4:00Z Fla Wind Gen
	This hummer was causing noise in the sodar (really nice winds!)
	disrupting data.  Shut down
	Running overnight on battery and 1-PV
	So, is the battery good/fully charged?  (had ~10 hours on windgen)

4/13 23:30Z Atl TEG
	Replaced TEG w/ cleaned unit from Cle. which was reassembledt
	his am.  Also added in-line filter.
	Examined Atl teg  atShip for contamination: yuk, much worse than
	Fla or Cle TEG.  Internal filter actually 1/2 clogged.  Lots of
	liquid.  Oozing out of pressure switch.  Need to clean fittings
	at ship.  Nozzle very dirty also.

4/15 Bal TEG
	Replaced TEG on 4/14.  Examined on 4/15.  Extremely gunked,
	perhaps worst of the lot.  Substantial amount oozed out of pressure
	switch and a little out of steel hose 'downstream' from the internal
	filter.  Cleaned, replaced nozzle, replaced filter.

We had such good results with both the heaters and fans on, that we decided to
try connecting the SOV back up.  At 1430 the SOV spade plug was reconnected with
no immediate ill effects.  Ten and half hours later it's still good. No noise or
high current.  The auto-spark did not come on either.  :-)

See ya soon .... ?

We were at Fla yesterday checking the levels and giving a little tour of the 
station. We noticed the current was high again.  About the same time the
tour was focusing on the TEG. We found that when the door was shut the current
was high and low when open.  Very repeatable. We noticed the white/red TEG
output wire was being pinched between the TEG door and TEG heat sink.  We moved
the wire back into what looked like a access slot in the side so it wouldn't 
get pinched when the door is closed.  The data has looked good since then.

Now the interesting part, the auto-ignitor whould spark whenever someone whould
bend near the TEG.  It could be some weird electro-magnetic field thing or 
maybe the way we were leaning on the sled might have moved some loose wires.
We'll to isolate the problem area if time permits.  Everything is currently 
connected ( SOV, heaters and fans )

Dr. Sheba