Looking for source of arcing
We've seen broadband rf bursts in the
gregorian dome receivers when the hf transmitters are on.
We've seen the bursts in all of the receivers we used to look for
them (more info). We do not think
the arcing is occurring in the transmitters or heliax lines
- 327 mhz receiver
- 430 mhz receiver
- alfa receiver
During the nov15 hf run, the problem was bad
enough that we had to limit the transmitted power so the arcing
would not contaminate the 430 Mhz data (we never got rid of it
What we know
Most of the following tests were done with the
ch and dome 430 Mhz receiver:
- az, ch, dome motion (more
- the arcing gets worse as we move the dome, ch , or azimuth.
- The arcing is repeatable with az, za position
- This would not be true if the problem was in the
transmitters, transmission lines, or dipoles.
- During the nov15 run, we melted one of the load cell cables on
td 8 from arcing..
- This cable was run inside a metallic protective hose (about
1/2 inch diameter)
- it shorted where the load cell cable/hose was tied to the
yoke (a few feet from the load cell).
- The loadcell cable has a shield that was connected at
the huffman box, and not connected at the load cell end
- the protective hose was grounded on the load cell end. it
was left open on the huffman box side.
- the cables were at least 30 to 40 feet long.
- when this cable was removed, the broad band interference in
the 430 receiver decreased appreciably.
- the arcing was seen to be worse with lcp (we use rcp during
heating runs to get the optimum heating).
- the arcing got worse when we used juha's pulsing mode (20 ms
on , 40.1 ms off). We had to reduce the transmitted power 3 to 6
- testing on 03dec15 showed that the radial dipoles closest to a
tiedown is the one that couples into the tiedown cables.
03dec15: looking for arcs under the dish.
We searched for sources of arcing under the dish on
03dec15. The setup was:
Mike and i went under the dish to search
for the arc. We tried using:
- run all 6 hf transmitters (freq at 5.1 mhz .. according
- bring the power up to around 75 kw each (16dbm)
- then turn on the 20ms, 40.1 ms pulsing mode. (later we
switched back to cw).
- put az at 285, dome at za=10, move ch in za 3 to 6dec at .02
deg per sec (tiltza routine)
- We used the 430 Mhz ch receiver to monitor the arcing..
- We had intermittent arcing...
- narda power meter set to 5.125 Mhz
- ar8000 radio tuned to 27 Mhz.. This did not show the arcing.
- ultrasonic detector.
- AO 150 Mhz communications radio.
What we found
We then decided to characterize the 5 mhz field strengths as we
turned on, off various dipoles. The measurement sequence was:
- the ultra sound detector.
- no reliable arcing detected.
- We did see strong signals coming from
- the heliax .. at 3a,3b .
- the heliax .. at 1a,1b
- and most other heliax
- We saw a strong signal coming from dipole 6x. this is an 8
- it was the balun that didn't have the short on it. Lots
of 5 mhz power is coming out of the balun.. probably
picked up by the dipole
- After the test we realized that the pulsing could cause the
pieces of the heliax to oscillate (center conducter, spacers).
If we want to use the ultra sound detector, we will need to
figure out how to generate consistent arcing without pulsing.
- ar8000 radio
- did not here anything on it..
- communications radio:
- we heard increased noise when we moved close to large fields
(as measured by the narda power meter).
- Narda power meter.
- This gave us reliable measurements of the power levels
at 5 Mhz. it probably did not tell us much about the arcing
(other than where the strong fields were located.
- set hf transmitters to cw.
- turn all transmitters on.
- Measure the field strength at the tiedown ground cables
- We stood on the td block, and mike held the narda meter
about 1/4 inch from the grounding cable (about 5 feet above
block). he moved it up/ down till he found the peak in the
standing wave along the cable.
- the measured field were up to 15 times higher than the
allowable restricted access exposure limits (34
- This occurred when the power meter probe was within 1/4"
of the grounding cable.
- backing the power head back by about 6 inches dropped the
level to well below the max allowable values.. say 5
- The table below shows the power values measured on the
grounding cables of the tiedowns:
|td 8 L
- L - left most object (grounding cable, etc)
when you look at the tiedown block from center of dish.
- R - rightmost object (grounding cable,etc)
when you look at the tiedown block from center of
- There were variations in the readings caused by :
- standing waves in the grounding cable,
- positioning of the power probe
- tx power variation
- The closest B dipole coupled the strongest. We
know that the transmitted power varies tx to tx.
- For tx 4 we read 250 mW/cm^2 and then a little later
read 500.. so we're probably good to 3db..
- the leftmost cable had 250 mW/cm^2. After turning
all of the tx off, then back on, we got a reading of
up to 500 mW/cm^2,
- the rightmost cable gave a reading of 150 mW/cm^2
(when the leftmost one read 250).
- When placing the power probe between the two plates
of the yoke (that holds the cable/loadcell pins)
- we got 150 mW/cm^2 on the right side,
200mW/cm^2 on the left side
- above the motor transmission (in free space) we got
- we only looked at the rightmost grounding cable
- when placing the probe above the motor transmission
we got about 50 mW/cm^2 (not close to any metal)
- The cables that connect the tiedown huffman box to the jack
have power near them
- encoder cable
- motor cables
- load cell cables.
- The heliax cables that feed the power to the dipoles run from
the tx building all the way down to the dipoles.
- The power meter showed that these heliax cable also
have power near them (don't remember the levels...)
- The adjacent radial dipoles couple into the tiedown cables.
- above the motor transmission and up toward the yoke we have
fields of 50 to 100 mW/cm^2
- tx 6 (between td 12 and td 8) is the 8 mhz dipole with the
- One of the baluns does not have the grounding short at the
bottom of the balun.
- lots of 5 mhz power is coming out.. we should put the short
<- page up