I am not sure this is exactly relevant, but here goes. When a Polar mount tracks an object, it solves for the center of the Earth's mass that is 4000 miles below your feet. There is compensation for this but never absolute. Remember long debates about this in the TVRO days. Ted WA8ULG From: moon-net-bounces(a)mailman.pe1itr.com [mailto:moon-net-bounces@mailman.pe1itr.com] On Behalf Of Roger Rehr W3SZ Sent: Monday, August 01, 2016 2:55 PM To: moon-net(a)mailman.pe1itr.com Subject: Re: [Moon-Net] 3 cm auto tracking error Interesting discussion! Somehow these emails landed in my spam folder, so I am replying a few days late :) The problem Gerald describes is similar to what I had to deal with on 10 GHz where I had an Az-over-El mount. The math is similar. I wrote a software correction based in part on a paper by Chuck MacCluer, W8MQW. However, I had to extend his solution. For those interested, I did a small web page on this with references and source code back in 2010. That can all be referenced at: http://www.nitehawk.com/w3sz/AzoverEl.htm The info on this page would provide a starting point for those, like me, who like to write their own code. 73, Roger W3SZ On 7/27/2016 1:13 PM, TexasRF(a)aol.com wrote: Tom, if the azimuth axis is not perfectly vertical (never is perfect) the actual azimuth direction has a built in error that is related to the amount and direction of the axis tilt and the elevation angle. Without getting into the details I will give you an extreme example of the obscure and not well understood problem: Assume you live at a latitude that allows the moon to reach an elevation of 90 degrees. Further assume that there is an azimuth axis tilt of .5 degrees toward the east and the moon is passing through a calculated direction of 180 degrees. The calculated azimuth axis would be straight up. The tilt toward the east changes the axis from a calculated value of 180 degrees to an actual angle of only 90 degrees. The elevation angle is reduced from 90 degrees to 89.5 degrees. This is an extreme example but the trend and concept is present for other latitudes and tilt directions. This is the same problem that is encountered with multiple side by side feeds in a dish. Working with Al, W5LUA in the past we have been able to mathematically correct his errors to a large degree. On my own system, the tower has a slight lean toward the southeast that requires a correction of about 2 degrees in azimuth as the moon passes through zenith. The correction varies as the declination and elevation angle varies. I hope this is the issue you were asking about. If not, please ignore this. 73, Gerald K5GW In a message dated 7/26/2016 4:41:53 P.M. Central Daylight Time, w2drz(a)ramcoinc.com writes: Hi, If on 3 cm auto tracking and must correct moon noise during a QSO? I am trying to understand why the auto tracking does not do 100% correct TRACKING, the cause may be various reasons?. Those on 3 cm or 5760 and auto tracking, drop a note on what you find and if have found a cure of the error problem. Try to change the W2DRZ controller and AB encoders to hi res (16 bit) on the encoders to see if that changes the tracking error. the USD encoders will then do about 14 bit resolution or "1 tenth a degree" (0.01) degree. tom W2DRZ http://www.w2drz.ramcoinc.com/ http://www.w2drz.ramcoinc.com/ProgNotes.htm _______________________________________________ Moon-Net posting and subscription instructions are at http://www.nlsa.com/nets/moon-net-help.html _______________________________________________ Moon-Net posting and subscription instructions are at http://www.nlsa.com/nets/moon-net-help.html

Dear Friends With great interest I'm following the discussion regarding some tracking problems especially in high bands (13cm and up). In HB9Q we experiment similar situation and most of the cases that errors are originated by how difficult is to reach the 100% vertical axis with no deviations. Depending on the mechanical system some can have more or less tilt angle and this will be translated in different EL angles at different AZ positions of the antenna, below 13cm that errors are no so critical but as soon you go up in frequency that errors increase and are especially hard in 6 and 3cm. Since the EME conference in Dallas I produce more than 200 antenna controllers OE5JFL (DRACS-G2), and many of them are in service in high bands. Just 2 years ago, with the help of Rene PE1L I place on my page the software DRITRACK, with that software you can control the antenna controller with a PC helping a lot specially for remote operation. During the deployment of my 13cm station in South Africa I experiment with a correction table capable to compensate any error in your vertical axis, the idea is you must plot your AZ at certain intervals and take note about the error introduced in EL due the tilt angel in your vertical axis, then that values are introduced in a correction table. The DRITRACK software will add or remove the correction in EL, keeping the antenna in the properly position. Well I'm not a programmer and the initial beta was working with some problems, until I discover my 3.6m has almost not tilt angle (at least in 13cm) then I didn't make any follow up to this project. Looks like is time to think in this correction table for the DRITRACK software and the OE5JFL controller, especially when now I decide to boost my 9, 6 and 3cm projects. What I will need for those who are operating in high bands (13cm and up) is a feedback about the nature of the problem, how many corrections we need? At how many degrees as interval, Could be different correction per band? How is the best way to test your EL errors in different AZ positions? Your feedback will give me a more clear idea and I will work on that to add a correction table for the OE5JFL controller. Your feedback is welcome. 73 de Alex, ZS6EME (HB9DRI) From: moon-net-bounces(a)mailman.pe1itr.com [mailto:moon-net-bounces@mailman.pe1itr.com] On Behalf Of James Malone Sent: Dienstag, 2. August 2016 01:51 To: Ted Hartson Cc: moon-net(a)mailman.pe1itr.com Subject: Re: [Moon-Net] tracking error Hello EME friends , I use the OE5JFL controller with MAB 25 encoders . It works ! I use a digital level capable of .01 degree resolution to set the mount as close as possible to level with the earth across the moon arc . On my 3 cm EME antenna which is a 2.4M parabolic made for SNG I find an error of approximately .5 degree on EL between my most easterly aim at the moon to my most westerly aim . So I use a R/S URV 35 power meter with a 4 MHz BW filter and 40 db gain in front of it to measure moon noise . I usually see 1.2 db noise from a cold sky measurement to the moon power . ( and 14 db sun noise) These measurements are my reference and in the OE5JFL controller I use the offsets dynamically during the course of the moon rise and set to keep my antenna aimed . When the beacon is running (DL0SHF on .025 ) I use a secondary receiver to keep an eye on the received signal level in JT4G . This also helps me know I am on the moon . ( when the beacon is operational) The skill developed for this operation was a big eye opener for me . I am enjoying the home brew solution and development to keep my antenna tracked on the moon or the sun . ( as well as Doppler Managment ) I am now in my spare time working on tuning TWTA and SSPA 3 cm amplifiers . W5LUA and VE4MA ,OZ1LPR and Per Dudek (DL0SHF beacon operator) have been very helpful to the success of my station . Soon I hope to conquer the "CW" hurdle .. As many stations are only CW ops . Best 73 and see you all in Circular Pol some day ! Jim WA3LBI On Monday, August 1, 2016, Ted Hartson &lt;awombat(a)earthlink.net&gt; wrote: I am not sure this is exactly relevant, but here goes. When a Polar mount tracks an object, it solves for the center of the Earth’s mass that is 4000 miles below your feet. There is compensation for this but never absolute. Remember long debates about this in the TVRO days. Ted WA8ULG From: moon-net-bounces(a)mailman.pe1itr.com <javascript:_e(%7B%7D,'cvml','moon-net-bounces@mailman.pe1itr.com');> [mailto:moon-net-bounces@mailman.pe1itr.com <javascript:_e(%7B%7D,'cvml','moon-net-bounces@mailman.pe1itr.com');> ] On Behalf Of Roger Rehr W3SZ Sent: Monday, August 01, 2016 2:55 PM To: moon-net(a)mailman.pe1itr.com <javascript:_e(%7B%7D,'cvml','moon-net@mailman.pe1itr.com');> Subject: Re: [Moon-Net] 3 cm auto tracking error Interesting discussion! Somehow these emails landed in my spam folder, so I am replying a few days late :) The problem Gerald describes is similar to what I had to deal with on 10 GHz where I had an Az-over-El mount. The math is similar. I wrote a software correction based in part on a paper by Chuck MacCluer, W8MQW. However, I had to extend his solution. For those interested, I did a small web page on this with references and source code back in 2010. That can all be referenced at: http://www.nitehawk.com/w3sz/AzoverEl.htm The info on this page would provide a starting point for those, like me, who like to write their own code. 73, Roger W3SZ On 7/27/2016 1:13 PM, TexasRF(a)aol.com <javascript:_e(%7B%7D,'cvml','TexasRF@aol.com');> wrote: Tom, if the azimuth axis is not perfectly vertical (never is perfect) the actual azimuth direction has a built in error that is related to the amount and direction of the axis tilt and the elevation angle. Without getting into the details I will give you an extreme example of the obscure and not well understood problem: Assume you live at a latitude that allows the moon to reach an elevation of 90 degrees. Further assume that there is an azimuth axis tilt of .5 degrees toward the east and the moon is passing through a calculated direction of 180 degrees. The calculated azimuth axis would be straight up. The tilt toward the east changes the axis from a calculated value of 180 degrees to an actual angle of only 90 degrees. The elevation angle is reduced from 90 degrees to 89.5 degrees. This is an extreme example but the trend and concept is present for other latitudes and tilt directions. This is the same problem that is encountered with multiple side by side feeds in a dish. Working with Al, W5LUA in the past we have been able to mathematically correct his errors to a large degree. On my own system, the tower has a slight lean toward the southeast that requires a correction of about 2 degrees in azimuth as the moon passes through zenith. The correction varies as the declination and elevation angle varies. I hope this is the issue you were asking about. If not, please ignore this. 73, Gerald K5GW In a message dated 7/26/2016 4:41:53 P.M. Central Daylight Time, w2drz(a)ramcoinc.com <javascript:_e(%7B%7D,'cvml','w2drz@ramcoinc.com');> writes: Hi, If on 3 cm auto tracking and must correct moon noise during a QSO? I am trying to understand why the auto tracking does not do 100% correct TRACKING, the cause may be various reasons?. Those on 3 cm or 5760 and auto tracking, drop a note on what you find and if have found a cure of the error problem. Try to change the W2DRZ controller and AB encoders to hi res (16 bit) on the encoders to see if that changes the tracking error. the USD encoders will then do about 14 bit resolution or "1 tenth a degree" (0.01) degree. tom W2DRZ http://www.w2drz.ramcoinc.com/ http://www.w2drz.ramcoinc.com/ProgNotes.htm _______________________________________________ Moon-Net posting and subscription instructions are at http://www.nlsa.com/nets/moon-net-help.html _______________________________________________ Moon-Net posting and subscription instructions are at http://www.nlsa.com/nets/moon-net-help.html