(Untitled)

[tau_iota_mu_c]

I had an eventful week of night shift when I was on 2 weeks ago (and have only just gotten time to write this up, at 2am, on a worknight when I have to be up in 4 hours. Sigh, sometimes I shit me). The first eventfulness started when I got up to the 'scope on my first night to discover the telescope was out of action. It had collided with the

Comments

[badpauly]

(I'm not surprised. The work that proceeded this found out that a dome zero switch that had been recently installed for the new computer control had been installed at not the useful azimuth of 0 degrees, but had been installed at the precise azimuth of 176 degrees).

A small degree of error - we will fix it next time.

[tau_iota_mu_c]

You can always redefine 0==176 in the software with regards to the calibration of the dome rotation, but surprisingly, we pass through a dome azimuth of 0 degrees more often than we go through 180 these days, because we now look at more targets in the North than in the South. If you spent more time looking at targets in the South (as you would expect of a Southern based telescope), then as you scanned across the meridian, you'd pass through 180degrees. Since we go through 0 more often now, the project specs specced a dome zero of zero so the dome is synced up more often. Still, 176 as opposed to 180 if such a choice was a deliberate subversion of the project specs? Maybe someone inadvertently rotated the dome between when the person fitting the switch checked where the dome was, and when they fitted it.

[badpauly]

Or maybe they just threw it in with a "late for lunch, this will do" attitude

[tau_iota_mu_c]

Yeah, we are public servants. Lunch is our core business (I left my lunch next to my bag this morning prior to leaving for the bus. I picked up my bag, put on my gloves and beany, and left my lunch on the floor. Didn't realise til I got to the bus stop just as the bus was rolling up. Made for a very sad TimC.).

[tcpip]

Except that someone came along and raised the floor but didn't raise the limit switches enough

...Now that's an oops.

[tau_iota_mu_c]

I don't know the full story behind it, but I imagine the floor was raised when the new topend for 2dF was built, and the switch surely would have been moved, but you have to calculate it based on an arc, and if the switch isn't struck properly, then it might trigger later than you expect, and I can very much understand why you would feel nervous about testing that it can stop in time at full speed (they probably only tested it under manual control at slow speed with radio commication between the person doing the testing, and the person at the control desk, and the striking of the switch may be dependant upon the speed of slewing if you're unlucky). You only get one unsuccessful attempt at such a test if the switch fails catastrophically.

[tcpip]

I can very much understand why you would feel nervous about testing that it can stop in time at full speed

There would be a hell of a lot of momentum on one of those things. Your pictures really display the enormity of the problem.

[tau_iota_mu_c]

You should come visit. A good insight into working 1970s era computing (before we decommision it within the next year!) and engineering (overengineer (almost) everything, and do it properly). I added the "retrocomputing" tags to my flickr photos last night (I am disappointed there is only a page of other photos with "interdata" tags