Monday, August 16, 2004

Why I'll never use BT again - A couple of years ago the company I was working for had to provide a video-over-IP terminal to a client for remote VHS duplication - I had to organise the installation of a leased 512kbit line into the client's premises (but we would pay for the install and lease). I ordered the install and felt very pleased with myself for calling the Post Office ahead of time to confirm the post-code. The day the job was due to happen I called BT to confirm and they told me the job had been cancelled by engineering because I'd given them the wrong post-code.

"But I got the postcode from the Post Office - surely they are the people who do post-codes?"

"Sorry sir - it's not the one engineering have on file for that building"

"OK, to avoid this happening again can you please tell me what post-code engineering have on file so I can re-schedule the job and be confident it won't get bounced back again?"

"I'm sorry sir, engineering won't disclose that information"

Eventually I did get the line installed and the day it was livened up I found myself sitting on the floor of the client's comms room with my laptop connected to the new router and no return pings from their end - I could see a carrier on the line, but no data. So, I called BT again and gave the various job numbers etc. and got the reply:

"I'm sorry sir, that line isn't due come on line until 2099"

He honestly said that without a hint of irony!

Thursday, August 12, 2004

Issues with the mains - when building TV facilities I always go to great lengths to keep the technical and domestic mains separate - well, to be more precise I ensure there is a clean technical earth derived from the same point in all areas where signals flow.
I had a worried couple of hours yesterday with one room that had huge hum over all the signals - running cables around the corridor in case the wiremen had run signal cables a long way along a mains tray made no difference - measuring earth discepency to the 16A C-Form connector on the wall showed no standing earth current yet when everything was plugged up the monitors looked like they didn't know the tune (humming!). Eventually I found an earth-neutral reverse in the Olsen block and the problem was solved.

Now, this client have their own substation, so I was even more suprised there was so much potential difference between neutral and earth, but was glad to have found the fault none the less.

Monday, August 09, 2004

Very wee hard drive!

I spotted this article in my institute's magazine and was blown away - how tiny! We'll be getting a Unity in a matchbox!

Sunday, August 08, 2004

Tektronix video site is a superb resource not only for their products but for all manner of info on test signals etc.

Saturday, August 07, 2004

RF distribution

When building TV facilities there is often the requirement to include an RF ring main of some sort so that office dwellers can see off-air channels and even channels modulated on in the building - it's good to know that channel 6 (for example) on every TV or VHS in the building is seeing the output of the studio (for example).
Having trained at the beeb I'm very familiar with the old method of having a launch-amplifier that sends RF on a passive distribution, typically sending initially at 100dBuV and using attenuators whenever a TV or VHS is connected to the "ring". This system is robust and once implemented is cheap to scale - all you need at each TV is a passive attenuator. The only design consideration is that you keep track of the attenuation introduced by the co-ax and use local attenuators to feed the TVs.
The only down-side to this method is that in the last twenty years everyone has a domestic VHS on their desk as well as a TV and even the best designed modulator input stages leak signal back out of their inputs - you'll see a tiny amount of the tape output of a VHS on it's RF input - not a problem at home because that tiny signal (typ. less than a dBuV) stops at the antenna. In an RF ring main it all accumulates and the overall ring gets a bit noisier. Go to any aging BBC facility and the engineers will tell you how noisy the RF is and it's not down to the quality of the components - it's all those VHS machines (and now PC TV cards, PlayStations etc.) leaking a tiny bit of signal back up their inputs.
And now I will show you a better way!
Hotels tend to use multi-output RF DAs that have individually isolated outputs - it by-passes this problem. The issue is that now you have to run RF cable to every desk or room but you can abuse the system a lot more with no impact on the quality of the RF. Since we've moved away from 10Base2 to 10/100/1000BaseT (or even fibre) networking we're used to running cables per person rather than big rings you tap off I don't see that (at install time anyway) it is such an issue.
Suffice to say I use the latter (at our current MTV install as well as every other facility I've built).

Friday, August 06, 2004

Intel SR2200 server keyboard/mouse conundrum!

Ben Davison and I were installing a dual-server Avid Unity rig at MTV (it's the fail-over config with two Media Managers seeing the fibre array) - both servers, once rack'ed up would boot once and then never boot again (couldn't even see the BIOS screen) - if we extracted the machine from the bay it would boot fine - back in the rack and it would boot once then nada. Very frustrating - the fault was found to be the little PS/2 breakout adaptor - the SR2200 has only one PS/2 socket on the back and the breakout for mouse and keyboard isn't marked - if you get it the wrong way round (keyboard plugged into mouse hole etc.) it does the above - crazy considering the server clearly sees the keyboard and mouse on that first good boot!

Thursday, August 05, 2004