PAL/NTSC to 48Ksample rate

Just about to get my workshop up and running again after a couple of years and I'm going to play with some digital audio, but I want to eventually genlock to a "professional" video system, ie not a crap VHS machine.

I can get a Elantec EL4585 PLL to generate 27MHz then use a BB PLL1700 to generate the audio clock. The ADC's also have their own PLL,( Alesis AL1101, not the greatest, but adequate for majority things) Three PLL's in series, sounds(sorry) like a disaster in the making for the jitter dept.

Any other/sensible options

martin

"Facts are stupid things.." -- Reagan, '88

Production facilities use a central *wordclock* and sync everything to that.

Only one PLL / device in the chain.

Graham

If you are referring to video production the proper term is "Sync generator" or Master Sync generator" to provide a composite sync along with separate vert, horiz, burst and other signals needed to sync all of the other equipment. This is usually followed by a full set of DAs to buffer the signals and give multiple outputs per signal. I've used Grass Valley and Tektronix at different studios and TV stations over the years. A neat early digital video manipulation system was the Squeeze Zoom made by Vital Industries in Gainsville years ago. It filled a full height relay rack, had a 1000 A 5 V power supply and needed three phase

208 to operate it.

--
Beware of those who suffer from delusions of adequacy!

Michael A. Terrell
Central Florida

Ye Gods, I remember the first Squeeze Zoom in the UK, at Molinare, but then I was just an audio guy. AFAIR it wasnt terribly reliable,

martin

"Facts are stupid things.." -- Reagan, '88

Was it new when you got it? Someone was selling refurbished units that were rebuilt with the cheapest low temp electrolytics they could find, and they cut corners on the voltages. I mean, who in their right mind replaces a bad 25 volt cap on a 12 volt buss with a 12 volt cap? We replaced the caps as they failed with the OEM specified parts and it ran for several years with very few problems. On the other hand, doing all those video effects one filed at a time in real time with a Z80B processor was pushing things quite a bit. Some of the parts were obsolete before they even got into full production, but TV stations were lined up with checks for $250,000 US dollars when it was introduced. Their studio switcher/router gave more problems than the Squeeze Zoom. The ICs had to be run hard to get the gain and bandwidth needed and the selection of video op amps wasn't very good when it was designed.

BTW, did you ever try zooming into infinity? It created a great star field effect as the random noise generated stars at the center of the screen and they grew as they headed for the edge of the CRT. I had the Cyron operator create a couple different colored screens that flashed, "Game Over!", overlaid it on the star field and fed it to the master monitor and waited for the audio guy to see it. He almost fell out of his chair when he did a triple take. :)

CH 55 replaced it after I left with something that took about 17.5 inches of space, but didn't have a couple of the better effects.

--
Beware of those who suffer from delusions of adequacy!

Michael A. Terrell
Central Florida

I wasnt really involed with video, I had my hands full maintaining 8 audio studios. It was new, but ISTR that the Squeeze zoom 5V supply was wel over

5.25V at the supply entry and well below 4.75 at the end of the rack. A rewire of the power distribution cured many probs, but the blacks kept drifting. Them were the days, the first ADO, 1" C format, first Satellite channel etc in Europe. We (Molinare) made a lot of money, then the bean counters took over. I left, havent had a real job since.

martin

"Facts are stupid things.." -- Reagan, '88

Well... the original query was about audio clocking, but audio wordclock is also synced to video in such applications.

This why you end up with slightly odd sample rates like 44.056kHz ( IIRC ) instead of 44.1 kHz in some applications.

And don't even start with 'drop frame' time code syncing !

Graham

snip

Hi Graham My OP was really about stability problems with so many PLLs in a system, ie your original SPG generating 525/625, which may possibly be locked to another master SPG, and then my system which will derive

48K, and the ADC's have an inaccesable internal PLL to generate bitclock. I was just wondering if there is a potential problem of jitter building up by going up to 27MHz from something that is already slightly jittery from previous PLL's and then down to get my 48K sample rate. I chose the 27Meg option, since it is a common sample rate to all TV systems, unlike trying to lock to subcarrier frequencies

It's probably not a problem, its just something I'd never thought of before.

martin

"An eye for an eye makes the whole world blind" Gandhi