AVS › AVS Forum › Audio › Receivers, Amps, and Processors › Laserdisc RF AC-3 Demodulator Question
New Posts  All Forums:Forum Nav:

Laserdisc RF AC-3 Demodulator Question

post #1 of 46
3/7/10 at 2:24pm
Thread Starter 
Could someone point me to a source to understand how RF-AC3 is converted to something useable on the S/PDIF input?

From http://en.wikipedia.org/wiki/Laserdisc
Quote:
...Unlike DVDs, which carry Dolby Digital audio in digital form, Laserdiscs store Dolby Digital in a frequency modulated form within a track normally used for analog audio. Extracting Dolby Digital from a Laserdisc required a player equipped with a special "AC-3 RF" output and an external demodulator in addition to an AC-3 decoder. The demodulator was necessary to convert the 2.88 MHz modulated AC-3 information on the disc into a 384 kbit/s signal that the decoder could handle. ...

Oh, I thought, sounds simple, just need some sort of demodulator, probably FM demodulator using a phased locked loop and you are done. However, after much searching, I found a block diagram and schematic for a Pioneer RFD-1 demodulator at http://www.kallhovde.com/pioneer/tri...-1_RRV2125.pdf

Studying the block diagram for the circuit given in this reference that includes a block diagram of the Pioneer PM4007A IC that does the heavy lifting, it is clear that this is not just a simple RF demodulator. It appears to use PLLs to demodulate the signal. It also has memory and appears to be reformatting the digital stream before relaying it. Also, the RF frequencies in the PLL circuit do not appear consistent with the wikipedia description (or some details in the circuit are not obvious to me from this diagram.)

Can someone enlighten me on the rf modulation scheme used and also the digital reformatting required?
Reply
Reply

AVS Top Picks

post #2 of 46
3/7/10 at 3:04pm
Wow, that's a lot of circuitry for converting analog to digital. Maybe it's hard to modulate 300+ kbs digital to FM reliably.

I was also confused by the frequencies being used.

I supposed the AC3 RF specification was never an open standard, either?

I found this stuff, but it was about modifying a player to output RF, not about decoding.

http://www.laaudiofile.com/ac3mod.html
Reply
Reply
post #3 of 46
3/7/10 at 4:34pm
Whoa, that is more complicated than I had expected. Sorry, I'm not really familiar with the details of the process. I had always assumed it was pretty straightforward, but there appears to be more going on than I was aware of.

Thanks for posting the service manual, though. It may be helpful if my RFD-1 ever dies.

I've also done the mod Michael referenced and it doesn't include a demodulator circuit. It simply adds a small board to extract the RF modulated AC3 signal from the right analog audio track and pass it out of the player. The kit I had didn't include any technical information or schematics; just installation instructions.
Reply
Reply
post #4 of 46
3/8/10 at 8:49am
If your player has the AC3 out? It's pretty simple. If you have to mod it to get the AC3 out? I'd just find one on the bay with AC3 output in all honesty.

Were you wanting to build the extrenal box? I am using the one I got off of some auction site back in the 90's. Always stays on and I don't do AC3 LD's that darn often. If you buying a player with AC3 output? See if you can also find one that has two 2video outputs. This makes it much simpler when doing automation with a remote. Like on my Harmony I have an activity defined as "Play LD wAC3" or something like that.
Reply
Reply
post #5 of 46
3/8/10 at 3:26pm
Thread Starter 
Well, I was thinking of building an external box (or putting it internal) and adding an AC3 output. Adding the RF AC3 output is easily doable. The external box looks problematic until I can understand the function.

My current thought is that it may be possible to demodulate and send to a computer interface to finish the job.

However, at this point, the interest may just be academic. My current guess is that the modulation may be some sort of PSK with a 2.88 MHz carrier frequency. Note that 46.08 MHz is just 16 x 2.88 MHz. Maybe someone can enlighten me further.
Reply
Reply
post #6 of 46
3/8/10 at 3:47pm
I probably knew somewhat how it worked at one time. This would have been easier in a by gone era of the past.

There was a guy selling for $75 on Satalite Guys forums recently. It's worth that.
Reply
Reply
post #7 of 46
3/8/10 at 3:50pm
wow.. I just checked and these still bring some cash on the bay. Surprising. If I ever get hard up for cash, I may list mine as I am generally fine with modern processors and old 2 channel ProLogic. Although for nastalgia sake, I do pull out DTS and DD LD's to screen sometimes in a row. Always reminds me of more hopeful / youthful times in my life.
Reply
Reply
post #8 of 46
3/8/10 at 5:18pm
Yeah, the RF demodulators often go for $300+ on eBay. Anyone that still uses their LD player needs one, since no new receivers or processors have them built-in anymore. I got lucky a couple of years ago when I semi-retired my DSP-A1 and snagged the RFD-1 for "only" $80 by somebody who I guess didn't check prices before listing his "Buy it Now" sale.
Reply
Reply
post #9 of 46
3/8/10 at 5:31pm
A while back Technics had one... Technics SH-AC500D. There it is... comes on fleabay every so often for about $50.
Reply
Reply
post #10 of 46
3/8/10 at 7:39pm
Thread Starter 
Quote:
Originally Posted by goobenet View Post

A while back Technics had one... Technics SH-AC500D. There it is... comes on fleabay every so often for about $50.

For the record, from my reading of the manual from http://service.us.panasonic.com/OPERMANPDF/SHAC500D.PDF and looking at the backpanel, it does not support RF AC3
LL
Reply
Reply
post #11 of 46
3/8/10 at 8:03pm
well i'll be. Could've swore technics had a model that supported RF...
Reply
Reply
post #12 of 46
3/8/10 at 8:34pm
They probably did have models that did; pretty much everyone did.
Reply
Reply
post #13 of 46
3/8/10 at 8:43pm
I recall Yamaha had a high end AVR that supported RF modulator for LD..

Just my $0.01..
Reply
Reply
post #14 of 46
3/9/10 at 7:22am
To tell you the truth, it's been so long now, I can't even remember the exact transaction of how / where I aquired mine. I think it's a Sony. I mean it istucked away in the equipment rack somewhere and I never even look at it.

I'd still just try to acquire one if I were the op. Make sure you understand the vast majority of LD's don't even have the AC3 on them. The few real DTS discs have the DTS encoded the audio goes right out the normal optical digital audio output jack.
Reply
Reply
post #15 of 46
3/9/10 at 7:33am
Not every LD used AC3-DD. Some still used LPCM. So it (the RF) had to be processed, and recognized before being sent through. Then there's error correction checksums, and so on and so on. To Process or Not To Process, that is the question.

Just buy the converter.
Reply
Reply
post #16 of 46
3/9/10 at 8:08am
But I thought you could have the SC27 apply processing to the analog DD signal output to get the 5.1 channel sound without the AC-3. I think it will be comparable if not better than the AC-3 option considering the plethora of 5.1 processing options on the Pioneer. I will try it tonight and report back how it sounds tho I dont have a AC-3. But I do have the same source material in DVD and Laserdisc so I am going to give it a try.
Reply
Reply
post #17 of 46
3/9/10 at 8:36am
It can't take it from the two channel analog output. Or not that I ever knew of anyways. I mean it gets stripped before hitting the actual analog outputs of the LD player.
Reply
Reply
post #18 of 46
3/9/10 at 9:41am
Quote:
Originally Posted by JOHNnDENVER View Post

It can't take it from the two channel analog output. Or not that I ever knew of anyways. I mean it gets stripped before hitting the actual analog outputs of the LD player.

???

The 5.1 DD discrete signal gets stripped, but you can use the anolog/digital two track DD signals and have them processed by the Sc27 into 5.1 format or 7.1 format (tho not 5.1 discrete DD)- see page 62 of the manual - into Pro Logic IIx Movie, Neo:6 Cinema etc. which should compare favorably with what the Laserdisc with AC-3 puts out (5.1 DD).
Reply
Reply
post #19 of 46
3/9/10 at 11:37am
As I mentioned above many brands supported the laserdisc format in previous years, and certain brands even had built-in RF modulators like Denon (5600), Yamaha and Pioneer in their AVRs. Also they sold the RF modulator as a separate component such as the Yamaha APD-1. And frequently these can be found on Craig's List and E-bagger. Here is a link to an APD-1 now on the bag..

http://cgi.ebay.com/Yamaha-APD-1-AC-...item20afccc8a7

Just my $0.01..
Reply
Reply
post #20 of 46
3/9/10 at 12:59pm
I thought there were at least some upper model recent Denon AVR's that still had it no?

Would be cool to have HDMI full bore and the legacy AC3 input available for sure.
Reply
Reply
post #21 of 46
3/9/10 at 4:44pm
Quote:
Originally Posted by surdy View Post

???

The 5.1 DD discrete signal gets stripped, but you can use the anolog/digital two track DD signals and have them processed by the Sc27 into 5.1 format or 7.1 format (tho not 5.1 discrete DD)- see page 62 of the manual - into Pro Logic IIx Movie, Neo:6 Cinema etc. which should compare favorably with what the Laserdisc with AC-3 puts out (5.1 DD).

There is no "two track DD signals" on LD. The only Dolby Digital on LD is the 5.1 track that takes up the space normally occupied by the right analog channel.

You can play the remaining mono track from the left analog track, or take the PCM stereo track and do ProLogic or similar post-processing on that, but if that "compared favorably" with the 5.1 AC-3 track, they wouldn't have bothered putting AC-3 on the discs to begin with. It's still processed stereo, rather than discrete separate 5.1 channels.
Reply
Reply
post #22 of 46
3/9/10 at 7:50pm
Quote:
Originally Posted by RoboRay View Post

Yeah, the RF demodulators often go for $300+ on eBay.

I had a couple brand new B&K DT-1's that I couldn't give away a few years back. Wish I knew where they were now.
Reply
Reply
post #23 of 46
3/19/10 at 7:32am
Quote:


Studying the block diagram for the circuit given in this reference that includes a block diagram of the Pioneer PM4007A IC that does the heavy lifting, it is clear that this is not just a simple RF demodulator. It appears to use PLLs to demodulate the signal. It also has memory and appears to be reformatting the digital stream before relaying it. Also, the RF frequencies in the PLL circuit do not appear consistent with the wikipedia description (or some details in the circuit are not obvious to me from this diagram.)
Can someone enlighten me on the rf modulation scheme used and also the digital reformatting required?

The block diagram is actually not that bad. You're right that the RF is processed via a PLL-based demodulator but first the right channel on a 2.8 MHz carrier must be picked off with the band-pass filter (BPF, Q2/3) on the front end while the left channel on 2.3 MHz is discarded. There is also a DC level detector (Q4/5) for the mute function. After the RF signal is demodulated it needs to be time-base corrected, using the memory and timing circuitry in the demodulator chip, to remove any jitter caused by mechanical vibration in the player (the video signal undergoes a similar correction). Finally, the resulting bitstream is logically ORed with the coaxial input at IC8 (2/2), squared-up and driven out the jack with IC6.

Quote:


My current guess is that the modulation may be some sort of PSK with a 2.88 MHz carrier frequency.

LaserDisc uses pure FM modulation with no keying for all signals (ie. video, analog L/R, digital PCM). This is because, to very simplify things, what's physically stored on the disc itself is the zero-axis crossings of the blended FM carrier. Any phase-shift keying would be lost with this scheme. (During mastering, the FM signal is highly amplified, then hard clipped and level shifted to form a "digital" bitstream of sorts used to create the land and pit patterns on the disc.) While keying allows for more efficient data packing, it's not used to maintain compatability with LD's simple storage mechanism.

Quote:


Note that 46.08 MHz is just 16 x 2.88 MHz. Maybe someone can enlighten me further.

I wouldn't worry about the internal frequencies matching up with the 2.8 MHz signal carrying the AC-3 bitstream. Typically, PLLs and oscillators run much faster than the "working" frequencies in both analog and digital circuits for the exact reason you mention above.
Reply
Reply
post #24 of 46
3/19/10 at 7:53am
Listed em on the bay? I can see that demand has to be real low by now. I mean most of us that wanted it got them more than a decade ago and they still work fine.
Reply
Reply
post #25 of 46
3/19/10 at 9:14am
Quote:


The few real DTS discs have the DTS encoded the audio goes right out the normal optical digital audio output jack.

That's because the DTS bitstream was carried by the PCM used for the digital soundtrack rather than the original analog soundtracks.

Quote:


It can't take it from the two channel analog output. I mean it gets stripped before hitting the actual analog outputs of the LD player.

The analog outputs are 1) demodulated and 2) limited to 20 kHz so it's not surprising these cannot be used for RF signals.

Quote:


There is no "two track DD signals" on LD. The only Dolby Digital on LD is...occupied by the right analog channel.

At one time, there was talk about also converting the left analog channel into another DD 2.0 stream carrying a Dolby Surround encoded matrixed soundtrack. This would allow the three-soundtrack possibility (ie. DTS, DD and Dolby Surround) on one disc the same as with 35mm film. I'm sure politics and competition between Dolby and Digital Theater Systems put the kiabosh on this. Also, there would be no analog soundtrack left (even if only mono) for backwards compatability.
Reply
Reply
post #26 of 46
3/19/10 at 4:41pm
Quote:
Originally Posted by JOHNnDENVER View Post

Listed em on the bay? I can see that demand has to be real low by now. I mean most of us that wanted it got them more than a decade ago and they still work fine.

I believe the recent spike in demand over the last couple of years resulted from people replacing their older receivers or pre-pros containing integrated demodulators with HDMI models which, of course, lack RF demodulators. That's why I bought an RFD-1 two years ago when I finally replaced my Yamaha DSP-A1 with an Onkyo 885.
Reply
Reply
post #27 of 46
3/19/10 at 10:03pm
Thread Starter 
Quote:
Originally Posted by marcusm750 View Post

The block diagram is actually not that bad. You're right that the RF is processed via a PLL-based demodulator but first the right channel on a 2.8 MHz carrier must be picked off with the band-pass filter (BPF, Q2/3) on the front end while the left channel on 2.3 MHz is discarded. There is also a DC level detector (Q4/5) for the mute function. After the RF signal is demodulated it needs to be time-base corrected, using the memory and timing circuitry in the demodulator chip, to remove any jitter caused by mechanical vibration in the player (the video signal undergoes a similar correction). Finally, the resulting bitstream is logically ORed with the coaxial input at IC8 (2/2), squared-up and driven out the jack with IC6.



LaserDisc uses pure FM modulation with no keying for all signals (ie. video, analog L/R, digital PCM). This is because, to very simplify things, what's physically stored on the disc itself is the zero-axis crossings of the blended FM carrier. Any phase-shift keying would be lost with this scheme. (During mastering, the FM signal is highly amplified, then hard clipped and level shifted to form a "digital" bitstream of sorts used to create the land and pit patterns on the disc.) While keying allows for more efficient data packing, it's not used to maintain compatability with LD's simple storage mechanism.



I wouldn't worry about the internal frequencies matching up with the 2.8 MHz signal carrying the AC-3 bitstream. Typically, PLLs and oscillators run much faster than the "working" frequencies in both analog and digital circuits for the exact reason you mention above.

Thank you very much. This explanation makes sense.

There appear to be a few additional nuances of the circuit in that it may be phase locked to the start of frame and not just to carrier.

However the explanation of the FM modulation and the reason for the memory were the key pieces I was looking for.
Reply
Reply
post #28 of 46
3/19/10 at 10:24pm
Quote:
Originally Posted by toups View Post

Studying the block diagram for the circuit given in this reference that includes a block diagram of the Pioneer PM4007A IC that does the heavy lifting, it is clear that this is not just a simple RF demodulator. It appears to use PLLs to demodulate the signal. It also has memory and appears to be reformatting the digital stream before relaying it. Also, the RF frequencies in the PLL circuit do not appear consistent with the wikipedia description (or some details in the circuit are not obvious to me from this diagram.)

Can someone enlighten me on the rf modulation scheme used and also the digital reformatting required?

It's QPSK modulation. The Pioneer chip has error correction in it, so it's more complex than basic demodulation.
Reply
Reply
post #29 of 46
3/20/10 at 6:23am
Thread Starter 
Quote:
Originally Posted by Roger Dressler View Post

It's QPSK modulation. The Pioneer chip has error correction in it, so it's more complex than basic demodulation.

QPSK would explain using the 46.08 MHz (16 x 2.88 MHz) oscillator.
Reply
Reply
post #30 of 46
3/20/10 at 11:42am
Quote:


It's QPSK modulation. The Pioneer chip has error correction in it, so it's more complex than basic demodulation.

Yeah, you're right, the AC-3 (and PCM) tracks are quadrature phase-shift keyed. (Damn! It's been too long since I studied this in school! ) I still believe the error correction is just a time-base corrector for the mechanical jitter from the player. The BPF, mute level detector and the mixing with the coaxial input are correct.

Quote:


QPSK would explain using the 46.08 MHz (16 x 2.88 MHz) oscillator.

Yes it does. However, it still gets cut down for demodulating the fundamental 2.88 MHz carrier.

Quote:


There appear to be a few additional nuances of the circuit in that it may be phase locked to the start of frame and not just to carrier.

You're right, the TBC needs to know where the start of AC-3 frames begin just as the video TBC needs to know where the start of the field raster begins.

In any event, building one of these does not appear difficult if and only if one is able to find the original Pioneer PM4007 ASIC which has everthing built-in. Simply add the BPF, level detect, SRAM and a power supply to create an AC-3/RF demodulator.
Reply
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Receivers, Amps, and Processors

AVS Top Picks

AVS › AVS Forum › Audio › Receivers, Amps, and Processors › Laserdisc RF AC-3 Demodulator Question