I will bring a thread from the Pass Labs forum to this place, where it belongs to (now 
).
Perhaps someone has an idea to bring this to an effective solution (used transmitter/receiver and not too extreme schematic). There do exist solutions with only 2-3 electronic parts...
Summing-up: Years ago there were these AT&T parts with ST-connectors (CTS 1361AAC) in some digital units like Wadia, ...
Actually there are i.e. these faster things (Optek, HP-Avago,...) on the market:
- HP with a rather wasteful schematic at figure 21 (for 1414 and 2418 - also from Optek)
- Optek OPF694-2 and OPF562
Optek recommends this more friendly schematic
Does anyone have an idea, how to go on ...
).Perhaps someone has an idea to bring this to an effective solution (used transmitter/receiver and not too extreme schematic). There do exist solutions with only 2-3 electronic parts...
Summing-up: Years ago there were these AT&T parts with ST-connectors (CTS 1361AAC) in some digital units like Wadia, ...
Actually there are i.e. these faster things (Optek, HP-Avago,...) on the market:
- HP with a rather wasteful schematic at figure 21 (for 1414 and 2418 - also from Optek)
- Optek OPF694-2 and OPF562
Optek recommends this more friendly schematic
Does anyone have an idea, how to go on ...
The Fig.21 schematic is a combined tx and rx with ECL logic, while the optek datasheet shows a TTL tx driver only. The HP App Bulletin has the same TTL TX driver shown as figure 8.
I've found a slightly more appropriate RX circuit in another HP datasheet designed for lower speed operation but should still be more than adequate for audio. It uses a 2416 RX module and a single LT1016 comparator. Ignore the HFBR-25x6 as that can be left out.
See Figure 4.
www.avagotech.com/docs/AV02-0723EN
I've found a slightly more appropriate RX circuit in another HP datasheet designed for lower speed operation but should still be more than adequate for audio. It uses a 2416 RX module and a single LT1016 comparator. Ignore the HFBR-25x6 as that can be left out.
See Figure 4.
www.avagotech.com/docs/AV02-0723EN
this is from the
EMM lab CDSD model player
aglient HFBR-1414T tx and HFBR-2416T rx
.
good link
http://www.avagotech.com/pages/en/o...ring/hfbr-1414/
EMM lab CDSD model player
aglient HFBR-1414T tx and HFBR-2416T rx
.
good link
http://www.avagotech.com/pages/en/o...ring/hfbr-1414/
Attachments
Yes Zoran. Thank you! That are exactly the things (Emitter and Receiver) we want to use. But please read in spzzzzkts -link (post #2) about "Arbitrary Duty Factor" and have a look at Figure 4. May be EMM is using any kind of "normal" circuit ... ?
Your link doesn't work. I think it should be this
Your link doesn't work. I think it should be this
Hi Zoran,
I agree that 1414/2416 pair is the way to go and have ordered one of each to try out. The older ODL 50 series II used in the wadia, pass d1, etc would be ideal - it's easy to implement but is extremely hard to get a hold of as appears to have gone out of production by around 2000. A couple of brokers list it as stocked but they don't bother replying to enquires on "diyaudio" quantities.
The driver circuit for the 1414 given in the HBRF-0400 series datasheet is fine for our purposes. Unfortunately there is no real detail given on RX circuits.
Application Bulletin 78 (linked in TJF's first post) initally looked promising but the rx circuits contained in it are designed for 155MBd operation over long distances, and utilise ECL logic with conversion to TTL levels.
Application Note 1121 (link in my post above) seems more appropriate to audio transmission speeds and the rx circuits are designed for lower speeds - up to 32MBd - and operate at TTL levels.
An observation on the drivers for the TX - the recommend circuit uses a 74ATCQ00. It's important to make the distinction as ACT logic is widely regarded as inappropriate for audio use due to it being quite noisy - a search on the archives will turn up plenty of warnings not to use ACT chips. The ACTQ series is based on FACT Quiet logic and is apparently a significant improvement over the older ACT logic design.
cheers
Paul
I agree that 1414/2416 pair is the way to go and have ordered one of each to try out. The older ODL 50 series II used in the wadia, pass d1, etc would be ideal - it's easy to implement but is extremely hard to get a hold of as appears to have gone out of production by around 2000. A couple of brokers list it as stocked but they don't bother replying to enquires on "diyaudio" quantities.
The driver circuit for the 1414 given in the HBRF-0400 series datasheet is fine for our purposes. Unfortunately there is no real detail given on RX circuits.
Application Bulletin 78 (linked in TJF's first post) initally looked promising but the rx circuits contained in it are designed for 155MBd operation over long distances, and utilise ECL logic with conversion to TTL levels.
Application Note 1121 (link in my post above) seems more appropriate to audio transmission speeds and the rx circuits are designed for lower speeds - up to 32MBd - and operate at TTL levels.
An observation on the drivers for the TX - the recommend circuit uses a 74ATCQ00. It's important to make the distinction as ACT logic is widely regarded as inappropriate for audio use due to it being quite noisy - a search on the archives will turn up plenty of warnings not to use ACT chips. The ACTQ series is based on FACT Quiet logic and is apparently a significant improvement over the older ACT logic design.
cheers
Paul
What are you going to use it for?
Remember that you may need some channel encoding for zero DC and enough signal transitions for clock recovery to work.
I suggest you use some of cheap SFP transceivers, they are easiest to use or if you find some "1x9" transceivers.
Ps. These optek transceivers look great, this would be FAR better than any plastic fiber connection.
My note on zero DC applies only to laser transmitters, also I think if you stick to SPDIF you don't need to worry about transitions frequency, because SPDIF clock/data recovery already knows how to recover it.
Remember that you may need some channel encoding for zero DC and enough signal transitions for clock recovery to work.
I suggest you use some of cheap SFP transceivers, they are easiest to use or if you find some "1x9" transceivers.
Ps. These optek transceivers look great, this would be FAR better than any plastic fiber connection.
My note on zero DC applies only to laser transmitters, also I think if you stick to SPDIF you don't need to worry about transitions frequency, because SPDIF clock/data recovery already knows how to recover it.
In case of Finisar SFP i.e. look at this reference circuit . Didn't find the FDB-1018 Spec for Finisar FTLF1424P2 this moment. Do you know any other (easy to use) SFP-alternative?
You can get the Avagos with metal-housing. Used are glass-cables (62.5/125 µm, multimode glass) ... What concrete types from Optek do you mean? I couldn't find any good working combination like the Avagos. But I had also difficulties, because the Optek-site is not so clear... and I didn't find any circuit working like the asynchronous way in the Avago Bulletin 1121.
TJF said:
Yes, you are right! I was getting a bit over-excited. There is some kind of ACT chip used however
The main criteria seems to be the higher output current of the ACT(Q) logic, so it might be possible to adapt the CEC T0 schematic and use 74ACTQ74's to reclock data. The 32MBd looks perfect - with manchester encoding it gives a 16Mhz data rate so plently of headroom.DF,
The aim is to find an ST compatible RX and TX. The up to 2000 it seems the ATT/CTS ODL 50 series II pair was used by most high end manufacturers but these are now extremely hard to source. Schematics posted to the ShigaClone thread by Peter Daniels are an example of the type of circuit being used.
http://www.diyaudio.com/forums/showthread.php?postid=1743945#post1743945
The current crop of RX modules which have sufficient bandwidth no longer feature the TTL outputs of the ODL 50, so require additional external gain. So we've been looking for a suitable circuit in the app notes. The circuit in AN 1121 appears to fulfill the requirements. The TX would be driven from the same data as the SPDIF outputs.
ps: the ic's near the TX modules look like they might be ACT114 - which would be dual JK flip flops.
I think that EMM lab CDSD
did not use any of fancy chips to drive...
I shall try to find and post out a bigger pict.
*
(we opened the damn thing to explore fault of transport mech
philips one cheep VAL used in some philips DVD players sacd style...
EMM put the laser in constant operation, so it died very fast, and after that tryed to charge 500usd w/o posting, to fix the problem... changed only lasr diode, now working good... Also PS is from chopper fashion... And pretty emputy inside... dissapointing from that amount of money)
*
cheers
did not use any of fancy chips to drive...
I shall try to find and post out a bigger pict.
*
(we opened the damn thing to explore fault of transport mech
philips one cheep VAL used in some philips DVD players sacd style...
EMM put the laser in constant operation, so it died very fast, and after that tryed to charge 500usd w/o posting, to fix the problem... changed only lasr diode, now working good... Also PS is from chopper fashion... And pretty emputy inside... dissapointing from that amount of money)
*
cheers
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.