Fibre optic - AT&T style (HP, Optek,...)

[darkfenriz]

Do you know any other (easy to use) SFP-alternative?

Try finding in "1X9" form. Some of them are TTL compatible, some are ECL.
Lots of cheap Chinese stuff, pretty simple to use.

 

[TJF]

Can you give me a link? Circuit?

Thanks 🙂

 

[TJF]

For the first I found things like this with simple circuits. May be you are right... 🙂 Still searching ...

 

[spzzzzkt]

And completely incompatible with existing ST gear which renders the exercise a waste of time imo

TJF I seem to remember you making the following comment in the other thread:

My problem: I have in all components ST-connectors.

 

[darkfenriz]

There are easily available SC/ST hybrid adatpers (passive female to female) as well as hybrid patchcords for the same price as ST/ST or SC/SC.

 

[TJF]

And completely incompatible with existing ST gear which renders the exercise a waste of time imo

Yes. But anyhow I must change the existing parts, because I don't know anything about this parts. There is a very, very simple circuit at this optic parts (looking similar to Optek products) and i am a little bit confused about that fact: This little circuits looks like the electronic for ODL 50 - but instead of this there could only be a LED (?) inside the metal-housing...

We have now a working solution with AN 1121. This different way won't probably work with my existing parts - or won't work the way, this new solution can do - used on both sides. Also 1121 recommends 62.5/125 µm cables. So my Belden 50/125 cable will be also changed ...

I will change everything now - for the best fibre optic solution. This moment I think it is on 1121. I am still looking to produce some PCBs.

But I am also inquisitive to darkfenriz' input ...

Hope you are still looking at that ... 🙂

 

[spzzzzkt]

I guess my project is slightly different. None of my gear is fitted with ST at the present, but I want a ST RX design that is fairly simple to include "on board" the Pass D1 clone I'm working on. The App Note 1121 circuit is probably the best option given readily available parts.

The HFBR-2416 is a $20US part from mouser or digikey, or you can pick up a NOS part from fleabay for half that 😉

The TX is separate issue and I'll probably put together something based on the wadia circuit George kindly posted, with the AN1121 TX circuit replacing the ODL50 part.

 

[nagaesan]

Optical Wavelength

Be carefull, optical wavelength AT&T ODL-50 SeriesII and hp HFBR14xx.

 

[nagaesan]

HFBR series Reciever Sensitivity using 50um fiver

Be carefull, optical fiver core size, when using 50um core fiver with HFBR series.
AT&T ODL50 II has AGC sercuit, it is no problem when using 50um fiver.
But hp is only LED, we must attach an outside circuit like hp's application note.

Now I produce ST-Link fiver cable, I recommend 62.5um core GI fiver cable.

Yahoo! auction(Japan)
http://page6.auctions.yahoo.co.jp/jp/auction/f63649490
http://page18.auctions.yahoo.co.jp/jp/auction/w29337056
http://page2.auctions.yahoo.co.jp/jp/auction/b88015016
http://page13.auctions.yahoo.co.jp/jp/auction/r50529598

 

[darkfenriz]

Hello, I here

First, 50/125 and 62.5/125 um are inter-compatible multimode fibers, 50/125 has generally lower dispersion and 62.5 has higher core diameter, which allows more light in from a LED. But the differences translate to around 10-20% of distance range which is probably in range of more than 1km for 100Mbit/s and much more for lower bitrate (bitrate*distance product is a main parameter for multi mode fibers).

Next, you can really use an encapsulated LED for MM fiber if it is close to 1st window transmission wavelength 850nm. CD player LED will work if you can do the encapsulation and put enough light into core.
However some 50/125 transmitters use a VCSEL laser, which is way better than any LED.

But for best results (lowest dispertion translating for lowest jitter) I would use singlemode fiber 9/125 and 1310nm wavelength fabry-perot laser transmitter capable of 155Mbit/s or more.
Also keep in mind that zero DC for laser is essential, because every laser transmitter has a monitor photodiode (to maintain thermal stability) which works as a (light) DC-servo.
I believe SPDIF already provides DC-balance, otherwise SPDIF transformers wouldn't work.

 

[georgehifi]

Hi darkfenriz, you seem to know your stuff, I use this ST Glass fiber lead, and it sounds better than the standard Wadia one.
I also use Indexing Gel at both ends

Cheers George

 

[georgehifi]

This is the only description I can find of the Wadia Glass lead

 

[georgehifi]

And here is Wadia's AT&T ODL50

Cheers George

 

[TJF]

@ nagaesan
@ georgehifi

Thank you both for the informations!

@ darkfenriz

Also 🙂 ... but could you please suggest concrete parts? I know -> you know...

 

[darkfenriz]

for example
http://www.laser2000.de/index.php?id=364073&L=0

 

[TJF]

Hi darkfreniz,

thank you. Funny: Laser2000 is near the place I live. Can go there by bike 🙂

Will have a look at these things ...

Thomas

 

[Zoran]

Hi, TJF
The chip before TX at EMM lab platyer is ACT11004...
...

 

[TJF]

Hi Zoran,

thank you! Not quite up to date - this part ... Like you said ...

Thomas

 

[TJF]

Hi,

today i had a meeting with a phycisist from Ratioplast-Optoelectronics Germany . Seems that the solution is near ... and I understand why my old parts are coming with this few electronic parts.

My "old unknown" parts are Ratioplast-Parts. Ratioplast builds LED and Laser (i.e from Optek, HP) into housings with electronics. They also have Tranceivers and more. Looking like the Optek-parts in metal-housing (see links above). There is an alternative of the same parts buildable on PCBs (like the Avagos).

The recommended parts for digital-audio 44,1 up to 192 kHz have this attributes:

Receiver:
Part No. 905 EM 850 ST 1Z3 (F-St with screws)
- 850 nm 25MBit/s
- with integrated Preamp and digital output, DC buffered
- no electronic is needed!

Transmitter:
Part No. 905 SE 850 ST 1Z1 (F-St with screws)
- LED 850 nm 50 MBit/s
- Very friendly circuit is needed

May be little count of parts will be only available from a distributor of Ratioplast. Who is interested could ask.

Because I have only a paper this moment, I will post the recommended schematic for the transmitter tomorrow.

Not recommended was the transceiver version over 155 MBit/s because of possible problems with our standard bitrate with less than 3 MBit/s (44,1 kHz). Only if using permanent upsampling on 176,4 kHz. Hope we don't discuss this theme 🙂

Thomas

 

[darkfenriz]

Maybe that's contrary to your hope, but as low as 1MHz can easily be transmitted via 1250Mbit/s channel. Have checked that in reality.

Good luck!