[Turniptom]

Hi all, I'm taking the leap and attempting to build the external dual optic data link for Mscaler/TT2 designed by @Reactcore.
He's been very helpful with questions about components and sourcing and thought it might be a good idea for me to post, encouraging any others who might be tempted to have a go.
I've always appreciated optical links for digital, smoother and darker in my system. I currently use two Oyaide DR-510 cables between my Mscaler/TT2. Good sound, but brighter with more top end zing than I'm used to.
I'm just a keen amateur with limited electronics experience/knowledge, and this project is on the limit of what I'd attempt to build. It's only due to Reactcore's generosity in sharing his design schematics and pcb files that I'm comfortable enough to make the attempt. So thanks to him and here's hoping I don't blow anything up!

P.S. the parts are all here and I've just started to populate the pcb.

 

[Reactcore]

Hi all, I'm taking the leap and attempting to build the external dual optic data link for Mscaler/TT2

Im sure you will succeed and get rewarded in the end

I will write an adjustment sequence if needed

 

[GuiltyRocker]

Hi all, I'm taking the leap and attempting to build the external dual optic data link for Mscaler/TT2 designed by @Reactcore.
He's been very helpful with questions about components and sourcing and thought it might be a good idea for me to post, encouraging any others who might be tempted to have a go.
I've always appreciated optical links for digital, smoother and darker in my system. I currently use two Oyaide DR-510 cables between my Mscaler/TT2. Good sound, but brighter with more top end zing than I'm used to.
I'm just a keen amateur with limited electronics experience/knowledge, and this project is on the limit of what I'd attempt to build. It's only due to Reactcore's generosity in sharing his design schematics and pcb files that I'm comfortable enough to make the attempt. So thanks to him and here's hoping I don't blow anything up!

P.S. the parts are all here and I've just started to populate the pcb.

Reactcore is a gentleman, I'm sure he will guide you and you'll be fine. Enjoy the tinkering!

 

[Turniptom]

Im sure you will succeed and get rewarded in the end

I will write an adjustment sequence if needed

Thanks, I'm sure that will be very useful indeed!

 

[Turniptom]

Reactcore is a gentleman, I'm sure he will guide you and you'll be fine. Enjoy the tinkering!

It is great fun and such a neat and clever design, thanks.

 

[Reactcore]

Thanks, I'm sure that will be very useful indeed!

Post some pictures of your progress

 

[Turniptom]

Post some pictures of your progress

Busy week this week, but will do so when I get the chance.

 

[Turniptom]

Pcb is ready to go.

Just about fits in the box. Will fit the bncs and dc connectors next and do all the off-board connections. Those tssop chips were a fiddle to solder! Good practice though.

 

[Reactcore]

Pcb is ready to go.

Just about fits in the box. Will fit the bncs and dc connectors next and do all the off-board connections. Those tssop chips were a fiddle to solder! Good practice though.

Nice job!!
Tip: screw the BNC chasis on TT2/scaler and make a scratch mark on the box after.

When testing use a series resistor as current limiter and feel the chips with a finger if theyre not heating up prior to fully power it..

 

[Turniptom]

Nice job!!
Tip: screw the BNC chasis on TT2/scaler and make a scratch mark on the box after.

When testing use a series resistor as current limiter and feel the chips with a finger if theyre not heating up prior to fully power it..

Thanks, pretty happy so far. Testing is where my ignorance kicks in. Would you suggest 5v and limiting current to 30/40 mA and testing at this stage? Presumably just looking for chip heat and the leds to light up?

 

[Reactcore]

As promised heres the adjustment description for the TX PCB stepwise.
This presumes the RX side is ready.
The voltage part applies for RX also ofc.

1. Before connecting the resistor regulator + wire, measure its output to be just under 5v using pot R5.
2. Turn all PCB pots counter clockwise to minimum setting.
3. Connect power and leave the data inputs not connected.
4. Connect + and measure the voltage again when powered, if its under 2v, feel the chips if theyre not warming up, (if they are there can be a short on the pins which can be solved using flux and reheating the pins).
5. If all is good regulate R5 for 5v output.
6. For both channels: NEVER turn pot R4 or both pots R2 and R3 clockwise to max position as this will overdrive and damage the LED! Pot R1 is safe in any setting.
7. To test the LED first turn R4 clockwise but not more than 75% or 3 quarter. The LED should lit up in red which can be seen as a small dot inside the connector. If it doesnt, theres something not right soldered or the led is faulty. (The led can be measured as a normal diode while unsoldered).
8. Turn R4 back to lowest, its normal if the led shows very low light residue.
9. Turn R2 and R3 to middle position, the led should remain off/very low.
10. Turn R1 clockwise until the led switches to on. This means the chip goes to logic '1' state. (You can steer the chip input bit by hand from 0 to 1 using R1)
11. Turn R1 counterclockwise until the led just goes off.
12. Connect one channel to DXdata BNC of Mscaler and set it to bypass (44.1k), connect your fiber cable too.
13. Leave R4 to min setting, set R2 to about 10%.
14. You might already get a link, even with R2 to R4 at minimum, if not turn R3 up until the DAC locks in. As expected the lower sample rates easier lock and work.
15. Once working increase the samplerate up to 1M (until 192k my sets were really easy and tolerant to work on most potmeter settings) On one channel you get stereo sound up to 192k, 1M uses 2 channels, one transmitting mono L and the other R at 352.8/384k making 705.6/768k stereo combined.
16. Now it gets more critical to lock, if it just works stable.. connect the second channel with the same R settings, if not make the 1st channel stable by slowly turning up R3 not further than 80%, then R2 also not further than 80% (R4 stays minimum until its might needed later)
17. If all goes normal it locks in at one R point and while turning further it unlocks again.. now find the middle of these ends.
18. If the 1st channel is good, do the same with the second and then connect both together to see if the DAC locks to dualdata (705.6/768k) It probably is flickering and making cricking sound meaning both channels are too far out of phase sync.
19. Now use R1 of either channel to phase shift the datastream. Also here you will find a middle setting on which the DAC locks to full scaling.
20. If it wont get stable.. turn R4 up a little (but not further than about half) and repeat the adjustments with the other potmeters.
21. It takes time and patience to optimally adjust the link.. you also might find after you play a 44.1 multiple at 705.6k and have it stable, and switch to a 48 multiple track at 768k its unstable on that.. youll have to find the middle of both optimum settings to reach the best result.

After the link adjustment is done.. its optimised for the fiber as connected.. if you swap fibers or connector sides it deviates from your optimum. So mark the fibers to remember best after a disconnect.

I hope this helps and brings you audio heaven!
i'm going back to mine

 

[Turniptom]

As promised heres the adjustment description for the TX PCB stepwise.
This presumes the RX side is ready.
The voltage part applies for RX also ofc.

1. Before connecting the resistor regulator + wire, measure its output to be just under 5v using pot R5.
2. Turn all PCB pots counter clockwise to minimum setting.
3. Connect power and leave the data inputs not connected.
4. Connect + and measure the voltage again when powered, if its under 2v, feel the chips if theyre not warming up, (if they are there can be a short on the pins which can be solved using flux and reheating the pins).
5. If all is good regulate R5 for 5v output.
6. For both channels: NEVER turn pot R4 or both pots R2 and R3 clockwise to max position as this will overdrive and damage the LED! Pot R1 is safe in any setting.
7. To test the LED first turn R4 clockwise but not more than 75% or 3 quarter. The LED should lit up in red which can be seen as a small dot inside the connector. If it doesnt, theres something not right soldered or the led is faulty. (The led can be measured as a normal diode while unsoldered).
8. Turn R4 back to lowest, its normal if the led shows very low light residue.
9. Turn R2 and R3 to middle position, the led should remain off/very low.
10. Turn R1 clockwise until the led switches to on. This means the chip goes to logic '1' state. (You can steer the chip input bit by hand from 0 to 1 using R1)
11. Turn R1 counterclockwise until the led just goes off.
12. Connect one channel to DXdata BNC of Mscaler and set it to bypass (44.1k), connect your fiber cable too.
13. Leave R4 to min setting, set R2 to about 10%.
14. You might already get a link, even with R2 to R4 at minimum, if not turn R3 up until the DAC locks in. As expected the lower sample rates easier lock and work.
15. Once working increase the samplerate up to 1M (until 192k my sets were really easy and tolerant to work on most potmeter settings) On one channel you get stereo sound up to 192k, 1M uses 2 channels, one transmitting mono L and the other R at 352.8/384k making 705.6/768k stereo combined.
16. Now it gets more critical to lock, if it just works stable.. connect the second channel with the same R settings, if not make the 1st channel stable by slowly turning up R3 not further than 80%, then R2 also not further than 80% (R4 stays minimum until its might needed later)
17. If all goes normal it locks in at one R point and while turning further it unlocks again.. now find the middle of these ends.
18. If the 1st channel is good, do the same with the second and then connect both together to see if the DAC locks to dualdata (705.6/768k) It probably is flickering and making cricking sound meaning both channels are too far out of phase sync.
19. Now use R1 of either channel to phase shift the datastream. Also here you will find a middle setting on which the DAC locks to full scaling.
20. If it wont get stable.. turn R4 up a little (but not further than about half) and repeat the adjustments with the other potmeters.
21. It takes time and patience to optimally adjust the link.. you also might find after you play a 44.1 multiple at 705.6k and have it stable, and switch to a 48 multiple track at 768k its unstable on that.. youll have to find the middle of both optimum settings to reach the best result.

After the link adjustment is done.. its optimised for the fiber as connected.. if you swap fibers or connector sides it deviates from your optimum. So mark the fibers to remember best after a disconnect.

I hope this helps and brings you audio heaven!
i'm going back to mine

Thanks very much Reactcore. I really appreciate the time you're putting in to help a dabbler like myself learn more, and hopefully build something i wouldn't have a hope of doing on my own. I'll digest all this and post some results when i can.

 

[Amberlamps]

Now playing a bit.. the unit gets hand warm which means the fpga is cooling itself on the housing..

Superb harsch free sound.. thick massive metal has good shielding.. also no porthole for RF to leak out

I noticed on one picture on the first page that the heatsink had drill marks all over it, is it still like that or is it a totally smooth surface ? Also, did you use any paste or liquid metal to make the heat transfer from fpga to heatsink/case easier and giving better cooling temps, or are the temps low enough as is.

 

[Amberlamps]

As promised heres the adjustment description for the TX PCB stepwise.
This presumes the RX side is ready.
The voltage part applies for RX also ofc.

1. Before connecting the resistor regulator + wire, measure its output to be just under 5v using pot R5.
2. Turn all PCB pots counter clockwise to minimum setting.
3. Connect power and leave the data inputs not connected.
4. Connect + and measure the voltage again when powered, if its under 2v, feel the chips if theyre not warming up, (if they are there can be a short on the pins which can be solved using flux and reheating the pins).
5. If all is good regulate R5 for 5v output.
6. For both channels: NEVER turn pot R4 or both pots R2 and R3 clockwise to max position as this will overdrive and damage the LED! Pot R1 is safe in any setting.
7. To test the LED first turn R4 clockwise but not more than 75% or 3 quarter. The LED should lit up in red which can be seen as a small dot inside the connector. If it doesnt, theres something not right soldered or the led is faulty. (The led can be measured as a normal diode while unsoldered).
8. Turn R4 back to lowest, its normal if the led shows very low light residue.
9. Turn R2 and R3 to middle position, the led should remain off/very low.
10. Turn R1 clockwise until the led switches to on. This means the chip goes to logic '1' state. (You can steer the chip input bit by hand from 0 to 1 using R1)
11. Turn R1 counterclockwise until the led just goes off.
12. Connect one channel to DXdata BNC of Mscaler and set it to bypass (44.1k), connect your fiber cable too.
13. Leave R4 to min setting, set R2 to about 10%.
14. You might already get a link, even with R2 to R4 at minimum, if not turn R3 up until the DAC locks in. As expected the lower sample rates easier lock and work.
15. Once working increase the samplerate up to 1M (until 192k my sets were really easy and tolerant to work on most potmeter settings) On one channel you get stereo sound up to 192k, 1M uses 2 channels, one transmitting mono L and the other R at 352.8/384k making 705.6/768k stereo combined.
16. Now it gets more critical to lock, if it just works stable.. connect the second channel with the same R settings, if not make the 1st channel stable by slowly turning up R3 not further than 80%, then R2 also not further than 80% (R4 stays minimum until its might needed later)
17. If all goes normal it locks in at one R point and while turning further it unlocks again.. now find the middle of these ends.
18. If the 1st channel is good, do the same with the second and then connect both together to see if the DAC locks to dualdata (705.6/768k) It probably is flickering and making cricking sound meaning both channels are too far out of phase sync.
19. Now use R1 of either channel to phase shift the datastream. Also here you will find a middle setting on which the DAC locks to full scaling.
20. If it wont get stable.. turn R4 up a little (but not further than about half) and repeat the adjustments with the other potmeters.
21. It takes time and patience to optimally adjust the link.. you also might find after you play a 44.1 multiple at 705.6k and have it stable, and switch to a 48 multiple track at 768k its unstable on that.. youll have to find the middle of both optimum settings to reach the best result.

After the link adjustment is done.. its optimised for the fiber as connected.. if you swap fibers or connector sides it deviates from your optimum. So mark the fibers to remember best after a disconnect.

I hope this helps and brings you audio heaven!
i'm going back to mine

The part where I have highlighted, I’ve wondered about that for years, does the mscaler give us full 768 on left and right channels or was it 384 on each to make a total of 768 combined. Cheers for breaking my heart on my birthday

Good work though.

 

[Reactcore]

I noticed on one picture on the first page that the heatsink had drill marks all over it, is it still like that or is it a totally smooth surface ? Also, did you use any paste or liquid metal to make the heat transfer from fpga to heatsink/case easier and giving better cooling temps, or are the temps low enough as is.

Theres paste between the housing heatblock and the FPGA. Just like in the original scaler housing.