M2Tech HiFace Mods and Discussions - Page 3

Here's the transformer that is recommended by Jocko - Newava s22083 http://parts.digikey.ca/1/1/324876-transformer-isolation-dig-audio-s22083.html

It's 1:1 so output stage has to be changed to bring down high level voltage

Current consumption:

Xilinx typically 30mA max,

Cypress current consumption 85mA max

Clock consumption 20mA?

Total current consumption 115mA for 3.3V

Quote:

Originally Posted by jkeny 

Here's the transformer that is recommended by Jocko - Newava s22083 http://parts.digikey.ca/1/1/324876-transformer-isolation-dig-audio-s22083.html

It's 1:1 so output stage has to be changed to bring down high level voltage

 

 

Current consumption:

 

Xilinx typically 30mA max,

 

Cypress current consumption 85mA max

 

Clock consumption 20mA?

 

Total current consumption 115mA for 3.3V

Nothing against Jocko but the datasheet doesn't show a 75 ohm secondary,   I emailed the manufacturer for clarification,  as we need to know what the impedance of the secondary is in order to spec the resistor divider ahead of it.

I think we can make the stock pulse transformer work if we use a 150 ohm divider roughly 75ohm in series and 75 ohms across the primary,   the secondary impedance is so high that it won't matter in the calculation.

Regal,

I don't think you need a 75ohm secondary - just use a resistor divider to bring it down to 75ohm.

Here's a schematic for double termination on primary & secondary & includes a 10db attenuator using the Newava 1:1 transformer

Yes and we can use the same technique to "fix" the stock pulse transformer.
 

Quote:

Originally Posted by jkeny 

Regal,

I don't think you need a 75ohm secondary - just use a resistor divider to bring it down to 75ohm.

 

Here's a schematic for double termination on primary & secondary & includes a 10db attenuator using the Newava 1:1 transformer

 

Quote:

Originally Posted by regal 

Yes and we can use the same technique to "fix" the stock pulse transformer.
 

 

I'm not sure as I've seen somewhere that the stock one is so far off the 75ohm that it can't be successfully brought down to 75 ohm impedance - you might know more about this than me though?

BTW, you'll need to correct that calculation on the last page 3.3V * 0.8 is correct for all digital lines on the DIT4192 EXCEPT T+ & T- line drivers which gives us the SPDIF signal where it's 3.3V - 0.7 I know this doesn't make much of a difference but for the sake of clarity if anybody is checking it!

Quote:

Originally Posted by jkeny 

I'm not sure as I've seen somewhere that the stock one is so far off the 75ohm that it can't be successfully brought down to 75 ohm impedance - you might know more about this than me though?

 

BTW, you'll need to correct that calculation on the last page 3.3V * 0.8 is correct for all digital lines on the DIT4192 EXCEPT T+ & T- line drivers which gives us the SPDIF signal where it's 3.3V - 0.7 I know this doesn't make much of a difference but for the sake of clarity if anybody is checking it!

Actually the datasheet says T+ is 5V-.4V,  Steve N's Mod comes out to right into the SPDIF spec (I edited the post.)

As far as the 75 ohm,  this transformer is 4:1 impedance ratio,  so assuming that the secondary impedance is very high if in paralell with 75 ohms it would be negligable.  So we would want 75x4=300 on the primary side.  So to get this we would want 75 in series to and 225 across the primary to give the 75 ohms at the secondary and give the proper voltage divider(reduction.)   I'm not 100% on this

 (used to dealing with tube transformers) so I emailed the maker of the stock transformer and will report back.

Here is a thread that explains how to get a proper 75ohm output from a transport (my last post is correct so I'm ordering from digikey):

http://www.head-fi.org/forum/thread/129132/adding-coaxial-s-pdif-to-the-m-audio-transit

Also read this for the calculatios (series/parallel section)

http://en.wikipedia.org/wiki/Resistor

Quote:

Originally Posted by regal 

 

Actually the datasheet says T+ is 5V-.4V,  Steve N's Mod comes out to right into the SPDIF spec (I edited the post.)

 

Yes you're correct - phew, eventually we got there :)

 

 

As far as the 75 ohm,  this transformer is 4:1 impedance ratio,  so assuming that the secondary impedance is very high if in paralell with 75 ohms it would be negligable.  So we would want 75x4=300 on the primary side.  So to get this we would want 75 in series to and 225 across the primary to give the 75 ohms at the secondary and give the proper voltage divider(reduction.)   I'm not 100% on this

 

 (used to dealing with tube transformers) so I emailed the maker of the stock transformer and will report back.

Quote:

Originally Posted by regal 

Here is a thread that explains how to get a proper 75ohm output from a transport (my last post is correct so I'm ordering from digikey):

 

http://www.head-fi.org/forum/thread/129132/adding-coaxial-s-pdif-to-the-m-audio-transit

 

 

Also read this for the calculatios (series/parallel section)

 

http://en.wikipedia.org/wiki/Resistor

 

 

Excellent, we only need the output driver impedance of DIT4192 & then we are good to go. For the output drivers, I've calculated it at 13.3ohms for High & 23.3ohms for Low so an average of 18ohms & maybe in the absence of actual measurements, these will do.

This 18ohms needs to be taken into account in the calcs. Does this mean our 75R in series on the primary side becomes:

75-18=57ohm?

Quote:

Originally Posted by jkeny 

Excellent, we only need the output driver impedance of DIT4192 & then we are good to go. For the output drivers, I've calculated it at 13.3ohms for High & 23.3ohms for Low so an average of 18ohms & maybe in the absence of actual measurements, these will do.

 

This 18ohms needs to be taken into account in the calcs. Does this mean our 75R in series on the primary side becomes:

75-18=57ohm?

I'm surprised I assumed that the DIT4192 with its drivers would have an output impedance under 1-2ohms,  the link I provided was for a transmitter that didn't have a driver/buffer.   Guess I need to look at the datasheet again (thanks btw almost placed a digikey order!)

For a cheap but decent +5V PS here is one that folks can build on a radioshack breadboard:

http://hubpages.com/hub/zpedia

I have one already built that I paid $5 for off ebay a few years ago,  I couldn't find it for sale anymore,  but there are LM317 boards on ebay that are probably of better quality.  For a transformer a 9-12V from digikey will work.

To get the 1.8V I'm using this:  http://search.digikey.com/scripts/DkSearch/dksus.dll?WT.z_header=search_go&lang=en&site=us&keywords=MCP1700-1802E%2FTO-ND&x=24&y=14 after the +5V power.

On a side note, I tried to do some testing yesterday with the Hiface output stage. I picked up a TASCAM MD-801R MinidisK professional recorder/player that was being thrown out to use for my testing as it has some nice digital readouts of SPDIF strength. I can'get a lock on the Hiface SPDIF out. The Tascam works with every other SPDIF source I throw at it. Any ideas what might be wrong? I've noticed this before in other situations but wasn't familiar with the equipment to make a judgement as to which was at fault. I think I got a  lock for a second when i was switching though the various settings on the Tascam as I heard music in the headphones plugged into it.

Any experience of this?

I know the Hiface is also non-standard in it's SPDIF coding output from this note I found on the Australia M2tech site:

This would hardly cause non-locking issues would it?

I sent an email to Marco to ask about it & if it would be fixed for the Hiface too but apparently not! I have a follow up mail to him

Well I can tell you that I discovered that my version has the DIT4192 set to low output,  so 0.4V before the pulse transformer and only 0.2V after it.  So they have been making changes.  They should have changed the output transformer to a 1:1 but didn't.

Quote:

Originally Posted by regal 

Well I can tell you that I discovered that my version has the DIT4192 set to low output,  so 0.4V before the pulse transformer and only 0.2V after it.  So they have been making changes.  They should have changed the output transformer to a 1:1 but didn't.

Are you sure? This seems very odd. What's the Vio voltage on pin 7? This is probably too low for a lock signal or would severely compromise the SPDIF receiver operation, I think?

This is not my problem as I scoped the output & it's about 2Vpp

Pin 7 Vio is 3.3 which is in spec.   Edit it looks like Steve N's mod of the 1k R across the transformer may have been too much of a load and damaged the DIT4192 TX drivers.  Luckily the DIT4192 is only $6 at digikey.

Quote:

Originally Posted by regal 

Pin 7 Vio is 3.3 which is in spec.   Edit it looks like Steve N's mod of the 1k R across the transformer may have been too much of a load and damaged the DIT4192 TX drivers.  Luckily the DIT4192 is only $6 at digikey.

Oops, that explains it - I don't envy you trying to unsolder & resolder that DIT4192 - it's tight in there!

I have this low temp solder that makes removal easy,  the soldering back on is easy.

Quote:

Originally Posted by jkeny 

Oops, that explains it - I don't envy you trying to unsolder & resolder that DIT4192 - it's tight in there!