Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › M2Tech HiFace Mods and Discussions
New Posts  All Forums:Forum Nav:

M2Tech HiFace Mods and Discussions - Page 2

post #16 of 247
6/21/10 at 2:15am

Here is Steve N's suggestion to fix the out of spec SPDIF output,  I'm going this route as it costs only $.50. The 470R resistor is a replacement for the stock 330R and the 1000R resistor is an addition.

 

 

steveN.jpg

Reply
post #17 of 247
6/21/10 at 3:21am

One last thing the 1.8V supply to the CPLD has to be tightly regulated their spect is +-1V.   This means a transformer and a linear or shunt reg,  meaning $'s and space,  no batteries will work here.

 

I can see doing this part of the mod if you are building a good 5V power supply for the USB receiver,  then you could just add a small reg to drop to 1.8V.  Big question is it worth the $'s ?

Reply
post #18 of 247
6/21/10 at 3:43am

deleted


Edited by jkeny - 9/17/10 at 1:20am
Reply
post #19 of 247
6/21/10 at 3:50am

deleted


Edited by jkeny - 9/17/10 at 1:21am
Reply
post #20 of 247
6/21/10 at 4:03am
 
Quote:
Originally Posted by jkeny View Post

 

By all means fix the SPDIF out to be within spec but not because the attenuators are a bad idea as I don't believe they are or because the Hiface does not follow the DIT4192 application guidelines, as it does follow them exactly.

 


Yes it does say Vio +2.7V to Vdd (max),  our Vdd is 5V+ but the Vio=3.3V so then output voltage is .8 VIO (in highlevel mode) and with the 2:1 pulse transformer we aren't even trying to meet SPDIF spec.  I hope you are right about the attenuators,  keep us posted.  I do think that a resistor divider (static pot) has more potential and is less expensive.  But opinions are like @-holes lets experiement and see what sounds best.
 


Edited by regal - 6/21/10 at 4:58am
Reply
post #21 of 247
6/21/10 at 4:17am

Here is the calc STOCK:  5V-.4V= 3.6V,    transformer voltage reduction= 1/2*3.6= .1.8V.  Output SPDIF = 1.8V

 

So doing Steve N's mod  3.6*470/(470+1000)*.5= 0.57V.   

 

Output impedance Steve N Mod 470*1000/(470+100)/4= 75 ohms  (assuming output impedance of DIT is negligible.)

 

Maybe is still a little high,  increase the 1000R to reduce further,  its the same principle as your attenuator idea just IMO more textbook and potentially better cause you don't have a wiper (read up on stepped attenuators vs potentiameters.)  Plus its cheaper!

 

 

Anyone know the package size of the little 330R resistor in series between TX and the pulse transformer?  I can't find my calipers.


Edited by regal - 6/30/10 at 4:55am
Reply
post #22 of 247
6/21/10 at 4:50am

deleted


Edited by jkeny - 9/17/10 at 1:22am
Reply
post #23 of 247
6/21/10 at 4:56am
Quote:
Originally Posted by jkeny View Post




Is the VDD on the DIT4192  5V (from the USB PS) & not 3.3V?

BTW, I believe that an independent, clean (battery) 3.3V supply to this Vio pin 7 could improve things more. So we would have three independent 3.3V supplies - one to the clocks, one to the CPLD clock handling output stage & one to the DIT4192 SPDIF generating output stage.

 



Yes I meant Vio  (post edited)  Vio is 3.3V from your battery CPLD mod. 

 

If they are fixed there should be no degredation.  So you're fixing it on the receiving end ,  I'm fixing it on the output side.  But its pretty similar.  You may want to try the resistor across the primary of the pulse transformer to determine the attenuation you need,  that way you only need to buy one attenuator (again I'm a cheap bastard.)

 


Edited by regal - 6/21/10 at 5:01am
Reply
post #24 of 247
6/21/10 at 5:18am

deleted


Edited by jkeny - 9/17/10 at 1:22am
Reply
post #25 of 247
6/21/10 at 5:20am

Bad news,

 

If you guys are concerned about proper 75 ohm transmission,   the stock pulse transformer is incorrect:

 

http://www.murata-ps.com/data/magnetics/kmp_786.pdf

 

See 78604/1C   .

 

The DIT datasheet calls for a Scientific conversions SC982.04 which would have the proper 75 ohms on the output.

http://www.scientificonversion.com/catalog.html

 

 

Don't know where we can buy these,  but me thinks the whole BNC 75 ohms quest is pointless with the stock pulse transformer.


Edited by regal - 6/21/10 at 7:59pm
Reply
post #26 of 247
6/21/10 at 5:35am
Quote:
Originally Posted by regal View Post

Bad news,

 

If you guys are concerned about proper 75 ohm transmission,   the stock pulse transformer is incorrect:

 

http://www.murata-ps.com/data/magnetics/kmp_786.pdf

 

See 78604/1C  primary is 2000 this means the sceondary is 500 ohm.

 

The DIT datasheet calls for a Scientific conversions SC982.04 which would have the proper 75 ohms on the output.

http://www.scientificonversion.com/catalog.html

 

 

Don't know where we can buy these,  but me thinks the whole BNC 75 ohms quest is pointless with the stock pulse transformer.


Yes Regal, this is what I was getting at in my post about the way I would do it - (in the methods below the signal level will be even higher, allowing for an even higher attenuator to be used.

 

Use a 1:1 transformer Newava 22083, I think is a highly regarded one

OR don't use an output transformer at all - these DIT4192 output driver (T+ & T-) lines are differential so galvanic isolation is still fine. I'm hoping they would be able to drive into one of these attenuators which would represent a mainly resistive 75R load

 

A third way & possibly the best & premium way is from a post of John Swenson who states:

 

Quote:
As a general purpose solution, I think the best approach would be a receiver that plugs in right on the output of the source with an adjustable input impedance that can be tuned to match the output as best as possible, followed by an amplifier and detector, which drives an LVDS transmitter, which goes over balanced cable to the receiver which is followed by an adjustable driver where the output level and impedance can be adjusted to get the best results out of the DAC receiver. With the input and output tuned to whatever equipment you have you would get close to as good as you can get.

Edited by jkeny - 6/21/10 at 6:13am
Reply
post #27 of 247
6/21/10 at 5:47am

Jenky you're always blazing a new trail and I admire this.  I just follow the classic methods (which Hiface failed to do.)   Keep us posted, the best transformer is no transformer so I really like you're #2 option, I  just don't have the background to say if it will work or not.

Reply
post #28 of 247
6/21/10 at 7:05am

A few months back I modified my USB cable to give the HiFace fresh 5V power from battery. Perhaps some improvement, but not much. 

 

Last Saturday I gathered the courage to "hack" my HiFace open for the 3.3V mod. I didn’t have the patient to wait for internet order, so I bought two 3.3V battery + charger from a local shop for $35.

 

WOW, best $35 upgrade ever!!!  Even though I’m using fairly low-end gear (Compass with RCA), for sure the result is a very clear improvement. Better detail and overall clearer sound from low to high. But best of all, more natural musical touch to the sound.

 

Thanks to jkeny for mod & pictures, and thanks to regal too (your recent posts in the other threads made me cross the line).

 

Now I’m very tempted to order DAC-19 and C-2....

 

My battery is 500mAh only. After approx 24h of use, voltage drop from 3.32 to 3.30. Anyone figured out the power consumption? How often do you need to recharge?

Reply
post #29 of 247
6/21/10 at 7:17am
Quote:
Originally Posted by ulfp View Post

A few months back I modified my USB cable to give the HiFace fresh 5V power from battery. Perhaps some improvement, but not much. 

 

Last Saturday I gathered the courage to "hack" my HiFace open for the 3.3V mod. I didn’t have the patient to wait for internet order, so I bought two 3.3V battery + charger from a local shop for $35.

 

WOW, best $35 upgrade ever!!!  Even though I’m using fairly low-end gear (Compass with RCA), for sure the result is a very clear improvement. Better detail and overall clearer sound from low to high. But best of all, more natural musical touch to the sound.

 

Thanks to jkeny for mod & pictures, and thanks to regal too (your recent posts in the other threads made me cross the line).

 

Now I’m very tempted to order DAC-19 and C-2....

 

My battery is 500mAh only. After approx 24h of use, voltage drop from 3.32 to 3.30. Anyone figured out the power consumption? How often do you need to recharge?



It sounds like you only did the clock mod?  I think the clocks use very little juice.  My 1500mAH battery lasted about 3 days (roughly)  but when they go they go fast then it is hard to recharge them.  Don't let the voltage go below 3V.  I have a feeling that the CPLD uses more mA's,  so you would need a bigger battery? (I haven't done that part yet.)

Reply
post #30 of 247
6/21/10 at 9:32am


 

Quote:
Originally Posted by regal View Post

 

It sounds like you only did the clock mod?  I think the clocks use very little juice.

 

Yes, I did the clock mod.

 

My 500mAH battery voltage is dropping faster now after 24h (about 24h effective playing over past 2 days). During first 24h, it was constant at 3.32V.

 

Are you disconnecting the battery when not using the HiFace, or do you simply keep it powered all the time?

 

Did anyone consider to put a few batteries in parallel to make them last longer, or is there some theory that parallel batteries would make it worse?

Reply
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › M2Tech HiFace Mods and Discussions