SPDIF balanced output

jkeny

Banned
2007-02-06 12:43 am
Dublin
I want to investigate the possibility of using a differential SPDIF line driver on the output of a device to:
- eliminate the need for a SPDIF output transformer as these require all sorts of contortion to get them right & I don't have the equipment or time or knowledge to do so.
- provide a purely resistive 75ohm load to the DIT for better impedance matching rather than the inductive/reactive load that a SPDIF transformer presents.

I'm using a DIT4196 Digital transmitter which has differential output line drivers. I'm not talking about AES, 110ohm balanced, I still want to use existing SPDIF 75 ohm cabling & the DAC will have standard SPDIF input circuitry - transformer receiver, etc. Would I lose out on the common mode noise rejection if not having twisted pair? How much of a drawback would this be?

Firstly, why isn't this done more often? Am I missing something? I know probably grounding issues would be a problem if not using a balanced driver but in this case are there other gotchas?

I know there are T-pad calculators for working out the resistive network needed to apply some attenuation & keep it 75ohm but they apply to single ended output - how do I adjust for balanced output?

Would a resistive divider be too much load for the DIT4196 line drivers? The datasheet specifies their characteristics as being able to output from min 2.6V to max 3.3V at 30mA when powered from 3.3V
 
Last edited:
Reasons you don't see it more often:

- active solution is at least as complex as passive; if you don't have equipment/experience to check a passive solution, you probably are in the same situation with an active one;

- a xformer with a resistive termination on the secondary looks like a resistor to the DIT, difficult to get it simpler and more reliable...

jd
 

stormsonic

Member
2008-08-31 4:08 pm
I know there are T-pad calculators for working out the resistive network needed to apply some attenuation & keep it 75ohm but they apply to single ended output - how do I adjust for balanced output?

Would a resistive divider be too much load for the DIT4196 line drivers? The datasheet specifies their characteristics as being able to output from min 2.6V to max 3.3V at 30mA when powered from 3.3V

Take a piece of paper, write down letter T. Then write down another letter T over previous, but turned upside down. After that rotate this piece of paper for 90 degrees, until you see letter H. This is what you want :D

LINK

Google for H pad attenuator or use this link H pad calculator.

As Jocko pointed out several times, you will have to look at datasheet for your device to find out output Z & use it when calculating resistor values.

For output loading just use Ohm law. If your device output 3.3V, with 100R resistor to ground there will be 33mA current through this resistor. If max. output is 30mA, you are overloading it and need to increase resistor to decrease current.
 
Last edited:

jkeny

Banned
2007-02-06 12:43 am
Dublin
Reasons you don't see it more often:

- active solution is at least as complex as passive; if you don't have equipment/experience to check a passive solution, you probably are in the same situation with an active one;
I'm not talking about an active stage - just a resistive divider Pi-pad (or H-pad) on the differential output line drivers of the DIT - I don't see why I need equipment for this, I've already calculated the output impedance of the DIT so this will be taken into account (more than most SPDIF transformer output stages do!)

- a xformer with a resistive termination on the secondary looks like a resistor to the DIT, difficult to get it simpler and more reliable...

jd
AFAIK, I think this is way too simplistic a statement :
- firstly you are assuming that the transformer can handle the frequencies involved here in a completely transparent manner & pass the signal unadulterated?
- secondly transformers have an inductive kick in them - this has to be catered for, I believe
- there are probably other issues with them
 

jkeny

Banned
2007-02-06 12:43 am
Dublin
Take a piece of paper, write down letter T. Then write down another letter T over previous, but turned upside down. After that rotate this piece of paper for 90 degrees, until you see letter H. This is what you want :D

Google for H pad attenuator or use this link H pad calculator.
Yes, stormsonic, I was posting ast the same time as you & figured that it must be a H-pad.

As Jocko pointed out several times, you will have to look at datasheet for your device to find out output Z & use it when calculating resistor values.

For output loading just use Ohm law. If your device output 3.3V, with 100R resistor to ground there will be 33mA current through this resistor. If max. output is 30mA, you are overloading it and need to increase resistor to decrease current.
Great, need to do some experiments so!

Are there any issues I'm overlooking? Like the DAC end of the SPDIF cable will be treated as normal (i.e not balanced) - is the coaxial outer conductor normally connected to ground on SPDIF input? = ground issue?

Why isn't this done more often?
 
Last edited:
I'm not talking about an active stage - just a resistive divider Pi-pad (or H-pad) on the differential output line drivers of the DIT - [snip]

OK, I probably misread you there. I thought you wanted an active solution.

[snip]- firstly you are assuming that the transformer can handle the frequencies involved here in a completely transparent manner & pass the signal unadulterated?
- secondly transformers have an inductive kick in them - this has to be catered for, I believe
- there are probably other issues with them

Well, the xformer would be designed for that freq range, wouldn't it?
'inductive kick', 'other issues' - possibly.
One extremely important advantage of the xformer is the galvanic isolation - no problems with ground currents and such that can lead to jitter.

jd
 

stormsonic

Member
2008-08-31 4:08 pm
jkeny said:
I'm using a DIT4196 Digital transmitter which has differential output line drivers. I'm not talking about AES, 110ohm balanced, I still want to use existing SPDIF 75 ohm cabling & the DAC will have standard SPDIF input circuitry - transformer receiver, etc.

Depends on your circuit. CS841x receiver chip will accept differential, DIR9001 only SE. Look at datasheet for your chip.
But you wrote your DAC will have transformer coupled input, why worry about ground?

jkeny said:
Why isn't this done more often?

What should be done more often?

jkeny said:
Are there any issues I'm overlooking?

Yes. Put active circuit infront your receiver chip. Jocko posted several schematics here and at Diyhifi.
 

jkeny

Banned
2007-02-06 12:43 am
Dublin
Depends on your circuit. CS841x receiver chip will accept differential, DIR9001 only SE. Look at datasheet for your chip.
But you wrote your DAC will have transformer coupled input, why worry about ground?
Yep, you're right



What should be done more often?
Dispensing with SPDIF transformer (at least one of them: transmitting or receiving?). It would seem to be so much easier (maybe I'll find out it isn't)? Allows the transmitter to see a resistive load, easier to keep impedance at 75ohm & no need for taming non-ideal transformers.



Yes. Put active circuit infront your receiver chip. Jocko posted several schematics here and at Diyhifi.
Yes, I remember this mentioned before - a nice linear front-end - I'll search!
 
Last edited:

jkeny

Banned
2007-02-06 12:43 am
Dublin
Moving on - Another question guys:
I'm looking to work out the impedance of the output stage using a multimeter.
Here's my proposed technique:
- measure the voltage out when the outputting digital silence of the output pins without any caps in-line
- using a multimeter across the pins adjust a potentiometer also connected across the pins until half the voltage is read across the pins
- the reading of the potentiometer is approx the impedance of the output stage
- better than guessing?

Will this work?
 
Last edited:

jkeny

Banned
2007-02-06 12:43 am
Dublin
Well if I want to try and get a 75ohm impedance match on the line which do I need to take into account the output stage impedance? What is the difference with Zout? I need to add this to a first series resistor that I'll be using as part of a 75ohm H-pad resistive network attenuator - is this correct?

Edit: Sorry, I re-read my post & see that I left out one important bit - I'm using a potentiometer as an adjustable R across the output pins & so acting as a voltage divider, doh

I've adjusted the post above but repeat it here:
Here's my proposed technique:
- measure the voltage out across when the outputting digital silence of the output pins without any caps in-line
- using a multimeter & potentiometer across the pins adjust the potentiometer until half the voltage is read
- the reading of the potentiometer is approx the impedance of the output stage

Does the output stage being balanced cause any problem here?
 
Last edited:

stormsonic

Member
2008-08-31 4:08 pm
Huh, there is a long way infront of you :)

Start with Zout for your chip.
Look into datasheet and search for VOH and VOL parameters (Voltage Output Low and Voltage Output High).
Take into consideration your power supply voltage (Vcc) and look into IOH and IOL values (output current).
You have voltage, you have current, use Ohm law to calculate resistance.
There will be different resistance for VOH and VOL.
Sum those 2 values together and divide with 2 to get average value.
 

jkeny

Banned
2007-02-06 12:43 am
Dublin
I've done this already but get crazy figures:
High-Level Output Voltage, VOH IO = –30mA VDD–0.7(min) VDD–0.4(typ) VDD(Max)
Low-Level Output Voltage, VOL IO = +30mA 0(min) 0.4(typ) 0.7(max)

I (mA) V R (ohm)
30 4.6 153.333333 for High
30 0.4 13.333333 for Low

So 83 ohm but this seems too high, no? I think I heard 20 to 30 mentioned somewhere as a reasonable figure?

What's wrong with my technique above?
 
Last edited:

jkeny

Banned
2007-02-06 12:43 am
Dublin
I'm sorry again the posting of my method is messed up (does anybody find this frustration with trying to type on a laptop that has a touchpad - my thumb seems to activate this touch pad & cause the cursor to jump somewhere while in the middle of typing causing gibberish as above)

It should read:
Here's my proposed technique:
- measure the voltage out across the output pins when outputting digital silence without any caps in-line
- using a multimeter & potentiometer across the pins adjust the potentiometer until half that voltage is now read
- the reading of the potentiometer is approx the impedance of the output stage

Does the output stage being balanced cause any problem here?