SPDIF to AES/EBU

[Charles Robertson, Psy.D.]

Does anyone have the circuit for an adapter between these formats? I have
seen a little in-line adapter and am wondering what's inside.

[Artie Turner]

Charles Robertson, Psy.D. wrote:
> Does anyone have the circuit for an adapter between these formats? I have
> seen a little in-line adapter and am wondering what's inside.
>
>
It's a simple transformer - 75 ohm unbalanced to 110 ohm balanced.

Artie Turner

[Scott Dorsey]

Charles Robertson, Psy.D. > wrote:
>Does anyone have the circuit for an adapter between these formats? I have
>seen a little in-line adapter and am wondering what's inside.

It's a transformer. And all it does it change the levels and impedance, it
doesn't alter the subcode.

I think the Schott Transformers folks have schematics for these things in
their catalogue and they might be on the website too.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[rich rookie]

"Artie Turner" > wrote in message
. ..
> Charles Robertson, Psy.D. wrote:
> > Does anyone have the circuit for an adapter between these formats? I
have
> > seen a little in-line adapter and am wondering what's inside.
> >
> >
> It's a simple transformer - 75 ohm unbalanced to 110 ohm balanced.
>
> Artie Turner
>
Isn't there a little more though. AES/EBU has no SCMS, it has some other
information in there.

S/P-DIF to AES/EBU works fine. AES/EBU to S/P-DIF will sometimes fail
because the S/P-DIF unit may interpret some bits as copy protection. That is
if the unit incorporates SCMS.

Richard

[Artie Turner]

rich rookie wrote:
> "Artie Turner" > wrote in message
> . ..
>
>>Charles Robertson, Psy.D. wrote:
>>
>>>Does anyone have the circuit for an adapter between these formats? I
>
> have
>
>>>seen a little in-line adapter and am wondering what's inside.
>>>
>>>
>>
>>It's a simple transformer - 75 ohm unbalanced to 110 ohm balanced.
>>
>>Artie Turner
>>
>
> Isn't there a little more though. AES/EBU has no SCMS, it has some other
> information in there.

The guy asked for a circuit, for what was inside, and it really is just
a transformer. You are right that the SPDIF subcode is not compatible
with AES/EBU, but the audio data is essentially the same.
>
> S/P-DIF to AES/EBU works fine. AES/EBU to S/P-DIF will sometimes fail
> because the S/P-DIF unit may interpret some bits as copy protection. That is
> if the unit incorporates SCMS.

That could be, but I've never seen it happen, and I transfer stuff to
and from my old SPDIF DAT recorder and my AES/EBU soundcard with a
Canare BULUN transformer. Works like a champ.



AT
>
> Richard
>
>

[rich rookie]

"Artie Turner" > wrote in message
. ..
> rich rookie wrote:
> > "Artie Turner" > wrote in message
> > . ..
> >
> >>Charles Robertson, Psy.D. wrote:
> >>
> >>>Does anyone have the circuit for an adapter between these formats? I
> >
> > have
> >
> >>>seen a little in-line adapter and am wondering what's inside.
> >>>
> >>>
> >>
> >>It's a simple transformer - 75 ohm unbalanced to 110 ohm balanced.
> >>
> >>Artie Turner
> >>
> >
> > Isn't there a little more though. AES/EBU has no SCMS, it has some other
> > information in there.
>
> The guy asked for a circuit, for what was inside, and it really is just
> a transformer. You are right that the SPDIF subcode is not compatible
> with AES/EBU, but the audio data is essentially the same.
> >
> > S/P-DIF to AES/EBU works fine. AES/EBU to S/P-DIF will sometimes fail
> > because the S/P-DIF unit may interpret some bits as copy protection.
That is
> > if the unit incorporates SCMS.
>
> That could be, but I've never seen it happen, and I transfer stuff to
> and from my old SPDIF DAT recorder and my AES/EBU soundcard with a
> Canare BULUN transformer. Works like a champ.
>
Oh, I agree. I've actually not seen it happen either. And I also have
transferred a lot of material between my S/P-DIF DAT and AES/EBU equipment.
It turns out I need to use a logic level shifter, so a simple transformer
won't work for me. But that's just the DAT manufacturer's fault.

It could happen and if you're not aware of the potential, it could drive you
crazy.

Richard

[Preben Friis]

"Charles Robertson, Psy.D." > wrote in message
...
> Does anyone have the circuit for an adapter between these formats? I have
> seen a little in-line adapter and am wondering what's inside.

http://www.rane.com/note149.html

/Preben Friis

[Dan Maas]

Just a note - before getting into anything fancy, you could try simply
using an RCA->XLR cable. Of course this ignores all the issues of
voltage, impedance, signal format, etc... But in some cases it works.
(I hook up my Delta sound card to a DAT deck this way).

Dan

[Arny Krueger]

"Charles Robertson, Psy.D." > wrote in
message

> Does anyone have the circuit for an adapter between these formats?

I picked up some transformers that do the job, from markertek for a
reasonable price.

> I have seen a little in-line adapter and am wondering what's inside.

A transformer. By the time you buy the transformer and two connectors, the
commercial adapters start looking pretty good.

[Kurt Albershardt]

Dan Maas wrote:

> Just a note - before getting into anything fancy, you could try simply
> using an RCA->XLR cable. Of course this ignores all the issues of
> voltage, impedance, signal format, etc... But in some cases it works.
> (I hook up my Delta sound card to a DAT deck this way).

Many recent hardware implementations have sufficient source current on
the output and receiver AGC on the input that they can handle this, at
least for short cables. If you're dealing with any serious amounts of
line length, it's worth investing in a balun or two to properly
terminate the line.

[Pooh Bear]

Artie Turner wrote:

> Charles Robertson, Psy.D. wrote:
> > Does anyone have the circuit for an adapter between these formats? I have
> > seen a little in-line adapter and am wondering what's inside.
> >
> It's a simple transformer - 75 ohm unbalanced to 110 ohm balanced.
>
> Artie Turner

Not possible.

AES/EBU works at a higher voltage as well as a slightly different impedance.

Any unit using just a transformer won't be a reliable source.


Graham

[Ty Ford]

In Article >, "Charles Robertson, Psy.D."
> wrote:
>Does anyone have the circuit for an adapter between these formats? I have
>seen a little in-line adapter and am wondering what's inside.


Graham Patten makes one. I think there's just a transformer inside. Not
certain though.

Ty Ford

**Until the worm goes away, I have put "not" in front of my email address.
Please remove it if you want to email me directly.
For Ty Ford V/O demos, audio services and equipment reviews,
click on http://www.jagunet.com/~tford

[Mike Rivers]

In article > writes:

> Isn't there a little more though. AES/EBU has no SCMS, it has some other
> information in there.

Yes, but the adapter just takes care of the physical interface. The
audio data is the same in both except where it comes to 24-bit word
length. I'd have to look up an old article of mine to give you the
details. Officially AES/EBU is defined for up to 20 bits of audio
data, so the other four bits have to replace some status bits that you
or your system probably doesn't care about.

Generally the receiver will either ignore bits that it doesn't
understand or there will be a switch to select between the two data
formats.



--
I'm really Mike Rivers - )

[Artie Turner]

Pooh Bear wrote:
> Artie Turner wrote:
>
>
>>Charles Robertson, Psy.D. wrote:
>>
>>>Does anyone have the circuit for an adapter between these formats? I have
>>>seen a little in-line adapter and am wondering what's inside.
>>>
>>
>>It's a simple transformer - 75 ohm unbalanced to 110 ohm balanced.
>>
>>Artie Turner
>
>
> Not possible.
>
> AES/EBU works at a higher voltage as well as a slightly different impedance.

I'm not sure if you know what a transformer does; it "transforms"
voltage and impedance to match the source to the load.
>
> Any unit using just a transformer won't be a reliable source.

Works every time for me.

Artie
>
>
> Graham
>
>

[Mike Rivers]

In article > writes:

> Not possible.

Very possible.

> AES/EBU works at a higher voltage as well as a slightly different impedance.

That's what the transformer does - rasies the voltage and impedance.

> Any unit using just a transformer won't be a reliable source.

It's as reliable when converted as the original source is. Of course
it's better to have matching hardware, but we always have to make
adjustments because we don't buy all our equipment together every time
we make a purchase.



--
I'm really Mike Rivers - )

[Rob Reedijk]

Dan Maas > wrote:
> Just a note - before getting into anything fancy, you could try simply
> using an RCA->XLR cable. Of course this ignores all the issues of
> voltage, impedance, signal format, etc... But in some cases it works.
> (I hook up my Delta sound card to a DAT deck this way).

Actually, the distinction between AES and SPDIF has nothing to do with
physical characteristics such as RCA and XLR connectors. It has to do
with how the information is encoded. It is very possible for
an AES signal to travel down a 75 ohm line with RCA connectors. In fact,
it probably happens more often than you think!

Rob R.

[Kurt Albershardt]

Rob Reedijk wrote:
>
> Actually, the distinction between AES and SPDIF has nothing to do with
> physical characteristics such as RCA and XLR connectors.

Both, actually.


> It has to do with how the information is encoded.

The vast majority of which is the same for both formats.




> It is very possible for
> an AES signal to travel down a 75 ohm line with RCA connectors. In fact,
> it probably happens more often than you think!

AES3id specifies 75 Ohm coax and BNC connectors and is preferred by some
users over the 110 Ohm balanced lines of AES3.

[Pooh Bear]

Artie Turner wrote:

> Pooh Bear wrote:
> >
> > AES/EBU works at a higher voltage as well as a slightly different impedance.
>
> I'm not sure if you know what a transformer does; it "transforms"
> voltage and impedance to match the source to the load.
> >
> > Any unit using just a transformer won't be a reliable source.
>
> Works every time for me.

You're lucky then.

I'm assuming a 1:1 transformer typically.

Trouble is, IEC 60958 specifies AES/EBU as 5V pk-pk and SPDIF as 0.5V pk-pk. Some
equipment may vary from this spec.

So your AES receiver will see 1/10 th of its intended input.

You're likely to see problems with the 'eye pattern' and likely corrupted bits and
frames, sesp when using any length of cable.

See this : http://www.murraypro.com/spdif.htm as an example of how to do it
properly.


Graham

[Pooh Bear]

Mike Rivers wrote:

> In article > writes:
>
> > Not possible.
>
> Very possible.

On a good day with the wind behind you maybe. Depends on the receiver performance.


> > AES/EBU works at a higher voltage as well as a slightly different impedance.
>
> That's what the transformer does - rasies the voltage and impedance.

Ok, AES/EBU has a specified operating level of 5V pk-pk whereas SPDIF is just 500mV.

A 10:1 step up transformer will raise the impedance by 100x.

therefore 75 ohms source > 7k5 ohms on the secondary, not 110 ohms !

Since you have to pay for IEC 60958, suggest you look here.

www.cirrus.com/en/pubs/proDatasheet/CS8405A-5.pdf

Section 14. Appendix A, page 35. Specifies operating levels etc for AES/EBU and SPDIF.


> > Any unit using just a transformer won't be a reliable source.
>
> It's as reliable when converted as the original source is. Of course
> it's better to have matching hardware, but we always have to make
> adjustments because we don't buy all our equipment together every time
> we make a purchase.

The data stream is compatible, the interface levels aren't.

You're playing with fire.

It's a bit like saying you can plug an electric guitar pickup into a low-Z mic pre-amp. It
kinda works.

Hope you don't drop too many bits, or frames !

Here's an example of the professional way to do it. It will also 'clean up' the transitions.

http://www.murraypro.com/spdif.htm


Graham

[Pooh Bear]

rich rookie wrote:

> It turns out I need to use a logic level shifter, so a simple transformer
> won't work for me. But that's just the DAT manufacturer's fault.

It's not the fault of the DAT manufacturer. They probably designed to IEC 60958.

AES/EBU and SPDIF logic levels are *not* compatible officially. Some equipment
may work. It's your risk.

> It could happen and if you're not aware of the potential, it could drive you
> crazy.

Yes ! Graham

[rich rookie]

"Pooh Bear" > wrote in message
...
> rich rookie wrote:
>
> > It turns out I need to use a logic level shifter, so a simple
transformer
> > won't work for me. But that's just the DAT manufacturer's fault.
>
> It's not the fault of the DAT manufacturer. They probably designed to IEC
60958.
>

All the other information about this not withstanding. On my DAT, a Sony,
the "1" volatage level is some strange voltage like 1.4V. I forget exactly
the levels though, it's been a long time since I looked.

Richard

[Mike Rivers]

In article > writes:

> Actually, the distinction between AES and SPDIF has nothing to do with
> physical characteristics such as RCA and XLR connectors.

Actually, it does. The connector types as well as the impedance,
voltage levels, and the balanced wiring (in the case of AES/EBU) are
all described in their respective standards.

> It has to do
> with how the information is encoded.

This is described in IEC-958. The "professonal" part of the standard
(Type 1, I believe, but I mix this up 50% of the time) describes the
data format we call "AES/EBU" and the "consumer" part describes what
we call "S/PDIF."

> It is very possible for
> an AES signal to travel down a 75 ohm line with RCA connectors. In fact,
> it probably happens more often than you think!

Yup. Even without transformers.


--
I'm really Mike Rivers - )

[Mike Rivers]

In article > writes:

> > > AES/EBU works at a higher voltage as well as a slightly different
> impedance.

> Ok, AES/EBU has a specified operating level of 5V pk-pk whereas SPDIF is just
> 500mV.

Fortunately this is one specification that isn't too specific. The .5V
level for S/PDIF is the minimum level at which the receiver will work.
Most S/PDIF outputs are higher, many are 5V open circuit. I have an
old Phillips CD player that puts out just under 1V and it's the lowest
output S/PDIF device I have ever seen. There's just no reason to put
out only .5V when there's 5V floating around all over the place, or at
least 3V in some of the modern chips.

So this is one of those things that works in practice because those
who build equipment build it to work, not to just barely work.

> The data stream is compatible, the interface levels aren't.

The audio data is compatible up to 20 bits, since that's all that's
defined in the original AES/EBU specification. There's room to tack on
the other four bits of a 24-bit word in the audio frame of the S/PDIF
(or more correctly, the "consumer") format, but those four bits have
to replace some other bits in the "pro" format. Generally some of the
user bits are used. There's probably an industry standard by now, but
in the early 24-bit days, your receiver had to know how to interpret
the odd bits so it wouldn't think that audio data was actually a
status.

> You're playing with fire.
>
> It's a bit like saying you can plug an electric guitar pickup into a low-Z mic
> pre-amp. It
> kinda works.

Not at all. Until you don't have enough level change to tell a one
from a zero, digital works. Rise time can get sloppy and you can have
degradation due to jitter if you have excessive loading, but that's
why you use a transformer. Plugging a guitar pickup into a low
impedance load makes significant changes to the frequency response due
to the inductance of the pickup becoming significant.

> Here's an example of the professional way to do it. It will also 'clean up' the
> transitions.

There's always a better way, but why not try the cheap way first?

PS - I don't mix down to a 1/2" ATR-102 with Dave Hill's electonics
either.




--
I'm really Mike Rivers - )

[Graham Hinton]

In article <znr1066874613k@trad>,
(Mike Rivers) wrote:

>The audio data is compatible up to 20 bits, since that's all that's
>defined in the original AES/EBU specification. There's room to tack on
>the other four bits of a 24-bit word in the audio frame of the S/PDIF
>(or more correctly, the "consumer") format, but those four bits have
>to replace some other bits in the "pro" format. Generally some of the
>user bits are used. There's probably an industry standard by now, but
>in the early 24-bit days, your receiver had to know how to interpret
>the odd bits so it wouldn't think that audio data was actually a
>status.

Not quite. The original PRO format could be defined as having 20 or 24 bit
word length with maximum audio length anything from 16 to 24 bits. The
difference of 4 Auxiliarly bits is not used for the Status, that is in bit
30 (C). User data is in bit 29 (U). The four Auxiliarly bits were usually
zero, they may have been used by someone, but I don't know of any examples.

The Channel Status has to be common between Pro and Consumer because that
is where the difference is defined.

Early compatibility problems were caused by receivers that insisted on all
the status information being correct. This is what happens when a spec is
designed by a commitee who haven't developed any working hardware. Later
devices just took the obvious practical approach and ignored most of it.
The original AES Journal proposal (specification is too strong a
description) put in lots of things that seemed like a good idea at the
time, but omitted isolation transformers. That is what the EBU added.

[Mike Rivers]

In article > writes:

> The Channel Status has to be common between Pro and Consumer because that
> is where the difference is defined.

I may be using the term "channel status bits" too generically. One bit
designates whether the other bits are to be treated as the consumer
format (two bits defining the copy restriction status) or professional
(I don't remember what the others do here).

> Early compatibility problems were caused by receivers that insisted on all
> the status information being correct. This is what happens when a spec is
> designed by a commitee who haven't developed any working hardware.

I have a TEAC DA-P20 DAT that insists on the consumer format.
Everything I have that I want to connect to it digitally has to go
through my trusty old Digital Domain format converter.



--
I'm really Mike Rivers - )

[Graham Hinton]

In article <znr1066956804k@trad>,
(Mike Rivers) wrote:

>I may be using the term "channel status bits" too generically. One bit
>designates whether the other bits are to be treated as the consumer
>format (two bits defining the copy restriction status) or professional

These bits that you are referring to are in the Channel Status block. This
is one bit (C) from every audio word that builds up over a 192 word frame
to make the block of 24 bytes. The first bit of this is PRO where 1 =
Professional and 0 = Consumer and the rest of the block structure depends
on that.

There are four bits at the end of every audio sub-frame V (valid), U (user
bits), C (channel status) and P (parity) and any of these could cause
problems receiving. C has to be read, but the others can be ignored. If
they are not set correctly by a transmitter, receivers that obey the rules
will mute, but ones that ignore them will receive it.

A similar thing happened with SMPTE LTC: the definition of the parity bit
was so indecipherable that implementers had a 50% chance of getting it
right. Once you get a few major products out that don't implement it or
invert it the only solution is to ignore it which defeats its purpose.