osd20 spec

33 views
Skip to first unread message

Michael Gao

unread,
Mar 16, 2008, 8:51:44 AM3/16/08
to Neuros-DM320
All,

For those who are interested, attached is the latest osd20 HW spec, in
fact, this spec came along with the inputs from many of you, as always,
comments are welcome.

Thanks,

/MG

OSD-2.0-spec-1.10.pdf

Fernando Cassia

unread,
Mar 16, 2008, 10:14:53 AM3/16/08
to Neuros-DM3...@googlegroups.com
Just read it. I want one!. :)

However the following line needs clarification:

"No video passthrough
support, live with the composite video degrade and component/HDMI output
will not be user noticeable"

What does it mean no video passthrough? I do like menus overlaid over live video. Is this what you mean?

FC
--
Dream of the Daily Mail
It is the Holy Grail
And then the BBC
Your life would be complete

-Manic Street Preachers, "Royal Correspondent"

Pavel Tkatchouk

unread,
Mar 16, 2008, 7:15:39 PM3/16/08
to Neuros-DM3...@googlegroups.com
Looks good, when it's going to be available, in Canada
in particular?

Few questions... can you elaborate on "playback, up to
1080i with certain contents" please?

Which Ethernet will it support, Gigabit?

So OSD-2 is 6446 based. What would be ETA for 6447
based one?

Thanks.


____________________________________________________________________________________
Looking for last minute shopping deals?
Find them fast with Yahoo! Search. http://tools.search.yahoo.com/newsearch/category.php?category=shopping

Kevin

unread,
Mar 17, 2008, 8:36:15 AM3/17/08
to Neuros DM320
Will the osd20 replace the currently released version of the Neuros
OSD? If so, how long will it be until it is end of lifed?

Thanks,
Kevin.
> OSD-2.0-spec-1.10.pdf
> 738KDownload

Adam Torgerson

unread,
Mar 17, 2008, 1:00:56 PM3/17/08
to Neuros-DM3...@googlegroups.com
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

It looks like audio is still limited to RCA jacks. Don't we get digital
audio pretty much for free with these DSPs? I'd love to see digital
audio in and out and not only for quality purposes, it would also
provide surround sound capability.

adam
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.7 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFH3qPI2ZcKn9labBsRAvr3AKCftL66PluUi8zmmhxFBMwgZQCnBwCfR6dk
ETvuapv9Pk0QQHsAPAqCGcQ=
=lE2F
-----END PGP SIGNATURE-----

Fernando Cassia

unread,
Mar 17, 2008, 2:27:42 PM3/17/08
to Neuros-DM3...@googlegroups.com


On Mon, Mar 17, 2008 at 9:36 AM, Kevin <kma...@gmail.com> wrote:

Will the osd20 replace the currently released version of the Neuros
OSD? If so, how long will it be until it is end of lifed?

Thanks,
Kevin.

I can´t speak for Joe but I´ve heard many times that the next-gen OSD will NOT replace the current one, in fact, I´d hate to see them stop producing the current low-cost device, which has plenty of applications for the embedded space, home automation, security, etc.

As Joe has said many times, the cost of the hardware decreases as time passes... it gets even more attractive, price-wise. Can´t wait for the $50 OSD v1. :))

FC

Joe Born

unread,
Mar 17, 2008, 3:40:05 PM3/17/08
to Neuros-DM3...@googlegroups.com

>
> I can愒 speak for Joe but I扉e heard many times that the next-gen OSD
> will NOT replace the current one, in fact, I悲 hate to see them stop
> producing the current low-cost device, which has plenty of
> applications for the embedded space, home automation, security, etc.
>
> As Joe has said many times, the cost of the hardware decreases as time
> passes... it gets even more attractive, price-wise. Can愒 wait for the
> $50 OSD v1. :))
>
This is exactly right, I think there will be a long lifetime for the OSD
1, I don't know about a $50 OSD, but certainly there are cost
reductions over time, there's no doubt of that.

Joe Born

unread,
Mar 17, 2008, 6:03:47 PM3/17/08
to Neuros-DM3...@googlegroups.com

>
> It looks like audio is still limited to RCA jacks. Don't we get digital
> audio pretty much for free with these DSPs? I'd love to see digital
> audio in and out and not only for quality purposes, it would also
> provide surround sound capability.
>
someone else can speak to the surround sound issue, which is, for
whatever reason, somewhat more complicated. WRT, digital out, it's
provided by HDMI, but I was wondering myself if there'd be any way to
use my older amp with toslink or spdif with it. I don't know the
answer, but presumably someone does. The other thing is USB since our
USB driver has hub and is supposed to be much better than the OSD 1 USB
driver, we can potentially use one of the USB to S/PDIF adapters which
are ubiquitous and cheap. I gather there are already drivers available
for many of these.

Adam Torgerson

unread,
Mar 17, 2008, 9:13:21 PM3/17/08
to Neuros-DM3...@googlegroups.com
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


Joe Born wrote:
| someone else can speak to the surround sound issue, which is, for
| whatever reason, somewhat more complicated. WRT, digital out, it's

| provided by HDMI...

Is there going to be audio over HDMI? The spec makes it sound like it is
video only since HDMI is not listed in the audio section.

If we do have digital audio out via HDMI, is there any chance of also
getting HDMI input? Then, we'd also get digital audio in...

If I understand things properly, I have a hard time seeing how the osd20
is going to fit into my system. Right now, everything I run is digital
over HDMI and/or toslink. To use the osd20 it looks like I'd need to
convert both the audio (via RCA) and the video (via component/composite)
to an analog source first. I suppose I could get an adapter for the PS3
to accomplish part of that, but that leaves me with no way of recording
broadcast TV, which goes into my mythtv box via a tuner and then out to
my TV via DVI/HDMI. I'd hate to have to fallback to composite video
there. Am I understanding this properly, or am I missing something?

, but I was wondering myself if there'd be any way to
| use my older amp with toslink or spdif with it. I don't know the
| answer, but presumably someone does. The other thing is USB since our
| USB driver has hub and is supposed to be much better than the OSD 1 USB
| driver, we can potentially use one of the USB to S/PDIF adapters which
| are ubiquitous and cheap. I gather there are already drivers available
| for many of these.

I guess that would work, but it sure would be nice to have native
digital audio in and out.

adam

PS - sorry if I missed previous discussion on this subject

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.7 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFH3xcx2ZcKn9labBsRAkrMAKCJNLxUEANcJSwZ7Im8WutvYi7lGQCfaizR
CBeajL8/f2f+p+r0XzbiUWU=
=Cq+i
-----END PGP SIGNATURE-----

Adam Torgerson

unread,
Mar 17, 2008, 9:23:35 PM3/17/08
to Neuros-DM3...@googlegroups.com
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


Adam Torgerson wrote:
| If we do have digital audio out via HDMI, is there any chance of also
| getting HDMI input? Then, we'd also get digital audio in...

I know how much everyone loves responding to themselves, but I just
realized that encrypted HDMI is probably part of the problem here. So
I'll guess that HDMI input is not an option.

adam


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.7 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFH3xmX2ZcKn9labBsRAgkIAJ4/sLpVd/XZqud17jqxHIjtQCgQ8gCePiV5
el1zCxvPi6saCnsjrtAFmDE=
=4UTc
-----END PGP SIGNATURE-----

Joe Born

unread,
Mar 17, 2008, 10:33:32 PM3/17/08
to Neuros-DM3...@googlegroups.com

> Adam Torgerson wrote:
> | If we do have digital audio out via HDMI, is there any chance of also
> | getting HDMI input? Then, we'd also get digital audio in...
>
> I know how much everyone loves responding to themselves, but I just
> realized that encrypted HDMI is probably part of the problem here. So
> I'll guess that HDMI input is not an option.
>
>
exactly. The fear is that HDMI input is simply not reliable due to the
potential for there to be encryption, so we left it out, for fear that
it would be too hard to explain that sometimes you can record from HDMI,
but sometimes you can't.

S/PDIF doesn't have that problem, I just didn't realize there was enough
demand for it.

But yes, to answer your other question there, there will be audio on the
HDMI output.

Joe

Michael Gao

unread,
Mar 18, 2008, 3:36:29 AM3/18/08
to Neuros-DM3...@googlegroups.com

On Sun, 2008-03-16 at 12:14 -0200, Fernando Cassia wrote:
> Just read it. I want one!. :)
>
> However the following line needs clarification:
>
> "No video passthrough
> support, live with the composite video degrade and component/HDMI
> output
> will not be user noticeable"
>
> What does it mean no video passthrough? I do like menus overlaid over
> live video. Is this what you mean?
>
"Video passthrough" means a hard-wired signal passthrough, which is what
the OSD1.0 relay designed for (where the annoying click noise comes
from ;)).

Menu overlay will surely be there.

/MG


Fernando Cassia

unread,
Mar 18, 2008, 3:50:27 AM3/18/08
to Neuros-DM3...@googlegroups.com

Thanks for the clarification.

With regards to the noise, I never considered it an annoyance.

And I wonder if covering the relay with a big mountain of silicone based resin would lower the "clicking noise" ??. In fact, that sounds like a nice project for testing... :)

FC

Fernando Cassia

unread,
Mar 18, 2008, 5:06:01 AM3/18/08
to Neuros-DM3...@googlegroups.com
On Tue, Mar 18, 2008 at 4:50 AM, Fernando Cassia <fca...@gmail.com> wrote:


And I wonder if covering the relay with a big mountain of silicone based resin would lower the "clicking noise" ??. In fact, that sounds like a nice project for testing... :)

I did it... "only" lowers noise less than 20%. I´d say 10-15%, or about 5-9 dBa.

Goes from 53dBa to 47-44.

Still, it´s better than nothing. I think the big issue is that the empty space inside the OSD case makes it act like a "resonance chamber" so so speak... or a big speaker box if you prefer.

Pics here
http://the13thfloor.org/osd-noise/

I have to be 100% sure if that the "expanding insulating foam" (which forms tiny air bubbles inside) of this kind
http://images.oldhouseweb.com/stories/bitmaps/13174/foam1.jpg
does or doesn´t attack plastics before trying "insonorization 2.0 - the final frontier" ;-).

If it doesn´t attack plastics, then filling the whole OSD interior with it, waiting for it to expand to its fullest, trim any excess foam and then  them putting the two halves together might be a good option. (but first make sure every slot and socket has a connector-memory card plugged in, you don´t want any connectors filled with foam. :)

FC

PS: If anyone is going to try the foam approach, keep in mind it often expands 2x to 3x its original volume. It could well break the plastic enclosure due to the pressure if you fill it at then screw it back in. it´s better to apply it in small layers, let it dry and expand to its fullest before applying the next. Also note that expanding foam should not touch your skin.

PS2: In the pics above, I used silicone based sealant, the kind used to fix leaks in windows, pipes, etc. It remains flexible as rubber, all the time, and can be easily removed without breaking anything.  The same can´t be said about expanding foam. That´s why I haven´t tried that yet.

pEEf

unread,
Mar 18, 2008, 9:52:39 AM3/18/08
to Neuros DM320
You are joking, right? That little occasional click is that annoying
that you'd actually try to muffle it?

If you are serious, you should NOT use insulating foam in the whole
OSD case. The components inside depend on convection to cool,
especially the DM320. If you cover it with any kind of insulating
material you risk failure due to overheating!

If the click really does drive you crazy, you have several options.
One is disconnect and bypass the relay, then switch between live
inputs another way, such as a S-Video input direct to the TV. I
suspect a lot of people do this anyway.

Another way is to replace the relay with a 4066. This is what is
normally used in this application anyway. It will probably take some
additional circuitry to pull off, but isn't that complicated. An
easier solution to use would be the MAX4616, as it can be installed
without any additional circuitry. (http://www.maxim-ic.com/
quick_view2.cfm/qv_pk/2071)

-Phil


On Mar 18, 2:06 am, "Fernando Cassia" <fcas...@gmail.com> wrote:
> On Tue, Mar 18, 2008 at 4:50 AM, Fernando Cassia <fcas...@gmail.com> wrote:
>
> > And I wonder if covering the relay with a big mountain of silicone based
> > resin would lower the "clicking noise" ??. In fact, that sounds like a nice
> > project for testing... :)
>
> I did it... "only" lowers noise less than 20%. I´d say 10-15%, or about 5-9
> dBa.
>
> Goes from 53dBa to 47-44.
>
> Still, it´s better than nothing. I think the big issue is that the empty
> space inside the OSD case makes it act like a "resonance chamber" so so
> speak... or a big speaker box if you prefer.
>
> Pics herehttp://the13thfloor.org/osd-noise/
>
> I have to be 100% sure if that the "expanding insulating foam" (which forms
> tiny air bubbles inside) of this kindhttp://images.oldhouseweb.com/stories/bitmaps/13174/foam1.jpg

Fernando Cassia

unread,
Mar 18, 2008, 10:03:09 AM3/18/08
to Neuros-DM3...@googlegroups.com
On Tue, Mar 18, 2008 at 10:52 AM, pEEf <ingin...@gmail.com> wrote:

You are joking, right?  That little occasional click is that annoying
that you'd actually try to muffle it?

Errr.. DID YOU ACTUALLY read this thread?? Did you miss it with I said "it never really bothered me".

But since people were ASKING about  IT (the "relay noise") I tried that little experiment to muffle it. And it reduced the "click" somewhat.

If you are serious, you should NOT use insulating foam in the whole
OSD case.  The components inside depend on convection to cool,
especially the DM320.

Well, Every time I touched it (and I left it operating on  table without the case several months once), the DM320 wasn´t particularly hot to the touch. What convection are you talking about if the OSD case doesn´t have ANY vent holes??. If the DM320 got really really hot the air temperature inside would raise to such a level that the plastic case would get hot to the touch. It never does.
 
 If you cover it with any kind of insulating
material you risk failure due to overheating!

I think I said in my post why I won´t do the insulating foam approach. The current silicone based approach worked for me. I just thought some people would appreciate my pics, rather than lecture me...

If the click really does drive you crazy,

It doesn´t. !!!!!!!!!!!!!!!!!!!!!!!!!
 
you have several options.
One is disconnect and bypass the relay,

Yeah, other is smashing it with a hammer.

then switch between live
inputs another way, such as a S-Video input direct to the TV.  I
suspect a lot of people do this anyway.

define "a lot"
 
Another way is to replace the relay with a 4066.

I don´t want to replace anything. The sound doesn´t bother me. I just wanted to check how much the click  could be reduced by simple hacking.

No cookie for you.

FC

Vladimir Pantelic

unread,
Mar 18, 2008, 10:14:09 AM3/18/08
to Neuros-DM3...@googlegroups.com
Fernando Cassia wrote:

> Well, Every time I touched it (and I left it operating on table without
> the case several months once), the DM320 wasn´t particularly hot to the
> touch. What convection are you talking about if the OSD case doesn´t
> have ANY vent holes??. If the DM320 got really really hot the air
> temperature inside would raise to such a level that the plastic case
> would get hot to the touch. It never does.

No, this is a "feature" the DM320 does not have, wait for the DM6446 to worry
about that :-)


pEEf

unread,
Mar 18, 2008, 10:54:40 AM3/18/08
to Neuros DM320
Fernando, Yes, I did read the thread, and did see where you didn't
consider it an "annoyance", yet you took signifigant steps to reduce
the noise. I just didn't want anyone foaming the inside of the OSD
using your post as a starting point and frying it.

There is convection cooling even in a "sealed" case. The parts
transfer heat to the air, the air transfers it to the large surface
area of the plastic (which does conduct heat), and the plastic
transfers it to the outside air.

This being said, the OSD is not "sealed". It has two large slots on
either side on the bottom, and a few big holes on the front.

Take a look with a thermal imager, and you will be suprised! I've
done a lot of thermal design and analysis, and have experience in this
area. The OSD working hard sucks down more than a few watts, and
almost all of that ends up as heat that must leave the board.

-Phil

On Mar 18, 7:03 am, "Fernando Cassia" <fcas...@gmail.com> wrote:

Fernando Cassia

unread,
Mar 18, 2008, 11:38:36 AM3/18/08
to Neuros-DM3...@googlegroups.com
On Tue, Mar 18, 2008 at 11:54 AM, pEEf <ingin...@gmail.com> wrote:

Fernando, Yes, I did read the thread, and did see where you didn't
consider it an "annoyance", yet you took signifigant steps to reduce
the noise.  I just didn't want anyone foaming the inside of the OSD
using your post as a starting point and frying it.

I don´t think foaming it would fry it, specially if you leave out the DM320.


There is convection cooling even in a "sealed" case.  The parts
transfer heat to the air, the air transfers it to the large surface
area of the plastic (which does conduct heat), and the plastic
transfers it to the outside air.

In theory, yes. In practice, the fact that the plastic case doesn´t even get warm indicates to me that heat dissipation is VERY LOW.

This being said, the OSD is not "sealed".  It has two large slots on
either side on the bottom, and a few big holes on the front.

Holes on the front?. With USB , SD and CF inserted in the slots, what extra holes are you talking about?

Take a look with a thermal imager, and you will be suprised!  I've
done a lot of thermal design and analysis, and have experience in this
area.  The OSD working hard sucks down more than a few watts, and
almost all of that ends up as heat that must leave the board.

-Phil

I was going to say that you´re full of hot air but I guess I can´t, after this statement from you. ;-)

In any case, I fell tempted of foaming my OSD now more than ever, if just to prove you´re wrong. ;-)

Ever seen sealed power supplies?. Commodore 64 ones comes to mind (the black ones). If you ever took one apart you´d realize it was FILLED WITH PLASTIC RESIN. So it was impossible to disassemble or repair. To open the power supply you had to use a hammer and break the plastic case to pieces. It was, for all intents and purposes, a BRICK. And it got really hot!.  Yet, it worked fine. :)

Moral of the story: modern electronics have a lot of tolerance to heat, and most of the time work at room temperature, which is WAY BELOW the maximum allowed temperature it´s designed to survive.

I´m not saying that´s the case with the DM320 in the OSD. I don´t know. Perhaps I´m full of hot air myself. ;-)
FC 

Pavel Tkatchouk

unread,
Mar 18, 2008, 12:09:08 PM3/18/08
to Neuros-DM3...@googlegroups.com
Gone un-noticed so far, still would appreciate if
someone can comment below...

Dan Fandrich

unread,
Mar 18, 2008, 12:45:29 PM3/18/08
to Neuros-DM3...@googlegroups.com
On Tue, Mar 18, 2008 at 12:38:36PM -0300, Fernando Cassia wrote:
> Ever seen sealed power supplies?. Commodore 64 ones comes to mind (the black
> ones). If you ever took one apart you´d realize it was FILLED WITH PLASTIC
> RESIN. So it was impossible to disassemble or repair. To open the power supply
> you had to use a hammer and break the plastic case to pieces. It was, for all
> intents and purposes, a BRICK. And it got really hot!. Yet, it worked fine. :)

Using the Commodore 64 power supply as an example was a bad idea. It was
notorious for failing due to overheating :-(

>>> Dan
--
http://www.MoveAnnouncer.com The web change of address service
Let webmasters know that your web site has moved

Joe Born

unread,
Mar 18, 2008, 4:11:48 PM3/18/08
to Neuros-DM3...@googlegroups.com
Pavel Tkatchouk wrote:
> Gone un-noticed so far, still would appreciate if
> someone can comment below...
>
>
Not so much un-noticed, just more wishing it would go away quietly :)

> --- Pavel Tkatchouk <ptk...@yahoo.com> wrote:
>
>
>> Looks good, when it's going to be available, in
>> Canada
>> in particular?
>>
Honestly, its hard to say at this point. I dread questions like this
because the software integration effort is hard to predict precisely.
We're going to have developer samples available probably as early as
May, but these are not much more than booting boards, certainly a far
cry from a "product," but we can make those available to Canada.

>> Few questions... can you elaborate on "playback, up
>> to
>> 1080i with certain contents" please?
>>
>>
Probably 1080i with something like MPEG-2 or the like. MPEG-4 and h.264
are too processor intensive to be able to playback 1080i

>> Which Ethernet will it support, Gigabit?
>>
>>
I believe 100 mbps for 6446

>> So OSD-2 is 6446 based. What would be ETA for 6447
>> based one?
>>
>>
I think you mean 6467:
http://www.ti.com/corp/docs/landing/davinci/dm6467.html That will
likely be developer samples later in '08 and again "product schedule"
will be determined by software.

Joe Born

unread,
Mar 18, 2008, 4:18:22 PM3/18/08
to Neuros-DM3...@googlegroups.com

>
> Moral of the story: modern electronics have a lot of tolerance to
> heat, and most of the time work at room temperature, which is WAY
> BELOW the maximum allowed temperature it愀 designed to survive.
>
> I惴 not saying that愀 the case with the DM320 in the OSD. I don愒
> know. Perhaps I惴 full of hot air myself. ;-)
> FC
I've done experiments with the DM320 and overclocking the arm for
relatively brief periods and never had a problem, and I've often
wondered if heat was the only issue could we put a heat sink and some
ventilation on the processor and be safe over clocking it? Some extra
cycles on either the DM320 or 6446 arm would surely be welcome.

Fernando Cassia

unread,
Mar 18, 2008, 4:20:35 PM3/18/08
to Neuros-DM3...@googlegroups.com
On Tue, Mar 18, 2008 at 5:18 PM, Joe Born <jb...@neurostechnology.com> wrote:

I've done experiments with the DM320 and overclocking the arm for
relatively brief periods and never had a problem, and I've often
wondered if heat was the only issue could we put a heat sink and some
ventilation on the processor and be safe over clocking it? Some extra
cycles on either the DM320 or 6446 arm would surely be welcome.

That´s a question the folks at TI will surely be able to answer to you. :)

FC

Pavel Tkatchouk

unread,
Mar 18, 2008, 6:47:27 PM3/18/08
to Neuros-DM3...@googlegroups.com
> Not so much un-noticed, just more wishing it would
> go away quietly :)

A-ha, just as I thought. Thanks Joe for taking your
time and bravery :-)

> We're going to have developer samples available
> probably as early as May, but these are not much
more > than booting boards, certainly a far cry from a

> "product," but we can make those available to
Canada.

That is fine as long as there won't be expected
hardware fixes.

> Probably 1080i with something like MPEG-2 or the
> like. MPEG-4 and h.264 are too processor intensive
> to be able to playback 1080i

Is 6467 any better in this regard?

> I think you mean 6467:

Sorry, 6467, of course.

> That will likely be developer samples later in '08
> and again "product schedule" will be determined by
> software.

I see. I'm building this house video distribution
system with intent to use OSDs as its clients. I'd
prefer all boxes to be of the same model so for me
question was whether to base it on 6446 or wait and
use 6467 instead.

____________________________________________________________________________________
Never miss a thing. Make Yahoo your home page.
http://www.yahoo.com/r/hs

bfa...@neuros.us

unread,
Mar 18, 2008, 8:30:11 PM3/18/08
to Neuros-DM3...@googlegroups.com, neuros-dm3...@googlegroups.com
Be careful with these kinds of thing.
The circuitry will generate some heat, and since these foams and sealants are thermal insulators, the heat will tend to build up inside, and not dissipate.
It can possible "cook" the OSD (most likely the DM320) to the point where it won't recover.

Bob


-----Original Message-----
From: Fernando Cassia <fca...@gmail.com>
Sent: Tuesday, March 18, 2008 5:06am
To: Neuros-DM3...@googlegroups.com
Subject: Re: osd20 spec

bfa...@neuros.us

unread,
Mar 18, 2008, 8:54:01 PM3/18/08
to Neuros-DM3...@googlegroups.com, neuros-dm3...@googlegroups.com
Overclocking the DM320 or DM6446 is a risky business.
It is not just the heat.
Electric signals take time to travel through gates, and if not all input signals have reached the output of some logic block when the clock edge comes then an error will occur.

Some errors are more benign than others.
For example, you may register a slightly wrong color for one pixel of a video.
Nobody will ever notice it.
However, it may also cause a bit in some instruction to flip, causing the processor to go out on the weeds and never return.

The errors are usually random.
Worse still every chip, even from the same batch, will be slightly different than the other and break down at different overclocking points and in a randomly different manner.

If you overclock for periods of time with no consequence, it may mean that the particular unit you are running is capable of supporting that clock. At least at that particular temperature (which constantly change), humidity (which constantly change), power suppy level (which constantly change), etc.

I am sure you get the point by now.


Bob




-----Original Message-----
From: Fernando Cassia <fca...@gmail.com>
Sent: Tuesday, March 18, 2008 4:20pm
To: Neuros-DM3...@googlegroups.com
Subject: Re: osd20 spec



bfa...@neuros.us

unread,
Mar 18, 2008, 11:08:52 PM3/18/08
to Neuros-DM3...@googlegroups.com, Neuros DM320
Hi Phil,

Circuits like the 4066 or MAX4616 because there are no guarantees about the DC level of the incoming video level.
It can be biased below ground level or above VCC level.
In each case the analog switch would be destroyed.
You cannot AC-couple the video into the switch for the exact same reasons.
Also, some video consumers (notably older TVs) bias the input. Sad but true.

Bob



-----Original Message-----
From: pEEf <ingin...@gmail.com>
Sent: Tuesday, March 18, 2008 9:52am
To: Neuros DM320 <Neuros-DM3...@googlegroups.com>
Subject: Re: osd20 spec


You are joking, right? That little occasional click is that annoying
that you'd actually try to muffle it?

If you are serious, you should NOT use insulating foam in the whole
OSD case. The components inside depend on convection to cool,
especially the DM320. If you cover it with any kind of insulating

material you risk failure due to overheating!

If the click really does drive you crazy, you have several options.
One is disconnect and bypass the relay, then switch between live

inputs another way, such as a S-Video input direct to the TV. I
suspect a lot of people do this anyway.

Another way is to replace the relay with a 4066. This is what is
normally used in this application anyway. It will probably take some
additional circuitry to pull off, but isn't that complicated. An
easier solution to use would be the MAX4616, as it can be installed
without any additional circuitry. (http://www.maxim-ic.com/
quick_view2.cfm/qv_pk/2071)

-Phil


On Mar 18, 2:06 am, "Fernando Cassia" wrote:

> On Tue, Mar 18, 2008 at 4:50 AM, Fernando Cassia wrote:
>
> > And I wonder if covering the relay with a big mountain of silicone based
> > resin would lower the "clicking noise" ??. In fact, that sounds like a nice
> > project for testing... :)
>
> I did it... "only" lowers noise less than 20%. I´d say 10-15%, or about 5-9
> dBa.
>
> Goes from 53dBa to 47-44.
>
> Still, it´s better than nothing. I think the big issue is that the empty
> space inside the OSD case makes it act like a "resonance chamber" so so
> speak... or a big speaker box if you prefer.
>
> Pics herehttp://the13thfloor.org/osd-noise/

>
> I have to be 100% sure if that the "expanding insulating foam" (which forms
> tiny air bubbles inside) of this kindhttp://images.oldhouseweb.com/stories/bitmaps/13174/foam1.jpg

pEEf

unread,
Mar 18, 2008, 11:50:24 PM3/18/08
to Neuros DM320
Fernando, you need to brush up on your thermodynamics. The dense
black epoxy resin used to pot a power supply has low thermal
resistance (lower than air). If it were an insulator, you would not
feel anything getting hot as it would prevent heat from being
transferred into your nerve endings.

However, expandable foam has trapped air cells, which almost anyone
knows is an excellent insulator. It's sold specifically to insulate.

The exterior of the OSD doesn't feel hot because convection is cooling
it and taking the heat away. The internal components radiate heat
into the air inside the OSD, then the internal air convects and
transfers some of that to the large surface of the housing which
conducts the heat outside to the surrounding air and is gone.

Measure the power consumption of your OSD playing a video, then tell
me where it is going. Remember, the energy cannot be destroyed, it
must go somewhere. Probably 99.8% of it ends up as unwanted heat on
the PCB. 0.1% is probably used to light the LED on the front, and the
other 0.1% is sent out the back as audio and video.

Do not foam your OSD to prove me wrong, that is silly and will only
cause your OSD harm.

I'm not going to reply to any more of your baiting on this issue.

-Phil

pEEf

unread,
Mar 18, 2008, 11:54:18 PM3/18/08
to Neuros DM320
Hi Bob,

The MAX4616 has integral schottky clamping diodes, and it's unlikely
there is going to be a bias high enough to cause the kind of current
that would be needed to destroy them. It is sold for this purpose!

In my younger years I worked on TV sets for quite a few years, and the
4066 (with clamping circuitry) was always in common use to switch the
inputs.

A typical input buffer in a modern TV consists of a 75 ohm termination
to ground with a DC blocking capacitor. (AC coupled)

I know this is a way off-topic discussion, and no one is likely to
care enough about the relay to want to replace it, but I couldn't
resist! =)

-Phil
> > about expanding foam. That´s why I haven´t tried that yet.- Hide quoted text -
>
> - Show quoted text -

bfa...@neuros.us

unread,
Mar 19, 2008, 12:03:43 AM3/19/08
to Neuros-DM3...@googlegroups.com, Neuros DM320
Phil,

Sure enough, the clamping analog switch will work for 99.99% of the time.
However, when you make enough devices the 0.01% blow-out rate could mean dozens of units returned for repair and warranty.
Not to mention the need to comply to the standard.

Also, in my endeavors designing both broadcast and consumer equipment, I saw my share of weird stuff happening because of this - none of it pleasant and some even hard to explain.

Also, the clamping diode mean that you are ac-coupling the video, which provides for cost and/or signal degradation (more likely a combination of both), or the clamping diode will be blown some day.

This is why I designed the OSD relay as it is.
Safe and pristine.


Bob

-----Original Message-----
From: pEEf <ingin...@gmail.com>
Sent: Tuesday, March 18, 2008 11:54pm
To: Neuros DM320 <Neuros-DM3...@googlegroups.com>
Subject: Re: osd20 spec


Fernando Cassia

unread,
Mar 19, 2008, 7:11:00 AM3/19/08
to Neuros-DM3...@googlegroups.com
On Wed, Mar 19, 2008 at 12:50 AM, pEEf <ingin...@gmail.com> wrote:

However, expandable foam has trapped air cells, which almost anyone
knows is an excellent insulator.  It's sold specifically to insulate.

We were talking about sound, so yes.

The exterior of the OSD doesn't feel hot because convection is cooling
it and taking the heat away.  The internal components radiate heat
into the air inside the OSD, then the internal air convects and
transfers some of that to the large surface of the housing which
conducts the heat outside to the surrounding air and is gone.

And yes, I understand that!

Measure the power consumption of your OSD playing a video, then tell
me where it is going.  Remember, the energy cannot be destroyed, it
must go somewhere.  Probably 99.8% of it ends up as unwanted heat on
the PCB.  0.1% is probably used to light the LED on the front, and the
other 0.1% is sent out the back as audio and video.

Will do. Probably dissipating as heat.

Do not foam your OSD to prove me wrong, that is silly and will only
cause your OSD harm.

You´ve decided me to do it!.

I'm not going to reply to any more of your baiting on this issue.

-Phil

Baiting?. You make it sound like I was questioning thermodynamics. I just ventured that soundproofing the relay with foam COULD work. And did you read the part about leaving a hole on top of the DM320?.

To end this stupid thread, I´ll include a quote from The Simpsons:

-Homer: Lisa! Get in here.
[Lisa walks in, chuckling nervously]
In this house, we obey the laws of thermodynamics!

;-P

FC

kiboko

unread,
Mar 19, 2008, 12:01:54 PM3/19/08
to Neuros DM320

On Mar 17, 8:33 pm, Joe Born <jb...@neurostechnology.com> wrote:
> > Adam Torgerson wrote:
> > | If we do have digital audio out via HDMI, is there any chance of also
> > | getting HDMI input? Then, we'd also get digital audio in...
>
> > I know how much everyone loves responding to themselves, but I just
> > realized that encrypted HDMI is probably part of the problem here. So
> > I'll guess that HDMI input is not an option.
>
> exactly. The fear is that HDMI input is simply not reliable due to the
> potential for there to be encryption, so we left it out, for fear that
> it would be too hard to explain that sometimes you can record from HDMI,
> but sometimes you can't.
>
> S/PDIF doesn't have that problem, I just didn't realize there was enough
> demand for it.

I would like to throw my vote in for S/PDIF audio out. (S/PDIF audio
in would be nice too). I would expect that those most interested in
OSD20 box already have other media devices in their entertainment
setup, and may already find that routing and switching RCA audio
sucks. Digital is better and very well supported these days. If you
able to driver digital audio out via HDMI, they is not a small step to
include a S/PDIF output port?

> But yes, to answer your other question there, there will be audio on the
> HDMI output.
>
> Joe

Anthony

nash...@googlemail.com

unread,
Mar 21, 2008, 12:17:31 PM3/21/08
to Neuros DM320
Thats an interesting read. Are there any rough estimates about a
potential release date of the OSD 2?

On Mar 16, 12:51 pm, Michael Gao <m...@neuros.us> wrote:
> All,
>
> For those who are interested, attached is the latest osd20 HW spec, in
> fact, this spec came along with the inputs from many of you, as always,
> comments are welcome.
>
> Thanks,
>
> /MG
>
> OSD-2.0-spec-1.10.pdf
> 738KDownload

Fernando Cassia

unread,
Mar 21, 2008, 4:00:50 PM3/21/08
to Neuros-DM3...@googlegroups.com
On Fri, Mar 21, 2008 at 1:17 PM, nash...@googlemail.com <nash...@googlemail.com> wrote:

Thats an interesting read. Are there any rough estimates about a
potential release date of the OSD 2?

Tuesday morning.

FC
PS: JOKE JOKE. I wish!. :)

Joe Born

unread,
Mar 24, 2008, 7:15:52 PM3/24/08
to Neuros-DM3...@googlegroups.com
nash...@googlemail.com wrote:
> Thats an interesting read. Are there any rough estimates about a
> potential release date of the OSD 2?
>
>
April/May for dev kits but it will be probably Q4 before it's a real
"consumer product." The dev kits will be just enough to boot, etc.
Reply all
Reply to author
Forward
0 new messages