Seeking the solutions of high speed interconnection for the long distance transmission of 3.3v/24MHz signals.

Seeking the solutions of high speed interconnection for the long
distance transmission of 3.3v/24MHz signals.

We need to connect a CMOS image sensor and a FPGA chip. The distance
between them is approximately 1 meter. The output signal of the sensor
is 3.3v and 24MHz. However, the length of the original cable of it is
only 1 centimeter, so it should be prolonged. If they are connected by
a cable directly, the signal would attenuate greatly which will cause
the system doesn't work. In my opinion, an interconnecting circuit
should be added between the sensor and the FPGA chip to process the
signal before and after the transmission. By now, the methods I have
found are as follows:
1, Using LVDS signal Transceiver and Receiver
The drawback of the method is: there are more than 10 bits signals,
which need many of this kind of chips. It will take too much place in
PCB and it also cost much.
2, Using chips of Serializer and Deserializer.
Both Maxim and National Semiconductor have such chips such as MAX9247
and DS90C241. It is a good method. Unfortunately the solution is fired
by my tutor because these chips are expensive.

My question is whether there are any other solutions? More cheap will
be better. Thanks a lot.

"X.Y." <Xieyu1219@gmail.com> wrote in message 
news:1176904353.399657.160000@p77g2000hsh.googlegroups.com...
> Seeking the solutions of high speed interconnection for the long
> distance transmission of 3.3v/24MHz signals.
>
> We need to connect a CMOS image sensor and a FPGA chip. The distance
> between them is approximately 1 meter. The output signal of the sensor
> is 3.3v and 24MHz. However, the length of the original cable of it is
> only 1 centimeter, so it should be prolonged. If they are connected by
> a cable directly, the signal would attenuate greatly which will cause
> the system doesn't work. In my opinion, an interconnecting circuit
> should be added between the sensor and the FPGA chip to process the
> signal before and after the transmission. More cheap will
> be better. Thanks a lot.
>
Hi XY,
Use a length of ribbon cable. Make every other conductor ground. Use source 
termination on your signals to match the cable impedance in the gnd-sig-gnd 
mode. Make sure you connect all the grounds properly at both ends of the 
cable. It works well for IDE ATA parallel disks up to 133MBytes/second. HTH, 
Syms.
p.s. BTW., ribbon cable makes great speaker wire. It's flat and fits under 
the carpet! 

Symon wrote:
> "X.Y." <Xieyu1219@gmail.com> wrote in message 
> news:1176904353.399657.160000@p77g2000hsh.googlegroups.com...
>> Seeking the solutions of high speed interconnection for the long
>> distance transmission of 3.3v/24MHz signals.
>>
>> We need to connect a CMOS image sensor and a FPGA chip. The distance
>> between them is approximately 1 meter. The output signal of the sensor
>> is 3.3v and 24MHz. However, the length of the original cable of it is
>> only 1 centimeter, so it should be prolonged. If they are connected by
>> a cable directly, the signal would attenuate greatly which will cause
>> the system doesn't work. In my opinion, an interconnecting circuit
>> should be added between the sensor and the FPGA chip to process the
>> signal before and after the transmission. More cheap will
>> be better. Thanks a lot.
>>
> Hi XY,
> Use a length of ribbon cable. Make every other conductor ground. Use source 
> termination on your signals to match the cable impedance in the gnd-sig-gnd 
> mode. Make sure you connect all the grounds properly at both ends of the 
> cable. It works well for IDE ATA parallel disks up to 133MBytes/second. HTH, 
> Syms.
> p.s. BTW., ribbon cable makes great speaker wire. It's flat and fits under 
> the carpet! 
> 
> 

Great idea about the speaker wire!  I'll have to try that next time


-Eli
[Someone who hasn't been into that low cost wire adds 'distortion']

X.Y. wrote:
> Seeking the solutions of high speed interconnection for the long
> distance transmission of 3.3v/24MHz signals.
> 
> We need to connect a CMOS image sensor and a FPGA chip. The distance
> between them is approximately 1 meter. The output signal of the sensor
> is 3.3v and 24MHz. However, the length of the original cable of it is
> only 1 centimeter, so it should be prolonged. If they are connected by
> a cable directly, the signal would attenuate greatly which will cause
> the system doesn't work. In my opinion, an interconnecting circuit
> should be added between the sensor and the FPGA chip to process the
> signal before and after the transmission. By now, the methods I have
> found are as follows:
> 1, Using LVDS signal Transceiver and Receiver
> The drawback of the method is: there are more than 10 bits signals,
> which need many of this kind of chips. It will take too much place in
> PCB and it also cost much.
> 2, Using chips of Serializer and Deserializer.
> Both Maxim and National Semiconductor have such chips such as MAX9247
> and DS90C241. It is a good method. Unfortunately the solution is fired
> by my tutor because these chips are expensive.
> 
> My question is whether there are any other solutions? More cheap will
> be better. Thanks a lot.

SERDES chips aren't very expensive. And the (12-bit) Fairchild Semi
FIN12AC are very cheap - under $2 each and your tutor can probably get a
couple of free samples.

Un bel giorno X.Y. digit�:

> 1, Using LVDS signal Transceiver and Receiver
> The drawback of the method is: there are more than 10 bits signals,
> which need many of this kind of chips. It will take too much place in
> PCB and it also cost much.

There are single chip with many channels, for example:

http://focus.ti.com/docs/prod/folders/print/sn65mlvd080.html

> 2, Using chips of Serializer and Deserializer.
> Both Maxim and National Semiconductor have such chips such as MAX9247
> and DS90C241. It is a good method. Unfortunately the solution is fired
> by my tutor because these chips are expensive.

I don't think so:

http://focus.ti.com/docs/prod/folders/print/sn65lv1021.html

-- 
emboliaschizoide.splinder.com

Attenuation over 1 m = 40 inches is not your problem. Reflections from
either end of this cable can be a very ugly problem, unless you have
some kind of termination.
As Symon suggested, the simplest termination is series termination
(since you have a "single-source, single-destination" case.
So: estimate the cable impedance (probably 100 Ohm) and put a series
resistor right at the dricing output, so that the IC sees twice the
characteristic impedance. This drives a half-amplitude signal along
the cable, but at the far end there is total reflection (since there
is no termination), which doubles the amplitude, This wave now travels
back to the source, where it is swallowed by the perfect termination.
Bingo, no more reflections.
If you have a fast scope, you can see the half-amplitude signal
driving the cable, and the additional half arriving back at the source
after about 10 ns.
Peter Alfke

On Apr 18, 6:52 am, "X.Y." <Xieyu1...@gmail.com> wrote:
> Seeking the solutions of high speed interconnection for the long
> distance transmission of 3.3v/24MHz signals.
>
> We need to connect a CMOS image sensor and a FPGA chip. The distance
> between them is approximately 1 meter. The output signal of the sensor
> is 3.3v and 24MHz. However, the length of the original cable of it is
> only 1 centimeter, so it should be prolonged. If they are connected by
> a cable directly, the signal would attenuate greatly which will cause
> the system doesn't work. In my opinion, an interconnecting circuit
> should be added between the sensor and the FPGA chip to process the
> signal before and after the transmission. By now, the methods I have
> found are as follows:
> 1, Using LVDS signal Transceiver and Receiver
> The drawback of the method is: there are more than 10 bits signals,
> which need many of this kind of chips. It will take too much place in
> PCB and it also cost much.
> 2, Using chips of Serializer and Deserializer.
> Both Maxim and National Semiconductor have such chips such as MAX9247
> and DS90C241. It is a good method. Unfortunately the solution is fired
> by my tutor because these chips are expensive.
>
> My question is whether there are any other solutions? More cheap will
> be better. Thanks a lot.

Thanks for your reply. Thank you!
To Symon and Peter Alfke, using ribbon cable is really a good ideal.
It's simple and cheap. However, our CMOS image sensor (OV9655) is from
OmniVision and is originally used on a mobile phone. It is very small
not only on volume but also on power (90mW). Are you sure the
attenuation is not a problem? And besides, I knew a little about the
method of source terminators from the book "High Speed Digital Design"
of Johnson & Graham. However, I do not understand the means of "Make
every other conductor ground" and "in the gnd-sig-gnd mode" as Symon
suggested. Sorry for my ignorance, could you tell me more about it?

To Tim and dalai lamah, thank you and I will try these chips later.

Ribbon cable:
Ideally you would like to use a coaxial cable for each signal, so that
there would be a controlled and defined impedance,and no crosstalk
between the signals.
A ribbon cable is an approximation of that, if you use the even-
numbered wires for the various signals, and all the odd-numbered wires
together  as common ground. The shielding is not perfect, but usually
acceptable, and you keep the characteristic impedance fairly high (I
assumed 100 Ohm, but it might be 70 Ohm).
The asumption is that the output driver has a lower impedance than
that. CMOS outputs ar often as low as 10 Ohm.
You can measure the output impedance by loading the output with
various resistors to ground (try 100, 47, and 22 Ohms or something
like it).
Never thinkof the cable as  a lumped capacitance. It is a transmission
line with distributed C and L plus some series resistance causing
losses , but I don't believe the losses will bother you much. 1 m
should be easy. If you had said 10 or 50 m, it would be another
matter...
Peter Alfke
=========================
On Apr 18, 9:28 pm, "X.Y." <Xieyu1...@gmail.com> wrote:
> Thanks for your reply. Thank you!
> To Symon and Peter Alfke, using ribbon cable is really a good ideal.
> It's simple and cheap. However, our CMOS image sensor (OV9655) is from
> OmniVision and is originally used on a mobile phone. It is very small
> not only on volume but also on power (90mW). Are you sure the
> attenuation is not a problem? And besides, I knew a little about the
> method of source terminators from the book "High Speed Digital Design"
> of Johnson & Graham. However, I do not understand the means of "Make
> every other conductor ground" and "in the gnd-sig-gnd mode" as Symon
> suggested. Sorry for my ignorance, could you tell me more about it?
>
> To Tim and dalai lamah, thank you and I will try these chips later.

On 4=D4=C219=C8=D5, =CF=C2=CE=E71=CA=B147=B7=D6, Peter Alfke <a...@sbcgloba=
l=2Enet> wrote:
> Ribbon cable:
> Ideally you would like to use a coaxial cable for each signal, so that
> there would be a controlled and defined impedance,and no crosstalk
> between the signals.
> A ribbon cable is an approximation of that, if you use the even-
> numbered wires for the various signals, and all the odd-numbered wires
> together  as common ground. The shielding is not perfect, but usually
> acceptable, and you keep the characteristic impedance fairly high (I
> assumed 100 Ohm, but it might be 70 Ohm).
> The asumption is that the output driver has a lower impedance than
> that. CMOS outputs ar often as low as 10 Ohm.
> You can measure the output impedance by loading the output with
> various resistors to ground (try 100, 47, and 22 Ohms or something
> like it).
> Never thinkof the cable as  a lumped capacitance. It is a transmission
> line with distributed C and L plus some series resistance causing
> losses , but I don't believe the losses will bother you much. 1 m
> should be easy. If you had said 10 or 50 m, it would be another
> matter...
> Peter Alfke
> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D
> On Apr 18, 9:28 pm, "X.Y." <Xieyu1...@gmail.com> wrote:
>
>
>
> > Thanks for your reply. Thank you!
> > To Symon and Peter Alfke, using ribbon cable is really a good ideal.
> > It's simple and cheap. However, our CMOS image sensor (OV9655) is from
> > OmniVision and is originally used on a mobile phone. It is very small
> > not only on volume but also on power (90mW). Are you sure the
> > attenuation is not a problem? And besides, I knew a little about the
> > method of source terminators from the book "High Speed Digital Design"
> > of Johnson & Graham. However, I do not understand the means of "Make
> > every other conductor ground" and "in the gnd-sig-gnd mode" as Symon
> > suggested. Sorry for my ignorance, could you tell me more about it?
>
> > To Tim and dalai lamah, thank you and I will try these chips later.- =
=D2=FE=B2=D8=B1=BB=D2=FD=D3=C3=CE=C4=D7=D6 -
>
> - =CF=D4=CA=BE=D2=FD=D3=C3=B5=C4=CE=C4=D7=D6 -

OK, I understand your means. Thank you very much! I will try it!

Thank you!

"X.Y." <Xieyu1219@gmail.com> wrote in message 
news:1176963245.037547.197000@b58g2000hsg.googlegroups.com...

> OK, I understand your means. Thank you very much! I will try it!
> Thank you!

Dear XY,
No worries. Here are some links to show Peter and I are not making this up!
Shows loss vs. length:-
http://www.csee.umbc.edu/~plusquel/650/slides/ribbon_cables.pdf

Shows impedance:-
http://www.daburn.com/1888.html

Let us know how you get on!

Cheers, Syms