How to send video signal from and power to a camera through more than 1000m of cable

[DrWhoF]

How to send video signal from and power to a camera through more than 1000m of cable which looks like coax but with very high attenuation and external armoured steel layer?

 

[IanP]

Re: Video and power

First of all, measure the resistance of the core and the shield to estimate dc-loses in the cable ..
Next, connect video from any source (video generator, camera, vcr ..) to this cable and observe (connect the other end to video-in of any TV, monitor ..) what is the quality of signal that passes through ..

Is the coax electrically isolated from the steel?

Regards,
IanP

 

[throwaway18]

Video and power

onnect video from any source (video generator, camera, vcr ..) to this cable and observe (connect the other end to video-in of any TV, monitor ..) what is the quality of signal that passes through ..

The quality is guaranteed to be terrible when tested like that.

The coax will attenuate the signal much more at higher frequencys than lower frequencys which will kill colour information and detail in the picture.

CCTV systems uses preemphasis or a video equalizer to compensate for the frequency response.

 

[echo47]

Video and power

Is your video bandwidth 5 MHz, or something else?

Try to determine the coax type. If you can't, then measure the attenuation at a few frequencies, such as DC, 100 kHz, 1 MHz, and 5 MHz. Don't forget proper termination. Tell us what you find, and maybe someone can suggest a cable equalization circuit or product.

Here's a description of an equalizer, but it's for differential cable instead of coax. If your attenuation is high, you'll have better luck putting the equalizer on the transmitting end.
https://www.analog.com/library/analogDialogue/archives/38-07/equalizer.html

The armor shouldn't affect coax performance.

 

[DrWhoF]

Re: Video and power

Resistance of the core is 68ohms.
Resistance of coax shield and steel wire shield connected together is 18ohms.
I used camera and monitor and connected them to this cable. Monitor can not synchronize to this signal and picture is black&white and has no details.

 

[throwaway18]

Video and power

If you can't get sync then I think you are going to need an amplifier at the camera end. Something like this

http://www.decadenet.com/cab/cab.html

The numbers for a kilometer of RG58 and RG6 come out to 20-25dB loss at 4.5MHz
I think you will need lots of preemphasis and eq to get a descent picture using composite video.
Another amplifer in the middle of the coax might be a good way to do it.

It might be better to use a modulator and a dsitribution amplifier to put the video on a low VHF channel. A receiver would be required to convert back to composite video.

I take it your camera has a seperate power input rarther than accepting power on its coax connection?

The next question re power supply is what is the volts/amps/power rating of your camera.

It might be best to seperate the coax shield and the armour and use them to send DC volts to the camera. Check if the outside of the video socket is connected to it's 0V power connection.
You will need a regulator at the camera end.
In theory you could put DC power on the ineer and shield of the coax with some DC blocking capacitors, some clamp diodes to avoid putting a big pulse into the equipment at switch-on when the caps are dischared, a well regulated power supply and a strign of chokes to avoid the video signal getting shorted to groung by the power supply.

 

[IanP]

Re: Video and power

The results of your test are moretheless what I have expected ..
Would it be possible to take a photo of this cable and post it here?
What I suspect is that your cable is a logging wire ..
I am aware of two types of these logging wires, both used in geological surveys:
- type one, older, data speed below 1200bps
- type two, more recent, possibly suitable for baseband video transmission (with low and high frequency compensation)
Let's hope you are dealing with "type two" as type one is completely useless for what you are trying to do ..

The primary gol will be to transmit good quality video signal; the issue of powering the camera can be dealt with later ..
To coduct further tests you will need at least one line driver with equalization; you can get it off-the-shelf (see throwaway18's post) or you can build one by yourself ..
Below is an example ..

Regards,
IanP

 

[xxargs]

Re: Video and power

First of all, measure the resistance of the core and the shield to estimate dc-loses in the cable ..
Next, connect video from any source (video generator, camera, vcr ..) to this cable and observe (connect the other end to video-in of any TV, monitor ..) what is the quality of signal that passes through ..

Is the coax electrically isolated from the steel?

Regards,
IanP

Power feed capability is important, but check video signal capability on this long distance first!!


1000 meter is long distance for video in base band and you can possible see unwanted effects as bluring or tex. 'H' -> '|+' depend of low resp high frequences have different loss and traveling speed in cable, also after using amplifier to composate loss in cable.

frequency range from 50 Hz to 1 MHz in coax is very vary depend of frequency both in loss and group delay and can give affect on video quality. This effect can only see on very long cable test and not see in testing short cable length in lab.

This effect is near impossible to fixing up afterward with filters or equlizer ie. part of distorsion is depend of group delay distorsion [1].

And take care of loss (attenuate) in cable, most case video input have not som much dynamic lower than video:s 1 Volt p-p

Test if you can make acceptable video quality on _whole_ cable length (without DC-feed in first try) connected to selected cable drum before costly cable installations job - is more or less catastrophic for you or companys economy if you find out cable is unusable for you application and need installation new other type of cable and/or equipment again....

Is not guarantee of enough video quality on 1000 meters also if looks nice testing on 500 meters cable drum - test whole length!! my experience with radio and leaky/radiating cable say me always latest few hundred meter from end have very big influences if projekt going succefully or not...

Remember: most thing in transmissions line world have ' y^x' in rules depend of distance. Nothing happend in first short distance (x) even if 'distorsion of any kind' is quite high per meters, but in long distance is closest possibly value to '1' on 'y' importent to make project succefully or not, and very much happens in latest few hundred meter from end side...

ex.

(ref 1 Volt p-p input on cable)

99.9 % power left per meter cable (att. 0.1 % power or -0.0043 dB/m)

0.999^10 = 0.99 power left or 0.995 volt p-p after 10 meter
0.999^100 = 0.948 or 0.9512 volt p-p after 100 meter
0.999^500 = 0.6064 or 0.7787 Volt p-p after 500 meter
0.999^1000 = 0.3677 or 0.6064 volt p-p after 1000 meter ( att. 4.3 dB )

99.7 % power left per meter cable (att. 0.3% power or -0.0132 dB/m)

0.997^10 = 0.9704 or 0.9851 volt p-p after 10 meter
0.997^100 = 0.705 or 0.8605 volt p-p after 100 meter
0.997^500 = 0.2226 or 0. 4718 volt p-p after 500 meter
0.997^1000 = 0.0496 or 0.2226 volt p-p after 1000 meter (att. 13.2 dB )


even small change of cable parameter can give costly result, i first case with 99.9% power left per meter possibly not need video amplifier before equipment after 1000 meter

but using cable at 99.7% power left per meter you need using extra videoamplifier depend of 0.2 Volt p-p video signal is not good for videomonitor after 1000 meter.
etc.

Very small change on 'y' parameter from estimated calculated value in real world, in wrong direction can increase cost very much or ruin whole installations/functions on very long cable distance project. Right cable select is very importent here...


[1] is possible with allpass-filters , but > 90 percent of filter books skip this chapter to how calculate and build analog all pass-filters with wanted parameters and mostly decribe only digital FIR-filters for equalzing already sampled digital modulated signals...




/Xxargs

 

[DrWhoF]

Re: Video and power

I have access to all 3 cable elements at both ends and I can send power without mixing it with video signal.
Is it possible to use more common components than the LTC1256?

 

[IanP]

Re: Video and power

I would try to transmit video through the centre core, +power through coax braid and 0V-GND through steel armour ..

If you have difficulty in obtaining the LT1256, here is another version of a video line driver with compensation, this time based on descrete components ..
The LH0002 can be as difficult to find as the LT1256, but its function in this circuit is unity gain "power" amplifier and it can be replaced by any "video" rated opamp such as EL2020, AD818, or similar .. or it can be re-constructed following the LH's internal schematic diagram (see picture below) ..

Regards,
IanP

 

[DrWhoF]

Re: Video and power

Thank you very much for this line driver.
I'm going to try it ASAP.
What is the resistance of input VR?
Thanks

 

[xxargs]

Re: Video and power

Thank you very much for this line driver.
I'm going to try it ASAP.
What is the resistance of input VR?
Thanks

hmm. I not check with spice, but input input seems high impedace and you need connect 75 ohm resistanse between input and ground for matching to cable impedances , and possibe 75 ohm in serial (low output impedances stage or to ground if a current mode stage and high impedance output) on output on same reason.

this stage is so easy and few component you can using demo version of spice-simulator a Pspice or microcap (latest can using smith chart for impedance view on port)


It not so importent if cable are very long and have high attenuate on all used frequency (>10, dB 50 Hz - 5.5 MHz for CVBS, RGB or S-Video) each side, but short distane cable on input and output give reflexes and shadows on picture if input and output in amplifier not have 75 Ohm impedances.

And use 75 Ohm attenuator (75 Ohm PI-pads) or adjust gain in amplifier block to adjust level to monitor around 1 Volt p-p on inport, normaly monitor without AGC on input have very small working dynamic on input (guess 0.5 - 1-5 Volt p-p Ohm)



/Xxargs

 

[IanP]

Re: Video and power

The typical output level from cameras is 1Vpp and it they will easily drive any load from ∞ down to 75Ω, or less ..
In this case you can use 1kΩ variable resistor and there is no need for 75Ω termination at that end ..

Regards,
IanP

 

[DrWhoF]

Re: Video and power

IanP, what transistor types will you recommend as Q1, Q2, Q3 and Q4?

 

[IanP]

Re: Video and power

This is only baseband video so any small signal NPN and PNP transistors will work ..
BC557, 2N3906, .. --> PNP
BC547, 2N2222, .. --> NPN

Regards,
IanP

 

[throwaway18]

Video and power

The typical output level from cameras is 1Vpp and it they will easily drive any load from ∞ down to 75Ω, or less ..

This is true of cameras intended for CCTV with video out on a BNC socket. I'v seen a minature camera module that could not directly drive a low impedance transmitter input.

 

[IanP]

Re: Video and power

I'v seen a minature camera module that could not directly drive a low impedance transmitter input.

That was the reason I suggested 1kΩ as the input variable resistor ..
All what the camera has to do is to deliver video signal of ≈0.5-1.5Vpp onto 1kΩ ..

Regards,
IanP

 

[DrWhoF]

Re: Video and power

Thank you very much for your help.
I built 2 line drivers and cionnected them at both ends.
I have almost perfect picture. Sometimes when a machinery is working nearby the colour "flickers" but it is not very often.
Once again thank you very much IanP and I donate 100 points for your help.