I’ve seen various products out there for extending USB cables beyond the usual 5 metre limit; they seem to consist of:
1 Ready-made 10 metre booster cables
2 Paired booster adaptors USB>Cat5 and Cat5>USB
3 Huge clunky KVM boxes.
Number 1 is too short for my requirements (I know they can be chained, but that won’t work for me either)
Number 2 is rather expensive for what you (seem to) get…
Number 3 is expensive as well as pointlessly overspecified for me.
What exactly is involved in boosting a USB signal; it can’t be that complex, can it?(there are only two conductors carrying signals) Is there some way I could make my own booster?
For the record, what I’m trying to do is build a birdbox with a webcam inside; it’s either a case of finding a way to extend the reach of a USB webcam (or the guts of one, anyway), or buying a camera module and a TV capture board…
If you are willing to go to the effort of creating your own cable:
Here is a UseNet thread where someone created their own USB cable to solve the same problem you are having (extending the usb cable on a webcam)
Towards the bottom of the page, one of the posters talks about how he made his own 50 foot USB cable using thin net (BNC) with HVAC thermostat cable. He then matched the PIN with this guide.
It is complex because the signals are bidirectional. The signal is being transmitted by both the host (PC) and the device (USB peripheral). So your signal booster would have to know the USB protocol and decide which direction to boost the signal from. This is one of the functions of a USB hub.
Starbury’s recommendation of just making a longer cable does not do any boosting of the signal. It will probably work as the cable length limit is just something the USB committee chose so they wouldn’t need to worry about what happens if the signal is degraded by long cable runs.
Note also that you’ll have more trouble extending cable length if the webcam is expecting to power itself from USB. If the webcam has it’s own power source, you can likely get away with the longer cable.
Hey! I just also discovered that the CMOS image chip in my webcam can see infra red - I pointed the TV remote at it and it looks (on the screen) like a white LED, although illumination with an IR source will only enable the camera to see in mono (I just tried a red gel over the end of the remote and the webcam still sees it as white.
Are you restricted to using USB? A FireWire cable can potentially be as long as 100 meters. Realizing that most FireWire products aimed at the Mac market (I assume you have a PC), I did a quick Google search, and it looks like PC FireWire webcams do exist.
I second Belrix’ suggestion. I’ve used X10 wireless cameras and they work nicely over a decent range. You can choose between batteries – impractical in your case – or wall power, too. The power module is separate from the camera, so you can put the camera in the birdhouse and the power arrangements underneath. Just make sure the power you set up is sheltered from the elements.
“Blue canary in the outlet by the lightswitch” indeed.
Ummm… why don’t you just get an inexpensive powered hub to bridge the signals - ie 5 meter to hub input > (hub) > five meter from one of hubs outlets. 10 meters total.
Interesting… is there somewhere that I can find out more about this?
Is it not possible to construct a simple bidirectional line amplifier with a few transistors?
A lot of digital signaling isn’t amplified, it’s repeated; rebroadcast for distance applications. There’s probably a really good reason for this but frankly, I forget majority of the details but here goes.
Digital signalling is composed of lots of different frequencies that propagate at slightly different speeds. After a certain distance, the signal shape kind-of smears. If you simply amplified the signal, the degradation is just amplified & the signal continues to smear until it arrives completely botched. By repeating, you start each new leg of the trip with a fresh, clean signal.
The twist in the cat5 cable extender above tends to combat this, the inductance of the cable changes the resitance according to the frequency of the digital signal components and makes the signal smear less over distance. A standard USB cable may not plan for this.
Here is what the inside of an active USB extension cable looks like - I think it must be fairly safe to assume that they wouldn’t have done it like this if it could be done with a few transistors hmpf!
Tonight I’m going to hack and splice this 5 metre cable with 20 metres or so of shielded cable and see if I can get anything through it.
Given that options 1 and 2 are not really scalable (if I want to add a second camera, I need more expensive hardware… it looks like firewire could be the way forward.
Please could somebody confirm the maximum range of firewire and whether a decent range of say 30 metres is achievable with nothing more than a long cable (i.e. no repeaters or hubs in the line).
Also, does firewire on a PC typically auto-detect devices in the same way as USB?
Awww nuts! I think the info on that page is wrong, look:
So the maximum bandwidth is available at the maximum cable length? - I think there’s been a decimal place slip there; I reckon it should say 100 centimetres for 800Mbps.
Everywhere else I have looked is citing similar extension limits to USB (without repeaters).
Looks like firewire may not be such a rosy option after all.
I’ve just cracked open a USB webcam that I had lying about and there are two PCBs inside one of them carries the imaging module and a couple of electrolytic capacitors, the other carries a couple of ICs and a few discrete components; the two boards are linked by two 8-pin connectors.
I’m wondering if I can get away with inserting a length of cable between the camera module and the main circuit board, if this is the case, it will certainly make the remote camera very small…
No go. You end up with each amplifier output feeding the input of the other - one great big feedback loop.
Extending the guts guts of the camera as you are describing them would probably not be possible, either. I’d expect that at least some of those connections carry high frequency signals, and you’d have a hard time extending that over thirty meters.
Your best bet is probably to go with a simple camera with a composite output. Use a good coax cable or a TV extender RF rig to bring the signal back to the PC. There you could use a Buzz box on USB to digitize the video. Being USB, you could add more Buzz boxes easily and they aren’t that expensive. They don’t provide real whoopy video quality, but then neither will a webcam.
**I thought of that (well, eventually), I suppose you’d have to have each one switching the other one off when it is active, but that probably just wouldn’t work anyway.
I think you’re right; it looks like I can pick up a USB video capture module for about £25 and I think I can get the camera board for about the same. the most expensive bit is probably going to be the cabling.
Yeah, that really torques my nut sometimes. I can buy a SCSI board with all of its complicated electronics for about about EUR 50, but a printer cable (a piece plain cable with a couple of connectors) with a mini-centronics will set me back over 60.
Why are you fixated on USB? As has been said before, a Firewire port would be an option, but why not roll back the years and go through the serial port? Here is a ready made $y$tem which would work up to 200 ft. You seem like the kind of guy who could rig up something similar for a lot less money.
