Page 1 of 1

Cableing

Posted: 24 Oct 2015, 05:16
by Mark Rowley
What is the best choice for a PMT cable?

Re: Cableing

Posted: 24 Oct 2015, 06:04
by iRad
Are we talking about the coax cables coming from single or dual jacks from the back of a scintillator to your bias adapter/soundcard/survey meter? Or are we talking about the type of wires to use when wiring a PMTs voltage divider?

If talking about the coaxial cable connection between devices, then any GOOD QUALITY 50ohm coax will do, shorter the better. What may matter more is the quality of the connectors and how well they are connected to the cable. These cables are used for supplying high voltage DC and carrying a low voltage DC pulse. Fidelity and impedance matching is not usually any issue. Signal loss of the low voltage pulse can be an issue on longer cable runs, where everything can become an issue. Most of us use cables short enough to avoid any issues associated with long runs, especially for hobby spectroscopy.

For PMT VD wiring, any good quality solderable wire will work. BUT, I always use PTFE insulated silver plated multicore 24/22 gauge wiring myself. Why skimp pennies and give yourself (or your customers) headaches. Cheers, Tom

Re: Cableing

Posted: 24 Oct 2015, 08:48
by Steven Sesselmann
Mark,

Yes, 50 ohm flexible coaxial cable is what you want, not always so easy to find. The typical thick one is called RG58C/U and the thinner one is RG174/U. If I am not mistaken, the /U at the end means it's a flexible multi core cable.

I carry both sizes in stock.

Steven

Re: Cableing

Posted: 24 Oct 2015, 18:08
by Mark Rowley
Thanks guys for the response. I have several RG179 cables that I can convert to BNC, but its 75 ohms. However, I'm not far from a shop that carries RG174. I'll stop by on Monday during my break.

Assuming I use good quality cable, what would be the maximum length before incurring signal loss?

Re: Cableing

Posted: 25 Oct 2015, 07:17
by iRad
Mark
In most cases a 75 ohm cable will work fine, but why would you use it when you have to change out the connectors anyway. Especially as most 75 ohm cables are cheaply made for cable/TV applications. As mentioned previously, quality connectors and their connection to the cable are usually what is most important. So why not just buy a quality pre-made 50 ohm BNC cable in the length you need?

Steven is correct in stating that "RG58C/U and the thinner one is RG174/U" are the cable types you should be looking for. Even then, quality can vary from brand to brand. Belden cable is usually of good quality, but they make many types. Pomona, Jameco, & Pasternack make great quality pre-made BNC cables that are easy to find and order. Get something cheap (or do-it-yourself) and you could risk 1000 volts or more at the loose end of a cable that's lost its connector, doing damage to your equipment, yourself, or someone else. I have nothing against do-it-yourself connectors, but make sure you know what your doing... Sometimes when dealing with different connectors on each end of a cable, do-it-yourself connectors are the only reasonably priced option. But then again, high quality name brand connectors can run $20, $30, even $50 or more each in some instances.

As for length, stick to under 5 to 6 feet and you should have no issues. Any longer lengths and it would require a lot more knowledge of the application to say at what point single loss or noise would affect the application. Signal loss and ringing/noise does not happen suddenly at a specific distance of cable, but increases at a steady rate over the cables length until it is no longer manageable at some distance (depending on the application). This is where impendence matching and low loss high quality cables could really come into play. But I'm feeling that you probably don't really have a worry there.

Cheers, Tom

Re: Cableing

Posted: 25 Oct 2015, 09:28
by Steven Sesselmann
Just one important point I would like to add to this thread, stay clear of foam core coaxial, these cables have a white foam core rather than the solid polypropylene core. The porous core cables won't hold high voltage, I discovered this with one of my neutron detectors which needed about 1400V bias, which was ticking away in the absence of neutrons, eventually I narrowed the problem down to the SHV cable which was an old one I had made myself from foam core cable obtained at the local electronics store.

Re: Cableing

Posted: 25 Oct 2015, 12:17
by iRad
Excellent point. Foam core cable is very typical of the cheap cable used/sold often for cable/TV and video applications. Those cables were never meant to handle the kind of voltages needed here. And of course, even the good stuff may not be able to safely carry current higher than the few mA we need to drive the typical PMT. Don't use BNC connectors for high voltage high current applications (~>5mA), where increased danger of exposure to heart-stopping current could be realized.

Re: Cableing

Posted: 25 Oct 2015, 22:22
by Mark Rowley
Hi Tom,
Think maybe I mis-typed. Never had any intention of using the 75 ohm cable. The shop near me stocks RG174/U. Not sure but its either Amphenol or Belden. Doubt I'll have to put on connectors as they have tons of pre-made lengths, but I am pretty good at doing it myself if needed. No worries there.
I'll double check on the foam core....totally bad news it sounds and it will be avoided.

A 10 foot cable length would be ideal, but I can make 4 or 5 feet work. Still trying to decide where to place the castle, but its home will be where a 4 foot cable will do.

If all goes to plan, everything should be here for a test run by Tuesday evening.

Re: Cableing

Posted: 26 Oct 2015, 08:19
by brehwens
For my first detector, I "accidentily" bought a 40€ BNC cable. Works great. I doubt it is any better than what ebay offers, you can easily find any cable type with any connector you could ever want... and for the price you can't go wrong....

Re: Cableing

Posted: 28 Oct 2015, 12:22
by pietkuip
The pulses from NaI detectors are not fast, so one does not need to keep the cables short.