How to Better Wire the BNC Connector and GHz issue.

[Mechatrommer]

i just got my male BNC connectors and female-T to interface my future project with my newly bought function generator. I bought quite many (10 for the male) so i can further interface to my rigol dso or do some funky BNC connections stuff. Since its only the head/connector, so i'm going to wire it. I can just wire it with simple 2 black and red wires and ignore about anything else, but for the sake of information and better knowledge, i want to make a public query. Since there are issues such as data integrity, attenuation, reflection at high/RF frequency etc which i dont really fully understand.

Variaty of cables, even coax alone. My satellite TV cable is very hard to bend. My FG cable quite flexible. My Rigol, i'm not sure if its even coax, you know? that thin Rigol's probe cable. So the question...
1) What is the best (or better) coaxial cable around and preferebly easy to get everywhere? if there is difference, what makes one coax better from another?
2) Whats with normal 2 wire setup? or twisted type? for RF design/analysis/probing?
3) How do i define and measure an impedance of a wire/transmission line (or two close together, coaxed or twisted) with relation to their length? (or is this just a stupid question to ask?)

hope to hear some answer. Thanx.

[jahonen]

1) Generally, stiffer the coaxial cable, better it is. Best "cables" are actually solid copper tubes (usable to 100 GHz or so), so called rigid coax. Then there is semi-rigid coax (usable to 30 GHz or so) and finally, the ordinary "flexible" coax, usable maybe up to 1-2 GHz, depending on the length. Flexible coax exhibits changes in cable geometry when moving it around, so it might be unusable in reflection loss measurements if you need repeatability (see below). Since you mentioned BNC, that probably itself limits usable frequency range to less than 1 GHz, so you might get away with something like RG-179 (about same than average scope probe coax), or RG-58 (a bit thicker), if your cables are short. See some datasheets of the cables to get idea of the losses involved. Basically, thicker cables tend to have less loss.

2) 2-wire setup (i.e. loose wires) is usually not an option since impedance is completely unpredictable. For RF-design, things you want to measure are S-parameters, like input/output reflection loss (S11, S22) and transmission loss (S21, S12), For amplifiers, noise figure, intermodulation distortion (IP3, ACLR etc) and 1 dB compression point are common measurements. For this, you'll need tightly controlled impedance all the way. Probing is much more difficult with RF, since you really can't stick just about any kind of probe to your circuit without affecting internal matchings. Yes, there are microprobes and such things with controlled impedance all the way to the tip of the probe but these are not really things that average hobbyist will get.. So you must usually provide controlled impedance test points with proper coaxial connectors. Of course this depends your requirements, but that probably is more like worst case.

3) Measuring the impedance is somewhat difficult, and would involve a VNA. Basically, it is determined by the conductor geometry (thats why twisting the coaxial cable will change its impedance!) and permittivity of insulator dielectric.

Regards,
Janne

[Mechatrommer]

Thanx master Janne. the description is really serious, but i dont think i will ever afford to design up to GHz system/project before i can collect substantial amount of fund. I just wondering. So let put the spec down a bit. Say for designing 100-400MHz system, is loose wire acceptable? or any rough figure of what MHz loose wires will still be applicable? but you already gave me a rough figure on the limit of "ordinary" coax cable capable of, so i should find the coax cable instead of loose wires, thanx. Will search my collection if i got any spare coaxial type cable, or even cut my satellite TV cable, its still got more length than what the TV needs.

Having learnt from this blog that we have to take harmonics into consideration, then for 400MHz system, we should consider for cable thats capable of 10th harmonics ie 4GHz, ie semi-flexible-coax, right? about the length and impedance measurement need vna to analyze which is expensive for me, any general rule about the length, my mind saying so far that shorter cable is better, or some multiple of frequency? a quarter of wavelength? isnt it?

[Wim_L]

Be careful with using TV cable for electronics. Most TV signals are sent over 75 ohm coax, while a lot of electronics equipment is matched to 50 ohm. Mixing the two is undesirable. The connectors usually are specified for one or the other too, not both. There's a table showing some common cable types at http://en.wikipedia.org/wiki/Coaxial_cable

[jahonen]

Thanx master Janne. the description is really serious, but i dont think i will ever afford to design up to GHz system/project before i can collect substantial amount of fund. I just wondering. So let put the spec down a bit. Say for designing 100-400MHz system, is loose wire acceptable? or any rough figure of what MHz loose wires will still be applicable? but you already gave me a rough figure on the limit of "ordinary" coax cable capable of, so i should find the coax cable instead of loose wires, thanx. Will search my collection if i got any spare coaxial type cable, or even cut my satellite TV cable, its still got more length than what the TV needs.

Having learnt from this blog that we have to take harmonics into consideration, then for 400MHz system, we should consider for cable thats capable of 10th harmonics ie 4GHz, ie semi-flexible-coax, right? about the length and impedance measurement need vna to analyze which is expensive for me, any general rule about the length, my mind saying so far that shorter cable is better, or some multiple of frequency? a quarter of wavelength? isnt it?

Do not use that TV coax, get 50 ohm coax. As for "loose wire", I suggest that you make some coaxial cables with BNC connector in the other end and leave other end free (or get some pre-made cables and then cut them to half, you get two cables then), you can then solder the other end to your board. Not perfect but it works quite often, especially for frequencies you are aiming.

Yes, harmonics must be considered, but it is not hugely important thing unless you want to measure harmonics All the measurements I mentioned are done in vicinity of the main frequency. The limits are not strict, so you can still see higher frequencies, but attenuation will increase, so upper harmonics will appear in lower level. It again depends how accurate measurements you want to make. Cable (and connectors) matter more when measuring those reflection losses, but if you are not going for a VNA, then don't care about it.

The cable length is not very significant, as long you have properly terminated system, i.e. your measuring equipment has 50 ohm input impedance, connected via 50 ohm connectors to 50 ohm cable. Of course, the attenuation will increase at higher frequencies, but is not usually prohibitively high for usual lengths for these kind of purposes. It only matters greatly if you do some tricks with it (like open end cable with length of 1/4 wavelength will look like a short circuit from feed end, due to reflection).

Regards,
Janne

[saturation]

Just my 2c to those already posted.  Yes, make sure cable used has a terminated impedance of 50 ohms; RG59 is 75 ohms and can look fairly close, specify the cable as 50 ohms RG58 which is as cheap and flexible, and typical test leads of 3' or 1m will not be significant.

Compare against RG8.

http://www.bcar.us/cablespec.htm

The differences between other 50 ohms cables is cost, size and stiffness; you'll pay a little more for reduced signal attenuation per foot.

Use pure coax whenever possible.  You'll note some have adapters for BNC to alligator or minigrabbers, or as in your photo, for basic AWG wire.  These can cause issues, so if you make your own keep the non-coax segment as short as is practical; like in those photos a good portion of the cable is still coax.  Its flexible to use them in lower frequency applications so long as you are aware of them; jahonen's recommendation is best; direct soldering of cable into the board.  



One thing you can do with your signal generator, is actually test the effect of various connectors [ banana, clips, alligators, BNC splitters etc., ] against one that is just purely coax; then you'll have some idea what artifacts its causing.  Also, the BNC connector and BNC adapters can produce artifacts too at higher frequencies.  For example, BNC  is useful to about 4-11 GHz, and how high you can take them before reflections become substantial can differential a brand name connector from Amphenol versus no-name Chinese brands that 'look' them same.

http://www.amphenolconnex.com/products/bnc.asp

For practical purposes, about ~ 2GHz is the limit of BNC.

i just got my male BNC connectors and female-T to interface my future project with my newly bought function generator. I bought quite many (10 for the male) so i can further interface to my rigol dso or do some funky BNC connections stuff. Since its only the head/connector, so i'm going to wire it. I can just wire it with simple 2 black and red wires and ignore about anything else, but for the sake of information and better knowledge, i want to make a public query. Since there are issues such as data integrity, attenuation, reflection at high/RF frequency etc which i dont really fully understand.

Variaty of cables, even coax alone. My satellite TV cable is very hard to bend. My FG cable quite flexible. My Rigol, i'm not sure if its even coax, you know? that thin Rigol's probe cable. So the question...
1) What is the best (or better) coaxial cable around and preferebly easy to get everywhere? if there is difference, what makes one coax better from another?
2) Whats with normal 2 wire setup? or twisted type? for RF design/analysis/probing?
3) How do i define and measure an impedance of a wire/transmission line (or two close together, coaxed or twisted) with relation to their length? (or is this just a stupid question to ask?)

hope to hear some answer. Thanx.

[Mechatrommer]

thanx guy. i got some picture now. and thanx Wim for reminding me of 75ohm vs 50ohm impedance cable. i know there are sort of thing, but it seem i just didnt get it. one last noob question... whats exactly the terminated system, terminated impedance? hihi.

[saturation]

Terminated impedance meant the final transmission line impedance from the effects of, source, line and load are matched.  I guess I should have said the cable has a 'characteristic impedance' of 50 ohms to refer to the cable only, my error, see below.

http://en.wikipedia.org/wiki/Electrical_termination



Zs and Zl can be resistors of 50 ohms each, but the cable's Zo, called the characteristic impedance, depends on the inductance and capacitance of the cable, per unit length, and bypassing all the math to summarize, its units are in ohms too.

http://en.wikipedia.org/wiki/Characteristic_impedance

If you use the wrong cable type, say Zo = 75 ohms, RG59, it will create an impedance mismatch even if you terminate each end as 50 ohms, because the cable is part of the system.

By example, the Hantek 3x25 has an source impedance of 50 ohms, Zs, and using '50 ohm' cable, Zo, the load end must also be 50 ohms, Zl, for the most efficient signal transfer.  If the Rigol 1052e is the load, you need to put a 50 ohm terminator, Zl,in parallel to the 1 megaohm so the terminated impedance is ~ 50 ohms. 

thanx guy. i got some picture now. and thanx Wim for reminding me of 75ohm vs 50ohm impedance cable. i know there are sort of thing, but it seem i just didnt get it. one last noob question... whats exactly the terminated system, terminated impedance? hihi.

[Mechatrommer]

Janne, the RG-179 is 75ohm i checked in wiki, i dont get why you make the suggestion as most device is 50ohm.
for my own later reference... http://en.wikipedia.org/wiki/Impedance_matching.
thanx saturation for the links, gave me some direction on what to read.

a.jpg) ok, made my first quick "invention" with bnc's, tried to make a 50ohm terminator (2x100ohm)
b.jpg) connected it from my Hantek 3x25 AWG to Rigol 1102E(mod) directly without terminator, 100Hz 3.5Vpp sine is generated, Rigol showed what is set in 3x25 Software. everything is fine!
c.jpg) with the terminator installed, then the peak went wrong. any explanation with this confusing 50ohm termination stuff?

[Mechatrommer]

well i think what i got is RG59 (or RG6) 75ohm connector, this is closest i found in ebay saying its RG59 (for video stuff)
http://cgi.ebay.com.my/6x-BNC-Male-RG59-cable-Crimp-75-OHMS-Connector-131-/300530023762?pt=LH_DefaultDomain_0&hash=item45f8fc3d52

i bought is local web here (no description whatsoever)
http://www.lelong.com.my/bnc-connector-10-pieces-pack-78810266-2011-03-Sale-P.htm

but it looks damned identical to the 3x25 bnc cable/connector. i dont know where is the physics that tells they are different.
from left (bnc to wire (screw type), bnc to coax (screw, solder/crimp type, 3x25 cable)

ps: i hate to buy another stuff, i'm not this professional. so i just assume they will not make any difference (RG6/RG59 75ohm to RG58 50ohm) and become ignorant

[jahonen]

Janne, the RG-179 is 75ohm i checked in wiki, i dont get why you make the suggestion as most device is 50ohm.
for my own later reference...

Sorry, I meant RG-174.

Regards,
Janne

[Mechatrommer]

thanx janne

[deephaven]

What sort of systems do you wish use with this high frequency stuff? This will have a large effect on how many harmonics (i.e. bandwidth) you should be considering.

If it's a modulated radio frequency system, then you don't need much more than the carrier frequency because the spectrum will be the carrier frequency +/- the frequency component of any modulation (with a multiplication factor on the modulation component depending on the form of modulation and it's spectrum).

If it's a data system, then you need to consider a higher bandwidth to accommodate some of the harmonics. Unless you really want to see the true nature of the signal, you can probably get away with a lot less than a 10X sample rate. This is because practical systems don't have vertical-sided square waves, but more curved looking waves. What is important here is the 'eye' which you may have seen elsewhere. What this means is that the 'eye' should have a good vertical and horizontal window relative to the (possibly recovered) clock so that the circuit can make a reliable decision on whether the signal is a logical 'high' or 'low' at that point. Also in a data system, matched impedances are important otherwise you will get reflections which can cause ringing and overshoots.

On the waveforms you showed, the signal will drop 50% when you terminate it.

[saturation]

The rated output of the 3x25 is 3.5Vp into 50 ohms, it will however produce more Vp into 1megaohm, but the amplitude will not be stable nor the waveform quality be maintained throughout its frequency range, to at least 20 MHz.  It depends on the reflections being sent back and the degree of interference from it, and it will vary by type of waveform and frequency.  At 100 Hz, that's pretty much just DC as far as transmission lines are concerned.

Set the output of the Hantek to its max, 3.5Vp, without the terminating resistors.  On the Rigol the output will be more that 3.5Vp.  Now put the terminator, it should now be 3.5Vp.

Now for the acid test, use the square wave output of the SYNC port, which as we discussed, is very sensitive.  You asked about harmonics earlier, well here is a way to test as the square wave Fourier components will push at least  9x the fundamental.

Check the output, unterminated at 100 Hz, note the Vp, it will be a high value, may near 4Vp.  Do the same, with terminated.  I think it should be 1Vp.

Now ramp the frequency up to 10 MHz.  You will see better edges as part of 90 MHz component of the fundamental, the limits of your modified Rigol.  You should have a stable square wave at 1Vp with some ringing depending on you cables and connectors but it will be noticeably square.  Now pull your resistors out and see what happens.  

Janne, the RG-179 is 75ohm i checked in wiki, i dont get why you make the suggestion as most device is 50ohm.
for my own later reference... http://en.wikipedia.org/wiki/Impedance_matching.
thanx saturation for the links, gave me some direction on what to read.

a.jpg) ok, made my first quick "invention" with bnc's, tried to make a 50ohm terminator (2x100ohm)
b.jpg) connected it from my Hantek 3x25 AWG to Rigol 1102E(mod) directly without terminator, 100Hz 3.5Vpp sine is generated, Rigol showed what is set in 3x25 Software. everything is fine!
c.jpg) with the terminator installed, then the peak went wrong. any explanation with this confusing 50ohm termination stuff?

[saturation]

[ edited for clarity]

BNC connectors will work for any cable you can stick it into.  The problem is more limited by the characteristic impedance of the cable, and less the connector, you choose.  However, since the characteristic impedance is really a property of inductance and capacitance, better manufacturers make separate 75 ohm vs 50 ohms BNC connectors, so they are not identical.

I posted a link for details earlier, but here it is again:

http://www.amphenolconnex.com/products/bnc.asp#specs75

I don't think it will make a big difference for under 100 MHz, using either connector, but whenever possible, buy the right spec.

well i think what i got is RG59 (or RG6) 75ohm connector, this is closest i found in ebay saying its RG59 (for video stuff)
http://cgi.ebay.com.my/6x-BNC-Male-RG59-cable-Crimp-75-OHMS-Connector-131-/300530023762?pt=LH_DefaultDomain_0&hash=item45f8fc3d52

i bought is local web here (no description whatsoever)
http://www.lelong.com.my/bnc-connector-10-pieces-pack-78810266-2011-03-Sale-P.htm

but it looks damned identical to the 3x25 bnc cable/connector. i dont know where is the physics that tells they are different.
from left (bnc to wire (screw type), bnc to coax (screw, solder/crimp type, 3x25 cable)

ps: i hate to buy another stuff, i'm not this professional. so i just assume they will not make any difference (RG6/RG59 75ohm to RG58 50ohm) and become ignorant

[jahonen]

For this context, perhaps I can refer to "1k probe" I once studied. That certainly gives idea how good such simple connection can be, if terminated properly.

Double termination is not absolutely necessary, but it matches both ends, suppressing reflections from either direction. The 1k probe is an example of the system where only load is matched. That means 1k resistor end is not matched, but in theory it does not have great significance since load (your 50 ohm terminated measuring equipment) absorbs the signal fully. In practice, there is some mismatch, thus there will be some ripple in the frequency response, as can be seen from my measurements.

Regards,
Janne

[Mechatrommer]

Set the output of the Hantek to its max, 3.5Vpp, without the terminating resistors.  On the Rigol the output will be more that 3.5Vpp.

no, the reading is less than 3.5, somewhere 3.3V (b.jpg)

Now put the terminator, it should now be 3.5Vpp.

no, the reading is halved, somewhere 1.8V (c.jpg)
the 50ohm i was making is parallel to dso, maybe thats a mistake?

i tried series 50ohm to dso (d.jpg), the reading is the same as direct connection, ie no terminator (~3.3V)
i tried 100Hz, assuming (or imagining) to minimize the reflection/loss/ringing effect.

Now for the acid test, use the square wave output of the SYNC port, which as we discussed, is very sensitive.  You asked about harmonics earlier, well here is a way to test as the square wave Fourier components will push at least  9x the fundamental.
Check the output, unterminated at 100 Hz, note the Vpp, it will be a high value, may near 4Vpp

no, still the same at ~3.3V

Do the same, with terminated.  I think it should be 1Vpp.

no with series 50ohm, still the same at ~3.3V (e.jpg). but yes with parallel 50ohm (1V). with both parallel and series 50ohm installed, its ~0.8V

[Mechatrommer]

now i think series 50ohm resistor is not the way. so i stick back to signal to ground 50ohm (parallel) and do some measurement... more confusing stuff
... 10MHz square (sync out) and sine (output) using 3x25 cable ...

[Mechatrommer]

more confusion... the same signal (10MHz square sync out) as above using rigol probe...

[Mechatrommer]

great! with rigol probe, its getting more and more complicated. now i dont know which reading to trust.
10MHz square synch out from 3x25. rigol set at 1V/div. notice the termination placement (ie close to the rigol and close to the 3x25)
the most notable difference is at 10X setting with 50ohm resistor termination, i dont know what the hell is happening.
maybe this is called ringing (or the reflection), but it will make it different if we put the termination on the different side, i dont know, maybe i'm just wasting time here.
ps: with 3x25 cable, whether the termination is close to 3x25 or rigol, the shape and amplitud is something like e&f in top part of efgh.jpg

[jahonen]

That BNC with screw terminals looks just horrible It is probably meant to be used for DC-stuff. Do not put it even in same room where high-frequency measurements are being done

Regards,
Janne

[Mechatrommer]

That BNC with screw terminals looks just horrible It is probably meant to be used for DC-stuff. Do not put it even in same room where high-frequency measurements are being done
Regards,
Janne

maybe i should make proper cable connection next time (bnc to bnc), proper coax cable, soldering, crimping and stuff. instead of that quick screw in connection. sigh

[Mechatrommer]

What sort of systems do you wish use with this high frequency stuff?

i want to make bnc connection from/to my hantek 3x25 signal generator or rigol dso to/from any project that i will be doing. anykind of project, i dont know, but i'm sure it not going to be GigaHertz, just a few or ten or hundred MHz. but i want to get some rough picture/advice on high quality cable connection diy stuff.

This will have a large effect on how many harmonics (i.e. bandwidth) you should be considering.

i guess my low tech tinkering/test/pictures earlier proved that.

[saturation]

now i think series 50ohm resistor is not the way. so i stick back to signal to ground 50ohm (parallel) and do some measurement... more confusing stuff
... 10MHz square (sync out) and sine (output) using 3x25 cable ...

You're procedures are confusing to me, mecha.  Some general comments.

efgh.jpg: I presume you mean the images are from left to right, up then down.
ef
gh

All quoted images seem to have your vertical setting at 1V/div. 

e = that's what the output looks like if unterminated, ~ +3.5V.  You cannot get a stable sync with that signal, but you can with f, ~ +1V.

g = that's ~ 7VACpp. 

h = as expected, at ~ half of g, but it looks like almost 4 V, it suggests something is wrong; maybe your resistors of 2x "100" ohm in parallel is <> 50 ohms.

Your reply below sounds like you've forgotten how to read graticules. 

Image b.jpg is ~7VACpp. 
Image c.jpg is correct, that is 3.5VACpp.

Series    Termination means parallel, as in the image I left for reference.

If you use 100 Hz to avoid ringing, you wont' see the effects of correct impedance matching, like it was demo'd with image e and f.  Its clearest when your cable is 1/10th the wavelength of the test frequency, if you use a 1 meter cable, the test frequency should be >= 35MHz.  Using the square wave from the sync output port at 10 MHz, the harmonics are above that, so you can see the effect more strikingly.

Enjoy.

Set the output of the Hantek to its max, 3.5Vpp, without the terminating resistors.  On the Rigol the output will be more that 3.5Vpp.

no, the reading is less than 3.5, somewhere 3.3V (b.jpg)
no, the reading is halved, somewhere 1.8V (c.jpg)
the 50ohm i was making is parallel to dso, maybe thats a mistake?
i tried series 50ohm to dso (d.jpg), the reading is the same as direct connection, ie no terminator (~3.3V)
i tried 100Hz, assuming (or imagining) to minimize the reflection/loss/ringing effect.

[Zero999]

Where do you live?

In the UK we have UHF PAL TV which are uses 50 Ohm co-axial cable and is fine for most purposes.

If I remember rightly, 75 Ohm is more common for VHF FM radio receivers.