Maker Pro
Maker Pro

Question for BNC terminator.

goodphy

Feb 19, 2015
2
Joined
Feb 19, 2015
Messages
2
Hello.

I have a Oscilloscope having selection for 1M, 75 and 50 Ohm. And all BNC cable in our lab has characteric impedance of 50 Ohm. Thus I simply thought OSC is always to be 50 Ohm for impedance matching. But when the device (ex: DG535 delay generator) has output impedance of HighZ, people say I have to set Oscilloscope as 1M ohm for not having signal interference, although BNC is still 50 Ohm. Could u tell me how this works? And people even say it is better to equip 50 Ohm terminator to the BNC cable end to prevent reflection. In this case T adapter (instead of in-line type) is used. It looks that T adapter is simply signal divider thus inserting additional load (terminator) to one side of T adapter for impedance matching is...hard to understand. Please help me to clear this confusion.
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,878
Joined
Jun 21, 2012
Messages
4,878
A lot depends on specific output characteristics of the device you are trying to observe with the oscilloscope. The DG535 delay generator appears to have several manufacturers. At least one of these has multiple output characteristics to match several popular loads, including high impedance, 50Ω impedance, TTL or ECL input impedances, and others.

The general objective is to preserve the rise and fall time characteristics of the delay generator by matching impedances at the source, the load, and in the cable connecting the two. A cable with BNC terminations and 50Ω characteristic impedance is common for interconnections of laboratory equipment. When connected to a 50Ω source driving a 50Ω load, maximum power transfer occurs and there are no reflections at the load.

Driving a 50Ω coaxial cable with a high impedance source will not cause reflections if the cable is terminated in a 50Ω load, for example at the oscilloscope. However, the high impedance source will be greatly attenuated by the 50Ω impedance presented by the cable and its attached 50Ω load. If the 50Ω coaxial cable is instead terminated at a high impedance oscilloscope input there will be pulse reflections... at both ends of the cable! Whether this causes any problems depends on the length of the cable and its attenuation.

The symptoms to look for are ringing at the pulse edges as observed on the oscilloscope, a sure sign that the impedance mismatch is having an observable effect. It is also possible that the ringing is low enough in amplitude and dampens quickly enough to not cause problems, but that is a poor approach to engineering. It is always best to match impedances at both ends of connecting cables when dealing with fast pulses. You would do this by applying the BNC tee at the source and connecting a 50Ω load to one arm of the tee, setting the oscilloscope input to 50Ω.

As mentioned earlier, if the source is high impedance, connecting the 50Ω load at the BNC tee will result in considerable attenuation of the source signal. This may or may not be acceptable, but at least there will be no reflections between the source and the oscilloscope. The best solution of course is to set the source to have a 50Ω output impedance if that is possible.

Finally, you should search the web for information on transmission line theory to see how this all works, both in theory and practice. Reflections on transmission lines are an age-old problem that was first observed on trans-Atlantic underwater telegraph cables. It has only become more important with modern signal processing proceeding at gigahertz data rates. Learn the theory, learn the math, and be prepared for the 21st century revolution in communications and data processing.
 

goodphy

Feb 19, 2015
2
Joined
Feb 19, 2015
Messages
2
A lot depends on specific output characteristics of the device you are trying to observe with the oscilloscope. The DG535 delay generator appears to have several manufacturers. At least one of these has multiple output characteristics to match several popular loads, including high impedance, 50Ω impedance, TTL or ECL input impedances, and others.

The general objective is to preserve the rise and fall time characteristics of the delay generator by matching impedances at the source, the load, and in the cable connecting the two. A cable with BNC terminations and 50Ω characteristic impedance is common for interconnections of laboratory equipment. When connected to a 50Ω source driving a 50Ω load, maximum power transfer occurs and there are no reflections at the load.

Driving a 50Ω coaxial cable with a high impedance source will not cause reflections if the cable is terminated in a 50Ω load, for example at the oscilloscope. However, the high impedance source will be greatly attenuated by the 50Ω impedance presented by the cable and its attached 50Ω load. If the 50Ω coaxial cable is instead terminated at a high impedance oscilloscope input there will be pulse reflections... at both ends of the cable! Whether this causes any problems depends on the length of the cable and its attenuation.

The symptoms to look for are ringing at the pulse edges as observed on the oscilloscope, a sure sign that the impedance mismatch is having an observable effect. It is also possible that the ringing is low enough in amplitude and dampens quickly enough to not cause problems, but that is a poor approach to engineering. It is always best to match impedances at both ends of connecting cables when dealing with fast pulses. You would do this by applying the BNC tee at the source and connecting a 50Ω load to one arm of the tee, setting the oscilloscope input to 50Ω.

As mentioned earlier, if the source is high impedance, connecting the 50Ω load at the BNC tee will result in considerable attenuation of the source signal. This may or may not be acceptable, but at least there will be no reflections between the source and the oscilloscope. The best solution of course is to set the source to have a 50Ω output impedance if that is possible.

Finally, you should search the web for information on transmission line theory to see how this all works, both in theory and practice. Reflections on transmission lines are an age-old problem that was first observed on trans-Atlantic underwater telegraph cables. It has only become more important with modern signal processing proceeding at gigahertz data rates. Learn the theory, learn the math, and be prepared for the 21st century revolution in communications and data processing.

Oh..thanks for giving very details info! I've actually read the Transmission line theory from wikipedia.

The cable impedance is fixed as 50 Ohm as usuall and source impedance is assumed as fixed as HighZ. Then could u tell me in a details that how adding 50 Ohm load with BNC tee at the source helps to prevent reflection? How it is different from the case without additional load? Maybe I have to know the structure of the BNC tee..
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,878
Joined
Jun 21, 2012
Messages
4,878
... Then could u tell me in a details that how adding 50 Ohm load with BNC tee at the source helps to prevent reflection? How it is different from the case without additional load? Maybe I have to know the structure of the BNC tee..
The load simply makes the source-end of the cable appear to be driven by its characteristic impedance, 50Ω, but at the expense of attenuating the signal of the high impedance source. The key criterion you should look for is ringing at the pulse transitions. No ringing? No problem. Use whatever connections work for you
 

Y2KEDDIE

Sep 23, 2012
259
Joined
Sep 23, 2012
Messages
259
I used to work evaluating video cameras. Proper termination is very critical and you can really see the effects visually of improper termination.

Starting at the camera which was 75 ohms, we used 75 Ohm cable through a BNC Tee to crosshatch/ bar generator, through another Tee to the monitor, and finally through a tee with a 75 ohm terminator resistor at the scope input connector.

You need to be mindfull there are 50 ohm BNC tee's and 75 ohm BNC tee's. They look alike but are different and can effect your signal greatly.
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,878
Joined
Jun 21, 2012
Messages
4,878
... You need to be mindfull there are 50 ohm BNC tee's and 75 ohm BNC tee's. They look alike but are different and can effect your signal greatly.
Absolutely! I forgot to mention that, maybe because I don't remember how you can distinguish between the two types if you don't have a VNA (Vector Network Analyzer) to measure them.
 

Y2KEDDIE

Sep 23, 2012
259
Joined
Sep 23, 2012
Messages
259
There are physical dimensional differences but not to my naked eye. I recently ordered both 50 and 75 ohm connectors and they have manufactuer's type numbers engraved on each piece. This is the only way I can tell them apart. There are many ddiffernt part numbers so I keep a cheat sheet of what I have in my stock for refernce.
 

davenn

Moderator
Sep 5, 2009
14,254
Joined
Sep 5, 2009
Messages
14,254
There are physical dimensional differences but not to my naked eye. I recently ordered both 50 and 75 ohm connectors and they have manufactuer's type numbers engraved on each piece. This is the only way I can tell them apart. There are many ddiffernt part numbers so I keep a cheat sheet of what I have in my stock for refernce.

usually the 75 Ω has a thinner centre pin. This becomes a problem when people accidently mix 50 Ω and 75 Ω connectors, as the fatter centre pin of the 50 Ω connector damages the centre part of the socket on the 75 Ω connector
 
Top