Connecting a Function Generator to an Oscilloscope

[ParkingLotLust]

Hey guys,

In lab class, we often need to feed a signal from a function generator into a circuit, then scope the signal going in, and the signal coming out (lets say Im doing a PID control lab and need to do step response).

The input signals we work with are usually in the 1Hz-10kHz range - we dont work with anything in the MHz range, at least for the moment.

The typical procedure is this: using 2 sets of really crappy alligator leads (like this), one set goes to the generator, and one set goes to the scope. The grounds and signal connections are tied together, and clipped to the input of the circuit. For all intents and purposes, this works just fine, albeit it gets a bit messy with lots of scope leads everywhere.

My proposed solution was this: I went out and got a BNC T, along with a short F-F cable. I added the T on the function generator, connected the F-F from one side to the scope, and connected to my circuit on the other side of the T. Everything worked perfect, exactly as expected, until my lab teacher walked over.

He went on about how because the scope is a 1Mohm input and the generator is a 50ohm output, by connecting them together how I have them, Im causing an impedance mismatch, Im causing signal reflections, and all these nasty things. He basically had me remove the T and my F-F cable because I was breaking things.

From my limited knowledge of the subject, it would seem to me the impedance mismatch and any reflections that may be caused would be there, regardless of using the alligator leads or the BNC cables. Is there any reason to his madness, and if so, can someone explain a little further why? Cause he sure didnt.

 

[Nigel Goodwin]

He went on about how because the scope is a 1Mohm input and the generator is a 50ohm output, by connecting them together how I have them, Im causing an impedance mismatch, Im causing signal reflections, and all these nasty things. He basically had me remove the T and my F-F cable because I was breaking things.

From my limited knowledge of the subject, it would seem to me the impedance mismatch and any reflections that may be caused would be there, regardless of using the alligator leads or the BNC cables. Is there any reason to his madness, and if so, can someone explain a little further why? Cause he sure didnt.

The reason to his madness is that he doesn't know what he's on about

Impedance matching is ONLY used under certain special circumstances, in the vast majority of applications it's not only not needed, but it would be disasterous to do so. In this case, while it probably wouldn't be 'disasterous' it certainly most not be done, and doing so would certainly greatly alter the results, and may prevent results entirely.

Impedance matching is for POWER TRANSFER, most things use VOLTAGE TRANSFER - and while I wouldn't have used a BNC T it's perectly valid (I'd have just clipped the scope probe on the input, and the other on the output).

 

[ParkingLotLust]

Yeah, we learned about impedance matching and reflections in previous semesters, but I was quite sure that was only at much higher frequencies than 10Hz... Come on now, the wavelength at 10Hz is (3x10^8m/s / 10)!

Glad to hear it wont cause any damage for my use. Thanks Nigel!

 

[ronsimpson]

If the device you are driving is not 50 ohms then you have reflections anyway.
Crappy alligator leads have an impedance of !$$$&^%*^&(^(*^&. If you move the clips slightly then the impedance changes to (*&)(&^%&. (exactly)
At audio frequencies we use extension-cord wire to run signals across the room.
Keep your wires short.

 

[KeepItSimpleStupid]

Just to not pic, an F-F cable is one with an F connector on one end and an F connector on the other. Other interpretations are Female to Female cable. F cables have a Z of 75 ohms.

BNC connctors have a Z of 50 ohms USUALLY unless you have one of these: Comprehensive - BP-2C7559 Connector

In professional video, you'll see BNC connectors. Z is 75 ohms. The connector is lilely 50. A real world example.

My circuit has a Z of 1000K and I connect a 10 M-Ohm DVM. The Z doesn't match. That's just as absurd as teaches objection.

Ask him how to measure the resistance of a sheet of paper. The amount of moisture in the paper affects the readings. I used to do it all the time.

Ask him what the voltage of a 1K resistor in an RF free environment? If he says zero, he's wrong.

Ask him if when your carrying a wire, is it generating a voltage. The answer is yes.

Ask him if flexing a BNC cable generates charge? The answer is yes.

 

[Reloadron]

This can get to be one of those things. While I agree with what you have been told here and disagree with your professor / teacher as to your application you as the student need to keep something in your mind. That classroom or lab area is your teacher's little oyster. To disagree with him sometimes is not the brightest thing to do. My guess, following his little lecture to you is it would be wise to let it slide. Choose your battles carefully and live to argue another day.

Just My Take
Ron

 

[KeepItSimpleStupid]

Battles like this can be hard to win. I waged them, won and was told if I had better things to do, I didn't have to come to class. Definately not the norm. GPA was 4.0.

 

[ParkingLotLust]

I'm not so much worried with proving the teacher wrong to his face, as I am with having the correct information about the expensive equipment Im using!

I have no intention of going back to him and telling him he was, in fact, wrong, but I plan on continuing to use the T and the cables (KISS, my apologies, I meant F-F in the Female-Female sense) whenever it makes my life easier. If he intervenes again, then I will respectfully disagree and take it from there.

Thanks again for the good info!

 

[JimB]

Everything worked perfect, exactly as expected, until my lab teacher walked over.

He went on about how because the scope is a 1Mohm input and the generator is a 50ohm output, by connecting them together how I have them, Im causing an impedance mismatch, Im causing signal reflections, and all these nasty things.

The rules of ETO prevent me from expressing my true feelings for your lab teacher!

If you were looking at fast, short pulses then reflections would be a problem.
As your set-up is with low khz signals and short cables, no problem.

Impedance matching is important where the cables are "long" compared with the frequencies you are examining. for example, a 1 metre long coax cable when short circuited at one end, will appear as an open circuit at ferquencies which are an odd multiple of a quarter wavelength, ie 75Mhz, 225Mhz, 375Mhz etc.
Welcome to the wonderfull world of standing waves!

Your lab teacher is just a muppet!

JimB

 

[KeepItSimpleStupid]

OP. Way to go. Most BNC cables have male pins and a female body. Hence PLUG. Some connectors, not BNC, are labeled plug, std sex and plug rev sex.

Female BNC cables are an ODD configuration. They would look like the BNC connectors on your scope.

Those BNC T's are about $15 so he may be concerned with them walking.

 

[Dean Huster]

As a side note, are you sure you're using F-F BNCs and not M-M? M-M BNC cables are common as dirt. F-F are rare. It's the center pin that determines the gender. Your scope inputs are female. Most of the BNC Ts in use are male splitting into two female.

 

[ParkingLotLust]

My mistake; having looked more closely at the cables I see my error in naming. To eliminate any further errors in naming, this is the cable connecting the scope to the T:
**broken link removed**
Which is (clearly) M-M.

This is the T (not this specific one, just for the visual):
**broken link removed**

The cables and T's I purchased myself, hence being afraid to harm the equipment; the school wouldnt provide anything like that simply in fear of it going missing.

 

[KeepItSimpleStupid]

Your learning. You have an excuse. The components you showed is what I had pictured in my head. I also know they are expensive. At work, someone ordered a bunch of name brand crimped BNC's. They all broke over time. It got to be annoying grabbing a cable and find intermittant connections. Got good at replacing them.

 

[Reloadron]

OK, I still advocate not getting in a pissing contest with the teacher. However, here is what I don't quite get, starting with:

The typical procedure is this: using 2 sets of really crappy alligator leads (like this), one set goes to the generator, and one set goes to the scope. The grounds and signal connections are tied together, and clipped to the input of the circuit. For all intents and purposes, this works just fine, albeit it gets a bit messy with lots of scope leads everywhere.

OK, so you have two sets of BNC to small alligator clips. Saw the picture in the link. Got it. You have a signal source (function generator or whatever) designed to work into a 50Ω load. OK, so the scope has a 1MΩ input impedance and what is the input impedance of your load? The circuit you are driving with the signal source? Unless the load impedance is 50Ω it's all moot isn't it?

From my limited knowledge of the subject, it would seem to me the impedance mismatch and any reflections that may be caused would be there, regardless of using the alligator leads or the BNC cables. Is there any reason to his madness, and if so, can someone explain a little further why? Cause he sure didnt.

To my way of thinking the moment we went from the 50Ω coax to the leads to the alligator clips there is a mis-match anyway?

Would he prefer you look at the signal right at the load?

Side Note: When it comes to BNC patch cables, connectors, adapters and everything like that I prefer and only buy Pomona.

Ron

 

[Dean Huster]

Trouble is, Pomona is about the most expensive stuff out there. The various "surplus" and import mail-order outfits have a decent array of cables, probes and adaptors suitable for audio and low RF work at a fourth the price and is great to get a beginner started with useful accessories. Will the cheap stuff handle 1 GHz with no problem? No. Few people need it to. and if you're careful with your stuff, it won't fall apart.

 

[mdanh2002]

A bit off topic, but many lab assistants, or "teachers" as the OP named them, are simply final year students, research assistants, or PhD students under the supervision of the lecturer handling the module. Many of them have no interest, much less expertise, in the particular subject that they are supposed to provide assistance. They are simply there because their lecturers pay them for being there. And this remains true even in the "top" universities in the world.

In my third year at university many years ago, I encountered a lab assistant who spoke little English (which is the teaching medium) and did not know how to handle some of the lab instruments...

 

[Reloadron]

Per my side note:

Side Note: When it comes to BNC patch cables, connectors, adapters and everything like that I prefer and only buy Pomona.

and Dean's comments I should point out I do not recommend buying Pomona products for the basic home enthusiast. They are very high prices adaters and connectors as well as patch cables. When I buy at work I am buying for the nuclear industry so we do look to buy the best we can.

Ron