50Ohm impedance of function generator probe

[mdanh2002]

Hi,

I just purchased a GW-INSTEK GFG-8019G 0-2MHz function generator for a cheap price on eBay. It comes with no probe. When I got it home, I used my oscilloscope 1x probe (1MOhm input impedance) and connect it to the 50Ohm output socket and try the waveforms. Everything seems to work fine (triangular, square and sine wave) even at high frequencies. Up to 2MHz triangular and sine wave looks fine and not distorted whereas square wave are fine up to around 500kHz, beyond which it will look like a distorted sine wave, which I suspect is simply the limitation of this device.

So does this mean that I can use the 1MOhm oscilloscope probe with this function generator? Will it cause any other problem in the future? I don't need high accuracy, as long as the generated waveforms look ok.

 

[ronv]

It will be ok until you want to "drive" something with the output of the probe. If you were for example to place a 1K resistor from the signal to ground the signal would disappear.

 

[mdanh2002]

Thanks ronv. That means I need to purchase new sets of 50 Ohm probe for my function generator.

One more question, with the 50 Ohm probe, will the oscilloscope still display the waveform properly if I connect the 50 Ohm probe output to my oscilloscope 1MOhm input probe, since the probe impedance is mismatched?

 

[MrAl]

Hi,

Because it only goes up to 2MHz you can probably get away with a jumper lead, without a probe. Connect the jumper to the thing you want to power, and of course another jumper for ground.
If you are worried and want to use 50 ohm cable then terminate it with a 50 ohm resistor before you use the scope.

The 50 ohm output means you can connect say a 500 ohm resistor to it and it will provide about 90 percent of the voltage as it does open circuit. In other words, the output is like a perfect amplifier with zero output impedance with a 50 ohm resistor in series with it.

For best matching you would use a 50 ohm cable and at the end connect a 50 ohm resistor unless you already have a circuit you want to power with it that has a 50 ohm input impedance. You dont usually need this though for these lower frequencies just for general purpose testing.

 

[Nigel Goodwin]

It will be ok until you want to "drive" something with the output of the probe. If you were for example to place a 1K resistor from the signal to ground the signal would disappear.

A 1meg probe is just a piece of screened wire, there's no resistance or impedance involved, the 1meg refers to the INPUT impedance it's designed to plug in to - not the probe itself.

So a 1K resistor to ground wouldn't have the slightest effect on it, his scope probe is perfectly fine to use on his generator.

As for his question about connecting it directly to a scope, again it's perfectly fine - as in almost all electronics you don't want impedance matching (maximum power transfer) you want a low impedance feeding a high one (maximum voltage transfer).

 

[JimB]

Let us sort out a bit of terminology here.

"Probe" infers some device which is used to connect to a device under test to sample a signal, for example an oscilloscope probe.

Scope probes come in a variety of types, x1, x10, switchable x1 x10, etc, etc.

As has already been mentioned, a x1 probe is usually just a straight through cable, although I am sure that I have seen some which had a low value resistor (about 1k ohm) at the tip.

The function generator, I would not expect to be supplied with a "probe".

What mdanh needs to do now is build up a collection of "test leads".
Some of the test leads will be BNC to BNC coax cables, some will be BNC to crocodile clips, and some will be just plain wire with croc clips on each end.
Keep them in good condition and store them where you can find them when you need them.

As time goes on and the test equipment increases, a few adapters can be useful as well,
BNC to BNC, BNC to 4mm sockets, the permutations are endless.

See the attachment for pictures of my collection (40+ years worth), and a pic of the OPs home area.

JimB

 

[mdanh2002]

Thanks all for the detailed replies. I learned a lot just by reading the replies and understand the difference between probes and test leads now Thanks JimB for the photos - look quite well organized

One more question, I have another frequency counter (GFC-8055G) that counts up to 550MHz. There are 2 input sockets, clearly marked as

(1) Input A, 1MOhm, 0.1-100MHz
(2) Input B, 50Ohm, 50MHz-550MHz

Currently I am using Input A with my oscilloscope probe to measure outputs of the frequency counter and the results look fine. If I am to use input B to measure higher frequencies (for example, the output of audioguru FM transmitter), can I still use the oscilloscope "probe"? Or do I need to modify it to allow proper impedance matching like MrAl suggests?

 

[Nigel Goodwin]

Thanks all for the detailed replies. I learned a lot just by reading the replies and understand the difference between probes and test leads now Thanks JimB for the photos - look quite well organized

One more question, I have another frequency counter (GFC-8055G) that counts up to 550MHz. There are 2 input sockets, clearly marked as

(1) Input A, 1MOhm, 0.1-100MHz
(2) Input B, 50Ohm, 50MHz-550MHz

Currently I am using Input A with my oscilloscope probe to measure outputs of the frequency counter and the results look fine. If I am to use input B to measure higher frequencies (for example, the output of audioguru FM transmitter), can I still use the oscilloscope "probe"? Or do I need to modify it to allow proper impedance matching like MrAl suggests?

You can't really use it to measure the output of a crude VHF transmitter at all, the load imposed would move the frequency too much to be of any use.

The 50Ohm high frequency input is really intended for connecting to the 50Ohm output on a dummy load (tapped off the dummy load so as to give a low level signal).

 

[ronv]

A 1meg probe is just a piece of screened wire, there's no resistance or impedance involved, the 1meg refers to the INPUT impedance it's designed to plug in to - not the probe itself.

Of course. My mistake.

 

[JimB]

One more question, I have another frequency counter (GFC-8055G) that counts up to 550MHz. There are 2 input sockets, clearly marked as

(1) Input A, 1MOhm, 0.1-100MHz
(2) Input B, 50Ohm, 50MHz-550MHz

Currently I am using Input A with my oscilloscope probe to measure outputs of the frequency counter and the results look fine. If I am to use input B to measure higher frequencies (for example, the output of audioguru FM transmitter), can I still use the oscilloscope "probe"? Or do I need to modify it to allow proper impedance matching like MrAl suggests?

My frequency counter is similar, and depending on what I am trying to measure, I often use a scope probe (set to x1 or sometimes x10) to connect into the device under test.

Another technique is to use a coupling loop.
A few turns of wire connected to a BNC connector. Connect the loop to the counter (or the scope, or the spectrum analyser) using a short (1 metre) BNC to BNC cable, and bring the loop near to the circuit under investigation and it will usually pick up enough signal for testing purposes.
Have a look at the attachment to see my coupling loops.

As for testing the output of a transmitter, it pains me to read some of the threads relating to these little self excited oscillators connected to a bit of wire.
The best you can do is to use a coupling loop and bring it near to one of the tuned circuits and hope that it does not load the coil too much and pull the signal frequency.

To measure the output of a real transmitter which works into a defined load impedance, have a look at the second photo attached here.
There are many variations of this set-up, but this one works well for me.

JimB

 

[Dean Huster]

A 1meg probe is just a piece of screened wire, there's no resistance or impedance involved, the 1meg refers to the INPUT impedance it's designed to plug in to - not the probe itself.

So a 1K resistor to ground wouldn't have the slightest effect on it, his scope probe is perfectly fine to use on his generator.

I don't know about crappy Chinese imported probes, but a decent 1X probe made by Tektronix uses resistance wire for the center (signal) conductor designed to improve the higher-frequency characteristics of the probe. The ONLY thing an oscilloscope probe should be used for is as an INPUT probe for an oscilloscope or other high-impedance INPUT. If the input impedance is other than 1M ohm for an attenuator probe, it won't have the stated attenuation factor.

If you use a non-attenuating ("straight-through") 1X probe as an "output cable" for a function generator or power supply, convenient though the connections may be, you risk damaging the probe. Any appreciable current through the probe's center conductor turns it into a resistive heating element with the expected results: damage. Connecting it to a lower current function generator output and then to a lower-impedance load will result in a serious loading effect, greatly reducing the generator output at the load.

Connect an ohmmeter between the probe tip and center conductor on the BNC connector and measure the resistance. It may change your thoughts on the construction of the probe. If it measures zero ohms, you probably have a pretty crappy probe in your hands and is why its bandwidth is so low. You can get as much as 11MHz or more from a Tektronix 1X probe because of its construction. That's also why they're a bit pricier.

 

[unclejed613]

Hi,

Because it only goes up to 2MHz you can probably get away with a jumper lead, without a probe. Connect the jumper to the thing you want to power, and of course another jumper for ground.
If you are worried and want to use 50 ohm cable then terminate it with a 50 ohm resistor before you use the scope.

The 50 ohm output means you can connect say a 500 ohm resistor to it and it will provide about 90 percent of the voltage as it does open circuit. In other words, the output is like a perfect amplifier with zero output impedance with a 50 ohm resistor in series with it.

For best matching you would use a 50 ohm cable and at the end connect a 50 ohm resistor unless you already have a circuit you want to power with it that has a 50 ohm input impedance. You dont usually need this though for these lower frequencies just for general purpose testing.

you really don't need a 50ohm termination. if you're not driving any signal cables that are more than 15 meters (1/10 wavelength at 2Mhz) long, then terminating your signal cable with a 50 ohm resistor isn't necessary. most function generator "test leads" consist of a BNC plug, a length (usually 3-6 ft) of shielded cable and a pair of alligator clips or ez-hooks at the end.

 

[MrAl]

you really don't need a 50ohm termination. if you're not driving any signal cables that are more than 15 meters (1/10 wavelength at 2Mhz) long, then terminating your signal cable with a 50 ohm resistor isn't necessary. most function generator "test leads" consist of a BNC plug, a length (usually 3-6 ft) of shielded cable and a pair of alligator clips or ez-hooks at the end.

Yes and i said that in my post so your post is superfluous.

 

[mdanh2002]

Hi JimB,

Thanks for your suggestions of connecting a few turns of wire to the frequency counter test leads and let it pick up the transmitter signal. I will try it and see how accurate the displayed frequency is