Using a function generator with a tube amp?

[jonnycat]

Hello, I am looking to use my function generator to play around with a tube amp. My concern is that the function generator has a 50 ohm output impedance, while the tube amp has a 1M ohm input impedance.

What's the trick to interfacing a function generator with a tube amp?

[demeterart]

unless we are talking about Rf that requires matched impedances there is no trick. you have to know the input sensitivity of your amp to adjust appropriately the output of the generator. one note though: usually, if there is a selection about the output impedance of 0/50 Ohms then the output with 50 Ohms is assumed to be across a 50 Ohms load at the receiver otherwise you get twice the 0 ohms selection.

[Mooly]

Theres no problem. All it means is the generator is considered an ideal voltage source that is in series with 50 ohms. In other words its output impedance

[jonnycat]

Okay, thanks guys, I appreciate it. I guess this makes sense, as I have no trouble plugging the 50 ohm output directly into my oscilloscope, which has a 1M ohm input.

[HollowState]

During square wave testing, as you increase the frequency of the square wave, it will suffer waveform degradation. To maintain the best waveshape, you must terminate the far end of the coaxial cable in 50 ohms where it enters the device under test. Amplitude will be somewhat reduced, but is a necessary tradeoff. Use 50 ohm characteristic cable like RG-58. Although below about 10KHz, none of this matters very much.

[tomchr]

"Somewhat reduced". Actually, the amplitude will be reduced by exactly a factor of two when the 50 ohm output is terminated in 50 ohm.

As others have said already, for high-frequency (>10 kHz is good SWAG) or square wave testing use the 50 ohm terminator. In fact, unless you absolutely need the 6 dB of amplitude lost due to the proper termination, I suggest leaving the terminator connected at all times. It should be connected at the "far" end of the coax - right where it meets with the amp input.

The most commonly used terminator (single male BNC with a resistor in it) can be used with a BNC Tee connector. Or you can get a feed-thru terminator. The latter is quite handy.

~Tom

[jonnycat]

I have a 50 ohm terminator and a tee, so I'll use them both just for the fun of it.

[tomchr]

Run a 10~20 kHz, 1 Vpp square wave into your oscilloscope (assuming you have one) with and without the terminator. Note the difference.

~Tom

[macboy]

Originally Posted by HollowState

During square wave testing, as you increase the frequency of the square wave, it will suffer waveform degradation. To maintain the best waveshape, you must terminate the far end of the coaxial cable in 50 ohms where it enters the device under test. Amplitude will be somewhat reduced, but is a necessary tradeoff. Use 50 ohm characteristic cable like RG-58. Although below about 10KHz, none of this matters very much.

If you are using square waves, then the frequency doesn't matter at all, it could be a 1 Hz square wave and still suffer from degradation due to mismatched impedance. What matters is the highest frequency component of the square wave, which is determined by the rise/fall time of the edges, not the period (repetition frequency) of the square wave.

[tomchr]

Originally Posted by macboy

If you are using square waves, then the frequency doesn't matter at all, it could be a 1 Hz square wave and still suffer from degradation due to mismatched impedance. What matters is the highest frequency component of the square wave, which is determined by the rise/fall time of the edges, not the period (repetition frequency) of the square wave.

True. However, you will notice the degradation (ringing) a lot more on a 10 kHz square wave than on a 1 Hz square wave as the decay time of the ringing starts approaching the period of the signal.

~Tom