I'm building a programmable analog signal generator to drive some scientific equipment. The equipment says it has an input impedance of 50 Ohms, and uses 50 Ohm coax to make the connection. The output stage of my generator consists of an instrumentation amplifier (INA2128).

Do I need to impedance match the output of the driver with the coax? Is it as simple as placing a series 50 ohm resistor, or will I need an additional buffer stage?

 

That chip can only provide about 4 ma to 6 ma on its output. That limits you to a few tenths of a volt of signal into 50 ohms. You will probably want a higher voltage, and that requires higher current. So, yes, I think you need a buffer.

 

okay thanks...but the other question remains....do I let the buffer drive the load directly, or do I need to 50 ohm series resistor?

 

I don't know. Probably need somebody else to answer this part, but he (she) will need to know the frequency range.

 

Okay thanks. The target frequency is 60 Hz or lower, so lets call it 0 -> 100Hz to be safe.

 

What is the frequency and shape of the signal?
What is the length of the 50-ohm cable?

The INA2128 is not designed to drive a low impedance load. You may want to use a 50-ohm cable driver such as LM6181 on the output of the INA2128.

Edit: For 100Hz you will not have problems with reflections but you will still have problems with such a low load impedance.

 

frequency is < 100 Hz. The shape of the signal is completely arbitrary. The target device is a fast steering mirror with analog inputs (=/- 10V) to control pointing angle. So the shape of the waveform depends on where it has to steer...but slew rates should be pretty slow...well within a few hundred mV per step.

 

Then you don't need any high frequency amplifier on the output but I would still use a low impedance driver.

 

Any op amp that has an output of 200mA or greater can be used as a non-inverting buffer amp to drive the 50 ohm load at +/- 10V. You will also need a +/- 15V dual supply for the op amp.

 

So I just looked up the data sheet for the device we have used to drive these guys in the past, and that driver is only capable of producing +-5 mA output current....yet the thing worked without a problem. I'm wondering if the actual input impedance is much higher, but they just balance it for 50 ohm coax.

 

So I just looked up the data sheet for the device we have used to drive these guys in the past, and that driver is only capable of producing +-5 mA output current....yet the thing worked without a problem. I'm wondering if the actual input impedance is much higher, but they just balance it for 50 ohm coax.

10V / 50 ohms is 200mA.
You can check the actual DC input impedance (which is the value of interest at your low frequencies) of the instrument with an ohmmeter.

 

Are you absolutely sure that the driven equipment has a 50Ohm input impedance? Even if it has a BNC connector, it might have a much higher input impedance, like a scope...

 

Are you absolutely sure that the driven equipment has a 50Ohm input impedance? Even if it has a BNC connector, it might have a much higher input impedance, like a scope...

I have the same suspicion. Devices that take a 10V input usually don't have 50 ohm impedance.

 

Hello,

Do you have more information of you scientific equipment?
That would make things more clear how to help you.

Bertus

 

The thing has a big label by the inputs that says "50 Ohm input", so that is what I was going off of. However, I did check the impedance at the input and it measured to be ~10k, which I was finally able to confirm with the documentation (it was hidden away in a block of text).

 

The thing has a big label by the inputs that says "50 Ohm input", so that is what I was going off of. However, I did check the impedance at the input and it measured to be ~10k, which I was finally able to confirm with the documentation (it was hidden away in a block of text).

In that case you should be able to drive the cable directly with your INA2128 instrumentation amp. At such low frequencies the characteristic cable impedance is not a factor. The cable capacitance could cause amplifier oscillations though. If so, try adding a small resistor (say 50 to 100 ohms) in series with the amp output.