Vector signal generator differential I and Q outputs not understood

[240RS]

I have this signal generator with 512MSamples baseband IQ memory and 160MHz maximum sampling clock. Where I normally use its function to upconvert this with its LO, I am also interested in getting the IQ signals out in baseband.

For this there is a 15 pin connector on the back with differential I and Q pairs and the rest GND. Baseband. I have attached the description of those outputs. I want to use the I output single ended to generate broadband 0 - 30MHz signal (first 2MHz and amateur bands notched). But I do not know what connector to make for it. Mostly since I do not understand this “standard” of differential IQ with controllable DC bias. Any explanation helpful.

[wasedadoc]

Why are you posting in the Repair section?

You just need to know which pins of the connector are I and Ground.  Set the offsets to zero if you want the signal to be AC (equal excursions above and below Ground).

[240RS]

Mistake on my side posting it here.

So just I and ground, combined with zero offset on the software side.

What are these outputs types normally used for?

[rfclown]

I don't believe that there is any "standard" for this. It is what the manufacturer tells you it is, and you deal with it as you please. IQ baseband output with 160 MHz clock is pretty impressive. I usually deal with HP/Agilent/Keysight units with max clocks of <= 100 MHz which output their baseband IQ using single ended connectors (like BNC). Whatever the case, you have to interface whatever they give you to what you want. If you want single ended, and they give you differential, then you have to use a differential to single ended converter which will usually be an op-amp circuit. When I've done this in the past it has been some RFIC that has differential I/Q outputs that I wanted to convert to single ended, and I dead bugged an op-amp circuit on a piece of scrap PCB to do the job.

[240RS]

I want to use it to test a 130MSPS 16bit DAC. If I can do that without an opamp that would be great.

[rfclown]

I want to use it to test a 130MSPS 16bit DAC. If I can do that without an opamp that would be great.

Do you mean test a ADC (not DAC)? If you look at the ADC's data sheet I would guess that they will show you an example application circuit with an op-amp at the input.

[240RS]

It is an HF zero IF software defined radio (RX-888). The actual ADC in it is differential, but my input to it is 50 Ohm single ended

[wasedadoc]

What does it say in the part of the spec that you have cut off at the bottom of the image in the first post?

I think you can use either I or its inverse by itself as a single ended source.

[240RS]

What does it say in the part of the spec that you have cut off at the bottom of the image in the first post?

I think you can use either I or its inverse by itself as a single ended source.

Attached

[240RS]

Got it, ran a test and below the results. Taking positive I against GND seems to work fine.

Instrument: Anritsu MG3700A Vector Signal Generator
Software to upload IQ files: Anritsu "IQ Producer"
Waveform file generated: 1kHz signal between +0.2 and 1.0 for AM modulation. I only, Q=0
Normalization by "IQ Producer": -1.0 => -8192, +1.0 => 2^14-1 (8192-1)
Modulation: AM with 250kHz carrier
Baseband IQ output:
    I+: ~0.3 to 1.5V (high impedance measurement)
    I-: ~ -0.3 to -1.5V

See screenshot for modulated signal, positive I output, negative I output.

[EEEnthusiast]

Using I+ and I- along with a diff amp will help reduce the 2nd and higher even order distortion components. If you are not worried about that, then using just the I+ or I- w.r.t. GND should be just fine. You may want to buffer this signal anyway as the output signals may be coming directly from the baseband DAC and these may not be able to drive high loads.

[240RS]

Thanks! Any suggestion for what opamp to use for wideband 2-30MHz?

You may want to buffer this signal anyway as the output signals may be coming directly from the baseband DAC and these may not be able to drive high loads.

Given that output impedance is given as 50Ohm, can I conclude there is no need to buffer?

[EEEnthusiast]

Thanks! Any suggestion for what opamp to use for wideband 2-30MHz?

You may want to buffer this signal anyway as the output signals may be coming directly from the baseband DAC and these may not be able to drive high loads.

Given that output impedance is given as 50Ohm, can I conclude there is no need to buffer?

Sorry... my bad. For 50 ohms output impedance you do not need any buffering. They are made to drive transmission lines and a 50 ohms load at the end.

[240RS]

Happy to hear. Quite convenient then.

[240RS]

Given that output impedance is given as 50Ohm, can I conclude there is no need to buffer?

Sorry... my bad. For 50 ohms output impedance you do not need any buffering. They are made to drive transmission lines and a 50 ohms load at the end.
[/quote]

All works fine. Have my converter plug made and tested. Two 50Ohm outputs by connecting plus I and Q outputs to BNC’s that are grounded.