Rigol DS1202Z-E bandwidth question

[user_x]

Hi,

 This is my first post so please excuse me if I am not expressing everything clearly!

 I just got a Rigol DS1202Z-E scope, 200 MHz, 2 Channels. Since I happen to have an EVAL-AD9912
https://www.analog.com/en/design-center/evaluation-hardware-and-software/evaluation-boards-kits/eval-ad9912.html#eb-overview
 DDS board with me, I decided to test out the scope bandwidth with the DDS board. This DDS evaluation board is 1GSPS, with a 400 MHz 7th-order elliptical low-pass reconstruction filter at its output. Therefore I expect the DDS output to have a flat magnitude below its bandwidth limit.

 I did the test by connecting the DDS filtered output (SMA on PCB board) with an SMA-BNC adaptor and a 1-foot BNC cable directly to the scope. I only turned the Ch1 on, therefore I am getting 1 GSPS on the scope.

 I measured the DDS output on the scope. I get 1.33 Vpp with its frequency set in between 10 MHz and 50 MHz. Then at 100 MHz I get about 1.17 Vpp (88% of the low-frequency value). At around 160 MHz I get 0.92 Vpp (69% of the low-frequency value). At 200 MHz I get only 0.68 Vpp (51% of the low-frequency value).

 So, based on this I would claim the scope bandwidth is about 160 MHz, as the 3 dB point would mean a 70.7% (oops, made a mistake at first) amplitude measured on the scope. I also tried measuring the DDS output with the Rigol PVP2350 probe on its 10x option and using the ground spring. The results are the same, thus I think my SMA-BNC adapter solution is good.

 So here comes the question: I wonder if my understanding above for the scope bandwidth is correct. Also, does the scope meet its 200 MHz spec in this case, or is something wrong?

 I guess there is a slight possibility that my DDS board output amplitude is frequency-dependent. However, I don't have a higher bandwidth scope at home to verify that. I don't have other high-frequency function generators for testing the scope. I don't have a convenient fast-rising edge of something either, so I can't test out the rising time of the scope myself right now. If you have suggestions on how I can build a simple circuit with a fast-enough rising edge, I would greatly appreciate that and be happy to try it out .

[M0HZH]

As it happens, I have a 1202Z-E on the way, should be delivered in a couple of days; I could try to recreate the measurements, with AD9910 as a signal source and HP 437B to measure signal level.

Your definition of scope bandwidth is correct, maybe there's a setup issue ? Termination, etc.

[user_x]

That would be great, thank you!

Actually, I did not terminate the scope with 50 Ohm. Instead, I just used the scope high Z input. Maybe that causes the problem? Looking forward to your tests!

[TK]

You need 50 ohm termination

[user_x]

Okay, thank you! What is the proper way to do that, since this scope does not have a 50 Ohm choice built-in?

[M0HZH]

Ideally, you want signal generator > 50ohm cable > termination > scope, with the termination at the scope input connector. Rigol sells one (see here) or you can just use a T junction and a simple 50ohm load.

[user_x]

Thank you very much! I'll go stock some BNC components then

[M0HZH]

Got the DS-1202Z-E today and I just ran a quick test with the RF signal generator from IFR COM-120B. -3dB seems to be around 220MHz or so (COM-120B last calibrated about 10 years ago), which is well within specs.