ADS54J64: how to do sub-sampling, image aliasing?

Hi Mike
I have notified the ADS54J64 expert regarding your question.
He will provide a more detailed response tomorrow.
Best regards,
Jim B

Hi Mike
What is the bandwidth of the signals at 500 MHz? ie. what is Fmin and Fmax?
Thanks,
Jim B

Hi Jim,

The primary signal of interest is the carrier at 500MHz, which represents both the amplitude of the charge of the beam in the synchrotron, and

via 4 channels of ratiometric measurement, the transverse position. There are coupled-bunch instabilities that we also measure that

modulate the carrier; the bandwidth is narrow, +/- 750 kHz.

Mike

Mike,

What mode do you plan on using? Do you plan on using real or I/Q data?

Regards,

Jim

Hi Jim,

I'm looking to get real data, I don't need I/Q.

I'm trying mode8 (I'm using the ADS54J64EVM and the TSW14J56), and it looks like I'm able to subsample with 487MHz sample  clock and a 500MHz input.

The real output shows up where I'd expect at 13MHz , Yayy!

However, it looks like the 4 inputs have a fixed phase delay between them.

Here's waveforms on the two positive channels 1 & 3, that have a power split time-aligned waveforms. The difference is a few sample clocks

Mike,

To make the routing easier on the PCB, channels A and C had the P/M signals swapped going to the ADC. This is the phase miss-alignment you are seeing.

Regards,

Jim

I know that there's a 180deg reversal of A&C w/rsp to B&D.

What I'm talking about is the relative phase of A to C (or B to D).

Is there some latency in the decimating input filter for each channel that might cause this?

Mike,

I am checking with the design team regarding your phase shift finding. Are you issuing a hard reset to the device before programming it?

In your last post you mentioned there was a waveform. This did not get attached. Can you send it again? What GUI settings are you using?

Regards,

Jim

ADS54J64_Mode_2.pptxMike,

After talking with design, they mentioned there should be no delay. I did a test with a 1MHz pulse that I sent to both CHA and CHC using a splitter and captured the data with the HSDC Pro GUI. When I looked at both channels, I did not see any sign of phase delay. Is there a chance you are getting a phase shift with your setup?

Regards,

Jim

 

Hi Jim,

I was able to duplicate your results for Mode 2, decimate x 4, ADS58J64_LMF_2241_mode24

However, when I try Mode 8, decimate x 2, ADS58J64_LMF_4421_mode38_16b,

I see that channels A&B are aligned (but 180 phase reversed as I expect), and C&D are aligned, but A&C and B&D are not.

It seems to be 4 samples off

Here's also a section showing the transitions:

2848	-2618	3052	-2376
2854	-2592	3022	-2368
2824	-2562	3018	-2354
2808	-2554	3000	-2342
2782	-2536	2988	-2314
2776	-2526	2946	-2294
2764	-2526	2960	-2276
2748	-2506	2898	-2256
2730	-2482	2906	-2240
2712	-2452	2894	-2224
2674	-2436	2888	-2198
2670	-2422	2874	-2200
2686	-2432	2832	-2182
2668	-2428	2802	-2136
2642	-2380	2784	-2130
2610	-2334	2754	-2140
2118	-1838	2786	-2132
-7534	7726	2750	-2096
-7730	7922	2744	-2082
-7960	8192	2706	-2058
-8198	8374	2222	-1550
-8102	8310	-7422	8112
-8020	8210	-7598	8320
-7822	8024	-7892	8600
-7574	7788	-8120	8824
-7476	7698	-8036	8760
-7416	7638	-7966	8672
-7362	7574	-7794	8492
-7214	7444	-7530	8254
-7152	7342	-7442	8176
-7048	7254	-7380	8112
-6958	7164	-7346	8060

Mike,

When I switched to mode 8 I did not see this issue. CHA and CHC were always aligned.

Regards,

Jim