Zmod Scope Calibration and Sample conversion

[Stefanski881]

Hello, i am using an Eclypse Z7 with the Zmod scope 1410-105. My implementation is working, but not really in the right way.

I was testing my implementation with a signalgenerator and a DC-source. Both ways I am receiving the correct Signal form, but the absolute value is not right. The value I get from the scope is always approximately twice as high as the signal I am seeing on an oszilloscope. Also it is very noisy and has an significant offset.

I cant get behind way that is. I am not sure if the conversion of the 2s complement I get from the scope is wrong or maybe I need to calibrate the device. My programm in vitis is only enabling the dma and the acquisition of the scope. Some help would be appreciated and I can provide more information if needed.

I was also trying to bypass the calibration block by using the test_mode, but when I read the calibration value from the flash I am only receiving 0s.

Thanks in advance.

Edited August 25, 2023 by Stefanski881

[artvvb]

Hi @Stefanski881

Per the calibration section (on page nine) of the ZmodScopeController manual:

Quote

Since the data port of the Data interface is 32 bit wide, regardless of the ADC resolution (kADC_Width), the most significant 16 relevant bits are selected out of the calibration result, as shown in Figure 3.

This means that data passed through into the DMA is effectively the full 16 bits, instead of just the lowest 14. The same shift is applied in test mode, where the incoming raw data is padded with two zero bits at the bottom of the vector. In both cases, this means the 14 bits you want to treat as the sample are the top 14 of your `number` variable, not the bottom 14.

There's a baremetal application in this under construction demo that uses dpmutil to read calibration coefficients out of DNA, which may also be helpful: https://digilent.com/reference/programmable-logic/eclypse-z7/demos/ddr-streaming

Thanks,

Arthur

[Stefanski881]

Oh yes that explains a lot.

Thank you very much, that was very helpful.

[Xband]

Could also be termination of the signal contributing to this.  Typically there can be a factor of 2 if you’re reading a 50ohm termination at high impedance 1Mohm.  I cannot recast what the zmod sma termination is, but I know it’s listed numerous places in the documentation.