ADA9007 EVM Measured Data

Hi Russell,

We don't have any data but can measure this early next week.  I believe that the equipment we have is limited to a modulation bandwidth of 20 MHz (that needs to be double-checked as I'm at another location).  I think that you will still get useful data from a 20 MHz measurement. Since we have to do this measurement fresh, can you provide a specifics on frequency, output power range and modulation (filtering, alpha, etc)?

Best Regards

Eamon 

Hi Russell,

I just want to double check that your carrier frequency is indeed 40 MHz. So in our measurement, the carrier will span either from 30 MHz to 50 MHz (carrier bw 20 MHz) or from 20 MHz to 60 MHz (if we can generate a 40 MHz wide QAM carrier). Am I understand your requirements correctly?

Thanks

Eamon

Thanks for clarifying Russell,

just wanted to double-check. We are still assembling all of the plumbing to measure this (bias tees, dc blocks, eval board, etc). I hope to have results by the end of the week. 

Eamon

Hi Russell,

my colleague Aubrey is doing the measurements at another location. He got the setup up and running this afternoon. The maximum symbol rate that the sig gen and analyzer can take is 25 MSPS. Also he setup still needs some optimization. I'd like to get better system EVM (i.e. SigGen drives analyzer directly) which is about 1% right now . Also, we need to improve the heat sinking on the setup. My colleague took two data points at +14 dBm and +17 dBm with 64QAM/20MSPS/0.15 and got an EVM of 1% for both.  We are going to order some better heatsinks over the weekend and have another colleague take a look at the setup. So we should have more results on Monday or Tuesday.

Best Regards

Eamon   

Hi Again Russell,

sorry for the delay in getting back to you. We are having some issues with our setup. Our signal generator is struggling to give us good EVM at the 5 dBm to 15 dBm levels that we need to deliver to the AMP. It's as high as 6% in some cases. So we are working to add an external amplifier to the setup so that we can give good EVM to the device under test. Btw, do you have any rough target values for EVM that you are hoping to hit (e.g. <1%, <5%)?  This will give us some guidance on how good our source needs to be.  Thanks. - Eamon