Hi Oliver,
Answers to your questions.
1) The use of the digital phase cal generator is two-fold. The first reason, I am told, is that the Tunnel diode responsible for the broadband performance of the legacy generator has become obsolete. The second reason is that the spectrum of the digital pcal generator produces more power at the higher frequencies (relative to the low frequencies) than does the tunnel diode generator. This is important for the new broadband hardware because we use the same receiver chain to detect 3-10 GHz and so we can’t set phase cal indeoendently for each band as is possible with the S/X receiver which uses separate S/X band receiver chains. For this reason we need a flatter phase cal spectrum for the broadband receiver.
2) The MHC672 chip requires that the potential difference between pins 1,4,5,8,9,12,13,16 and pin 15 is -3.3v. This is in fact the case for the digital pcal generator that Alan designed but Alan derives the -3.3 v difference by putting the chip’s gnd pins on +3.3 v and the supply pin to supply gnd. This is not a problem according to the manufacturer.
3) I agree that anything you put in the main signal path will only introduce more phase disturbance so if you need flexibility in terms of the rail spacing an fpga or programmable counter in front of the switch driver would be a desirable option.
4) I have attached a powerpoint presentation that I put together on the subject of the new phase cal generator. Maybe it provide answers your questions - by all means let me know if this is not the case.
I look forward to meeting LiBing.
Cheers,
Chris
From: Wang Jinqing [mailto:jqwang@shao.ac.cn] Sent: Wednesday, April 22, 2009 8:05 PMTo: cbeaudoin@haystack.mit.eduSubject: Some detail questions about the new pcal
Hi Professor Beaudoin:
I am Oliver Wang,a new engineer at SeShang station Shanghai.I know you from Professor Rogers. Now I have some detail questions about the new pcal design.
1. Could you explain that why we are using digital chips now instead of the tunnel diode which we used in the past? I still believe the latter one provides better performances in the pulse generation such as even narrower pulse width and the pulse’s frequency coverage scope. The advantage I can see is just that digital chip provides better thermal stability and smaller board size. If the tunnel diode is available easily, do we still adopt the digital chips?
2. From the HMC672LC3C datasheet, we see the chip’s supply voltage is -3.3V on pin 15, and pin14 (VR) is usually connected to GND through a resistor. Pin1, 4,5,8,9,12,13,16 are connected to GND. But in the new pcal calibrator schematic diagram, we see pin14 to +3.3V,pin15 to GND, and Pin1,4,5,8,9,12,13,16 are connected to +3.3V.So can we do that?Is that OK?
3. Because our recording bandwidth is 2MHz, we need 1 MHz pcal signal. And in the future the bandwidth of satellite communication, we need 500 kHz pcal signal. So we need different modes. In my opinion, adding a frequency divider in the main signal path is not a good design owning to phase instability. So we want to add a counter or FPGA behind AD8611AR to wait 5 or 10 positive signals to generate a microwave-switch enable signal. What about your opinion?
4. We want get some detail results about the new pcal such as the test result,schematics, the pcb boards(Is the schematic on the memory web is just the latest version? And it is just for the pcbs?) and so on. My colleague LiBing will attend the TOW in the end of this month.He will visit to you and ask you for your helps.
These are the questions.Waiting for your reply.
Thanks a lot.
Best Regards.
Oliver Wang
