footprint check

footprint checking is an arduous task.

The spartan-3 sitting on my floorplan.









AD8022:

Bill Of Materials

A simple BOM.

components ordered

hmm..hmm and more hmmm.

Heaps of stuff to write about.

1. Designed and implemented a usb prototype board. I got the board made from uni workshop. Just a usb type A header, ft245rl and a parellel header to 
be connected 
to a ucontroller devel board or an fpga board. Works absolutely 
perfect!
2. Implemented software needed for project using java. Just a shell version for now; may be a skin using swing laterz. Multi-threading and what not.
3. Finalized components and placed an order through digikey.
4. AT4521 replaced by hrf-at4610; parellel 31.5db attenuator from honeywell. The parellel interface has a Vih of 4.2V - so had to configure the z8 outputs as open-drains; and tie to 5V using 10k's. A bit slow - but it'll do.
5. All tantalum caps to D casing.
6. Finalizing PCB - place an order Insha Allah by Friday. I checked with the PCB manufacturer - the story 
regarding COD's are true. So no payment 
until I get the prototype pcbs in hand.
7. Redesigned 10th order filter for 0.01R input to 50R output. Implemented using E-24 range of inductors. Unfortunately three of the values are not to be found with digikey.
8. AD8022 step up redesigned and recalibrated.

and heaps more.

PCB update

posting after a while now.

PCB is almost finished. A few loose ends to tie up.

RF2420 has been replaced by AT-4521 from honeywell; as RFMD discontinued the production. The only attenuator that i can get from digikey would be AT4521 with a serial interface; as opposed to the parellel rf2420. no probs thr; since i'll just have to live with a few more z8 ios freed up.

LPF has to be redesigned for 0.01R input impdeance and 50R output impedance. may be a matching network?

finally decided no output buffer; since the current from adc output driver ths3201 is above the magical 100ma mark. I had to replace AD8003 by ths3201 (from TI); as you can only get AD8003 in multiples of 250. heh. Long live TI and their fed-exed samples. I will definitely use TI for industrial projects;cos i have more experience with them.

have to check with darrell regarding the 2.5V step up.

and loads of other changes. I'll add them here whn i remember them.

a quick glance at the PCB. I shaved off about 30 sq inches. this one here is 24 sqinches.

A/C and Monte Carlo Analysis on LPF:

back to work

Darrell finally returned on the 19th.

There will be a few changes in the current schematics. I had overlooked some important facts.

DDS : no I-Q summing. Only the I ouputs will be used. Transformer solution for summing complementary currents is no good; as the LF response is not good enough. I will be looking for an opamp solution now. No other major changes here.

ADC: The inputs to the AD8022 needs a step-up by 2.5V. I used a non-inverting summing configuration. Darrell suggested a voltage divider configuration. Have to look into that.

SPARTAN3-Z8: the bus is finally there.

JTAG: confirmed daisy chain configuration. Need to fix schematics from master-slave.

Power supply: major changes. Found TPS75003 from TI - IC power supply for S3. Moreover, seems TI is dishing out samples. I will change my other two power ic's to TI.

USB: No change.

Attenuator: Darrell vaguely suggested some other config than using an IC - due to the LF response of RF2420. I am still convinced rf2420 will work - given right choice of input blocking capacitors. More work on that.

Footprints for the buttons need to be confirmed and designed.

The LPF response should be simulated and confirmed with Darrell.

PCB needs an update.

FPGA and ADC interfaced