Project description

The FmcAdc100M14b4cha is a 4 channel 100MSPS 14 bit ADC card in FMC (FPGA Mezzanine Card) format. By default it uses only signals from the LPC rows of the HPC connector that is mounted. The gain can be set by software in three steps: +/-50mV, +/-0.5V, +/-5V. An advanced offset circuit is used in the front-end design of the ADC board, and allows a voltage shift in the range of +/- 5V that is independent on the chosen gain range. To see how this mezzanine can be combined with a carrier and turned into a complete system, see the system specifications page.



3D drawing
Bottom side
Front panel

Specifications

Parameter Value
max. sample rate 105 MSPS
analog bandwidth 30 MHz. DC-coupled (40 MHz possible by changing eight capacitors)
bits/sample 14 bit
ENOB 11, 11.5, 11.7 bit (@ +/-50mV, +/-0.5V, +/-5V range)
channels 4
connectors 4 x LEMO 00 for signals, 1 x LEMO 00 for trigger
input impedance 1 kOhm / 50 Ohm - software selectable
gain steps +/-50 mV, +/-0.5 V, +/-5 V for full scale
offset correction range +/- 5 V for every input voltage range
max. gain error +/- 1 %
SNR 67.7 dB, 70.8 dB, 72.2 dB (@ +/-50mV, +/-0.5V, +/-5V range)
FMC to carrier interface FMC high pin count connector (HPC only used if external clock is selected)
ADC interface Serial LVDS, 2 pairs for each channel
Clock source Internal: from programmable on-board oscillator. External: from dedicated FMC connector pins (HPC) when changing a resistor.

Project documents


Releases


Contacts

Commercial producers

General question about project


Status

Date Event
01-10-2009 Start working on project
25-10-2009 Functional specification written.
23-11-2009 Selection of components, waiting for schematic symbols.
02-12-2009 Work on the FmcAdc100M14b4ch_a is suspended, because estimated performance wasn't reached.
18-12-2009 Specification and schematic of new version has been added.
22-01-2010 FmcAdc100M14b4cha upgrade. Design revised and modified, better parameters expected.
10-02-2010 New design made with 3 input ranges and programmable offset.
Need to finalise local clock circuit before PCB layout can start.
24-02-2010 Draft PCB layout done. Design document being written. Design review planned for 2 March.
02-03-2010 Design review of schematics done. Review02032010
04-03-2010 Improvements made, responding to design review. Review02032010_improvements
19-03-2010 PCB review planned for 24 March 2010.
24-03-2010 Design review of PCB layout done. Review24032010
25-03-2010 Small changes involved to the schematic (to make routing easier),
PCB partially updated. Review24032010_improvements
29-03-2010 PCB updated. Waiting for Design Office's review.
06-04-2010 12 PCBs ordered. 3 will be mounted for prototyping.
12-04-2010 For availability reasons two resistor values need to be modified (120->121, 510->511). See issues.
06-05-2010 3 cards assembled, SMC connectors ordered but not yet available.
28-05-2010 SMC connectors given to workshop. Boards look really nice.
09-06-2010 3 assembled boards received.
12-07-2010 One board plugged in on Xilinx development kit and powered. Debugging start.
13-07-2010 Two channels working! Limited by Xilinx development kit clock connections.
30-07-2010 Current consumption has been measured, document is now available on the wiki.
30-07-2010 Changes for the second version of the PCB documented, mainly to solve thermal problems with regulators.
Required change of DC/DC converter and inductor.
30-07-2010 ENOB & SNR measured: >11 bits in all ranges.
03-08-2010 First harmonic distortion measurements made. Needs further investigation as with -60dB is higher than expected.
03-09-2010 The distortion source has been found.
For further information go to Investigating the distortion source...
03-09-2010 Thermal stability measurements made.
03-09-2010 Frequency response of the anti-aliasing filter will be verified and possibly corrected.
Needs development of software that shows frequency response.
07-09-2010 Both I2C buses are working. Accessing memory, setting the sampling frequency and
making temperature measurements are now possible.
08-09-2010 Issue with linearity in the 30-90 kHz region caused by offset compensation circuit being investigated.
20-09-2010 The schematic has been updated. All the differences for the second version are given here :PDF
29-09-2010 Updates to PCB made (EDA-02063-V2). Ten boards ordered. Designer left CERN.
22-10-2010 Found inconsistencies in BOM. Requires modifications to schematics and production files.
08-12-2010 Production of ten boards of V2 will arrive before 17 December. Update 16-12-2010: will become 7 January.
07-01-2011 Ten V2 boards arrived.
23-02-2011 First board powered up on a SPEC carrier, no smoke!
03-03-2011 Trial of wave soldering of LEMO connector instead of SMC looks OK.
PCB redesign needed if want to use wave soldering.
08-03-2011 Basic HDL code written (single shot, no time stamps). Needs testing.
28-03-2011 Basic HDL code tested. We're able to DMA raw adc data from DDR3 memory to the host and gnuplot them.
21-06-2011 Made schematics and layout modifications for V3 (see changelog).
22-06-2011 Project sent to DEM for front panel modification and production file generation. Ready on 24-06-2011.
27-06-2011 V3 ready for review.
04-07-2011 V3 reviewed. V3 review. 3 prototypes will be built.
01-08-2011 Price Enquiry sent out for first Open Hardware production.
12-08-2011 Firmware being finalised. Made re-usable for use in the fmc-adc-100k16b8cha, a slower ADC mezzanine.
18-08-2011 Design of a tester board.
30-08-2011 Production test software being written.
12-09-2011 Price Enquiry finished. Received replies from several companies. Order for 40 cards being made.
15-09-2011 V4 design made. Corrected some textual problems and one BOM item order number.
20-09-2011 Order for 40 cards placed with INCAA Computers (V5). Delivery in January 2012.
10-10-2011 Useful feedback from INCAA Computers made us improve the design as V5.
02-12-2011 Firmware getting stable. Demo oscilloscope program in Python written.
20-12-2011 Measured bandwidth is 30 MHz. 40 MHz bandwidth is possible by changing 2 capacitors per channel.
10-01-2012 Expect delivery of 40 cards in first week of March end March (update 7-3-2012).
04-04-2012 First production received: 40 V5 cards produced by INCAA Computers.
05-02-2013 Firmware development on-going. In parallel a driver is being written. Expect a release in April 2013
12-03-2013 Release 1.0 of the fmc-adc gateware for SPEC is available.
28-03-2013 Release 1.1 of the fmc-adc gateware for SPEC is available.
19-07-2013 Firmware ported to work on the SVEC VME carrier.
29-07-2013 Release 2.0 of the fmc-adc gateware for SPEC is available.
29-07-2013 Release 1.0 of the fmc-adc gateware for SVEC is available.
17-01-2014 Release 3.0 of the fmc-adc gateware for SPEC and SVEC is available.
24-03-2014 Cards deployed in CERN's accelerator complex. Start of development to deployement: 4.5 year.

Maciej Firmiarz, Matthieu Cattin, Erik van der Bij - 27 March 2014

PCB_bot_side_soldered_03_s.jpg - Wave solder trial with LEMO connector (76.6 kB) Erik van der Bij, 14/03/2011 10:05