FMC Analog to Digital Converter developments
CERN's user needs are compared to define the specifications of the ADC cards to be developed. Main input is coming from Stephane Deghaye.
Specification
FmcAdc100M14b4cha | FmcAdc100M14b4chb | FmcAdc100M14b2cha | FmcAdc100k14b8cha | FmcAdc100k14b2cha | |
4-channel, progr. gain | 4-channel, progr. offset | 2-channel, progr. gain&offset | 8-channel, progr. offset | 4-channel, progr. gain&offset | |
jumper settable gain | autocal (offset, gain) | jumper settable gain | autocal (offset, gain) | ||
Project aborted | On-going | Not started | Not started | Not started | |
Parameter | Value | Value | Value | Value | Value |
max. sample rate | 105 MSPS | 105 MSPS | 105 MSPS | 100 kSPS | 100 kSPS |
bits/sample (raw) | 14 bit | 14 bit | 14 bit | 16 bit | 16 bit |
bits/sample (effective) | |||||
channels | 4 | 4 | 2 | 8 | 4 |
analog bandwidth | 25 MHz | 25 MHz | 25 MHz | 25 kHz | 25 kHz |
Input impedance | 50 Ohm | 50 Ohm/ 10 KOhm jumper selectable | 50 Ohm/ 10 KOhm jumper selectable | 50 Ohm/ 10 KOhm jumper selectable | 50 Ohm/ 10 KOhm jumper selectable |
Offset adjustment | not possible | +/- 10V ? | +/- 10V ? | +/- 10V ? | +/- 10V ? |
Offset step size | |||||
min. input voltage range | +/- 40 mV for full scale | +/- 0.1V | +/- 40 mV for full scale | +/- 0.1V | +/- 40 mV for full scale |
max. input voltage range | +/- 10.24 V for full scale | +/- 10 V for full scale | +/- 10V for full scale | +/- 10 V for full scale | +/- 10V for full scale |
gain steps | 1,2,4...256 | +/- 0.1V, 1V, 10V jumper selectable | 1, 2, (4), 5, (8) | +/- 0.1V, 1V, 10V jumper selectable | 1, 2, (4), 5, (8) |
max. gain error | +/- 8.5% | +/- 2% | +/- 2% | +/- 2% | +/- 2% |
max. input voltage offset | max. 15 mV | ||||
Calibration | manual 0V | manual 0V | autocal (offset, gain), or only manual 0V | manual 0V | autocal (offset, gain), or only manual 0V |
External trigger | TTL (0.7V) | TTL (0.7V) | TTL (0.7V) | TTL (0.7V) | |
External clock | via FMC carrier | via FMC carrier | via FMC carrier | via FMC carrier | |
Connectivity | FMC/FMA/MMCX? | FMC/FMA/MMCX? | Fanout cable | FMC/FMA/MMCX? | |
SNR | |||||
Crosstalk | |||||
Notes | Cancelled development | Should be able to replace SIS modules | Priorities: 1) input impedance programmable, 2) autocal, 3) offset autocal, 4) gain autocal | ||
Notes |
Would be nice if SW could detect jumper settings. |
Would Stephane need a sweeping (unstable) clock? This would not be possible to implement with serial ADC |
> -----Original Message-----
> From: Stephane Deghaye
> Sent: Thursday 03 December 2009 17:36
> To: Erik Van Der Bij; Maciej Fimiarz; Javier Serrano
> Subject: RE: FMC ADC OHR page
>
> Without meeting:
>
> The offset could dependent on the full-scale (this is the
> usual case). However, if that can be done easily, a offset
> bigger than the full-scale is very convenient. Remember the
> example of the signal where the operation is interested by
> the full shape and the crest value. The bigger offset allows
> us to use a small full-scale (e.g. 1V) and still move the
> acquisition window on the top of the signal (e.g. +9V).
>
> I'll try to find out:
> - Typical full-scale (need to have a look at the global
> configurations)
> - Typical offset
> - Typical values of RF signals (not LLRF), CVORB/G, GFAS, PO
> (pulsed converters), kickers...
> - Typical sweep for external sampling frequencies (e.g. 96
> MHz in CTF).
>
> Cheers,
> Stephane
Status
- 30-11-2009 - FmcAdc100M14b4cha had too much SNR. Rethink user needs (e.g. programmable offset instead of programmable gain
- 2-12-2009 - Meeting with S. Deghaye on needs for ACCOR
-- ErikVanDerBij - 04 Dec 2009