-
X-Received: by 10.49.14.132 with SMTP id p4mr48076qec.2.1382003203385; Thu, 17 Oct
2013 02:46:43 -0700 (PDT)
X-Received: by 10.49.14.132 with SMTP id p4mr48076qec.2.1382003203385; Thu, 17 Oct
2013 02:46:43 -0700 (PDT)
Path: news-archive.icm.edu.pl!agh.edu.pl!news.agh.edu.pl!news.cyf-kr.edu.pl!news.nask
.pl!news.nask.org.pl!news.unit0.net!news.glorb.com!o2no9513707qas.0!news-out.go
ogle.com!9ni45434qaf.0!nntp.google.com!i2no17068572qav.0!postnews.google.com!gl
egroupsg2000goo.googlegroups.com!not-for-mail
Newsgroups: pl.misc.elektronika
Date: Thu, 17 Oct 2013 02:46:43 -0700 (PDT)
Complaints-To: g...@g...com
Injection-Info: glegroupsg2000goo.googlegroups.com; posting-host=81.219.220.14;
posting-account=67yd9woAAAAHUu8VHyA7Js47M98NE3m3
NNTP-Posting-Host: 81.219.220.14
User-Agent: G2/1.0
MIME-Version: 1.0
Message-ID: <8...@g...com>
Subject: FPGA - Xilinx
From: s...@g...com
Injection-Date: Thu, 17 Oct 2013 09:46:43 +0000
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Xref: news-archive.icm.edu.pl pl.misc.elektronika:653370
[ ukryj nagłówki ]Witam,
Mam problem, pierdylnełem opis na grupę dyskusyjną Xilinxa. Czasami dają sensowne
odpowiedzi, czasami nie. A problem kurde pilny!! Poniżej daję kopię tego co do nich
nabazgrałem po hamerykańsku. Jeżeli macie jakieś pomysły dot. rozwiązania problemu,
to weźcie coś podpowiedzcie. Bo q..va tydzień+ już nad tym ślęczę i nie mam do Pani
Nędzy pomysła jak to rozkminić..
Poniżej szczegóły po hamerykańsku:
Hi,
I've got a problem on my custom board. Based on XC6SLX45-324. OK, let's make you know
what I whant to do. Step by step:
1) Deserialization of data from just only one channel of AD9272. In fact, the logic
works fine according to my idea (2 shift registers+bits aligment). However, it works
fine up to 20MHz main (frame) clock frequency.
2) In fact, I should capture the data from AD9272 reliably @80MHz!!
3) For sure I've got the timing problems, because the design works fine @20MHz and at
higher frequenies fails. NOT a simulation, a real world!! LVDS signals are well
routed on my PCB and equally terminated by both TX/RX devices (AD9272/XCS6SLX45)
3) I've tried to use SelectIO Interface Winard 4.1 . Let's go page by page:
Page 1:
- Component name : del_line (name says what I intend to do. Just a delay of some tap
data delay [ps] in order to match the clocking "eye")
- Data Bus direction: Configure inputs to the device
- I/O signaling : differential
- I/O signaling stardant: LVDS 2.5
Page 2:
All the checkboxes are unchecked, External data width is set to "1".
Page 3:
- Delay Inserted Into Data Routing : I select *FIXED*, initialy set to *0*
Page 4:
- Clock Signaling : Differantial/LVDS 2.5
- Clock Buffer : BUFIO2
- Active Clock Edge : Both Rising and Falling
- Input DDR data aligment : Both Clock Edges: none
- IDDR reset type : ASYNC
Page 5:
Delay inserted Into clock routing : 0
=================
OK, then I click OK, wait a little, report says that the IP has been succesfuly
genrated!!
Well, now I put my new generated module into schematics. New generated module has
such a ports:
Outputs:
DATA_IN_TO_THE_DEVICE(1:0) - no question, understand it as 2 incoming data on pos and
neg clock edge.
Inputs:
DATA_IN_FROM_PINS_P(0:0) - LVDS(+) one wire "bus" as specified on core requirement
DATA_IN_FROM_PINS_N(0:0) - LVDS(-) one wire "bus" as specified on core requirement
CLK_IN_P/N - LVDS input clock pair from external device (AD9512)
======================
ATTENTION!! Here are problematic inputs and outputs:
1) IP core generator makes 2 additional inputs:
- CLK_RESET and IO_RESET. I've GND'ed both of them on my sch. Implementation
process says ERROR about the unknown CLK_RESET pin!
- OK, I've created a new sch module according to IP Core VHDL generated module.
- Implementation : succesfull!!
2) NOTHING on CLK_OUT !! Constant logical '0'.!!
3) For sure, there is no problem on my PCB. When I implement a test design with
CLK+/CLK- into the FPGA => IBUFGDS => CLK any counter => any bit into the output,
then I see on my scope, everything works fine.
4) I'm affraid, that something is wrong with the IP wizard. Can you help me in this
matter?
Następne wpisy z tego wątku
- 17.10.13 12:58 Adam Górski
- 17.10.13 15:09 s...@g...com
- 17.10.13 17:35 Adam Górski
- 17.10.13 21:24 s...@g...com
- 18.10.13 00:39 Adam Górski
- 18.10.13 09:39 s...@g...com
- 18.10.13 10:09 Adam Górski
- 18.10.13 13:37 s...@g...com
- 18.10.13 13:49 Adam Górski
- 18.10.13 18:00 s...@g...com
- 19.10.13 10:26 Adam Górski
- 20.10.13 02:09 s...@g...com
- 20.10.13 02:22 s...@g...com
Najnowsze wątki z tej grupy
- DS1813-10 się psuje
- Taki tam szkolny problem...
- LIR2032 a ML2032
- SmartWatch Multimetr bezprzewodowy
- olej psuje?
- Internet w lesie - Starlink
- Opis produktu z Aliexpress
- No proszę, a śmialiście się z hindusów.
- Zewnętrzne napięcie referencyjne LM385 1,2V -> 100mV dla ICL7106, Metex M-3800
- karta parkingowa
- Wl/Wyl (On/Off) bialy/niebieski
- I3C
- Pytanie o transformator do dzwonka
- międzymordzie USB 3.2 jako 2.0
- elektronicy powinni pomysleć o karierze elektryka
Najnowsze wątki
- 2024-11-25 Karty przedpłacone (podarunkowe) Google Play - pytanie do korzystających
- 2024-11-26 wina Tóska
- 2024-11-26 Rewolucja/Rewelacja!
- 2024-11-25 grupa ożyła ;)
- 2024-11-24 Być jak Clint
- 2024-11-24 Rura kanalizacja konceptu Franke = problem
- 2024-11-25 Wrocław => Lead Java EE Developer <=
- 2024-11-25 Warszawa => Business Development Manager - Network and Network Securit
- 2024-11-25 Kraków => Programista Full Stack (.Net Core) <=
- 2024-11-25 Lublin => Senior PHP Developer <=
- 2024-11-25 Karlino => Konsultant wewnętrzny SAP (FI/CO) <=
- 2024-11-25 Warszawa => ECM Specialist / Consultant <=
- 2024-11-25 Katowice => Regionalny Kierownik Sprzedaży (OZE) <=
- 2024-11-25 Warszawa => Senior Frontend Developer (React + React Native) <=
- 2024-11-25 Lublin => Inżynier Serwisu Sprzętu Medycznego <=