Site Home   Archive Home   FAQ Home   How to search the Archive   How to Navigate the Archive   
Compare FPGA features and resources   

Threads starting:
1994JulAugSepOctNovDec1994
1995JanFebMarAprMayJunJulAugSepOctNovDec1995
1996JanFebMarAprMayJunJulAugSepOctNovDec1996
1997JanFebMarAprMayJunJulAugSepOctNovDec1997
1998JanFebMarAprMayJunJulAugSepOctNovDec1998
1999JanFebMarAprMayJunJulAugSepOctNovDec1999
2000JanFebMarAprMayJunJulAugSepOctNovDec2000
2001JanFebMarAprMayJunJulAugSepOctNovDec2001
2002JanFebMarAprMayJunJulAugSepOctNovDec2002
2003JanFebMarAprMayJunJulAugSepOctNovDec2003
2004JanFebMarAprMayJunJulAugSepOctNovDec2004
2005JanFebMarAprMayJunJulAugSepOctNovDec2005
2006JanFebMarAprMayJunJulAugSepOctNovDec2006
2007JanFebMarAprMayJunJulAugSepOctNovDec2007
2008JanFebMarAprMayJunJulAugSepOctNovDec2008
2009JanFebMarAprMayJunJulAugSepOctNovDec2009
2010JanFebMarAprMayJunJulAugSepOctNovDec2010
2011JanFebMarAprMayJunJulAugSepOctNovDec2011
2012JanFebMarAprMayJunJulAugSepOctNovDec2012
2013JanFebMarAprMayJunJulAugSepOctNovDec2013
2014JanFebMarAprMayJunJulAugSepOctNovDec2014
2015JanFebMarAprMayJunJulAugSepOctNovDec2015
2016JanFebMarAprMayJunJulAugSepOctNovDec2016
2017JanFebMarAprMayJunJulAugSepOctNovDec2017
2018JanFebMarAprMayJunJulAugSepOctNovDec2018
2019JanFebMarAprMayJunJulAugSepOctNovDec2019
2020JanFebMarAprMay2020

Authors:A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Custom Search

Messages from 114300

Article: 114300
Subject: EDIF generation from C
From: "quad" <fyp.quadruples@gmail.com>
Date: 10 Jan 2007 23:35:28 -0800
Links: << >>  << T >>  << A >>
Hello
As I'd mentioned in one of my previous mails, I'm working on intrinsic
evolution. I'd like to know if its possible to generate netlists in
EDIF format from C itself for circuits, or will it be too complicated?
Regards
quad


Article: 114301
Subject: Quick question on Coolrunner II IO voltages
From: "PeteS" <PeterSmith1954@googlemail.com>
Date: 11 Jan 2007 01:25:59 -0800
Links: << >>  << T >>  << A >>
Hi chaps

I'm fairly sure I can power down the IO buffers of a Coolrunner II with
the internal logic powered, but I'm looking to see if anyone does this
or has done it just for confirmation.

>From the datasheet:

--------------------------------------------------------

Mixed Voltage, Power Sequencing, and
Hot Plugging

As mentioned in I/O Banking, CoolRunner-II CPLD parts support mixed
voltage I/O signals. It is important to assign signals to an I/O bank
with the appropriate I/O voltage.
Driving a high voltage into a low voltage bank can result in negative
current flow through the power supply pins.
The power applied to the VCCIO and VCC pins can occur in any order and
the CoolRunner-II CPLD will not be damaged.
***For best results, we recommend that VCCINT be applied before
VCCIO.*** This will ensure that the internal logic is correct before
the I/Os are active. CoolRunner-II CPLDs can reside on boards where the
board is inserted into a "live" connector (hot plugged) and the
parts will be well-behaved as if powering up in a standard way.

---------------------------------------------------------

Highlight mine. That implies I can power (should!) the internal logic
before powering the IOs, but can I do that indefinitely?

Any comments?

Cheers

PeteS


Article: 114302
Subject: Re: VHDL Model of a stepper motor
From: Jonathan Bromley <jonathan.bromley@MYCOMPANY.com>
Date: Thu, 11 Jan 2007 09:33:18 +0000
Links: << >>  << T >>  << A >>
On 10 Jan 2007 08:05:18 -0800, "fpgauser"
<fpgaengineerfrankfurt@arcor.de> wrote:

>Anybody allready designed a VHDL model of a stepper motor to simulate
>in modelsim ?

Do you really mean a model of a _stepper motor_?  If so, you will
need to decide how you map from the physical properties and 
operating conditions of the motor - shaft angle, externally
applied torque etc - to VHDL data types.  And you will need to
decide how fast to sample and update your model in order
to mimic its continuous-time behaviour in the discrete-time
VHDL simulator.

It's not even so simple to decide how to model the electrical
inputs to such a motor.

On the other hand, if you simply want to assume that the
motor is very lightly loaded and is driven by on/off
digital signals, you may be able to write a very crude
discretized model that looks at the electrical input, 
decides what the settled shaft angle should be for that
input, and causes the shaft angle to progress towards
that position over time.  I suspect, though, that you will
still need to model the motor's speed in some way.

For unusual applications like this, it's important to 
specify the level of modelling accuracy you require -
it makes a huge difference to the difficulty of creating
a model.

Note that the mixed-signal simulator people solved 
these things a long time ago; if you have a complete
mixed-signal simulator then you will probably get 
quite a few such models bundled with it.

Finally, since you are working in ModelSim, it is left as
a trivial exercise for the reader to write a little Tk graphical
widget that will display the motor's position visually as the
simulation proceeds :-)
-- 
Jonathan Bromley, Consultant

DOULOS - Developing Design Know-how
VHDL * Verilog * SystemC * e * Perl * Tcl/Tk * Project Services

Doulos Ltd., 22 Market Place, Ringwood, BH24 1AW, UK
jonathan.bromley@MYCOMPANY.com
http://www.MYCOMPANY.com

The contents of this message may contain personal views which 
are not the views of Doulos Ltd., unless specifically stated.

Article: 114303
Subject: Interlock and stall in CPU design?
From: "Davy" <zhushenli@gmail.com>
Date: 11 Jan 2007 01:35:09 -0800
Links: << >>  << T >>  << A >>
Hi all,

I am new to CPU design and confused with two CPU term. That is
interlock and stall. What's their difference? Does interlock and stall
all do insert NOP and remove the data dependency?

I have copied a sentence from a CPU document "The interlock is
responsible for detecting read-after-write hazards and stalling the
pipeline until the hazard has been resolved. This avoids the need to
insert nop directives between dependent instructions, thus keeping code
size to a minimum, as well as simplifying assembler-level programming."

Best regards,
Davy


Article: 114304
Subject: Re: Transport Delays in Modelsim
From: Jonathan Bromley <jonathan.bromley@MYCOMPANY.com>
Date: Thu, 11 Jan 2007 09:43:12 +0000
Links: << >>  << T >>  << A >>
On Wed, 10 Jan 2007 18:41:44 -0700, Kevin Neilson
<kevin_neilson@removethiscomcast.net> wrote:

>Has anyone ever been able to get Modelsim to model transport delays in 
>Verilog?  Verilog simulators, by default, use inertial delays, so if you 
>have an assignment such as this:
>
>assign #4 sig_out = sig_in;
>
>then any pulse on sig_in that is less than 4ns will get swallowed. 
>Modeling transport delays prevents this from happening.  Modelsim claims 
>to model transport delays using the +transport_int_delay option for 
>vsim, but this just doesn't seems to work.

I think you'll find that this option applies only to interconnect
delays that have been backannotated from an SDF file.  The Verilog
language defines continuous driver delays and net delays to be
inertial, and I was under the impression that simulators aren't
supposed to disobey that.

If you want transport delay in your own Verilog model, use
intra-assignment nonblocking delays - try tnis...

always @(sig_in) sig_out <= #4 sig_in;

(Of course, sig_out needs to be a variable now, not a net.)
-- 
Jonathan Bromley, Consultant

DOULOS - Developing Design Know-how
VHDL * Verilog * SystemC * e * Perl * Tcl/Tk * Project Services

Doulos Ltd., 22 Market Place, Ringwood, BH24 1AW, UK
jonathan.bromley@MYCOMPANY.com
http://www.MYCOMPANY.com

The contents of this message may contain personal views which 
are not the views of Doulos Ltd., unless specifically stated.

Article: 114305
Subject: Re: Quick question on Coolrunner II IO voltages
From: "-jg" <Jim.Granville@gmail.com>
Date: 11 Jan 2007 01:49:45 -0800
Links: << >>  << T >>  << A >>

PeteS wrote:
<snip>
> Highlight mine. That implies I can power (should!) the internal logic
> before powering the IOs, but can I do that indefinitely?

I'd say yes. but you may want to also check the IccINT, as
I presume your motivation is to actually save power.
With VccIO 'removed' you may need to ensure all pins
are true CMOS levels ( within 500mV of GND ) .
Check it by varying the IO float voltage,and measuring 
the IccINT 
-jg


Article: 114306
Subject: Re: EDIF generation from C
From: "Amit" <amit3007@gmail.com>
Date: 11 Jan 2007 02:57:22 -0800
Links: << >>  << T >>  << A >>

quad wrote:
> Hello
> As I'd mentioned in one of my previous mails, I'm working on intrinsic
> evolution. I'd like to know if its possible to generate netlists in
> EDIF format from C itself for circuits, or will it be too complicated?
> Regards
> quad

Ohh.... I think it will be too complicated to generate EDIF netlist
from C. We can look at the things this way.....Generating EDIF from C
is like first generate VHDL/Verilog code from C and then generate EDIF
from VHDL/Verilog RTL. Now generating EDIF from RTL itself is a big
task as this is nothing but synthesis.

Anyways, there's no word like 'Impossible' in some dictionaries.

Regards,


Article: 114307
Subject: Re: Interlock and stall in CPU design?
From: "Jon Beniston" <jon@beniston.com>
Date: 11 Jan 2007 03:09:45 -0800
Links: << >>  << T >>  << A >>
Hi,

> I am new to CPU design and confused with two CPU term. That is
> interlock and stall. What's their difference? Does interlock and stall
> all do insert NOP and remove the data dependency?
>
> I have copied a sentence from a CPU document "The interlock is
> responsible for detecting read-after-write hazards and stalling the
> pipeline until the hazard has been resolved. This avoids the need to
> insert nop directives between dependent instructions, thus keeping code
> size to a minimum, as well as simplifying assembler-level programming."

In this use, interlock is refering the logic that tracks dependencies
between instructions, and then stalls (pauses the pipeline from the
decode stage backwards) when a dependency is detected that cannot be
resolved by bypassing.

There are other causes of stall that are not related to the interlock,
such as stalls that occur on cache misses.

Cheers,
Jon


Article: 114308
Subject: Re: First Picture of Craignell Modules
From: "John Adair" <g1@enterpoint.co.uk>
Date: 11 Jan 2007 03:32:57 -0800
Links: << >>  << T >>  << A >>
Not set in stone yet but one off prices are likely to be GBP=A325-30,
US$ 40-50 on current exchange, for the base level versions with the
XC3S100E. If it starts to ship in reasonable numbers it is likely we
will drop that price back a bit. Certainly discounts will be there for
orders of 10+ even at the initial production run rate.

The 3 versions we have done are 32,36 and 40 pins. We will probably do
a 28pin and maybe 48pins as well. Anything up to 42 pins is easy as we
don't change the core design. We just hook up more signals to the bus
switches than provide the 5V protection. Beyond that we add another bus
switch but there is area available in the bigger pinouts.

CFCARD merely there to show size for those wondering.

We are aiming for these to be a stock item towards the end of Feb but
that is subject to usual spanners in the works.

John Adair
Enterpoint Ltd.

-jg wrote:
> Dave Pollum wrote:
> > Not from John,but -- http://www.enterpoint.co.uk/
> > John, have you set prices yet for the Craignell boards?
> > -Dave Pollum
>
> Thanks, but on the home page Craignell Modules does not appear,
> and google for Craignell Modules finds nothing, also news has only
> quite old infos, and not until I take a guess, and dig into
> BoardProducts
> do I see Craignell mentioned, and finally here find a link
>
> http://www.enterpoint.co.uk/component_replacements/craignell.html
>
> Tip for John:
> Post photos in two resolutions: Lower for fast loading, and then
> click-on
> any photo, to allow higher res.
>
> There seem to be 3 variants, but all labeled FD2 ?, and these all seem
> to have two
> 'Pin1 Square pad convention' - also no photo of the PCB rear ?
> I can see a compact flash card, but no mention of how that relates to
> Craignell Modules - is there a CF socket on the rear ?
>=20
> -jg


Article: 114309
Subject: Re: inserting text into a video stream (from a pre-existing video source)
From: Martin Thompson <martin.j.thompson@trw.com>
Date: 11 Jan 2007 11:56:30 +0000
Links: << >>  << T >>  << A >>
"wallge" <wallge@gmail.com> writes:

> I would like to superimpose some text into a video stream coming from a
> camera into my fpga.
> Lets say I would like to print the values of several registers to the
> screen as simple ascii text. This text would be superimposed over the
> incoming video stream. Does anyone know if there exists a core, or some
> free vhdl floating around somewhere to do this kind of thing?
> I have seen some cores that do VGA timing generation, then insert pixel
> values values from a look up table (but these dont do text from what I
> can tell, and must be controlled by a microprocessor).
> 

What format is your incoming video in?

Assuming it's got something like an LVALID and FVALID (and DVALID if
the clk is not the same as the pixel clock), I would use a BRAM as a
character frame buffer, decode the registers and stick the cahracters
into that.  Then another BRAM can have an array of bits indexed by
character (make them 8x8 for ease :-) that can be used by  a process
which keeps track of the location of the pixel and which of the
character locations you are in.

You could have a small state machine run at the end of each frame to
update the character frame buffer with the values you want for next
time around during the vblank.  Or double buffer.

Or do you really want to have something which looks like
ENTITY reg_to_screen
(
      clk
      reset
      registervalue (31 downto 0)
      xloc, yloc

      vsync, hsync
      video_out);

which you can then instantiate a number of times and then combine the
video_out signals together?  If so, you'll end up duplicating the
character->pixels map unnecesarily.

There's some thoughts for you :-)

Cheers,
Martin

-- 
martin.j.thompson@trw.com 
TRW Conekt - Consultancy in Engineering, Knowledge and Technology
http://www.conekt.net/electronics.html

   

Article: 114310
Subject: Re: EDIF generation from C
From: "acd" <acd4usenet@lycos.de>
Date: 11 Jan 2007 04:28:15 -0800
Links: << >>  << T >>  << A >>

quad wrote:
> Hello
> As I'd mentioned in one of my previous mails, I'm working on intrinsic
> evolution. I'd like to know if its possible to generate netlists in
> EDIF format from C itself for circuits, or will it be too complicated?
> Regards
> quad

You can of course write a C-program that generates an EDIF netlsit.
This is not very hard, as EDIF is a straight-forward ASCII-format.
The standard is openly available and you can look at EDIF-files from
standard examples in you FPGA toolkit.
For circuits with a high degree of regularity or with many
sophisticated
parameters it makes sense to do so.
But I would recommend you first to do some examples directly in
VHDL/Verilog first so that you get a feeling, which structures are well
suited for the tools. 

Regards, 
Andreas


Article: 114311
Subject: Re: Interlock and stall in CPU design?
From: "Ben Jones" <ben.jones@xilinx.com>
Date: Thu, 11 Jan 2007 13:00:26 -0000
Links: << >>  << T >>  << A >>
Hi Davy,

"Davy" <zhushenli@gmail.com> wrote in message 
news:1168508109.724917.173270@i39g2000hsf.googlegroups.com...
> Hi all,
>
> I am new to CPU design and confused with two CPU term. That is
> interlock and stall. What's their difference? Does interlock and stall
> all do insert NOP and remove the data dependency?

The data dependency is a property of the program being executed, and is 
never "removed" except by maybe changing the program.

In a machine where each instruction completes before the following 
instruction starts, data dependencies don't cause any problems. However, 
when a machine is pipelined, data dependencies and anti-dependencies can 
lead to "hazards".

An interlock circuit detects these hazards. An interlock circuit may stall 
one or more stages of the processor pipeline, often introducing a NOP-like 
"bubble", to avoid the hazard and ensure correct program execution.

Note that there are other ways to manage hazards that do not involve an 
interlock circuit, and that stalling the processor is only the simplest, 
most basic way to avoid hazards. You may wish to read Hennessy and Patterson 
for a proper description of more advanced techniques.

Cheers,

        -Ben- 



Article: 114312
Subject: Re: Interlock and stall in CPU design?
From: "JoshforRefugee" <ankitks@yahoo.com>
Date: 11 Jan 2007 05:12:34 -0800
Links: << >>  << T >>  << A >>
Hi Davy,
Interlock in this case refers to 'dependencies solver'. It might be a
simple stall logic or could be  register renaming logic


Article: 114313
Subject: Re: Interlock and stall in CPU design?
From: Yao Qi <qiyaoltc@gmail.com>
Date: Thu, 11 Jan 2007 22:01:56 +0800
Links: << >>  << T >>  << A >>
"Davy" <zhushenli@gmail.com> writes:

> Hi all,
>
> I am new to CPU design and confused with two CPU term. That is
> interlock and stall. What's their difference? Does interlock and stall
> all do insert NOP and remove the data dependency?

I am not a CPU guru, and here are some my cents.  Hope they are useful.

1. "stall" is a term for pipeline, and "interlock" is a method to
"stall" pipeline.

2. Both interlock and stall DO NOT insert NOP.  It is the work done by
compiler.  Compiler could insert some NOPs in executable to avoid data hazard.
>
> I have copied a sentence from a CPU document "The interlock is
> responsible for detecting read-after-write hazards and stalling the
> pipeline until the hazard has been resolved. This avoids the need to
> insert nop directives between dependent instructions, thus keeping code
> size to a minimum, as well as simplifying assembler-level
> programming."

Interlock logic could lock the pipeline when data hazard is detected to
make sure the right result.  So, compiler or assembler do not insert
NOPs to avoid hazard, so that code size is minimized.

>
> Best regards,
> Davy
>

-- 
Yao Qi
GNU/Linux Developer

Article: 114314
Subject: picoblaze RS-232 using 62.5 MHz
From: "axr0284" <axr0284@yahoo.com>
Date: 11 Jan 2007 08:41:32 -0800
Links: << >>  << T >>  << A >>
Hi,
 I was currently using picoblaze with RS-232 with the FPGA running at
50 MHz and it worked fine. I now need to run it at 62.5 MHz. From the
information I hve gathered, to attain a baud rate of 115200, I would
need a clock frequency of 115200 * 16  = 1843200 Hz or something close
to that.

In the picobalze hdl file, there is a counter that counts 27 cycle at
50 MHz to obtain 1851851 Hz and it seems to work fine with the RS-232.
So I figure 33 cycles at 62.5 MHz to obtain 1893939 Hz should also work
but I keep getting garbage on the terminal program on the computer. I
have tried 32 cycles as well as 34 cycles without any luck.

I was wondering if anybody has tried this before and can shed some
light on the matter. Thanks a lot,
Amish


Article: 114315
Subject: Re: picoblaze RS-232 using 62.5 MHz
From: "John_H" <newsgroup@johnhandwork.com>
Date: Thu, 11 Jan 2007 09:20:54 -0800
Links: << >>  << T >>  << A >>
UARTS are good for a bit of a frequency mismatch but going from .47% error 
to 2.75% error appears to be too much for your system to handle.  Consider 
dividing 62.5 MHz by 34 instead to get .27% error.  It shouldn't be 
important that 1838235 Hz is less than 1843200 Hz but that it's much closer 
than 1893939 Hz.


"axr0284" <axr0284@yahoo.com> wrote in message 
news:1168533692.658912.327530@i56g2000hsf.googlegroups.com...
> Hi,
> I was currently using picoblaze with RS-232 with the FPGA running at
> 50 MHz and it worked fine. I now need to run it at 62.5 MHz. From the
> information I hve gathered, to attain a baud rate of 115200, I would
> need a clock frequency of 115200 * 16  = 1843200 Hz or something close
> to that.
>
> In the picobalze hdl file, there is a counter that counts 27 cycle at
> 50 MHz to obtain 1851851 Hz and it seems to work fine with the RS-232.
> So I figure 33 cycles at 62.5 MHz to obtain 1893939 Hz should also work
> but I keep getting garbage on the terminal program on the computer. I
> have tried 32 cycles as well as 34 cycles without any luck.
>
> I was wondering if anybody has tried this before and can shed some
> light on the matter. Thanks a lot,
> Amish 



Article: 114316
Subject: Re: EDIF generation from C
From: "quad" <fyp.quadruples@gmail.com>
Date: 11 Jan 2007 09:59:29 -0800
Links: << >>  << T >>  << A >>
Hello
Actually we have a timeframe constraint of about 3 months for the
project. So wondering if generation of EDIF from C is possible within
that timeframe, considering that this is just a module in the whole
project. Would it be a better idea to pass parameters from C and use
JHDL (www.jhdl.org) to generate the netlist directly? Can JHDL code be
made generic for all kinds of circuits?
Regards
Quad


Article: 114317
Subject: Xilinx Synchronous FIFOs
From: "Brad Smallridge" <bradsmallridge@dslextreme.com>
Date: Thu, 11 Jan 2007 10:00:38 -0800
Links: << >>  << T >>  << A >>
Hello,

I have been looking at my FIFO designs and considering
if they are backward compatible from Virtex4 to Spartan.

I have been using the FWFT First_Word_Fall_Through feature
on the Virtex4 and when I compared waveforms of this with
those from a CoreGen FIFO got different results. I am still
using ISE 7.1.04 and the CoreGen models were Version 5.

I did not see any FWFT switches on my CoreGen wizard.
Are there any FWFT features in ISE 8.2?

Also it seemed that without the FWFT feature, the Virtex4
FIFO required 4 clock cycles from the wr_en to the rd_en
signals, while the CoreGen and the Virtex4 FIFO with FWFT
on requires only 3.

Brad Smallridge
brad at AiVision dot com



Article: 114318
Subject: Re: inserting text into a video stream (from a pre-existing video source)
From: "Brad Smallridge" <bradsmallridge@dslextreme.com>
Date: Thu, 11 Jan 2007 10:14:11 -0800
Links: << >>  << T >>  << A >>
> Really
> what I want is a function that would
> convert the value of a register into say, hex or base 10 ascii text,
> and then count the number of
> rows/columns of video coming in, and insert black or white pixels into
> the video stream as appropriate to make those characters appear
> (superimposed) on the output video stream.

Yeah. I might be able to help although I don't know what you are
using. Xilinx? Spartans? Virtex? VHDL? Verilog?

Can we assume that you already have a VGA display that is outputting?
And so can we assume that you have row and column registers in your
design?

Brad Smallridge
AiVision dot com





Article: 114319
Subject: Re: picoblaze RS-232 using 62.5 MHz
From: "axr0284" <axr0284@yahoo.com>
Date: 11 Jan 2007 10:31:42 -0800
Links: << >>  << T >>  << A >>
I have tried that and I still get garbage. This is weird. I am
wondering if the rest of the UART has something to do with this. Maybe
additional changes need to be made.
Amish

John_H wrote:
> UARTS are good for a bit of a frequency mismatch but going from .47% error
> to 2.75% error appears to be too much for your system to handle.  Consider
> dividing 62.5 MHz by 34 instead to get .27% error.  It shouldn't be
> important that 1838235 Hz is less than 1843200 Hz but that it's much closer
> than 1893939 Hz.
>
>
> "axr0284" <axr0284@yahoo.com> wrote in message
> news:1168533692.658912.327530@i56g2000hsf.googlegroups.com...
> > Hi,
> > I was currently using picoblaze with RS-232 with the FPGA running at
> > 50 MHz and it worked fine. I now need to run it at 62.5 MHz. From the
> > information I hve gathered, to attain a baud rate of 115200, I would
> > need a clock frequency of 115200 * 16  = 1843200 Hz or something close
> > to that.
> >
> > In the picobalze hdl file, there is a counter that counts 27 cycle at
> > 50 MHz to obtain 1851851 Hz and it seems to work fine with the RS-232.
> > So I figure 33 cycles at 62.5 MHz to obtain 1893939 Hz should also work
> > but I keep getting garbage on the terminal program on the computer. I
> > have tried 32 cycles as well as 34 cycles without any luck.
> >
> > I was wondering if anybody has tried this before and can shed some
> > light on the matter. Thanks a lot,
> > Amish


Article: 114320
Subject: RocketIO, MGT documentation. Does MGT clcok have to be 50% duty cycle?
From: "Dale" <dale.prather@gmail.com>
Date: 11 Jan 2007 10:44:49 -0800
Links: << >>  << T >>  << A >>
Is it just me or is the documentation for the RocketIO (MGT) for the
Xilinx Virtex4 very bad?  I'm trying to find duty cycle requirements
for the MGT clock.  Is it OK to use a clock with a 40% duty cycle?

Also, if anyone can point me to some better RocketIO (MGT)
documentation for the hardware guys I'd appreciate it.  I already have
ug076.

Thanks,
Dale


Article: 114321
Subject: Re: picoblaze RS-232 using 62.5 MHz
From: "axr0284" <axr0284@yahoo.com>
Date: 11 Jan 2007 11:03:59 -0800
Links: << >>  << T >>  << A >>
I think the issue might be in my code. I'll have to debug it.
Amish

axr0284 wrote:
> I have tried that and I still get garbage. This is weird. I am
> wondering if the rest of the UART has something to do with this. Maybe
> additional changes need to be made.
> Amish
>
> John_H wrote:
> > UARTS are good for a bit of a frequency mismatch but going from .47% error
> > to 2.75% error appears to be too much for your system to handle.  Consider
> > dividing 62.5 MHz by 34 instead to get .27% error.  It shouldn't be
> > important that 1838235 Hz is less than 1843200 Hz but that it's much closer
> > than 1893939 Hz.
> >
> >
> > "axr0284" <axr0284@yahoo.com> wrote in message
> > news:1168533692.658912.327530@i56g2000hsf.googlegroups.com...
> > > Hi,
> > > I was currently using picoblaze with RS-232 with the FPGA running at
> > > 50 MHz and it worked fine. I now need to run it at 62.5 MHz. From the
> > > information I hve gathered, to attain a baud rate of 115200, I would
> > > need a clock frequency of 115200 * 16  = 1843200 Hz or something close
> > > to that.
> > >
> > > In the picobalze hdl file, there is a counter that counts 27 cycle at
> > > 50 MHz to obtain 1851851 Hz and it seems to work fine with the RS-232.
> > > So I figure 33 cycles at 62.5 MHz to obtain 1893939 Hz should also work
> > > but I keep getting garbage on the terminal program on the computer. I
> > > have tried 32 cycles as well as 34 cycles without any luck.
> > >
> > > I was wondering if anybody has tried this before and can shed some
> > > light on the matter. Thanks a lot,
> > > Amish


Article: 114322
Subject: Re: Xilinx Synchronous FIFOs
From: "Andrew Holme" <andrew@nospam.com>
Date: Thu, 11 Jan 2007 19:04:26 -0000
Links: << >>  << T >>  << A >>

"Brad Smallridge" <bradsmallridge@dslextreme.com> wrote in message 
news:12qcuru8aagv82d@corp.supernews.com...
> I did not see any FWFT switches on my CoreGen wizard.
> Are there any FWFT features in ISE 8.2?
>
Yes. 



Article: 114323
Subject: Re: Transport Delays in Modelsim
From: Kevin Neilson <kevin_neilson@removethiscomcast.net>
Date: Thu, 11 Jan 2007 12:04:50 -0700
Links: << >>  << T >>  << A >>
Jonathan Bromley wrote:
> On Wed, 10 Jan 2007 18:41:44 -0700, Kevin Neilson
> <kevin_neilson@removethiscomcast.net> wrote:
> 
>> Has anyone ever been able to get Modelsim to model transport delays in 
>> Verilog?  Verilog simulators, by default, use inertial delays, so if you 
>> have an assignment such as this:
>>
>> assign #4 sig_out = sig_in;
>>
>> then any pulse on sig_in that is less than 4ns will get swallowed. 
>> Modeling transport delays prevents this from happening.  Modelsim claims 
>> to model transport delays using the +transport_int_delay option for 
>> vsim, but this just doesn't seems to work.
> 
> I think you'll find that this option applies only to interconnect
> delays that have been backannotated from an SDF file.  The Verilog
> language defines continuous driver delays and net delays to be
> inertial, and I was under the impression that simulators aren't
> supposed to disobey that.
> 
> If you want transport delay in your own Verilog model, use
> intra-assignment nonblocking delays - try tnis...
> 
> always @(sig_in) sig_out <= #4 sig_in;
> 
> (Of course, sig_out needs to be a variable now, not a net.)

That is wonderful.  It totally works.  I wish I'd known this before.  My 
Palnitkar book makes no reference to this (unsurprisingly) that I can 
see though I did find a reference in a book by Bhasker.  I would like to 
be able to adjust  the delay on-the-fly, in order to model pad/trace 
delays that change over time, but I don't think that's possible.  I 
suppose maybe I could get a switchable delay doing something like this:

always@(sig_in, dly_sel) sig_out <= dly_sel ? #4 sig_in : #5 sig_in;

-Kevin

Article: 114324
Subject: Re: crossing clock domain ??
From: "Matthew Hicks" <mdhicks2@uiuc.edu>
Date: Thu, 11 Jan 2007 13:49:08 -0600
Links: << >>  << T >>  << A >>
The method you just described sounds similar to what is done with MGT 
protocols.  Send a special K character that can either be processed to waste 
time or thrown away to catch up.  For small unknown differences that is a 
great idea.  As for the FIFO solution I mentioned earlier, our group is 
using it to cross clock domains.  We are using a NACK to control the sender 
and are also using extensive calculations to ensure that all FIFOs can 
handle a given data valid rate from the sender (because part of our system 
requirements is that we need to gurantee a minimum throughtput).


---Matthew Hicks


"Peter Alfke" <alfke@sbcglobal.net> wrote in message 
news:1168489015.772863.289540@i39g2000hsf.googlegroups.com...
> One solution is to have the transmitter include padding, extra bits or
> words that the receiver recognizes and can, if necessary,  "throw away"
> = skip over. That way, the receiver does not need a back communication
> to the transmitter. The price is of course a loss of throughput.
> There is no free lunch.
> Peter Alfke
>
> On Jan 10, 7:18 pm, "Rob" <robns...@frontiernet.net> wrote:
>> As John stated, make sure you do the analysis and assure yourself you can
>> use a FIFO, and if so, that it is large enough.  We ran into the very
>> problem--transmit clock faster than recieve clock--and our FIFO wasn't 
>> big
>> enough--ooops!  Furthermore, it was an ASIC--double oops!!  The solution 
>> was
>> to force the transmitter to use a slightly lower clock.  We weren't off 
>> by
>> much, less than 1%, but it was enough at times to cause problems.  Had we
>> not been able to slow the transmitter we would have been screwed!!
>>
>> "John_H" <newsgr...@johnhandwork.com> wrote in 
>> messagenews:12qak264aekr04b@corp.supernews.com...
>>
>> > Do you know what "large enough" is given the speeds of your two clock
>> > domains?  If your downstream clock domain is always slower than your
>> > upstream time domain, you cannot get a FIFO large enough.
>>
>> > If you can determine the maximum sustained upstream rate (less than or
>> > equal to the upstream clock frequency) and know the minimum sustained
>> > downstream clock frequency, you can determine the FIFO size that will
>> > guarantee a system that communicates all the information.  If your 
>> > output
>> > frequency isn't fast enough to keep up - ever - you're hosed from the
>> > start.
>>
>> > Keep in mind there will be a maximum delay encountered when the FIFO
>> > reaches full.
>>
>> > "Matthew Hicks" <mdhic...@uiuc.edu> wrote in message
>> >news:eo3h81$id4$1@news.ks.uiuc.edu...
>> >> Peter,
>>
>> >> Would your Asynchronous FIFO core handle this situation correctly 
>> >> given a
>> >> large enough FIFO?
>>
>> >> ---Matthew Hicks
>>
>> >> "Peter Alfke" <p...@xilinx.com> wrote in message
>> >>news:1168449451.035500.190590@p59g2000hsd.googlegroups.com...
>> >>>I wrote a TechXclusives paper a while ago, on this exact subject:
>> >>> You can click on the (ridiculously long) URL:
>>
>> >>>http://www.xilinx.com/xlnx/xweb/xil_tx_display.jsp?iLanguageID=1&cate...
>>
>> >>> When the transmit clock is faster than the receive clock, there is of
>> >>> course the danger of "overrunning" the receiver, and thus losing
>> >>> data...
>> >>> Peter Alfke, Xilinx
>>
>> >>> On Jan 10, 8:12 am, "Matthieu Cattin" <matthieu.cat...@cern.ch> 
>> >>> wrote:
>> >>>> Signal can change on each rising edge or less often.
>>
>> >>>> "John_H" <newsgr...@johnhandwork.com> wrote in
>> >>>> messagenews:12qa3g2otrg95e1@corp.supernews.com...
>>
>> >>>> > How often does the signal change relative to the two clocks?
>>
>> >>>> > "Matthieu Cattin" <matthieu.cat...@cern.ch> wrote in message
>> >>>> >news:eo30bh$c3j$1@cernne03.cern.ch...
>> >>>> >> Hi all,
>>
>> >>>> >> I have to transfert signals from a clock domain to another.
>> >>>> >> First clock domain is fixed, but the second can be faster or 
>> >>>> >> slower
>> >>>> >> than
>> >>>> >> the first one.
>>
>> >>>> >> Does somebody can give me some help.
>>
>> >>>> >> Thanks
>> >>>> >> Matthieu
> 





Site Home   Archive Home   FAQ Home   How to search the Archive   How to Navigate the Archive   
Compare FPGA features and resources   

Threads starting:
1994JulAugSepOctNovDec1994
1995JanFebMarAprMayJunJulAugSepOctNovDec1995
1996JanFebMarAprMayJunJulAugSepOctNovDec1996
1997JanFebMarAprMayJunJulAugSepOctNovDec1997
1998JanFebMarAprMayJunJulAugSepOctNovDec1998
1999JanFebMarAprMayJunJulAugSepOctNovDec1999
2000JanFebMarAprMayJunJulAugSepOctNovDec2000
2001JanFebMarAprMayJunJulAugSepOctNovDec2001
2002JanFebMarAprMayJunJulAugSepOctNovDec2002
2003JanFebMarAprMayJunJulAugSepOctNovDec2003
2004JanFebMarAprMayJunJulAugSepOctNovDec2004
2005JanFebMarAprMayJunJulAugSepOctNovDec2005
2006JanFebMarAprMayJunJulAugSepOctNovDec2006
2007JanFebMarAprMayJunJulAugSepOctNovDec2007
2008JanFebMarAprMayJunJulAugSepOctNovDec2008
2009JanFebMarAprMayJunJulAugSepOctNovDec2009
2010JanFebMarAprMayJunJulAugSepOctNovDec2010
2011JanFebMarAprMayJunJulAugSepOctNovDec2011
2012JanFebMarAprMayJunJulAugSepOctNovDec2012
2013JanFebMarAprMayJunJulAugSepOctNovDec2013
2014JanFebMarAprMayJunJulAugSepOctNovDec2014
2015JanFebMarAprMayJunJulAugSepOctNovDec2015
2016JanFebMarAprMayJunJulAugSepOctNovDec2016
2017JanFebMarAprMayJunJulAugSepOctNovDec2017
2018JanFebMarAprMayJunJulAugSepOctNovDec2018
2019JanFebMarAprMayJunJulAugSepOctNovDec2019
2020JanFebMarAprMay2020

Authors:A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Custom Search