UPGRADE YOUR BROWSER

We have detected your current browser version is not the latest one. Xilinx.com uses the latest web technologies to bring you the best online experience possible. Please upgrade to a Xilinx.com supported browser:Chrome, Firefox, Internet Explorer 11, Safari. Thank you!

cancel
Showing results for 
Search instead for 
Did you mean: 
Contributor
Contributor
10,584 Views
Registered: ‎10-11-2011

Clocking wizard for SP601

Jump to solution

Hi there!

I have just acquired a Xilinx SP601 Eval board and have received the ISE software ver. 11.3 with the same. I am going through the In-depth tutorial of the ISE but am stuck at one point. In the 'Creating a DCM Module' section, I need to use the Clocking wizard. In that when I select FPGA Features and Design > Clocking, I cannot find the clocking option for the Spartan 6 board in the software.

Could someone help me as to what values are needed to be changes so that I can use this module on my Spartan 6 board. Can I directly use the FPGA Features and Design > Clocking > Clocking Wizard. ?? If so, what values do I need to change?

 

Regards,

Jack

1.jpg
0 Kudos
1 Solution

Accepted Solutions
Advisor joelby
Advisor
13,176 Views
Registered: ‎10-05-2010

Re: Clocking wizard for SP601

Jump to solution

How far did you get with the clocking generator? It will do everything you want, other than deciding what to do with the 100 MHz signal - that's where you come in.

 

If you need to forward it to an external pin, use an ODDR2 clock forwarding scheme. Have a look in the language templates for details about ODDR2.

 

Depending on your needs, you may also be able to just use the 200 MHz system clock with a divider to produce a 100 MHz output.

View solution in original post

23 Replies
Advisor joelby
Advisor
10,582 Views
Registered: ‎10-05-2010

Re: Clocking wizard for SP601

Jump to solution

Yes, just use Clocking Wizard.

 

You'll need to correctly set the input clock frequency based on the SP601 clock you wish to use (I think you've got a couple to choose from). The SP601 manual will tell you what the clocks are, which pins they're present on, and if they're single-ended or differential.

 

Contributor
Contributor
10,554 Views
Registered: ‎10-11-2011

Re: Clocking wizard for SP601

Jump to solution
Thanks a lot there joelby!
I now need to generate a simple square signal of 100 Mhz using the FPGA. I tried looking for any simple programs but to no avail!!!
Any help???

Regards,
Jack
0 Kudos
Advisor joelby
Advisor
13,177 Views
Registered: ‎10-05-2010

Re: Clocking wizard for SP601

Jump to solution

How far did you get with the clocking generator? It will do everything you want, other than deciding what to do with the 100 MHz signal - that's where you come in.

 

If you need to forward it to an external pin, use an ODDR2 clock forwarding scheme. Have a look in the language templates for details about ODDR2.

 

Depending on your needs, you may also be able to just use the 200 MHz system clock with a divider to produce a 100 MHz output.

View solution in original post

Contributor
Contributor
10,530 Views
Registered: ‎10-11-2011

Re: Clocking wizard for SP601

Jump to solution

Wooohooo!!!! I've done it!!!

Thank you thank you thank you Joelby!

0 Kudos
Advisor joelby
Advisor
10,523 Views
Registered: ‎10-05-2010

Re: Clocking wizard for SP601

Jump to solution

Excellent! I'm glad you sorted it out :)

Contributor
Contributor
10,515 Views
Registered: ‎10-11-2011

Re: Clocking wizard for SP601

Jump to solution

I have managed to generate the required signal as instructed and have got the output both at the SMA connector as well as the LEDs on the SP601 board (the code is attached). Now I have an FMC XM105 debug card and I want to select one of the pins on the debug card as my output. I wanted to get the output on the LEDs on the FMC board.

I checked the Pin connections and the LED corresponds to the following pin connections:

FMC connector J17 pin G37 corresponds to the LED DS1 on the board.

Checking the FMC LPC Pin Connections of the SP601 board, I found out that this pin (G37) corresponds to the (N9) pin on the FPGA board.

 

Now when I connect my output to the (N9) pin on the board, I should be getting the output on the FMC card, right? But that's not the case! I am not getting any output on the FMC board. The XM105 card is working properly (I hope), since on Power Good LEDs i.e. DS5, DS6, DS7 are on.

 

Are there some more steps to configure the XM105 board? I even tried changing the Jumper setting J4 on the SP601 board (in which it says to include FMC) but then I got an error while downloading the program on the board. The ISE iMPACT window gave an error "Program Failed" in that case.

What do I do??

0 Kudos
Advisor joelby
Advisor
10,510 Views
Registered: ‎10-05-2010

Re: Clocking wizard for SP601

Jump to solution

J4 is used to route the JTAG chain via the FMC LPC port. It's only necessary if your FMC card contains additional JTAG devices that you want to put in the chain.

 

Pin N9 looks like it's correct. Are you sure the LED is not working correctly? Have you probed it with a fast oscilloscope? (at least 100 MHz, since you're just checking for the existence a 100 MHz signal)

 

If you're toggling an LED at 100 MHz with a 50% duty cycle, perhaps you just can't see it lighting up. Start small, with a 1 Hz pulse, or just assign the LED output to '1' or something.

 

0 Kudos
Contributor
Contributor
10,480 Views
Registered: ‎10-11-2011

Re: Clocking wizard for SP601

Jump to solution

Hi there again Joel!
I've been trying many different permutations and combinations with my signal and have been steadily increasing the frequency from 20MHz up to 400MHz.
The only problem is that the signal starts degrading as the frequency increases. Up until 20MHz, I get a proper square wave with a 50% duty cycle. Thereafter the wave starts getting more and more noisier. After 100 MHz I cant see any square wave anymore, it simply becomes a sine wave. Why is that??
Also, how do I counter that effect? I need a clean signal at least until 100MHz wherein I get a clean square wave without much percievable noise.

I am taking the output at Pin H18 of the SP601 which corresponds to the SMACLK_N on to the oscilloscope. Have attached my program for reference.

 

Regards,

Jack

0 Kudos
Instructor
Instructor
10,474 Views
Registered: ‎07-21-2009

Re: Clocking wizard for SP601

Jump to solution

Are you using an oscilloscope to probe your signal?

 

If you want to measure a 100MHz digital signal with a 'scope, the scope and probes should be rated for at least 350MHz bandidth.  A 1GHz scope will give you more accurate measurements than a 350MHz scope, but 350MHz should be (barely) adequate.

 

A 100MHz scope (and probes) is entirely inadequate for measuring or displaying amplitude, risetime, shape (e.g. ringing or reflections) or duty cycle of a 100MHz digital signal.

 

-- Bob Elkind

SIGNATURE:
README for newbies is here: http://forums.xilinx.com/t5/New-Users-Forum/README-first-Help-for-new-users/td-p/219369

Summary:
1. Read the manual or user guide. Have you read the manual? Can you find the manual?
2. Search the forums (and search the web) for similar topics.
3. Do not post the same question on multiple forums.
4. Do not post a new topic or question on someone else's thread, start a new thread!
5. Students: Copying code is not the same as learning to design.
6 "It does not work" is not a question which can be answered. Provide useful details (with webpage, datasheet links, please).
7. You are not charged extra fees for comments in your code.
8. I am not paid for forum posts. If I write a good post, then I have been good for nothing.
0 Kudos
Contributor
Contributor
9,819 Views
Registered: ‎10-11-2011

Re: Clocking wizard for SP601

Jump to solution
Yes I am using a 500MHz,1GSa/s,4ch scope (Tektronix TDS540). So I dont really think that the problem lies with the Scope. Even with 50MHz. the signal shows significant noise on the upper and lower crests of the signal.
Does the FPGA change the characteristic of the signal with the increase in frequency??
0 Kudos
Instructor
Instructor
9,811 Views
Registered: ‎07-21-2009

Re: Clocking wizard for SP601

Jump to solution

Do you have the probes properly GNDed to a point very near the signal being probed?

You have the scope's vertical BW limit disabled, yes?

 

1.  zoom scale to barely fit a single rising edge in the first 3 horizontal divisions, and barely fit vertically.

2.  without changing H or V scale, grab photos of a 100MHz signal rising edge and a 20MHz (or so) rising edge.

 

Let's compare displayed waveforms with all scope settings and displays equal.

 

-- Bob Elkind

SIGNATURE:
README for newbies is here: http://forums.xilinx.com/t5/New-Users-Forum/README-first-Help-for-new-users/td-p/219369

Summary:
1. Read the manual or user guide. Have you read the manual? Can you find the manual?
2. Search the forums (and search the web) for similar topics.
3. Do not post the same question on multiple forums.
4. Do not post a new topic or question on someone else's thread, start a new thread!
5. Students: Copying code is not the same as learning to design.
6 "It does not work" is not a question which can be answered. Provide useful details (with webpage, datasheet links, please).
7. You are not charged extra fees for comments in your code.
8. I am not paid for forum posts. If I write a good post, then I have been good for nothing.
0 Kudos
Contributor
Contributor
9,807 Views
Registered: ‎10-11-2011

Re: Clocking wizard for SP601

Jump to solution

The Probe is an SMA probe and hence I don't think that grounding is the issue. Have attached the pics as you wished!

Board.jpg
0 Kudos
Contributor
Contributor
9,806 Views
Registered: ‎10-11-2011

Re: Clocking wizard for SP601

Jump to solution
 
20MHz.jpg
0 Kudos
Contributor
Contributor
9,805 Views
Registered: ‎10-11-2011

Re: Clocking wizard for SP601

Jump to solution

50 MHz.

50MHz.jpg
0 Kudos
Contributor
Contributor
9,804 Views
Registered: ‎10-11-2011

Re: Clocking wizard for SP601

Jump to solution

100 MHz.

100MHz.jpg
0 Kudos
Contributor
Contributor
9,803 Views
Registered: ‎10-11-2011

Re: Clocking wizard for SP601

Jump to solution

100 MHz. at 10GS/s

 

Now you see the problem!!!

How do I get around this??

100 MHZ~.jpg
0 Kudos
Instructor
Instructor
9,797 Views
Registered: ‎07-21-2009

Re: Clocking wizard for SP601

Jump to solution

Turn on the 50 ohm termination in the scope's Vertical menu > DC Coupling selections.

 

You are probing half of a diff pair.  What is the output drive (IOSTANDARD selections) from the FPGA?

What differences do you see with real-time sampling instead of equivalent time sampling?

 

The crappy waveforms look like the results of a poorly grounded scope probe or a poorly terminated signal (which this is).

 

I assume the coax between the SMA connector and the scope is 50-ohm (and not 75-ohm), yes?

And the cable is roughly 18-24 inches long?

 

-- Bob Elkind

SIGNATURE:
README for newbies is here: http://forums.xilinx.com/t5/New-Users-Forum/README-first-Help-for-new-users/td-p/219369

Summary:
1. Read the manual or user guide. Have you read the manual? Can you find the manual?
2. Search the forums (and search the web) for similar topics.
3. Do not post the same question on multiple forums.
4. Do not post a new topic or question on someone else's thread, start a new thread!
5. Students: Copying code is not the same as learning to design.
6 "It does not work" is not a question which can be answered. Provide useful details (with webpage, datasheet links, please).
7. You are not charged extra fees for comments in your code.
8. I am not paid for forum posts. If I write a good post, then I have been good for nothing.
0 Kudos
Contributor
Contributor
9,787 Views
Registered: ‎10-11-2011

Re: Clocking wizard for SP601

Jump to solution

OK! I turned on the 50 ohm termination in the scope's selection and yes my waveform did improve dramatically. Thanks a lot for that Bob! 

And yes you are right again, I am using a 50 ohm coax (RG- 58) instead of the 75 ohm one. Also it is around 2 m or 78 inches. I know these are quite silly mistakes and I hope you aren't too mad at me for this!!!

Also, I understood what you meant by half a differential pair. But what did you mean by IOSTANDARD selections of the FPGA? Are you talking about the standard I/Os available to the user by SP601? According to the manual, there are 5.

1. LED's

2. DIP switch

3. Push Button switch

4. CPU Reset pushbutton switch

5 GPIO male pin header

 

Of these, I tried taking the output at the LED's but since the space is too tiny for the probe and the ground, my readings were inconsistent!

I think I should get a better reading with a 75 ohm cable, right?

Thank you for the help though! Really appreciate it!

Photo0010.jpg
0 Kudos
Instructor
Instructor
9,785 Views
Registered: ‎07-21-2009

Re: Clocking wizard for SP601

Jump to solution

I hope you aren't too mad at me for this!!!

 

Not even close.

 

But what did you mean by IOSTANDARD selections of the FPGA?

 

For each IO pin you can select from a wide range of signaling standards.  The voltage levels vary, and (for outputs) the output drive current can vary.  For Spartan-6 family, these IO options are listed in DS162, Tables 7, 8, 9, and 10.

 

I think I should get a better reading with a 75 ohm cable, right?

 

It's way more complicated than that.  You need to study up on transmission line principles and circuit board layout concepts.  You are clearly not (yet) a hot-shot circuit board designer.

 

I'm too tired from an all-nighter to expound on transmission line theory right now, but that's the simple category of your problem.  Another label for your type of problem is "signal integrity".  What it comes down to is this:

 

  • under certain conditions, you can just hook up outputs and inputs with simple wires.
  • under other conditions, you must "manage" the interconnect so that signal reflections (yes, the signals bounce off of load pins like billiard balls on a pool table, sometimes right side up and sometimes upside down) don't cause the system to crash and burn.
Make sure you switch the scope input back to 1MegOhm when you are using a normal everyday passive scope probe instead of coax.

 

Thank you for the help though! Really appreciate it!

 

You are very welcome.  Grab a hold of a board designer, and ask him/her what "transmission line" means.  You may need to buy lunch.

 

-- Bob Elkind

SIGNATURE:
README for newbies is here: http://forums.xilinx.com/t5/New-Users-Forum/README-first-Help-for-new-users/td-p/219369

Summary:
1. Read the manual or user guide. Have you read the manual? Can you find the manual?
2. Search the forums (and search the web) for similar topics.
3. Do not post the same question on multiple forums.
4. Do not post a new topic or question on someone else's thread, start a new thread!
5. Students: Copying code is not the same as learning to design.
6 "It does not work" is not a question which can be answered. Provide useful details (with webpage, datasheet links, please).
7. You are not charged extra fees for comments in your code.
8. I am not paid for forum posts. If I write a good post, then I have been good for nothing.
0 Kudos
Instructor
Instructor
7,039 Views
Registered: ‎07-21-2009

cell phone photos

Jump to solution

This is completely off-topic, but...

 

You did OK with the photos from your Samsung model GT-S5620 cell phone.  If you want your photos to be a bit sharper, try bracing or supporting your hand -- or the phone -- with something more solid.

 

The "shutter speed" for the photos was 1/8 second.  That's long enough time for your hand to twitch or move, and that caused blurring in the photos.  In the last photo, you can see the scope trace smear from left to right.

 

So, if you're interested in sharper photos, rest your hand or cell phone on a stack of books or something, to help keep the camera/phone steady.  If you are already resting the phone to steady it, keep practicing your technique.  Perhaps bracing the camera/phone from behind, as well as from below, would help.

 

Thanks for indulging me...  :)

 

-- Bob Elkind

SIGNATURE:
README for newbies is here: http://forums.xilinx.com/t5/New-Users-Forum/README-first-Help-for-new-users/td-p/219369

Summary:
1. Read the manual or user guide. Have you read the manual? Can you find the manual?
2. Search the forums (and search the web) for similar topics.
3. Do not post the same question on multiple forums.
4. Do not post a new topic or question on someone else's thread, start a new thread!
5. Students: Copying code is not the same as learning to design.
6 "It does not work" is not a question which can be answered. Provide useful details (with webpage, datasheet links, please).
7. You are not charged extra fees for comments in your code.
8. I am not paid for forum posts. If I write a good post, then I have been good for nothing.
0 Kudos
Instructor
Instructor
7,028 Views
Registered: ‎07-21-2009

forensic analysis

Jump to solution

I've taken two of the scope traces you posted and marked them up for illustration -- forensic analysis, if I'm allowed to be pretentious.  One trace is supposed to be "clean", and a higher-frequency trace is inexplicably corrupted.

 

The evidence suggests otherwise, and is quite consistent.  Take a look at the two traces:

forums_sig_ringing.png

 

The slow, clean signal is on top.  The 50MHz deranged signal is below.

 

Here's the breakdown...

 

  • Signal hits the "infinite" impedance of the scope input, and a 100% amplitude positive reflection travels back toward the FPGA
  • At the FPGA, the apparent impedance of the output circuit is lower than the 50-70 ohm circuit board transmission line impedance, resulting in a ~25% amplitude negative reflection back to the 'scope
  • The reduced amplitude, inverse reflection hits the scope input roughly 23nS after the initial signal edge
  • If the round-trip transit time is ~23nS, the one-way trip between FPGA and 'scope is ~11.5nS.  At 1.8nS per foot, this corresponds to roughly 6.3 foot long transmission line length.  This is close to the reported 2m coax cable length plus some circuit board trace length.

 

It all fits.

 

The scope traces are marked at the midpoint of a waveform edge and at the midpoint of reflected signal edge -- after the round trip from scope to FPGA and back.

 

The supposedly "clean" waveform is only "clean" because the reverse reflection of the rising signal edge is roughly coincident with the falling edge of the unreflected signal.  When you know what to look for and where to look, the 'hitch' in the falling edge of the trace is clear.  The 'hitch' is the effect of the reflection.  If the period of the lower frequency signal was 20nS longer, the effects of the reflection would have been as clear as it is in the 50MHz waveform.

 

In both waveforms, the reflection is roughly -25% of the original signal amplitude.  If the SP601 board trace impedance is ~50 ohms, a -25% amplitude reflection suggests a ~30 ohm FPGA output driver impedance.  The reflection coefficient is calculated by this equation:

 

 Zterm - Z0

------------- = ρ

 Zterm + Z0

 

Does this make sense to everyone?  I hope this helps illustrate how a "good" set of scope measurements can obscure problems rather than reveal them.

 

Comments, suggestions, and corrections are welcome.

 

-- Bob Elkind

SIGNATURE:
README for newbies is here: http://forums.xilinx.com/t5/New-Users-Forum/README-first-Help-for-new-users/td-p/219369

Summary:
1. Read the manual or user guide. Have you read the manual? Can you find the manual?
2. Search the forums (and search the web) for similar topics.
3. Do not post the same question on multiple forums.
4. Do not post a new topic or question on someone else's thread, start a new thread!
5. Students: Copying code is not the same as learning to design.
6 "It does not work" is not a question which can be answered. Provide useful details (with webpage, datasheet links, please).
7. You are not charged extra fees for comments in your code.
8. I am not paid for forum posts. If I write a good post, then I have been good for nothing.
Tags (1)
Contributor
Contributor
7,005 Views
Registered: ‎10-11-2011

Re: forensic analysis

Jump to solution
Hail BOB!!!!!
Mannn!!! Once you bite in, you sure don't let go, do you? Not that I'm complaining! I cant really thank you enough for the complete explanation of the topic that you have give here. I'm still going through the entire write up now.
N ya, thanks for the lesson on 'Photography of Signals, for Dummies' Should help me immensely next time I do something of this sort!

Regards,
Jack
0 Kudos
Instructor
Instructor
7,002 Views
Registered: ‎07-21-2009

Re: forensic analysis

Jump to solution

Once you bite in, you sure don't let go, do you?

 

Well... yes, to a point... occasionally...  I must confess you made the opportunity, by providing a classic test case and a very nice set of example scope trace photos.  All I needed to add was the text and a little bit of red-pen marking.

 

Remember, this was an explanation of a problem, and nothing more.  You still have an "action item" to learn transmission line concepts, as applied to circuit board traces.  This is essential stuff for hardware engineers.

 

thanks for the lesson on 'Photography of Signals, for Dummies'

 

If you liked that little bit, you should like this massive discourse proportionately more.  :)

 

-- Bob Elkind

SIGNATURE:
README for newbies is here: http://forums.xilinx.com/t5/New-Users-Forum/README-first-Help-for-new-users/td-p/219369

Summary:
1. Read the manual or user guide. Have you read the manual? Can you find the manual?
2. Search the forums (and search the web) for similar topics.
3. Do not post the same question on multiple forums.
4. Do not post a new topic or question on someone else's thread, start a new thread!
5. Students: Copying code is not the same as learning to design.
6 "It does not work" is not a question which can be answered. Provide useful details (with webpage, datasheet links, please).
7. You are not charged extra fees for comments in your code.
8. I am not paid for forum posts. If I write a good post, then I have been good for nothing.
0 Kudos