Possibility of high frequency interrupt handle with MCF5223x

classic Classic list List threaded Threaded
11 messages Options
Reply | Threaded
Open this post in threaded view
|

Possibility of high frequency interrupt handle with MCF5223x

David Pang
Dear all,

I want to use MCF5223x to read a 8bit digital signal and forward the sample data to UDP packet.
The signal consists of 8bit data and 1 clock, data fetch at rising edge or falling edge. However, the clock frequency is pretty high, max frequency is about 12MHz.
Are there possible to use the CLOCK input to trigger an interrupt and sample the data in each interrupt?

I want to do the connection as below:
CLOCK             ===>  nIRQ1
DATA[8:0]         ===>  PORT AN

So the system would include 3 tasks:
1. Interrupt for data input, interrupt trigger at about 12MHz
2. UDP Packet forwarding, each packet contains 200 samples
3. HTTP or TCP service for system configuration

Can a 60MHz ColdFireV2, 5223x, capable to do it?


B.R.
David
Reply | Threaded
Open this post in threaded view
|

Re: Possibility of high frequency interrupt handle with MCF5223x

Yevgeni Tunik
I use more powerful 5270 (Fsys=147.456 MHz). It cannot withstand interrupt trigger at about 12MHz.

I guess, only specially developed periphery (FPGA?) can help to implement the described task.

Regards,
Yevgeni Tunik
+972-52-448-9271 (mobile phone)
___________________________________

--- On Mon, 12/6/10, David Pang <[hidden email]> wrote:

From: David Pang <[hidden email]>
Subject: [ColdFire] Possibility of high frequency interrupt handle with MCF5223x
To: "Yevgeni Tunik" <[hidden email]>
Date: Monday, December 6, 2010, 2:16 PM

Dear all,

I want to use MCF5223x to read a 8bit digital signal and forward the sample data to UDP packet.
The signal consists of 8bit data and 1 clock, data fetch at rising edge or falling edge. However, the clock frequency is pretty high, max frequency is about 12MHz.
Are there possible to use the CLOCK input to trigger an interrupt and sample the data in each interrupt?

I want to do the connection as below:
CLOCK             ===>  nIRQ1
DATA[8:0]         ===>  PORT AN

So the system would include 3 tasks:
1. Interrupt for data input, interrupt trigger at about 12MHz
2. UDP Packet forwarding, each packet contains 200 samples
3. HTTP or TCP service for system configuration

Can a 60MHz ColdFireV2, 5223x, capable to do it?


B.R.
David

Reply | Threaded
Open this post in threaded view
|

AW: Possibility of high frequency interrupt handle with MCF5223x

Mark Butcher
In reply to this post by David Pang

Hi David

 

12MHz seems rather fast for an IRQ.

I would look into the use of DMA to copy the data input to a buffer and then using its ‘end of transfer’ interrupt to send the data on. For example, you may be able to use the trigger input to start a DMA timer sequence that then reads the data from the port at the correct rate (there are various other permutations in the chip which may also be able to solve it).

 

Note however that your data rate is 96Mb/s and it will not be possible to send this over a 100M LAN (there is overhead and collisions etc.). Probably the V2 will be able to achieve about 15Mb/s UDP so this is probably the main bottleneck.

 

 

Regards

 

Mark

 

 

**************************************************

Mark Butcher Bsc (Hons)                 MPhil MIET

 

M.J.Butcher Consulting              

Birchstrasse 20f,                  CH-5406 Rütihof

Switzerland          www.mjbc.ch / www.uTasker.com

++41 (0)79 402 26 37            Skype: M_J_Butcher

**************************************************

 

 

 

Von: [hidden email] [mailto:[hidden email]] Im Auftrag von David Pang
Gesendet: Montag, 6. Dezember 2010 13:16
An: [hidden email]
Betreff: [ColdFire] Possibility of high frequency interrupt handle with MCF5223x

 

Dear all,

I want to use MCF5223x to read a 8bit digital signal and forward the sample data to UDP packet.
The signal consists of 8bit data and 1 clock, data fetch at rising edge or falling edge. However, the clock frequency is pretty high, max frequency is about 12MHz.
Are there possible to use the CLOCK input to trigger an interrupt and sample the data in each interrupt?

I want to do the connection as below:
CLOCK             ===>  nIRQ1
DATA[8:0]         ===>  PORT AN

So the system would include 3 tasks:
1. Interrupt for data input, interrupt trigger at about 12MHz
2. UDP Packet forwarding, each packet contains 200 samples
3. HTTP or TCP service for system configuration

Can a 60MHz ColdFireV2, 5223x, capable to do it?


B.R.
David

Reply | Threaded
Open this post in threaded view
|

Re: Possibility of high frequency interrupt handle with MCF5223x

David Pang

Thanks Yevgeni and Mark for your prompt reply.

Seams I have to consider other processor of higher frequency, and other means to handle data input.
DMA is a good idea, but I don't see CF MCU has a digital input peripheral could do DMA transfer. An FPGA/CPLD conversion with parallel EMI bus to MCU is a good idea.
In fact 12MHz is the clock frequency, there is interval between packets. the max data rate would be 8MB/s approximately.

As Mark mentioned, CFV2 MCU UDP can only achieve 15Mb/s, that is the real bottleneck. I have to consider to discard redundant data, but it is still difficult to get through the bottleneck.


B.R.
David


On Mon, Dec 6, 2010 at 9:05 PM, Mark Butcher <[hidden email]> wrote:

Hi David

 

12MHz seems rather fast for an IRQ.

I would look into the use of DMA to copy the data input to a buffer and then using its ‘end of transfer’ interrupt to send the data on. For example, you may be able to use the trigger input to start a DMA timer sequence that then reads the data from the port at the correct rate (there are various other permutations in the chip which may also be able to solve it).

 

Note however that your data rate is 96Mb/s and it will not be possible to send this over a 100M LAN (there is overhead and collisions etc.). Probably the V2 will be able to achieve about 15Mb/s UDP so this is probably the main bottleneck.

 

 

Regards

 

Mark

 

 

**************************************************

Mark Butcher Bsc (Hons)                 MPhil MIET

 

M.J.Butcher Consulting              

Birchstrasse 20f,                  CH-5406 Rütihof

Switzerland          www.mjbc.ch / www.uTasker.com

++41 (0)79 402 26 37            Skype: M_J_Butcher

**************************************************

 

 

 

Von: [hidden email] [mailto:[hidden email]] Im Auftrag von David Pang
Gesendet: Montag, 6. Dezember 2010 13:16
An: [hidden email]
Betreff: [ColdFire] Possibility of high frequency interrupt handle with MCF5223x

 

Dear all,

I want to use MCF5223x to read a 8bit digital signal and forward the sample data to UDP packet.
The signal consists of 8bit data and 1 clock, data fetch at rising edge or falling edge. However, the clock frequency is pretty high, max frequency is about 12MHz.
Are there possible to use the CLOCK input to trigger an interrupt and sample the data in each interrupt?

I want to do the connection as below:
CLOCK             ===>  nIRQ1
DATA[8:0]         ===>  PORT AN

So the system would include 3 tasks:
1. Interrupt for data input, interrupt trigger at about 12MHz
2. UDP Packet forwarding, each packet contains 200 samples
3. HTTP or TCP service for system configuration

Can a 60MHz ColdFireV2, 5223x, capable to do it?


B.R.
David


Reply | Threaded
Open this post in threaded view
|

AW: Possibility of high frequency interrupt handle with MCF5223x

Mark Butcher

Hi David

 

I have used the DMA timers to read GPIOs and ADCs – the source for reads can be set to any address (including FLASH, RAM or peripherals).

 

Regards

 

Mark

 

 

Von: [hidden email] [mailto:[hidden email]] Im Auftrag von David Pang
Gesendet: Montag, 6. Dezember 2010 14:49
An: [hidden email]
Betreff: Re: [ColdFire] Possibility of high frequency interrupt handle with MCF5223x

 


Thanks Yevgeni and Mark for your prompt reply.

Seams I have to consider other processor of higher frequency, and other means to handle data input.
DMA is a good idea, but I don't see CF MCU has a digital input peripheral could do DMA transfer. An FPGA/CPLD conversion with parallel EMI bus to MCU is a good idea.
In fact 12MHz is the clock frequency, there is interval between packets. the max data rate would be 8MB/s approximately.

As Mark mentioned, CFV2 MCU UDP can only achieve 15Mb/s, that is the real bottleneck. I have to consider to discard redundant data, but it is still difficult to get through the bottleneck.


B.R.
David

On Mon, Dec 6, 2010 at 9:05 PM, Mark Butcher <[hidden email]> wrote:

Hi David

 

12MHz seems rather fast for an IRQ.

I would look into the use of DMA to copy the data input to a buffer and then using its ‘end of transfer’ interrupt to send the data on. For example, you may be able to use the trigger input to start a DMA timer sequence that then reads the data from the port at the correct rate (there are various other permutations in the chip which may also be able to solve it).

 

Note however that your data rate is 96Mb/s and it will not be possible to send this over a 100M LAN (there is overhead and collisions etc.). Probably the V2 will be able to achieve about 15Mb/s UDP so this is probably the main bottleneck.

 

 

Regards

 

Mark

 

 

**************************************************

Mark Butcher Bsc (Hons)                 MPhil MIET

 

M.J.Butcher Consulting              

Birchstrasse 20f,                  CH-5406 Rütihof

Switzerland          www.mjbc.ch / www.uTasker.com

++41 (0)79 402 26 37            Skype: M_J_Butcher

**************************************************

 

 

 

Von: [hidden email] [mailto:[hidden email]] Im Auftrag von David Pang
Gesendet: Montag, 6. Dezember 2010 13:16
An: [hidden email]
Betreff: [ColdFire] Possibility of high frequency interrupt handle with MCF5223x

 

Dear all,

I want to use MCF5223x to read a 8bit digital signal and forward the sample data to UDP packet.
The signal consists of 8bit data and 1 clock, data fetch at rising edge or falling edge. However, the clock frequency is pretty high, max frequency is about 12MHz.
Are there possible to use the CLOCK input to trigger an interrupt and sample the data in each interrupt?

I want to do the connection as below:
CLOCK             ===>  nIRQ1
DATA[8:0]         ===>  PORT AN

So the system would include 3 tasks:
1. Interrupt for data input, interrupt trigger at about 12MHz
2. UDP Packet forwarding, each packet contains 200 samples
3. HTTP or TCP service for system configuration

Can a 60MHz ColdFireV2, 5223x, capable to do it?


B.R.
David

 

Reply | Threaded
Open this post in threaded view
|

Spam:**********, RE: Possibility of high frequency interrupt handle with MCF5223x

Baker, Eric
In reply to this post by David Pang
X-SpamDetect-Info: ------------- Start ASpam results ---------------
X-SpamDetect-Info: This message may be spam. This message BODY has been altered to show you the spam information
X-SpamDetect: **********: 10.6 sd.6  [394]100%13.3(!129,300) [191]99%12.5(!7,60) [46]12%-5.9(X-Verify-Helo:+OK) [27]42%-0.1(X-LangGuess:English)
X-SpamDetect-Info: ------------- End ASpam results -----------------

Can you compress the data before you send it out UDP?

 

From: [hidden email] [mailto:[hidden email]] On Behalf Of David Pang
Sent: Monday, December 06, 2010 8:49 AM
To: Baker, Eric
Subject: Re: [ColdFire] Possibility of high frequency interrupt handle with MCF5223x

 


Thanks Yevgeni and Mark for your prompt reply.

Seams I have to consider other processor of higher frequency, and other means to handle data input.
DMA is a good idea, but I don't see CF MCU has a digital input peripheral could do DMA transfer. An FPGA/CPLD conversion with parallel EMI bus to MCU is a good idea.
In fact 12MHz is the clock frequency, there is interval between packets. the max data rate would be 8MB/s approximately.

As Mark mentioned, CFV2 MCU UDP can only achieve 15Mb/s, that is the real bottleneck. I have to consider to discard redundant data, but it is still difficult to get through the bottleneck.


B.R.
David

On Mon, Dec 6, 2010 at 9:05 PM, Mark Butcher <[hidden email]> wrote:

Hi David

 

12MHz seems rather fast for an IRQ.

I would look into the use of DMA to copy the data input to a buffer and then using its ‘end of transfer’ interrupt to send the data on. For example, you may be able to use the trigger input to start a DMA timer sequence that then reads the data from the port at the correct rate (there are various other permutations in the chip which may also be able to solve it).

 

Note however that your data rate is 96Mb/s and it will not be possible to send this over a 100M LAN (there is overhead and collisions etc.). Probably the V2 will be able to achieve about 15Mb/s UDP so this is probably the main bottleneck.

 

 

Regards

 

Mark

 

 

**************************************************

Mark Butcher Bsc (Hons)                 MPhil MIET

 

M.J.Butcher Consulting              

Birchstrasse 20f,                  CH-5406 Rütihof

Switzerland          www.mjbc.ch / www.uTasker.com

++41 (0)79 402 26 37            Skype: M_J_Butcher

**************************************************

 

 

 

Von: [hidden email] [mailto:[hidden email]] Im Auftrag von David Pang
Gesendet: Montag, 6. Dezember 2010 13:16
An: [hidden email]
Betreff: [ColdFire] Possibility of high frequency interrupt handle with MCF5223x

 

Dear all,

I want to use MCF5223x to read a 8bit digital signal and forward the sample data to UDP packet.
The signal consists of 8bit data and 1 clock, data fetch at rising edge or falling edge. However, the clock frequency is pretty high, max frequency is about 12MHz.
Are there possible to use the CLOCK input to trigger an interrupt and sample the data in each interrupt?

I want to do the connection as below:
CLOCK             ===>  nIRQ1
DATA[8:0]         ===>  PORT AN

So the system would include 3 tasks:
1. Interrupt for data input, interrupt trigger at about 12MHz
2. UDP Packet forwarding, each packet contains 200 samples
3. HTTP or TCP service for system configuration

Can a 60MHz ColdFireV2, 5223x, capable to do it?


B.R.
David

 

_________________________

LECO Corporation Notice: This communication may contain confidential information intended for the named recipient(s) only. If you received this by mistake, please destroy it and notify us of the error. Thank you.
Reply | Threaded
Open this post in threaded view
|

Re: Possibility of high frequency interrupt handle with MCF5223x

Yevgeni Tunik
In reply to this post by David Pang
X-SpamDetect-Info: ------------- Start ASpam results ---------------
X-SpamDetect-Info: This message may be spam. This message BODY has been altered to show you the spam information
X-SpamDetect: ******: 6.6 sd=6.6  [280]100%13.3(60,!225) [43]01%-13.0(X-SpamContent:clean) [96]02%-12.5(Accept Orbs) [226]98%12.3(!46,66) [47]98%12.0(X-Verify-Helo:failed) [45]93%8.9(X-Verify-Helo:-ERR) [29]22%-2.5(DomainKey-Signature:) [27]42%-0.1(X-LangGuess:English)
X-SpamDetect-Info: ------------- End ASpam results -----------------

David,

see the following forum thread for actually reached UDP performance:
http://forum.embeddedethernet.com/viewtopic.php?f=5&t=415&view=next

Regards,
Yevgeni Tunik
+972-52-448-9271 (mobile phone)
___________________________________

--- On Mon, 12/6/10, Yevgeni Tunik <[hidden email]> wrote:

From: Yevgeni Tunik <[hidden email]>
Subject: Re: [ColdFire] Possibility of high frequency interrupt handle with MCF5223x
To: [hidden email]
Date: Monday, December 6, 2010, 2:28 PM

I use more powerful 5270 (Fsys=147.456 MHz). It cannot withstand interrupt trigger at about 12MHz.

I guess, only specially developed periphery (FPGA?) can help to implement the described task.

Regards,
Yevgeni Tunik
+972-52-448-9271 (mobile phone)
___________________________________

--- On Mon, 12/6/10, David Pang <[hidden email]> wrote:

From: David Pang <[hidden email]>
Subject: [ColdFire] Possibility of high frequency interrupt handle with MCF5223x
To: "Yevgeni Tunik" <[hidden email]>
Date: Monday, December 6, 2010, 2:16 PM

Dear all,

I want to use MCF5223x to read a 8bit digital signal and forward the sample data to UDP packet.
The signal consists of 8bit data and 1 clock, data fetch at rising edge or falling edge. However, the clock frequency is pretty high, max frequency is about 12MHz.
Are there possible to use the CLOCK input to trigger an interrupt and sample the data in each interrupt?

I want to do the connection as below:
CLOCK             ===>  nIRQ1
DATA[8:0]         ===>  PORT AN

So the system would include 3 tasks:
1. Interrupt for data input, interrupt trigger at about 12MHz
2. UDP Packet forwarding, each packet contains 200 samples
3. HTTP or TCP service for system configuration

Can a 60MHz ColdFireV2, 5223x, capable to do it?


B.R.
David


Reply | Threaded
Open this post in threaded view
|

MCF52235CAL60s on Ebay

Sam Saprunoff-2
In reply to this post by Yevgeni Tunik
Good day All,

I do not know if this is allowed on this site and if not, then my
apologies and delete this post.  However, I thought that it might be of
interest to other Coldfire users, as we have 120 of MCF52235CAL60s that
are still in factory pouches.  We bought these for a project that never
went forward and so we are selling them on Ebay.  Just do a search on
the part number and you will find them if you are interested.

It is a great processor, however, for our upcoming designs we need more
horsepower, memory, etc and so we are going to be using an iMX and so
these parts will not be used.

Cheers,

Sam
---
[hidden email]              Send a post to the list.
[hidden email]        Join the list.
[hidden email]    Join the list in digest mode.
[hidden email]     Leave the list.

Reply | Threaded
Open this post in threaded view
|

AW: MCF52235CAL60s on Ebay

Mark Butcher
Sam

Are these with the silicon fix to the PHY bug or are they previous to that?
I don't actually know how to recognise this...
Put a plug on the uTasker forum if you like ;-)

Will the Kinetis be an option or are these also too small (up to 256k SRAM /
1Meg FLASH / 150MHz Cortex M4 - but availability may be an issue at the
moment) ?

Regards

Mark



-----Ursprüngliche Nachricht-----
Von: [hidden email] [mailto:[hidden email]] Im
Auftrag von Sam Saprunoff
Gesendet: Montag, 24. Januar 2011 19:27
An: [hidden email]
Betreff: [ColdFire] MCF52235CAL60s on Ebay

Good day All,

I do not know if this is allowed on this site and if not, then my apologies
and delete this post.  However, I thought that it might be of interest to
other Coldfire users, as we have 120 of MCF52235CAL60s that are still in
factory pouches.  We bought these for a project that never went forward and
so we are selling them on Ebay.  Just do a search on the part number and you
will find them if you are interested.

It is a great processor, however, for our upcoming designs we need more
horsepower, memory, etc and so we are going to be using an iMX and so these
parts will not be used.

Cheers,

Sam
---
[hidden email]              Send a post to the list.
[hidden email]        Join the list.
[hidden email]    Join the list in digest mode.
[hidden email]     Leave the list.

---
[hidden email]              Send a post to the list.
[hidden email]        Join the list.
[hidden email]    Join the list in digest mode.
[hidden email]     Leave the list.

Reply | Threaded
Open this post in threaded view
|

Re: AW: MCF52235CAL60s on Ebay

Sam Saprunoff-2
Good day Mark,

I am not sure, as I have not taken them out of the factory sealed
pouches to check the date codes.  I will check my documentation for date
codes to see if they have a silicon fix on them.  If they are the old
silicon that has issues, then they we will sell them cheaper and/or
donate them to the Local Electrical Engineering students for projects.

The new Kinetis would be a consideration, however, we have a small
production build in the next month and so availability is a big
consideration.  For this one we are using the 52255 and for another
upcoming build, a 52259 because of the extra Flash, extra RAM, USB, etc.

For another project that requires a lot more resources, we will be using
the iMX283.  These are really inexpensive and are a very powerful
processor.  We were to use a iMX357 (have parts for this too), but the
iMX283 has a better fit overall.  Sadly, production quantities of the
iMX283 will not be available until June or so...

Thanks for your offer to post on uTasker!

Cheers,

Sam




-------- Original Message  --------
Subject: AW: [ColdFire] MCF52235CAL60s on Ebay
From: Mark Butcher <[hidden email]>
To: Sam Saprunoff <[hidden email]>
Date: Monday, January 24, 2011 11:55:13 AM

> Sam
>
> Are these with the silicon fix to the PHY bug or are they previous to that?
> I don't actually know how to recognise this...
> Put a plug on the uTasker forum if you like ;-)
>
> Will the Kinetis be an option or are these also too small (up to 256k SRAM /
> 1Meg FLASH / 150MHz Cortex M4 - but availability may be an issue at the
> moment) ?
>
> Regards
>
> Mark
>
>
>
> -----Ursprüngliche Nachricht-----
> Von: [hidden email] [mailto:[hidden email]] Im
> Auftrag von Sam Saprunoff
> Gesendet: Montag, 24. Januar 2011 19:27
> An: [hidden email]
> Betreff: [ColdFire] MCF52235CAL60s on Ebay
>
> Good day All,
>
> I do not know if this is allowed on this site and if not, then my apologies
> and delete this post.  However, I thought that it might be of interest to
> other Coldfire users, as we have 120 of MCF52235CAL60s that are still in
> factory pouches.  We bought these for a project that never went forward and
> so we are selling them on Ebay.  Just do a search on the part number and you
> will find them if you are interested.
>
> It is a great processor, however, for our upcoming designs we need more
> horsepower, memory, etc and so we are going to be using an iMX and so these
> parts will not be used.
>
> Cheers,
>
> Sam
> ---
> [hidden email]              Send a post to the list.
> [hidden email]        Join the list.
> [hidden email]    Join the list in digest mode.
> [hidden email]     Leave the list.
>
> ---
> [hidden email]              Send a post to the list.
> [hidden email]        Join the list.
> [hidden email]    Join the list in digest mode.
> [hidden email]     Leave the list.
>
>

---
[hidden email]              Send a post to the list.
[hidden email]        Join the list.
[hidden email]    Join the list in digest mode.
[hidden email]     Leave the list.

Reply | Threaded
Open this post in threaded view
|

Re: AW: MCF52235CAL60s on Ebay

Sam Saprunoff-2
Good day again Mark,

After reviewing the errata and the date codes on the part pouches, I
believe the parts are of mask set 2M23E, as this was the first mask and
there is only one PHY errata that is not addressed in the latest mask
(3M23E).  Interestingly, other FEC and PHY errrata have still not been
addressed even with the latest mask.

That being said, the issue is annoying, but can be addressed in one's
firmware(stack).

Thanks for the heads up!

Cheers,

Sam


SECF025: EPHY Not Accepting Valid LTP Greater Than 100 ns

Errata type: Silicon
Affected component: EPHY

If a link partner's LTP is greater than 100 ns, using auto-negotiation,
the EPHY does not recognize it. This results in a failure to auto-negotiate.

Description:
Disable auto-negotiation and configure the EPHY or link partner to
operate in 100TX or 10BaseT mode.

Workaround:
Software implementation is available in the ColdFire_Lite stack, which
is available at http://www.freescale.com/coldfire.

Fix plan: This erratum is fixed in devices labeled with the A suffix or
3M23E mask.


---
[hidden email]              Send a post to the list.
[hidden email]        Join the list.
[hidden email]    Join the list in digest mode.
[hidden email]     Leave the list.