8051 suggestion

Hi Folks

I am looking at developing a new application, that I hope I can do using an
8051 derivative.

Two main requirements that I have are that it be pretty damn fast
(50-100MHz) and have I2C. Also around 6-ish external interrupts and port
for RS232...
Have been looking also at using a RTOS....but this is not fixed in stone.

I am quite familiar with the 8051 and the keil uvision IDE

Looked at DS80C400, but its a lot of overkill on other features, so far I
dont need TCP/IP, CAN and so on.

Any suggestions o learned people? ;)

	   
					
---------------------------------------		
Posted through http://www.EmbeddedRelated.com

>
> Two main requirements that I have are that it be pretty damn fast
> (50-100MHz) and have I2C. Also around 6-ish external interrupts and port
> for RS232...
> Have been looking also at using a RTOS....but this is not fixed in stone.

Look at www.silabs.com  -> C8051F120 runs at 100MHz, one cycle per clock.
At the moment, fastest on the market (except ASIC variants).

Also there are some other variants that also run at 100MHz, but with less peripherals...

On 01/03/11 19:26, wolf99 wrote:
> Hi Folks
>
> I am looking at developing a new application, that I hope I can do using an
> 8051 derivative.
>
> Two main requirements that I have are that it be pretty damn fast
> (50-100MHz) and have I2C. Also around 6-ish external interrupts and port
> for RS232...
> Have been looking also at using a RTOS....but this is not fixed in stone.
>
> I am quite familiar with the 8051 and the keil uvision IDE
>
> Looked at DS80C400, but its a lot of overkill on other features, so far I
> dont need TCP/IP, CAN and so on.
>
> Any suggestions o learned people? ;)
>

Well, the obvious learned answer is to drop the idea of using an 8051. 
Join the 21st century and use a proper microcontroller - ARM's are the 
new 8051's.

If you are somehow tied to having to use such an old and limited 
processor architecture, then I'm sure others will give you suggestions - 
there are a few fast 8051's.  But it's like using a turbo-charged kids 
tricycle rather than a car - except that in this case, the car is 
smaller and cheaper.

On 03/01/2011 12:03 PM, David Brown wrote:
> On 01/03/11 19:26, wolf99 wrote:
>> Hi Folks
>>
>> I am looking at developing a new application, that I hope I can do
>> using an
>> 8051 derivative.
>>
>> Two main requirements that I have are that it be pretty damn fast
>> (50-100MHz) and have I2C. Also around 6-ish external interrupts and port
>> for RS232...
>> Have been looking also at using a RTOS....but this is not fixed in stone.
>>
>> I am quite familiar with the 8051 and the keil uvision IDE
>>
>> Looked at DS80C400, but its a lot of overkill on other features, so far I
>> dont need TCP/IP, CAN and so on.
>>
>> Any suggestions o learned people? ;)
>>
>
> Well, the obvious learned answer is to drop the idea of using an 8051.
> Join the 21st century and use a proper microcontroller - ARM's are the
> new 8051's.
>
> If you are somehow tied to having to use such an old and limited
> processor architecture, then I'm sure others will give you suggestions -
> there are a few fast 8051's. But it's like using a turbo-charged kids
> tricycle rather than a car - except that in this case, the car is
> smaller and cheaper.
>
Even though I'd rather suck a green lemon, I feel like I need to defend 
the 8051.  I really don't think it's dead just yet.  Old and creaky, 
maybe, but not to be rejected out of hand.  Unfortunately.

But if you're talking about a 100MHz 8051, then you may be able to do 
the same thing with a 25MHz ARM, and they _are_ becoming available in 
some pretty attractively priced processors.

Have you looked?

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" was written for you.
See details at http://www.wescottdesign.com/actfes/actfes.html

Tim Wescott wrote:

>>
> Even though I'd rather suck a green lemon, I feel like I need to defend 
> the 8051.  I really don't think it's dead just yet.  Old and creaky, 
> maybe, but not to be rejected out of hand.  Unfortunately.
> 
> But if you're talking about a 100MHz 8051, then you may be able to do 
> the same thing with a 25MHz ARM, and they _are_ becoming available in 
> some pretty attractively priced processors.
> 
> Have you looked?
> 

While I would agree, the si labs device's speed does compensate for the
inefficient code on such an architecture and they are very cheap in
quantity, not to mention the low cost of development kits, less than
$100 for most. A far better set of devices than for example, pic, which
still seems to have a very loyal following. You can be out of the box
and running a simple program with the si labs kits in less than an hour.

Arm are obviously far more capable and prices are good as well, but dev 
tools
are more of a problem to pull together, unless you want to pay K$...

Regards,

Chris

On 03/01/2011 01:37 PM, ChrisQ wrote:
> Tim Wescott wrote:
>
>>>
>> Even though I'd rather suck a green lemon, I feel like I need to
>> defend the 8051. I really don't think it's dead just yet. Old and
>> creaky, maybe, but not to be rejected out of hand. Unfortunately.
>>
>> But if you're talking about a 100MHz 8051, then you may be able to do
>> the same thing with a 25MHz ARM, and they _are_ becoming available in
>> some pretty attractively priced processors.
>>
>> Have you looked?
>>
>
> While I would agree, the si labs device's speed does compensate for the
> inefficient code on such an architecture and they are very cheap in
> quantity, not to mention the low cost of development kits, less than
> $100 for most. A far better set of devices than for example, pic, which
> still seems to have a very loyal following. You can be out of the box
> and running a simple program with the si labs kits in less than an hour.
>
> Arm are obviously far more capable and prices are good as well, but dev
> tools
> are more of a problem to pull together, unless you want to pay K$...

I looked down my nose at PICs for years, until I had a project where I 
needed to put on my hardware engineer's hat and choose a micro to go 
onto a board.

The PIC won, for their current drive capabilities, their peripherals, 
and the fact that Microchip has a good reputation for delivering, while 
you can't always get Atmel chips.  The fact that the customer was 
insisting on PICs may have had something to do with it, too :-).

So the hardware guy (me) and the project manager (me) told the software 
guy (me) to just suck it up and write good code for a PIC.

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" was written for you.
See details at http://www.wescottdesign.com/actfes/actfes.html

On Mar 1, 1:51=A0pm, Tim Wescott <t...@seemywebsite.com> wrote:
> On 03/01/2011 01:37 PM, ChrisQ wrote:
>
>
>
> > Tim Wescott wrote:
>
> >> Even though I'd rather suck a green lemon, I feel like I need to
> >> defend the 8051. I really don't think it's dead just yet. Old and
> >> creaky, maybe, but not to be rejected out of hand. Unfortunately.
>
> >> But if you're talking about a 100MHz 8051, then you may be able to do
> >> the same thing with a 25MHz ARM, and they _are_ becoming available in
> >> some pretty attractively priced processors.
>
> >> Have you looked?
>
> > While I would agree, the si labs device's speed does compensate for the
> > inefficient code on such an architecture and they are very cheap in
> > quantity, not to mention the low cost of development kits, less than
> > $100 for most. A far better set of devices than for example, pic, which
> > still seems to have a very loyal following. You can be out of the box
> > and running a simple program with the si labs kits in less than an hour=
.
>
> > Arm are obviously far more capable and prices are good as well, but dev
> > tools
> > are more of a problem to pull together, unless you want to pay K$...
>
> I looked down my nose at PICs for years, until I had a project where I
> needed to put on my hardware engineer's hat and choose a micro to go
> onto a board.
>
> The PIC won, for their current drive capabilities, their peripherals,
> and the fact that Microchip has a good reputation for delivering, while
> you can't always get Atmel chips. =A0The fact that the customer was
> insisting on PICs may have had something to do with it, too :-).
>
> So the hardware guy (me) and the project manager (me) told the software
> guy (me) to just suck it up and write good code for a PIC.
>

Same here.  Customer insists on a PIC.  But a PIC24 is nothing like
PIC18 or PIC16.  And with 96K SRAM, it is almost pretty good.  I would
have gone to ARM, but I never argue with the check (or the one signing
it).

Tim Wescott wrote:

> I looked down my nose at PICs for years, until I had a project where I 
> needed to put on my hardware engineer's hat and choose a micro to go 
> onto a board.
> 
> The PIC won, for their current drive capabilities, their peripherals, 
> and the fact that Microchip has a good reputation for delivering, while 
> you can't always get Atmel chips.  The fact that the customer was 
> insisting on PICs may have had something to do with it, too :-).
> 
> So the hardware guy (me) and the project manager (me) told the software 
> guy (me) to just suck it up and write good code for a PIC.
> 

Si Labs != Atmel, right ?. To be honest, it's years since I last looked 
at pic, but at the time, the absence of a proper stack really put me off 
in terms of compiled language suitability and the fact that they were 
some sort of knockoff from a late 70's Gen Inst tv remote contol device 
said it all.

Later versions may be better, but why bother ?. You invest quite a bit 
of time in a new cpu and develop libraries over time, buy toolchains 
etc, so it makes sense to standardise if there no overwhelming 
advantage. The si labs devices are a natural development of the old 8051 
in compatability terms and even have a variant with 24 bit adc + 8 line 
mux, something no other device in it's class can offer afaik...

Regards,

Chris

> Si Labs != Atmel, right ?.

Not until they merge.

>  To be honest, it's years since I last looked
> at pic, but at the time, the absence of a proper stack
> really put me off

That's PIC16 and PIC18.

> in terms of compiled language suitability

PIC24 has GCC port, so don't care how ugly it is internally.
Actually, it's not too bad.

> and the fact that they were
> some sort of knockoff from a late 70's Gen Inst tv remote contol device
> said it all.
>
> Later versions may be better, but why bother ?. You invest quite a bit
> of time in a new cpu and develop libraries over time, buy toolchains
> etc, so it makes sense to standardise if there no overwhelming
> advantage.

Sometimes, they are external non-technical factors.

On Mar 2, 7:26=A0am, "wolf99" <toby.mole@n_o_s_p_a_m.gmail.com> wrote:
> Hi Folks
>
> I am looking at developing a new application, that I hope I can do using =
an
> 8051 derivative.
>
> Two main requirements that I have are that it be pretty damn fast
> (50-100MHz) and have I2C. Also around 6-ish external interrupts and port
> for RS232...
> Have been looking also at using a RTOS....but this is not fixed in stone.
>
> I am quite familiar with the 8051 and the keil uvision IDE
>
> Looked at DS80C400, but its a lot of overkill on other features, so far I
> dont need TCP/IP, CAN and so on.

You did not mention Pin Count and likely Code Size ?
SiLabs would be a strong candidate,
or the new Atmel variants with a ALU/DSP type support

http://www.atmel.com/dyn/resources/prod_documents/doc4088.pdf?doc_id=3D1021=
9&family_id=3D604

The Ramtron variants are also reasonably quick, and they also have an
ALU.

Cypress have a PSoC3 in sampling, which is 8051 + CPLD

Or, if you want a middle-ground part, with 5V operation, an ALE pin,
and similar peripherals, but a Cortec M0 core, look at the Nuvoton
M0516.
In stock at Digikey, and an Eval board is sub $20,

-jg