overclocking, well sorta!

Go To Last Post
8 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I need some sage advise from the freaks!
We make a sensor based on the M48PA that runs on 5v VCC and that uses a 16MHz xtal. A new client has asked us to produce a 3.3v version for their application. Looking at the datasheet, the fastest recommended speed for the 48 at 3.3v is about 12MHz. Although, I've been doing some testing, so far only at room temperature, and all seems to be working fine, even with a VCC as low as 2.5v. The M48 is still cool and current draw is about half what it is at 5.0v.
All that I've had to do is lower the BOD level to allow the sensor to work. If it were you, would you just run it at 3.3v, or add a dc to dc converter to boost the supply up to 5.0v and add an open collector output with a pull up to 3.3v as the client only needs a go/no-go digital output from our sensor. The client may not be interested in the increased cost of the voltage converter.
TIA
JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If its going to run hot, probably shouldnt go faster than 12Mz. Chnage the xtal and the vreg and the dash number.

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

My question is this... is the application so critical that it has to run at 16.0MHz and not at 12.0MHz?

If this is simply a G0-No-Go application, will a couple microseconds longer really make or break the usefulness of the system?

You can avoid reality, for a while.  But you can't avoid the consequences of reality! - C.W. Livingston

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks Bob, the specs have a max temp of 40C or 104F and lowering the xtal frequency will reduce the resolution of the timing measurement w/ T1.... oh, I just noticed the temp sensor (lm34) is low as well, seems it needs 5v to work. I guess that answers that! Thanks for the advice, looks like a boost converter/regulator is the way to go.
JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

lowering the xtal frequency will reduce the resolution of the timing measurement w/ T1....

Indeed, I have a production app (a high-speed high-res ultrasonic sensor) where I'd need to work through a 20% main clock speed reduction. What type of app is this? If this is a critical parameter, why not run at 20MHz?

A glance at the graph would indicate 13.3MHz at 3.3V. Now, Brutte will ask what happens when the e.g. battery supply voltage dips to 3.0V...

Perhaps there is an equivalent temperature sensor that is 3V-friendly?

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Now, Brutte will ask what happens when the e.g. battery supply voltage dips to 3.0V...

m48pa does not have BOD triggered at 3.0V so I am asking about Vcc = 2.5V actually :twisted:

No RSTDISBL, no fun!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What are the overall spec's for the project?

Is it just the output level that needs to change, or is the V+ supply also going to change?

The LM34 isn't known for its accuracy. Especially when one looks at the graphs in the data sheet, instead of the marketing B.S. in the data sheets header/bullet items.

Would a redesign with an NTC and an Xmega E series be cheaper or "better"?

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks to all who replied! Yes its a high res ultra-sonic app. We are looking at Xmega for the next gen of the product but need to fill the gap until its developed. I found a 3.3v in to 5v out boost reg that may do the trick if there is space for a coil on the board. That would let us use the existing design at 16MHz or even 20MHz while running from the supplied 3.3v rail.
JC