9
\$\begingroup\$

We have a legacy 5V system that makes heavy use of analog sensors and various other 5V digital IOs. We're looking at moving to an ARM MCU to bring this design in line with our newer systems which are all Cortex-M3 based. I would prefer to stay with a 5V MCU so I don't lose accuracy on the ADC inputs and so I won't have to run an extra 3.3V power supply. I've been searching for 5V compatible ARM MCUs and only found the FM3 series from Fujitsu, which it seems like almost no one stocks. My questions are:

  • Is it worth it to continue attempting to use a 5V micro or should we move to voltage-dividers for the ADC and level shifters/transistors for the digital IO?

  • Does anyone have experience with the Fujitsu FM3 line?

  • Are there any other 5V compatible ARM microcontrollers out there?

\$\endgroup\$

closed as off-topic by PeterJ, Daniel Grillo, user17592, placeholder, Chetan Bhargava Oct 21 '14 at 19:15

This question appears to be off-topic. The users who voted to close gave this specific reason:

If this question can be reworded to fit the rules in the help center, please edit the question.

  • \$\begingroup\$ Update: The TMPM380 from Toshiba looks like it may work. I'm a bit concerned about its ability to write to flash at 85C though. \$\endgroup\$ – QuestionMan Sep 13 '12 at 15:15
11
\$\begingroup\$

A dual 3.3V/5V power supply + level shifters will cost you more than they're worth. Resistor dividers will get you 3.3 V levels cheap, but you'll need the level shifters to go to a logic 5 V. 74HCTxx buffers will do this cheap, but they're extra space on your PCB.

For the ADC it would mean a 3.6 dB higher noise level. If this is not acceptable you can probably decrease the value by better decoupling, which still will be a cheaper solution.

I would go for a 3.3 V only system.

\$\endgroup\$
  • 1
    \$\begingroup\$ Unfortunately, I still have to provide a 5V supply for powering the sensor board, and it won't work with 3.3V. \$\endgroup\$ – QuestionMan Sep 12 '12 at 16:57
  • \$\begingroup\$ @QuestionMan - what does the interface to the sensor look like? I2C? SPI? Or is that the analog output for the ADC? Is the 3.6 dB a problem? \$\endgroup\$ – stevenvh Sep 12 '12 at 17:02
  • \$\begingroup\$ The sensor interface is a 0 to 5V analog output, unfortunately. I don't think 3.6dB will be too much of a problem, or at least I hope not. Otherwise I may be stuck with an external ADC chip. \$\endgroup\$ – QuestionMan Sep 12 '12 at 17:26
4
\$\begingroup\$

NXP's LPC line (and probably many others) has 5V-tolerant GPIOs, however output is only 3.3v. A common workaround is to pull up the pin output to a 5V supply and set it to open drain mode (tristate) when you need to "output" 5V:

https://github.com/ytai/ioio/wiki/Digital-IO

However, this won't solve your ADC issue.

\$\endgroup\$
  • \$\begingroup\$ Yeah, it's really the ADC issue that's bugging me, there are many tools to achieve the digital translation. \$\endgroup\$ – QuestionMan Sep 12 '12 at 17:27
  • 1
    \$\begingroup\$ "LPC" is very wide, it includes ARM7TDMI and even 8051 compatibles. So you'll have to be more exact. The NXP microcontroller selection guide only mentions 5 V tolerance for the Cortex-M0 LPC1100 series, IIRC. \$\endgroup\$ – stevenvh Sep 12 '12 at 17:29
  • \$\begingroup\$ The page you link to mentions open-drain outputs, but since most microcontrollers don't have those that seems to be of little relevance. \$\endgroup\$ – stevenvh Sep 12 '12 at 17:32
  • 1
    \$\begingroup\$ I don't know about "most" microcontrollers, but the LPC17xx (and I think other Cortex-Ms from NXP) certainly have open drain. Maybe not all GPIOs on the chip, but most of them. \$\endgroup\$ – Igor Skochinsky Sep 12 '12 at 17:40
  • 2
    \$\begingroup\$ Every GPIO ever made is capable of behavior equivalent to an open-drain output. That's inherent in the ability to act as an input. Typically you write your data to the "direction register" instead of the "data register". \$\endgroup\$ – Ben Voigt Sep 12 '12 at 17:59
3
\$\begingroup\$

Level shifters are the easiest and simplest way to solve your problem. You could use a 5 V ADC and convert its output to 3.3 V. I'm assuming you are mainly using the MCU and it is low power, so a 5 V to 3.3 V LDO regulator could work fine.

I would avoid a part that isn't stocked; that may because it isn't popular, or you need to buy a whole bunch. Either way I would worry about support.

\$\endgroup\$
1
\$\begingroup\$

The Nuvotron NuMicro NUC100 series can operate from 2V5 to 5V5, but they are Cortex M0. This is ARMv6-M and not v7-M, but the tools and libs often support this architecture too.

\$\endgroup\$
  • \$\begingroup\$ Very good find, I could probably get away with using an M0. Looks like they are a spin-off of Winbond, anyone know about their product lifetimes? They claim ten years, but it would be nice to get longer. \$\endgroup\$ – QuestionMan Sep 13 '12 at 12:42
1
\$\begingroup\$

A lot of Texas Instrument MCU are 5v tolerant, see their Cortex m3 series.

[EDIT]

As said in the question this page links to the Cortex M3 family MCU from Texas Instruments. According to their datasheets (section 20 Electrical Characteristics, Maximum Ratings), these MCU accept 5.5v max input.

This is also true for the Stellaris Launchpad family (ie TI Cortex M4) i'm using.

\$\endgroup\$
0
\$\begingroup\$

A lot of the AVR UC3 chips are able to run on a 5V supply. I'm assuming they are 5V I/O-tolerant by virtue of their supply voltage.

\$\endgroup\$
  • 1
    \$\begingroup\$ But OP says he wants a Cortex-M3. Which the UC3 isn't... \$\endgroup\$ – stevenvh Sep 12 '12 at 17:23
  • 1
    \$\begingroup\$ Well, the question doesn't say he requires a Cortex-M3. Any of the families using the ARM architecture would seem to be acceptable. (That still excludes AVR) \$\endgroup\$ – Ben Voigt Sep 12 '12 at 18:02
  • \$\begingroup\$ Ah, you're right, sorry. I saw ARM and made an incorrect mental link to 32-bit generically. My bad! \$\endgroup\$ – Toby Lawrence Sep 12 '12 at 21:17
0
\$\begingroup\$

Cypress PSoC 5 LP (Cortex M3) can operate from 0.5 V to 5.5 V. Moreover it has 4 Vdio power pins for 4 I/O groups. Each group of I/O can operate with different voltages. This makes it possible to connect circuits with voltages such 3.3 V, 1.8 V, 5 V at the same time without level shifters.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.