This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

MSP430FR5969 won't operate @ 1,8V supply voltage?

Other Parts Discussed in Thread: MSP430FR5969, MSP430FR5959, MSP430FR5949, MSP430L092, MSP430FR5739

Hello!

As the title said I want to operate the Launchpad MSP430FR5969 @ 1,8V supply voltage. However it first starts to work @ 1,9V. That's very unfavourable when you want to use a linear regulatar to power the board. In the forum is already another thread http://e2e.ti.com/support/microcontrollers/msp430/f/166/t/282823.aspx with the same topic but without a real solution for the problem. Because the thread was started in august 2013  i hope there is a solution now. As I understand from the datasheet I have to change the SVS levels. But after reading the users guide it doesn't seem to be possible.

Thanks for your help!

Simon

  • Simon,

    Take a look at the markings of the chip you are testing. Is it really a MSP430FR5969? If not, read the fine prints of TERMS and CONDITIONS that came with the board.

    I got an old Launchpad and the chip is X430FR5969.

    -- OCY

  • You're right! it is written "X430FR5969" on the chip. So that means it's one of pre released samples and one issue of this production run is that the chips won't work at 1,8V supply voltage? Does anybody know when the MSP430FR5969 will be available again?

  • It is quite common (and understandable) that for demo boards and samples, existing stocks of experimental silicon are sent out first, until depleted.
    If you really need the latest silicon version, then you need to buy the part, directly. If it is still an XMS430 or older version (shouldn’t happen, but some distributors want to clear their stock first), then you can complain and demand replacement.

    I don't know off any limitations of the XMS silicon, but errata of these are usually not release din the device errata sheet.

  • Hello everyone!

    Today I received my new MSP430FR5969-Launchpad from TI and have still the same issue that the board doesn’t start up with a supply voltage of 1,8V. When I set my power supply to 1,9V and turn it then down to 1,8V the microcontroller seems to work properly. In the datasheet it is written that the minimum supply voltage depends on the SVS levels. As I understand the Users Guide the SVS Levels can’t be changed neither disabled in LPM0. So how can the MSP430FR5969 operate at a supply voltage of 1,8V like it is specified in the datasheet?

    Thanks in advance!

    Simon

  • Right. I have MSP430FR5949 and MSP430FR5959, and both don't start after power up at 1.8 V with P1.0 blink example. They start at 1.9 V just fine.

  • Hi Simon,

    The footnote (4) on the Vcc spec in www.ti.com/lit/gpn/msp430fr5969 points you to checking the SVS levels in Table 5-2. While the device may be designed to be capable of running at minimum 1.8V, the part will not start up unless the SVS releases the device from reset. You can see in Table 5-2 that the SVS levels are listed:

     VSVSH+ is the level where the SVS will release at power-up and let the device start running. You can see it can be anywhere between 1.77-1.99V worst case, with typical 1.88V. This explains why on your device you are having to get near to 1.9V to get the part to start up.

    After the device has started, VSVSH- will tell you the level where the SVS will trigger a reset if you dip back down below it. You can see that this value can be 1.75-1.85V on this device, with a typical value of 1.8V. In addition you could have the SVS disabled while in some LPMs to save power (but this also carries some risks).

    Basically, you will need to get your voltage high enough at startup to get the SVS to release, and then can dip your voltage a little lower after the part is running and the SVS won't kick in - think of battery powered applications where you start at one voltage and it may decrease a bit over time. In order to cover worst case on both ends of the spectrum, in your power design you'll want to account for the Max SVSH+ at power-up and Max SVSH- at power-down.

    The reason the SVS is designed in this way was to make sure the SVS level keeps you from having the part operating below 1.8V which is the point where the operation of the device is guaranteed - the device isn't designed to work below 1.8V. So the SVS level was set in this way to make it unlikely for the SVS to release at too low of a level to ensure you were above 1.8V at startup, or to reset the part at too low of a level for a falling voltage - the typical values have been targeted to keep you in a good and allowed operating range above 1.8V. This is to help prevent the part running out of spec for some time which could mean erroneous operation.

    Regards,

    Katie

  • Problem is that due to the low power requirements, the threshold voltages for the SVS comparators aren’t exact. So the SVS won’t release the device at exactly 1.8V. It should allow the MSP to run at 1.8V, so the threshold is targeted a bit below with a hysteresis, assuming that the voltage will not settle between the threshold and 1.8V (user circuit design issue). To bridge the gap between reaching the trigger level and reaching the assumed sufficient voltage, there is a delay in which the MSP is held in reset after the trigger level has been reached.

    However, if the voltage is only slowly rising, then the delay may expire before the supply has settled. Which may prevent the MSP from starting (it crashes right when trying it). Especially if the rising current demand causes the voltage to slightly drop again, causing a crash.
    To extend the delay if the supply requires it, an R/C combination can be added to the ST pin, which holds the MSP in reset while VCC is still rising. Now this conflicts with the use of SBW, which limits the capacitor value and therefore the timing constant.

    Well, finding the optimum circuitry for the existing supply is an engineer’s job. But since this is E2E here, we all are engineers, aren’t we? :)

  • So Katie,

    I completely understand why the MSP430FR5969 would be designed this way with the SVS voltage set-points given in the data-sheet if the chip cannot reliably operate at a voltage at all below 1.8 V.  What I can't understand, however, is why the chip is advertised to work with a 1.8 V supply voltage.  If the max values of the two SVS voltage set-points are 1.85 V and 1.99 V respectively, then it seems to me that if I wanted to design an application that works reliably I would need to either have a supply voltage that is constant at a minimum of 2.0 V or a supply voltage that starts out higher and then lowers to a minimum of 1.85 V.  Even though some chips could work some of the time with a supply voltage of 1.8 V, it seems that I could in no case expect a reliable system incorporating this microcontroller and a 1.8 V supply.  Is this correct, or am I missing something?

    I guess I just feel like the data-sheet is a little misleading by saying that the supply voltage can be anywhere from 1.8 V to 3.6 V.  I know it has the footnote that clarifies after a bit of digging, but in that case it seems to me like the datasheet is saying "Works at a supply voltage of 1.8 V (Footnote: does not actually work with a supply voltage of 1.8 V)."  I hope I am not sounding overly rude in this post.  I just got done assembling a prototype system I designed and was sad to see that the microcontroller will not power on with the 1.83 V LDO on the board.  Unfortunately I cannot just bump up the voltage because the microcontroller is interacting on many lines with other chips with tight input tolerances centered on 1.8 V.  I don't have the real-estate to add level shifting to all of the connecting traces.

    So now I am wondering if I will be forced to move to another microcontroller.  This would be very sad for me because I absolutely love the MSP430FR5969.  Are there any FRAM microcontrollers from TI that can reliably operate with a supply voltage of 1.8 V?  Or do you have any other suggestions on what I could do?

    Thanks for taking the time to consider my situation.

    With respect,

    - Steven

  • Steven Noyce said:
     What I can't understand, however, is why the chip is advertised to work with a 1.8 V supply voltage.

    Why not? Two different specs: startup voltage and operation voltage. TI "conveniently forgot" about SVS tresholds :) Startup at higher voltage is common for battery-operated devices w/o regulator.

    I am interested in the answer to your question as well - which msp430 not only operate at 1.8V but also can start reliably at 1.8V (besides msp430L092)? Maybe MSP430F6-series?

    [edit] It is kind of funny that chips w/o SVS started fine (according to specs), but had brownout problems :) Well, they still do.

  • Yeah, I can see that startup voltage and operation voltage could be two different specs, but with a max V_SVSH- of 1.85 V I am not even sure it can be claimed that the chip can work with a 1.8 V supply voltage even in a battery powered application where the voltage started out higher and decreased slowly. Maybe some chips could reach the 1.8 V mark some of the time, but it doesn't sound reliable to me. :)

    On a slightly different note as I am thinking about this a bit more: is the V_SVSH+ value required only when the controller is first powered on and after a SVS triggered reset, or is this voltage value also required after other forms of resets such as watch-dog timer resets or RST pin triggered resets?
  • Ilmars said:
    which msp430 not only operate at 1.8V but also can start reliably at 1.8V

    The L092 runs from 0.9 to 1.65V. But the absolute maximum is 1.9V. So 1.8V operation is possible without permanent damage, but with perhaps reduced reliability. Well 0.65V operaiton would tolerate 1.8V signals on the input (there's a 0.2V margin), while 1.65V output shoudl be accepted by 1.8V devices.
    The L092 isn't an easy device anyway, as it has no internal flash and requires an external memory which proably won't run that low. (most serial flash chips require higher operating voltages)

    The 5438A, fo rexample (the first datasheet I looked at) has a brownout level of  1.5V max. A tlowest core voltage (1.4V), it can run from 1.8V supply reliably. For an MCLK speed of up to 8MHz.
    The SVS, however, has either a 1.57V -1.78V or 1.79V-1.98V range. But it isn't active by default. If you use an external precision SVS, you can reliably go down to 1.8V. Else, you need to ensure that the time your VSS rises from 1.5V to 1.8V is shorter than the brownout reset delay. And that your VSS is  falling below 1.4V if it ever starts falling below 1.8V. Bad for battery.powered devices with a slowly failing supply.
    And of course it is not an FRAM device.

    Well, you can't eat the cake and keep it. (very low power operation - 1.5µA in performance mode and 200nA in low power mode -.and high-precision SVS)

  • Ilmars said:
    I am interested in the answer to your question as well - which msp430 not only operate at 1.8V but also can start reliably at 1.8V (besides msp430L092)? Maybe MSP430F6-series?

    From my experience, "problem" is related only to new FR devices (for example not to MSP430FR5739). They can not start at 1.8V ( SBW / JTAG too).

  • Hi Steven,

    Sorry for the delay - I was out sick all of last week and am just getting caught up now.

    Thanks for your feedback about the SVS and I understand your frustration. The datasheet is correct because you can disable the SVS after startup and run at 1.8V, which you understand, but I also see your point - I will provide this feedback to marketing. Is the footnote on the front page of the datasheet not quite enough to make this clear as the user still has to go to the later section to actually see the SVS levels?  

    Steven Noyce said:
    On a slightly different note as I am thinking about this a bit more: is the V_SVSH+ value required only when the controller is first powered on and after a SVS triggered reset, or is this voltage value also required after other forms of resets such as watch-dog timer resets or RST pin triggered resets?

    This is a little bit of a tricky question. If you check the user's guide SVS registers, you should be able to tell which bits are reset at different levels of reset (PUC, POR, BOR). In the preface of the user's guide, under register bit conventions, it lists a convention for distinguishing which bits will reset for each reset type, using different kinds of brackets (e.g. [] for BOR reset). You can see that SVSHE is reset by a BOR reset - so you'd have to look up what type of reset each source (e.g. watchdog) causes to determine if the SVS would still be enabled if your code had cleared the enable bit. I'm unsure, however, at the reset if only VSVSH- or VSVSH+ are checked at the reset - I believe that VSVSH+ is only mattering at actual device power-up, having nothing to do with resets, but I will have to confirm this.

    Regards,

    Katie

  • V_SVSH+ is only related to the VCC voltage level and is not affected by any (digital) reset. The SVS is initialized by a BOR, so other monitoring settings are kept across any reset (including SVS reset!) except a BOR.
    Espeically the PUC is a pure CPU reset, not touching anything else by itself. However, the boot code that is executed after a CPU reset will initialize some things manually. See Katie's explanation about how to determine what is reset when.

**Attention** This is a public forum