CCS/MSP430FR5969: MODOSC vs DCO for the ADC

Ferran Reverter

Part Number: MSP430FR5969

Tool/software: Code Composer Studio

Dear colleagues,

I have a doubt with regard to the use of the DCO and MODOSC for the ADC.

According to the MSP430FR5969 datasheet, the DCO seems more accurate and less temperature dependent (see Table 5-6 vs 5-8). In table 5-8, the current consumption of the MODOSC is specified (25 uA) but not that of DCO in Table 5-6. On the other hand, the current consumption of the ADC (table 5-22) is specified considering that the MODOSC is employed, nothing is said if the DCO is used.

I’ve been testing the ADC with DCO and MODOSC (both operating at 1 MHz) and I’ve seen that the current consumption with the DCO is significantly lower than that with the MODOSC, to be precise 128 vs 213 uA. I was wondering why the current consumption of the ADC is specified for the MODOSC if it turns out that the DCO offers more advantages in terms of accuracy, thermal stability and even current consumption.

Best regards and thanks for your time,

Ferran

over 7 years ago

0 Tomasz Kocon over 7 years ago

Guru 26200 points

Hello Ferran,

could you, please explain what you did to have MODOSC running @ 1MHz?

0 Xiaodong LI over 7 years ago in reply to Tomasz Kocon

TI__Mastermind 23105 points

Hello Ferran
MODOSC has the quick turn on and off advantage comparing with DCO. Hence, the system power consumption should be lower using MODOSC on ADC because of the shorted ADC "wake-up ready" time.

0 Ferran Reverter over 7 years ago in reply to Tomasz Kocon

Intellectual 645 points

Hi Tomasz,
Actually the frequency of MODOSC is 1.2 MHz, the nominal one divided by 4.
Best!
Ferran

0 Ferran Reverter over 7 years ago in reply to Xiaodong LI

Intellectual 645 points

Hi Xiaodong,
First, many thanks for the reply.
I expected what you say, but experimentally I've seen that the current consumption of the ADC is lower with the DCO than that with the MODOSC (128 vs 213 uA). The measurements of current are performed with a digital multimeter and having the ADC always converting on LPM1 and without interrupting the CPU.
Do you believe that these results make sense or am I doing something wrong?
Best and thanks for your time,
Ferran

0 Dietmar Walther over 7 years ago in reply to Ferran Reverter

TI__Genius 14165 points

Hi Ferran,

if you really set the DCO to 1 MHz it operates at 1 MHz. The MODOSC always operates at 4.8 Mhz. This is were the difference comes from.
You can see that the AM current for 1 MHz 100% out of Cach cache is 110 uA. This is close to the 128 uA you see when sourcing the ADC with 1 MHz DCO instead of the MODOSC.

For better 1:1 comparison take the DCO to 5.33 MHz and compare it to the 4.8 MHz of the MODOSC then you should be much better and faster with MODOSC. It strongly depends what sample and hold requirements you have. If time is no problem and only current is important operate the ADC with the 32 kHz which is anyway running in LPM3.

Best regards,
Dietmar

0 Ferran Reverter over 7 years ago in reply to Dietmar Walther

Intellectual 645 points

Hi Dietmar,
Many thanks for the explanation. Clear!!
Ferran

0 Tomasz Kocon over 7 years ago in reply to Ferran Reverter

Guru 26200 points

MODOSC always operates @ about 5 MHz.
My hope was that you would see you were comparing 5 to 1 MHz.

Worth to mention:
To conserve power, MODOSC is powered down when not needed and is enabled only when required. When the MODOSC source is equired, the respective module requests it. MODOSC is enabled based on unconditional and conditional requests. Setting MODOSCREQEN enables conditional requests. Unconditional requests are always enabled. It is not necessary to set MODOSCREQEN for modules that utilize unconditional requests; for example, ADC or fail-safe.

0 Ferran Reverter over 7 years ago in reply to Tomasz Kocon

Intellectual 645 points

The MODOSC was divided by four, so the ADC was operating at 1.2 MHz, whereas with the DCO, the ADC operated at 1 MHz. For this reason, I thought that the comparison was fair.

0 Tomasz Kocon over 7 years ago in reply to Ferran Reverter

Guru 26200 points

I was referring to your words:
I’ve been testing the ADC with DCO and MODOSC (both operating at 1 MHz)
Fixed. O.K.

**Attention** This is a public forum

MSP low-power microcontrollers

MSP low-power microcontroller forum

CCS/MSP430FR5969: MODOSC vs DCO for the ADC