LMK04832: LMK04832

Hello,

Can the LMK04832 be configured to output a clock as soon as it is powered-up without the need to program it using its SPI interface?

Kind Regards,

Rajan

Hello Rajan,

No the LMK04832 does not have an EEPROM or ROM to enable this.

Some of our commercial parts such as the LMK05318 have this feature.

I will get back to you regarding the schematic.

Thanks,

Vibhu

Hello Rajan,

Following up on the schematic and power supply  capacitors. To understand the decoupling caps and ferrite beads needed a good source is section "10 Power Supply Recommendations" and "Figure 40. Example Application – Power System Schematic" of the LMK04806 datasheet. This explains some concepts of why certain capacitors and ferrite beads are necessary. Keep in mind that the VccXs are jumbled between the parts and this is not an apples to apples comparison. For example: on the LMK04806 Vcc4 is the digital supply while on the LMK04832 Vcc5 is the digital supply.

I think Figure 40 on this datasheet should be a good reference when looking at simultaneously with the LMK04832EVM I pointed to before. Here are some of the key points I'd like to point out on the LMK04832:

1. Power planes for the LMK outputs Vcc12_CG0, Vcc11_CG3, Vcc4_CG2, Vcc2_CG1 and the Vcc3_SYSREF should have ferrite beads to reduce crosstalk between outputs with unrelated frequencies. This is in addition to a filtering cap that you have on your schematics. The EVM uses a 1 uF bypass capacitor at each of these pins.
2. No ferrite beads are needed to be placed between the supplies Vcc1_VCO, Vcc5_DIG, Vcc6_PLL1 and Vcc10_PLL2. These pins produce little noise, and a ferrite bead may cause phase noise disturbances. The EVM uses a 0.01 uF bypass capacitor on Vcc1_VCO, Vcc5_DIG, Vcc6_PLL1 and a 0.1 uF capacitor on PLL2
3. Using a Vcc9_CP2 requires  a ferrite bead and a 0.1 uF capacitor is recommended to be added here
4. Finally for the Vcc7_OSCout and Vcc8_OSCin, please use a ferrite bead and a 0.1 uF capacitor as the EVM suggests. This supplies functionality is different between the two parts.
5. You may also want to consider a set of filtering capacitors at the output of the LDO if you predict there will be noise from other devices but it looks like you already use C343 and C18 in your schematic for this purpose.
6. Since you do not use any of the clocks in Clock Group 3, Vcc11_CG3 does not need the ferrite bead I mentioned earlier.

I do not see any issues with the LMX device, I looked through the unused pins briefly, they seem to be terminated correctly, if you haven't already maybe compare them with section "Table 11. Recommended Treatment of Pins" of the datasheet to double check that you have the functionality you require. The caps to Vcc and GND seem to have followed the EVM.

For future reference please do not share confidential information on e2e as this is a public forum, please contact the field engineer who supports you for sensitive inquiries.

Thanks,

Vibhu

what value of inductor (ferrite bead) are you recommending?

Hello Rajan,

This is what we use on our EVMs:

MMZ1005Y121CT000

Ferrite Bead, 120 ohm @ 100 MHz, 0.4 A, 0402

Ferrite beads with excessive impedance at high frequency (> 200-ohm) can degrade crosstalk performance, especially on the output and SYSREF supplies.

Thanks,

Vibhu