PLL does not work due to incorrect R8

Development Boards
1

Hello, SparkFun people.

Last week I bought a SparkFun RED-V Things plus to port NuttX.

I finished initial porting of NuttX and it works with an external OSC (e.g. 16MHz) and now I’m trying to enable PLL but it does not work. So I checked SparkFun_Red-V_Things_Plus-Schematic.pdf carefully and soon found that R8 is 100kohm which is NOT recommended by SiFive. Actually, voltage was significantly dropped from 1.79V to ** 0.68V ** after enabling PLL. I think the R8 should be ** 100ohm ** (not 100kohm) because HiFive1 board (including HiFive1. Rev.B) uses 100ohm for PLL_AVDD.

Actually I replaced the 100kohm R8 with 100ohm, I confirmed that PLL finally worked.

Thanks,

Masayuki Ishikawa

2

Hi Masayuki,

Thank you for bringing this to our attention! I have informed the engineer who designed the Red-V boards and we will get that fixed ASAP.

3

masayuki2009:
Hello, SparkFun people.

Last week I bought a SparkFun RED-V Things plus to port NuttX.

I finished initial porting of NuttX and it works with an external OSC (e.g. 16MHz) and now I’m trying to enable PLL but it does not work. So I checked SparkFun_Red-V_Things_Plus-Schematic.pdf carefully and soon found that R8 is 100kohm which is NOT recommended by SiFive. Actually, voltage was significantly dropped from 1.79V to ** 0.68V ** after enabling PLL. I think the R8 should be ** 100ohm ** (not 100kohm) because HiFive1 board (including HiFive1. Rev.B) uses 100ohm for PLL_AVDD.

Actually I replaced the 100kohm R8 with 100ohm, I confirmed that PLL finally worked.

Thanks,

Masayuki Ishikawa

Good work. I am curious - the SiFive datasheet specifies an FMAX of 320MHz for the -G002 part; what frequencies have you used the FE310 at? Can you confirm it operates to 320MHz?

Thank you,

Dana

4

masayuki2009:
Hello, SparkFun people.

Last week I bought a SparkFun RED-V Things plus to port NuttX.

I finished initial porting of NuttX and it works with an external OSC (e.g. 16MHz) and now I’m trying to enable PLL but it does not work. So I checked SparkFun_Red-V_Things_Plus-Schematic.pdf carefully and soon found that R8 is 100kohm which is NOT recommended by SiFive. Actually, voltage was significantly dropped from 1.79V to ** 0.68V ** after enabling PLL. I think the R8 should be ** 100ohm ** (not 100kohm) because HiFive1 board (including HiFive1. Rev.B) uses 100ohm for PLL_AVDD.

Actually I replaced the 100kohm R8 with 100ohm, I confirmed that PLL finally worked.

Thanks,

Masayuki Ishikawa

This is old news now, SparkFun is correcting R8, but yes, according to the FE310-G002 datasheet, section 6.5:

The PLL requires dedicated 1.8 V power supply pads with a supply filter on the circuit board.

The supply filter should be a 100 Ω resistor in series with the board 1.8 V supply decoupled with

a 100 nF capacitor across the VDDPLL/VSSPLL supply pins. The VSSPLL pin should not be

connected to board VSS.

I wasn’t surprised to brick my Red-V when I enabled the PLL with:

#include <metal/machine.h>

	long cpu_clock;

	cpu_clock = metal_clock_get_rate_hz(&__metal_dt_clock_4.clock);
    new_cpu_clock = metal_clock_set_rate_hz(&__metal_dt_clock_4.clock, 32000000L);

But I un-bricked it by re-programming while shorting R8 with a meter (100uA, BTW) to not enable the PLL

Anyone with a bricked Red-V should see it start working when R8 is replaced with 100 ohms

Cheers,

Dana

5

FWIW: I got a board about a week before Christmas 2019, and it arrived already reworked for the proper 100 Ohm value of R8. I guess it is all under control now.

The schematic on the SparkFun web documentation still shows R8 as 100K though.

6

Hey everyone,

Robin is correct here. All new RED-V’s (both RedBoard and Thing Plus) have this issue fixed and any users who received RED-V’s with the incorrect resistor values should have been contacted by us regarding the issue. We have updated the PDF schematics as well as the ZIP downloads of the Eagle files on both of the RED-V product pages.