N-counter issue ADF4001

Question

I designed and tested a PCB with a PLL circuit. The goal is to lock on the reference signal which is a square wave. The problem I am having is that the signal of the N-counter shows no response, just 0V dc. The N-counter signal is measured via the MUXOUT of the ADF4001. I can correctly measure the R-counter with the MUXOUT.

The VCO I am using is the CVCO33CL-0110-0150. I have measured the VCO signal with a scope. It is about 1V_PP. According to the spec sheet the VCO delivers -3 dBm to +3 dBm. The ADF4001 specifies an RF input sensitivity of -10 dBm to 0 dBm. I got this working before with a higher power VCO with 5 V supply (CVCO55CL-0060-0110, output power 5 dBm to 12 dBm). To be honest, I am not quite sure what these specs mean and if it is ok or not. Also I've been looking into whether I should do more effort in matching the impedance of the line (it is only a short line on the PCB of 5 mm). But I figured that for a 100 MHz signal this should not be too much of a problem.

The register settings:

cp = 7
muxout = 1
r0 = (cp << 18) + (cp << 15) + (1 << 7) + (muxout << 4) + 2
self.devA.spiMaster_SingleWrite(struct.pack(">I", r0)[1:], True)
# set R
ldp = 1
R  = 1
r0 = (ldp << 20) + (R << 2)
self.devA.spiMaster_SingleWrite(struct.pack(">I", r0)[1:], True)
# set N
N = 18
r0 = (N << 8) + 1
self.devA.spiMaster_SingleWrite(struct.pack(">I", r0)[1:], True)

The loop filter is bandwidth 10 kHz.

Link to data sheets:

• ADF4001
• CVCO33CL

Answer 1

You're probably OK with the impedance matching between the VCO and the RF input port. When you say the MUXOUT with the N divider is death, what do you mean? Are you not seeing a signal there at all? with Vtune limited to 3.3V, you're not going to be able to get above 150 MHz or so, the tuning voltage on the VCO is 0 < VTune < 5Vdc. My initial thought is that you're not setting the registers correctly, so I'd double check that you're setting that correctly.

Answer 2

Why no DC blocking cap between the VCO and the ADF4001? It is worth checking that the CVC033 DC output impedance is not upsetting the input bias of the ADF4001. This would certainly produce these symptoms - works with high signal levels, fails with lower ones.