TRF3750 Issues

Kathryn:

Are you using the TRF3750 in a closed loop synthesizer application?  If so, are you able to get the synthesizer to lock at the target frequency?  The CPout will set the voltage to the VCO to maintain proper frequency.  If you probe this output in a closed loop system then the voltage should not be railed (either 0V or max VCP) in a locked condition.  The description seems to indicate that the device is not locked and is railed low.

I am not sure about the burst condition on the muxout and the odd measurements you are seeing.  What is the 5 MHz input signal?  Perhaps I can give you further guidance if I understand the desired frequency and your configuration.

--RJH

Hi RJH,

Thanks for your response.  At the moment I am just trying to test our TRF3750 circuit.  I originally built a VCO that I was running around 400kHz but the TRF3750 wasn't locking on a frequency.  But then I read in the forum that it won't work below 4MHz with a power supply of 3.3V.  So then I got a frequency generator and input 5MHz to it to see if the charge pump would do anything or if the N divider would output an expected signal, but it still wouldn't work.  The frequency generator is set to make a 5MHz square wave which I voltage divide down to around 1Vp-p and connect to RFIn through a capacitor.  The RFIn- is through a capacitor to GND.  CPOut is going to the loop filter circuit copied from figure 26 of the TRF3750 datasheet for testing.  Rset is 4.7kohm.  I have the same decoupling as in figure 26 on every power pin.  Refin goes to a 20MHz crystal through a 1nF cap.  Vcp the charge pump power pin is connected to 7 to 8V and the other power pins are connected to 3.3V powered off a USB port that has a 3.3V regulator on it with 100mA available.  Clk, Data and LE are connected to a teensy 3.1 board and seem to be working as when I change the R divider register the output spikes change to the R frequency set, although they are not continuous but come out in bursts which seemed a bit strange.  I am pretty sure the Function latch is set correctly too as I tried the power down bits and they made a difference and when I change Muxout settings that seems to work as well or at least the output changes.  I haven't been able to prove the N divider is working but it uses the same code to send the data as the other registers.  I tried setting Muxout to 6 serial data output but this did not seem to resemble the input data and I couldn't find any information on what that setting actually does.  We have Muxout connected to a LED with 330 ohms so I might try disconnecting the LED to see if it makes any difference.

regards,

Kathryn.

Hi Again,
I was able to get the N divider working. I had to have an RFIn input amplitude of at least 130mV p-p and then the N divider started working. Taking the LED off made no difference. I can now see the charge pump doing things as the 5MHz signal gets in and out of phase with the crystal reference. And if I set the frequency to 4.9MHz the CPOut signal increases to 6.9V so it is all working now. The problem was my RFIn signal being too small and too low in frequency.
regards,
Kathryn.

Excellent. Good debugging skills on your part. --RJH