Page 1 of 1

Image rejection problem that I can't figure out

PostPosted: Sun Nov 20, 2016 9:54 pm
by kb1mvx
I've had a Peaberry V2 kit on my shelf for two years and I am finally building it. I have it assembled through the the receiver section. The receiver works and tunes in HDSDR. I've tested it on 20, 30, 40, and 60 using a signal generator. The problem is the the image is as big as the the fundamental and the receiver sensitivity seems low. I checked all the solders, tried direct injection on C27, and rewound T2. No change.

Using the recording function in HDSDR I recorded some data with a tone 1 KHz away from the LO frequency. I read the data into Matlab for analysis. The amplitudes of the both I and Q are low but they are very close. The phase difference between I and Q is 50 degrees. It should be 90 degrees. Any idea where I am getting a 40 degree phase shift?

Jim - KB1MVX

Re: Image rejection problem that I can't figure out

PostPosted: Mon Dec 05, 2016 3:00 am
by kb1mvx
I did more testing. I collected IQ data with the following LO settings:

5.060 MHz
6.060 MHz
7.060 MHz
8.060 MHz
9.060 MHz
10.060 MHz
11.060 MHz
12.060 MHz
13.060 MHz
14.060 MHz
15.060 MHz

In each case in injected a tone 1 kHz above the LO. The IQ phase difference ranged from 51 to 54 degrees across all of these frequencies so the phase shift is not happening at RF. It's happening at baseband. So, I don't have the answer yet but I'm getting there.


Re: Image rejection problem that I can't figure out

PostPosted: Wed Dec 07, 2016 3:25 pm
by kb1mvx
I'm going to switch gears and try to address the low sensitivity issue. In order to get ADC counts in the 400 range I have to push the RF drive level up to -10 dBm. That's obviously not right. I'm going to borrow an o-scope and work through the receive path.

Re: Image rejection problem that I can't figure out

PostPosted: Tue Nov 07, 2017 2:51 am
by kb1mvx
I figured it out....a year didn't take a year to figure it out it just took a year to get back to it. The problem ended up being the op amp gain stage U5. I did a lot of troubleshooting with a signal generator and an o-scope but I finally stumbled onto the problem with a volt meter. U5 takes +5V on pin 8 and pin 4 is grounded. When I looked between the board ground and pin 4 I got +3V. That's not right so I removed U5 and checked for voltages on the pads for pin 4 and 8. Pin 4 was 0V and pin 8 was +5V as expected. I installed a new LT6231CS8 for U5 and all is well. I tested the rx sensitivity on all 4 bands (20m,30m,40m, and 60m) and it was within a dB of what I expected.