Good news, bad news. My wideband loop initial build sort of works, but the transimpedance amplifier has stability problems.
Over the past few months, I have been tacking the most complex project I have ever undertaken. Namely, designing a pair of wideband magnetic loops for diversity reception up to 30 MHz. Take a look at the project overview. My design of the mechanical side was pretty easy using aluminum core PEX for the loops. On the electronic side, I designed a head end amplifier and dual channel base amplifier using operational amplifiers.
The whole project was modeled in LTSpice and the simulation looked pretty good. So, I designed the printed circuit boards in KiCAD and got them fabricated at OSH Park. Finally, I ordered the parts from Mouser. Learning to hot air solder the SMD components was challenging.
In the next few articles, I will describe how well the wideband loop initial build performed. If you want to follow along, you can download the following package which contains the schematics as well as the LTSpice simulation models.
The download contains LTSpice models for AC and Transient analysis showing how things “should work” in a perfect world. But as I found out, the world is not perfect. In the case of the Head End amplifier, I discovered instability likely caused by an imperfect PCB design and insufficient bypassing. More on that later.
Right now, let’s look at my Siglent 1052 DSO.
Wideband Loop Initial Build – First Look at Performance
In the picture above, you can see an input signal (yellow trace) and output signal (blue trace), as well as the FFT of the results. My signal generator was set to 5.0 mV output on 3 MHz. This small signal was fed into the Head End amplifier, through 30 feet of CAT cable into the Base Amplifier, which produced 75 mV output.
Overall, the numbers show 24 dB overall system gain. Looks great.
But my Head End amplifier is not stable, demonstrating a lot of motor boating. Instability appears to be a function of power supply voltage. I had to insert a trimmer pot to reduce V+ to 2 volts rather than 5V in order to get stable performance. On top of that, gain was not flat, decreasing below 1 MHz and above 10 MHz.
In the next few articles, we will examine performance in more detail.