I recommend the use of buck converters to supply 5.2v to the RPI, LCD (through the RPi), ADF4351 module, LO filter RF Output switch and 4-band switch. No extra filtering required. I use a linear converter (the 7805) to supply 230 mA to the Filter-modulator board (needs heatsinking).
You can use a buck converter to supply the filter-modulator board, but the switching ripple does introduce some detectable sidebands on the transmitted signal. These have very little adverse effect, but offend me! We have tested on-board switching regulators, and the pads are there for L10 to provide extra filtering. Again, they introduce low-level sidebands that offend me, but are probably not significant.
The most important thing about the power supplies is that the RPi should be fed with a low impedance 5.2v supply which does not drop voltage under load. Most commercial micro-USB leads should be cut at 5 cm from the plug and connected to real wiring going to the buck converter.
Hope that helps
G8GKQ wrote:I use a linear converter (the 7805) to supply 230 mA to the Filter-modulator board (needs heatsinking).
You can use a buck converter to supply the filter-modulator board, but the switching ripple does introduce some detectable sidebands on the transmitted signal. These have very little adverse effect, but offend me!
Dave, I've gone for 7805 on the filter mod board fed with 8V (to reduce power dissipation in 7805) from a buck converter (down from 12V overall supply). Do you think the 7805 will mitigate the ripple from the buck converter?
Thanks and 73
For this reason I set my switching supply to 8V output.
The LO is based on an ADF4351 is a 3 - 3.6V device. A 3.3V regulator would appear appropriate. I think it takes 100-200mA depending on how it is configured. If the on-board regulator is AMS1117-3.3 (like in the $15 board from Banggood), these operate down to 1V dropout, maximum 1.3V. This is well suited to a 5V power supply. Running it of a typical 13.8V supply means the regulator could end up dissipating 2W and the SOT-223 package could not handle this. Running it of 8V reduces that to 1W but its still a bit marginal on dissipation. I would recommend 5V. As suggested above, it seems you can get away with running it off the same supply as the PI. However, if you have a different board with a different regulator, it may need more.
The filter modulator board uses a 7805. I simply used the larger heatsink and ran it off 12V. It seems to work. I originally had a DC-DC converter in there, but concluded it just wasn't worth worrying about the ripple. The old trick of an appropriate series power resistor will deal with any dissipation problems without generating extra noise. Also putting a series diode in-line with the 12V supply is a good plan as it will provide some protection against supply reversal. None of us ever wire our supplies incorrectly, but someone else might.
For the PI, I use a DC-DC converter set to 5.1V. It is wired directly to the PI, on the underside of the PCB bypassing the USB connection but not the polyfuse. Easy to do and this has completely fixed the issue of the PI complaining about undervoltage.
Who is online
Users browsing this forum: No registered users and 2 guests