IRF510 2×2 Push Pull

Well this one has been a long time coming. From the very first amp i built till now has been close on 5 years. Those first amps were rather scary contraptions dead bug on scrap pcb and had more than a few fets explode on me. And after the debacle that the 2×4 amp was, this time around i decided to cut it in half and make sure it is all working right and to have the mechanical aspects of the build right also, so if something does break and lets be honest something always breaks, that its easy to fix. I have not really started to push this yet, but atm with conservative biasing, drive and voltage i am getting a respectable 36 watts out. Its been key down for the last 15mins and other than the heatsink needing to be bigger to dissipate the heat its going well.

Here is the schematic.

Here is the PCB as laid out in my ecad of choice.

So rather than try and go the home made SMD route, i just went old school with the mechanical design. What i did was mount the fets loosely to the heat sink, then i snug them up, bent the legs with the pcb and then tightened them all down. It meant i could slide the PCB on and off the legs really easily until i was ready to solder them on permanently. If i fet dies, it should make replacing it a lot easier.

Setup for testing. I kind of just snuck up on it with voltage and drive. 13.8v and 200mw drive and I was getting about 7 watts out. Slowly and surely i brought the voltage up and the drive as well. In the end i was giving it 1 watt of drive at 24v and was getting 36w out. Its got way more to go, but its nice and stable.

A step in the chain, was about 25w here.

60v peak is 36 watts. Now i need to put some 200v caps in the low pass filter and make that look like a sine wave LOL

PDF of the schematic : irf510-2×2-amp

Gerber files for the pcb : REMOVED found an error when I ran 10A through it.

 

Facebooktwitterredditpinterestlinkedin

DDS VFO CW Exciter

Here is the next part of my home brew CW transmitter. An Arduino and AD9850 are providing the DDS VFO signal. Q1 is a buffer to isolate the AD9850. I am not sure that it is needed, but its there to stop the gain stage from loading it up. Q3 is an amplifier that takes the 400mV P-P DDS signal and amplifies it to 8/9V P-P or there abouts into a 50ohm load. Thats about 200mW, enough drive to supply something like the QRP Labs 10w Linear. My plan is to design an amp to follow on from this that will provide 30 to 50w. More on that to come. Q2 is a PNP switch, pulling the base to ground turns lets power flow to Q3 thus keying the transmitter on and off. Simple but works.

Here is the schematic.

The PCB as laid out in Ecad.

Here is the board as built, if you are not colour blind and can read the colours on R5 you will notice i stuffed up here HAHAHA. Also Q1 silk screen is backwards.

Signal output into 50 ohm load at J2. This jumps up to about 10V P-P into the QRP Labs Linear amp. Not sure why yet.

Schematic PDF: CW-Buffer-Amp

Board Gerber Files: DDS-CW-Buffer-Amp

Facebooktwitterredditpinterestlinkedin

7th Order LPF Boards

Ok, I have still been working at building stuff. Here is the latest piece of what will become a CW transmitter. These boards are 7th order Chebychev low pass filters.

This is the PCB as laid out in ecad. 4 boards per PCB and you can get 5 PCB from JLCpcb for $2 so that is 20 low pass filter boards. A life times supply.

My current project is a 40m CW transmitter, here is the LPF sitting on top of the TX/RX switch.

Filter response plotted with the NanoVNA.

If you want to have your own LPF boards made, here are the gerber files:  LPF_Singles

Facebooktwitterredditpinterestlinkedin

Strawberry Wine

Strawberries are currently in oversupply and the shops are practically giving them away. So what do you do when you can buy berries for nix, well make wine of course. I have one batch in demijon already and another that has just started in the fermenter. 8Kg of berries to make 5L of wine. Works out at about $4 per bottle.

Facebooktwitterredditpinterestlinkedin

LPF Board Measurements

Here are some measurements of 3 filters off the low pass filter board I designed and built. First harmonic of each band is at least -50db down. The 4th band which in my case was for 15m, was well out and I need to look at it to work out why, cap values is going to be my initial guess LOL. I built this thing and have never used it LOL. Though I got some new boards coming from JLCpcb this week and i hope to have a transmitter happening that I will use in combination with the HackRF One as a receiver.

Facebooktwitterredditpinterestlinkedin

Xtal Filter Nano VNA

I had reason last night to use the Nano VNA for something and while i was at it, i grabbed down an Xtal filter i had built ages ago and put it on for measuring. If i recall right, and I would have to go and look in old blog posts, when I measured this using the Bode Plotter in the Red Pitaya it looked alot like Bart Simpson and had 5db of ripple in the pass band and very lossy. Well, better tool, better measurements and this does not look all that bad to me. The ripple is acceptable and actual loss in the filter is around 3db which is more than fine. Its not quite narrow enough for a CW filter, but its good enough to get started with.

Facebooktwitterredditpinterestlinkedin

Air Core Inductors

There was a post over on EEVblog that got me thinking. What do you do when you are unable to, for various reasons not be able to procure iron core toriods. Well, obviously air core inductors are a thing and highly practical for the low values of inductance used at VHF and UHF. But, what about at HF?

Well it turns out that as long as you do not mind a little bit of size to the inductors, they are still a vary practical option. So what i did was grab out some 1mm enamelled wire and wind 9 turns on a bit if 22mm OD pipe i had laying about to see what sort if inductance such a coil would create. Turns out about 2uH. That is highly usable value in low pass and band pass filters in the lower HF range. And with some thought on the actual construction and mechanics of the coil, very usable inductors could be made this way if you are unable to buy iron core toriods for whatever reason.

Anyway, that is some food for thought. There is nothing magical about iron core toriods, air cores will work just as well and should have a reasonable Q equal to or better than a toriod. So if you are stuck, there is a solution.

Facebooktwitterredditpinterestlinkedin