So today I populated the 2nd IF mixer, the 2nd mixer and some of the switching circuitry and gave all those elements a little bit of a test.
To follow on from the test of the first IF amp, I put a 10mv signal into the first IF amp and tested the whole IF strip through the Xtal Filter and out the 2nd IF amp buffer and it all seems to be working quite well.
But there is one issue and I am not sure what the cause is. The 470R resistor from voltage divider on Q6 does not go straight to ground and trying to find its ultimate path to ground has been elusive.
To make matters worse there are q Q10 in the schematic and Q10 and 21 on the board, there are 2 diode on the board not on the schematic and I cannot for the life of me work out what the actual cause is here. Either one of the Q10’s is not being biased on or there is some part missing or either VR7 or VR2 are not providing the emitter ground path on the emitters of those Q10 transistors.
Anyway, I think what i will do is pull both the Q10’s out and related circuitry, part of which is for a signal meter and just hard wire the 470R voltage divider to ground. That will at least leave me to be able to complete the receiver and actually test it.
At one point today I had fed a 9mhz signal into the IF strip and was getting a 600Hz tone out the audio amp, so that in itself is promising as I was worried about the dodgy design 2nd mixer. Which looks like a ring mixer but does not have trifilar windings on the inputs and outputs, but some crazy bifilar winding on one side and a variable resistor and cap divider on the other. I am not sure about this mixer design and I have not done any google foo to see if i can find it, but it does work
Oh and for the diodes in the mixer i did not use the typical 1n4148 diodes that everyone uses, but rather I used a Schottky Diode 1N5711, it probably means i have way to much drive from the BFO as the switch on of these are much lower, but its working and for now, I call that a win, particularly so when i really do not understand the mixer topology being used.
Oh and as far as audio goes, of the 2 AF stages, I am only using the LM386 first stage because the TDR2003 is motor boating or oscillating or just acting like a tard and I am not sure why here either
I do hope the next post will be a video of it receiving signals. Failure is always an option here, but so far this build has been very challenging, but its been rewarding and even if it ends up being a pile of junk, i will have learned a lot.
I did a bit more on this today, went through and added in the first IF amp and tested it to ensure it was working. After sorting out some probe issues with my scope, it was all systems go.
After working out where the input to the first IF amp was and also buzzing it out back to the first mixer, I fed a 100mv signal into the amp and scoped the output on the other end of the xtal filter to ensure signal was going in and coming out. As you can see from the oscilloscope output above it works.
The blue trace is the 100mv signal from function generator and the yellow trace is the signal on the other side of the xtal filter. Now to build the 2nd IF amp and test it also.
I thought before i added in the IF Amps i would build the xtal filter first and see how it looked. The values of 33pf seemed very small and would end up with quite a wide filter. Typical values for a 2K wide filter are usually around the 100pf mark, so i started there.
Obviously i was wrong, but i am thinking 50pf might be closer to the mark after looking at the bode plot. Its not going to break any performance records and the skirts are kind of garbage for a 5 xtal filter. But, I will see how it goes on receive once i have the rest of the radio built.
As far as xtal filters go, this is one of the better looking ones I have built HEHE, so I guess the others have been really crap. It is kind of symmetrical looking and only 3db-ish of ripple. Might be ok. We will see.
Ok, so after playing with cap values i got the filter about the right width, but there is a lot of variance in the passband. 8db to be exact. Way to much. But as there is no way to impedance match this, there is not a lot i can do. It is what it is as they say and as I am only building this as a receiver, I can live with it.
Ok, this has been a fun section to build. I spent the better part of an hour scratching my head over this and wondering why it was not working. Well, this is how I approached this section of the build.
First thing i did was rummage around in my parts bins for canned inductors. I do not have a lot of these things, but i do have a few and those are of a few different types. From what I understood from the few bits in information i have been able to glean from the internet, they used 10.7Mhz IF cans. This then narrows it down to 3 or 4 actual IF Transformers.
Of the ones i have in my parts trays, only one looked like it had suitable values for a 7mhz filter, with windings that measured 120nH and 2.2uH. That was going to get me ballpark and allow me to simulate the required capacitor values.
The fun begins when you go to buzz out the actual pcb, because the input series cap is not on the board, but its on the schematic, first issue that did my head in. The output series cap is on the board and on the schematic. Simulating without the input cap seemed to make no difference at all. So i just left it out rather than cutting traces and soldering it to the bottom side of the board. This is not the first time in this build that i have found such discrepancies either. In fact, it would be a miracle indeed if someone could stuff the board with parts and it just worked.
Not feeling very confident at this point, I thought it would be best to simulate the lowpass filter just to make sure it was ballpark and surprise it was actually quite ok. Perfect in fact.
Ok now this is where the fun began. First thing i do when building something is buzz it out to make sure there are no shorts. Well the low value windings on the canned inductors are only 100nh. They look like a short to ground at DC. So sticking my continuity meter on both the antenna pins caused it to go beep and I spent the next 30mins trying to track down the mystery short to ground, before i worked out what was happening
So finally I wired up some power and injected a signal and started to trace it though the board all the way to the input port of the mixer footprint. Initially there was nothing, so i probed around and found RF at a mystery pad showing a 6.8uh inductor, which is in series with the 7pf coupling cap. After some head scratching, I think that this meant to be a capacitor so that you can series 2 caps together to get a lower value. I dropped a wire link in there and things finally worked.
0.02V signal injected at the antenna port, through the lowpass filter, through the bandpass filter into the RF Preamp and we have 50mV. I do not think the losses in the filters are all that high.
So that end this nightmare 🙂 Its all worked so far, just some IF amps, IF filter and 2nd mixer to go to have the receiver working. Still plenty that can go wrong from here, but its getting there.
So the RF preamp has about 5x gain. On 40m its probably not really needed unless there is a lot of loss in the filters. Which in this case might actually be the case as the RF path from the antenna goes through the LPF, into the Relay and then into the BPF. So a little gain ahead of the mixer might be a good thing.
One thing that I left out for now were the back to back clipping diodes. I am only building this as a receiver pretty much to begin with and I am not sure clipping diodes are really necessary. Next I will be working on the LPF and BPF. Need to do some LTSpice here to make sure my values are correct.
Ok, back to doing some work on the Bebek. In this block the only real issues are that there is an extra DC blocking cap on the board that is now shown on the schematic and that the Cap values on the schematic are not the same shown on the board. In this case I have defaulted to using the values on the board. 10nf and 100nf on the board are 1nf and 10nf on the schematic. The former values just seem more logical to use and the latter are most obviously an error on the part of the designer.
So the LO Buffer Amp has a 10x gain. I injected it with a 250mV signal to get 2.5V out. Now obviously this is not good if you are using a SI5351A as it puts out 1.5V. But this should work quite well with an AD9850 which puts out 200mV. It still might be a little to much drive for a -7dB mixer, but it should be ball park upper limit.