Diplexer for 2m and 70cm

There are plenty of recipes for one of these so here is my version. I can't show you a picture since it's been in service for several years now without any issues. But it all fitted into a small diecast case just large enough for the 3 N connectors I used to sit alongside each other.

If you're unsure what a diplexer is then don't read the wikipedia entry:

"A diplexer is a passive device that implements frequency-domain multiplexing. Two ports (e.g., L and H) are multiplexed onto a third port (e.g., S). The signals on ports L and H occupy disjoint frequency bands. Consequently, the signals on L and H can coexist on port S without interfering with each other."

Clear as mud!

You can use a diplexer to combine signals from aerials on two ham bands, say 2m and 70cm, onto one feedline. Here I am using it to combine a 2m and a 70cm aerial onto one feedline which is connected to a dual band radio. A picture is worth a lot of words here:

You can use a diplexer to connect two radios on different bands to one run of coax. However, the diplexer must have good isolation when connecting two radio's together. You don't want the transmitted signal from one radio appearing at the aerial connection of the radio on the other band. For example, if you were transmitting 50W on 2m would you be happy if 0.5W was making it's way to the 70cm radio? I wouldn't connect a handheld directly to a radio and press transmit on the handheld. Even 0.05W, or 50mW, would make me uncomfortable. 50mW represents 30dB of isolation and even this diplexer does not achieve that. I consider this diplexer, and all other 3 and 5 element designs, to be for diplexing aerials only.

So isolation is important. How much you need depends on the application. My goal was to build something that needed no tuning once assembled. So no trimcaps.

I used some diplexer software from Tonne Software, www.tonnesoftware.com, to generate the filter values and settled on a 5th order design. After fiddling with the crossover frequency and ripple I had a filter that was close to the values of transmitting capacitors I had to hand.

From this point on the challenge was to wind the inductors. Which really wasn't that hard with my test jig described here.

73's
Richard

23cm Low Noise Amplifiers

I was quietly collecting parts to build a 23cm transceiver for local terrestrial use. I had a NE32684 and loading the S parameters into RFSim99 and using similar matching as other designs produced the following:

Circuit, Gain and Input return loss.

But I wasn't entirely happy with it. While existing designs had enjoyed great success, I was concerned about the stability. It's hard to convey in words but sweeping the frequency slider in RFSim99 showed that while the input was unconditionally stable, the potential existed for sub 1GHz oscillations depending on the load.

RFSim99 - about as easy as it gets to test. Here showing potential for instability with the output load.

Since I don't have a lot of 23cm test gear I decided to see how negative feedback would help. I was surprised to see that even with just a little resistive feedback the potential for oscillation disappeared and input matching with LC networks could be used. That gave me the following circuit:

This would not be optimum from a noise figure perspective, since we are both matching to 50 ohms instead of using a deliberate mis-match for lower noise figure, and we have used resistive feedback which generates noise.

However, on balance I was happy that my front end filter would see a reasonable termination at the image rejection frequency and this would prove to be suitable.

Then I chanced upon two cellular base stations: one at 900MHz and one at 2100MHz. Pulling htem apart and trawling the internet showed they had low noise front ends using ATF54143 and ATF58143 transistors. Both had comparable noise figures to the NE32684 at 23cm, but not as good at 10GHz.

So I decided to spin the wheel again (and again) to see what these transistors could deliver. With a bit of fiddling in RFSim99 I found that the ATF54143 gave me a useful front end RF amplifier:

Yes, it looks wrong.  With feedback between the gate and the capacitor you get quite a different response. What I have here is something with a good return loss and gain from LF to 23cm.  The ATF58143 was not vastly superior at first blush so I put that to one side.

Now I have something that works like a mmic. I know there are a couple of low noise mmics that have good return loss and better noise figure than this "theoretical" circuit, but I didn't have one to hand. So I fired up another software package to see what the noise figure might be. The simulation results were near identical, always a comfort, and the overall predicted noise figure is 1.09dB. I have yet to try to see if this could be improved.

It's time to build a board and add this part of the radio to the front end filter that is now in place. Nothing like building something to prove if the maths was correct.


Update:
I decided not to pursue this. Firstly, it looked wrong and when I checked it still had stability issues. Secondly, I found something better. A GPS LNA which I will write up soon. Board layout for that device is progressing as time permits. It looks like the best device yet.


73's

Solder - further testing of Chinese brands

Well, I tried a few more brands of Chinese solder. One was actually very good but I haven't time tonight to track down a picture for the summary.. I will though, because I will be buying more of it.

In the meantime I do not recommend this brand:

For a moment this looked like a surprisingly good solder.

I was wrong. What appeared to be good joints too often turned out to be dry joints.And this was a thick 0.8mm solder wire. I shudder to think how poor a thinner wire from this manufacturer would have been.

Solder - The Good, Bad and Ugly of Chinese Solders

So I've been really busy for too long now on work matters. One aspect of that was hand soldering SMD boards. As I looked at my dwindling spools of solder I thought it would be useful to order some more. But yikes, the cost was surprising. So I thought I would test a few varieties from Chinese sources.

The results were generally disappointing. Here then is a brief summary of what I found:

 I do wonder how the Chinese can manufacturer so many electronic products without good quality solder. It must exist, but so far I have not found anything that matches what the manufacturers must be using.

I will try a few more rolls from different manufacturers in due course. If you have found something that works, I'd like to hear about it as would anyone else reading this.

73's

1296MHz or 23cm Beacon

A nearby Ham, and a good friend of mine, VK6KDX Jack is working on an EME station for 23cm. Locally our 23cm beacon has been off air for eons. For testing purposes Jack needed a distant signal source. Here then is the 23cm beacon, measuring 10cm x 5cm, under construction:

The beacon is now complete and tested. The ident is in CW every 60 seconds. The PLL chip is loaded at start-up by a STM8S1003F, or STM8S003F, micro running the fantastic STM8 forth with more details here. This micro also gates the amplifier chain to produce the morse ident.

I used parts I had in the junk box. The PLL was recovered from some TV modulator boards I have. The varicap came from another TV demodulator board that I had. The mmics were all recovered too. Except for the final mmic. I had planned to use the AH1 mmic shown in hte phot above which is a 100mW output device. However, both my recovered mmics were broken. So I used an AH101G instead which is a 500mW device. The standing current of 200mA from an 8V supply means 1.6watts is being dissipated which my board wasn't designed for (the AH1 is only 0.75W ). To solve this I used some copper tape to better connect thermally the two sides of the board and create a small fin. I also attached a heatsink to the opposite side of the board. So far, so good. The device hits 75 degrees Celsius and should survive the abuse it will receive.

I'll just have to buy some AH1's and either retrofit one, or build another beacon up. I have a lot of boards left over. Please contact me if you want one. I can furnish a partially populated and programmed board for you.

The code can be found here.

Regards
Richard VK6TT

40m Direct Conversion Receiver - Completed

Update: Also read this for fixing AGC for weak signals

It doesn't matter how good something appears on the test bench, it's the on-air baptism that counts. This weekend there was an ARRL contest. A proper baptism of fire. So imagine the smile when I turned the receiver on with a dipole connected. Wall to wall signals. North American stations, JA's, VK's.  They were all there. Strong, loud and clear.

This morning, Sunday, I listened to the local new broadcast on 40m. I could hear every station that checked in. And I could just cope with the strong signal of Chris VK6JI who runs the broadcast and call-backs. Chris's station is perhaps 5km away so it is a full scale signal on every radio I have. I hope to simultaneously  record the call backs one day, my homebrew receiver on one channel, my IC746pro on the other. I know there will be a difference, but not much.

Now that I have the bugs of this approach sorted out I'm using one of the commercial boards for an 80m version.

There is one thing that I learned from this version: how a voltage inverter can cause trouble. I used an ICL7662 to create a negative rail for the op amps.  What I failed to catch until I tested the unit was the ripple from the ICL7662 was large enough in amplitude to upset the squelch circuit. That was my fault for not bypassing the V+/2 rail the squelch used. One capacitor and problem solved.

Now all of my design goals have been met.

• lots of volume from a speaker since headphones don't work if you're moving about the workshop while you listen to a net,
• Passed - I can easily listen to a net while moving around my very large workshop.
• no audio instability, 
• Passed. Micro-phonics solved by using tantalum's instead of MLCC's in certain parts of the circuit.
• no hum, 
• Passed.
• good AGC so you are not jockeying the volume control, 
• Passed
• good audio filtering
• Passed
• a squelch. 
• Passed.

That's 6 from 6. I'll finish the 80m version soon but it's time to focus on updating the synthesiser and transmitter to accompany this receiver. Then I'll have a working homebrew 40m transceiver again.

Complete schematics except for the squelch. I'm happy to pass on details of the squelch circuit privately. But as far as I know it is copied from a Codan radio so I'm mindful of copyright.

Regards
Richard VK6TT

40m Direct Conversion Receiver - Testing successful

Today I finished assembly and correction of a few minor mistakes of the receiver. The end product performs really well on the bench. In summary:

• about 80dB of dynamic range
• no instability whatsoever, even at "crazy" loud volumes
• no microphonics

I'm looking forward to on-air testing this week.

Regards
Richard VK6TT

40m Direct Conversion Receiver - Yes, I'm doing another one!

OK, so I saw a blog post by someone who suggested they were getting a batch of boards done at a great price. I won't name them because their bargain was way too expensive once freight was included. But by the time I discovered that it was too late, I'd made the mental commitment to one final, best I could build version.

Since then I've sat on my hands for almost a year. But this weekend I needed a break from the maths of loop filters and modifying 10GHz equipment. I wanted something I could do without much thought.  And since all the planning for this was done around a year ago it seemed like it was time to start.

I had wanted to see if toroids would offer a significant advantage over the tuned inductors in the RF bandpass filter.  While the existing filter works really well,  I considered moving from adjustable inductors, with a Q of perhaps 70, to toroids with a Q of say 200.

The tighter filter passband became less attractive when I worked out the maximum inductance I could wind on the toroids was around 1uH. A larger inductor has wire too fine to allow for compression and expansion of the coil turns to tune the filter. The advantages of the higher Q are offset by the lower impedances within the filter and the net result is not much difference. Had I had a larger toroid then things would have been different. So I will stick with the existing filter and this iteration will really be about incorporating all the modifications I made on the fly.

Here is a piece of advice I agonised over putting to print. I'm not comfortable denigrating other people's designs. But I was gob-smacked to see someone recommend an AM broadcast filter for a 40m receiver as a way of fixing receiver overload from short-wave transmitters. I'd like to know how that works. Because I can see no way that it can work. Remember, AM broadcast signals are around 1MHz. Those pesky short-wave signals run from 40m upwards to perhaps 16MHz.

And another matter that galls me is people still promoting simplistic designs as being worthwhile. If the project doesn't have at least a double tuned band pass filter between the aerial and first mixer then it's a glorified crystal set. I do wonder if those that propose the "easy" way know what they are doing.  For all the effort that goes into building a receiver reward yourself many times over with another 1% of effort and put in two or three extra components in one of the most critical parts of your receiver.

Here's a picture of progress to date. I hope to have this finished by the end of the month. Then I'll think about what to do with all the spare boards. Any takers??

Regards
Richard VK6TT

EM 10.5-10.68GHz microwave modules - Mods to use on 10.368GHz - update

The first RX module which down converts 10GHz to 144MHZ was finally finished today. Initial results were disappointing. However, after some modifications to improve the noise characteristics of the first local oscillator everything is running nicely now.

The conversion takes a while but the result is a sensitive down converter suitable for SSB and FM use. I'm not posting full details of all the changes required at this point. Please get in touch if you'd like to know more and I'll consider blogging further details.

In the meantime I have a few more modules to convert. The RX module has the following model identifiers:

em                   FM RECEIVER
10.5-10.68 GHz, Synth
P/N 036D                    S/N 111xxx

EM 10.5-10.68GHz microwave modules - Mods to use on 10.368GHz

This started out as what appeared to be an ambitious project but it is so far straightforward. I was fortunate enough to be given the chance to modify a few sets of the FM Tx and Rx boxes for ham use. It transpires that just about every set of modules is the same but different! They are all broadly similar but subtle variations exist so if you have some of these modules and yours don't match exactly then this shouldn't deter you.

Here's what I have done so far:

1. Used a cheap logic analyser to check what was being loaded into the PLL chip by the micro. 
2. Inserted my own micro to control the 2.5GHz PLL
3. Achieved lock on several TX and RX units

PLL chip
The units I modified had the LM2326 PLL chip. It has a different data format than the LM2325 chip I saw on the units I have not yet modified.  In a few words the 2.5GHz PLL uses a 250KHz reference frequency. It is multiplied by 4 on both tx and rx units. On the tx units it is mixed with the IF, presently 69MHz but I am modifying this to use a 144MHz IF on tx. On Rx the multiplied signal is mixed with a 692MHz oscillator down to a 140MHz IF. At present I'm leaving this oscillator alone since the saw filter appears wide enough to be serviceable with a 2m IF.

Micro
Once I knew what the existing data stream was it was relatively simple to work out the data stream needed to load the PLL chip with data to put it on the desired frequency. See the chart below for an overview. I used an ATTINY13 and a few lines of assembly to do this.

Achieving lock.
This should be straightforward but sometimes the soldering proves difficult due to the heatsinking the board has. I moved the short until the PLL locked with around 4V-6V on the tuning voltage link.

I'll post some more pics soon but for now the following may prove useful.

Regards
Richard

An unmodified receiver showing what is required to modify for 10.368GHz use

Details of how the IF signal on Tx gets to the mixer diode for up-conversion to 10GHz

Data loaded into PLL chips to shift to 10.368GHz