General Purpose Transistors in RF Amplifiers - Driver Test Results

Having reworded the amplifier chain to use 2:1 transformers instead of 22uH collector chokes I have commenced testing. My approach was:

1. Build the amplifier with transistors that were known to work and measure
2. Transfer the output and buffer transistors to a know board and start substituting SOT-89 packaged driver transistors
   
3. With a chosen driver transistors transfer everything to another new board and after confirming a SOT23 buffer would work start substituting final transistors.

Being a known good design the amplifier worked with the original transistors. I adjusted the output for 10v pp at 1MHz. The roll-off at 20MHz was 2.9dB. Looking at the inter-stage levels I concluded that most of the roll-off was split equally between the buffer and the driver. The output stage was almost flat (-0.2dB). 

After testing 10 different SOT89 driver transistors I found the average of the roll-off at 20MHz for the entire amplifier was now -3.7dB, a deterioration of 0.8dB. I explain that by the difference in Ft: the original driver, a 2N4427, has a transition frequency of at least 500MHz, compared with the transistors tested having a typical transition frequency of 100MHz. 

While I only tested one device from each packet, the best transistor was the BCX56 which was just 0.4dB worse than the 2N4427. I'm comfortable using that transistor because it also appears to have the highest Ft at the collector current being used (~180MHz at 42mA from the datasheet chart).

The worst transistor had a claimed 210MHz typical Ft at 500mA. But clearly the Hottech 2SC4672 does not perform very well at a 42mA collector current. That's probably why there is no chart! 

And be wary of just buying any BCX56. I checked a few datasheets and in this application I would only use a BCX56 where there was a chart of Ft versus current, or a stated Ft in the vicinity of 50mA. Of course, it you buy from Aliexpress then you have entered a lottery.

Conclusion:

In a class A amplifier around 30mW output you can use a 10cent transistor at HF. This BCX56 is now my standard and should work very well in my 80m Weaver transceiver being developed. This removes the need for a 2N4427 for low-mid HF applications.

Now to test some higher power alternatives.

73's

Richard

 

Surface Mount Soldering - Light- Box

This morning I was struggling with soldering a TQFP64 package. As I was inspecting for soldering defects I by chance discovered something something many will know - shining a bright light from the reverse side of the PCB can be very helpful.

I quickly rested the board on a frame over the bright light and attempted some touch up work. Wow! So much easier to see if those pesky solder bridges were being wicked away or otherwise being corrected. Later in the day I went to the shop with my better half, planning to swing by the hardware shop for some timber to make something more suitable. Left to my own devices I wandered around and stumbled on a bamboo organiser type box for around $10. Dimensions 12.5cm (H) x 17.8cm (W) x 31cm (D)

 

Flipped upside down with a hole and homemade LED light source mounted below the hole I now had a light-box for SMD soldering. 

No light

 

Light On

With the light on I can easily find solder bridges, whiskers and remnant flux. I can apply solder wick, or iron plus flux, to a bridge and easily see if it has corrected the problem.

Well worth the effort to make and it will be a huge timesaver for years.

73's

Richard

4:1 Transmission Line transformers - Which core?

 Right. So I know have a rule of thumb and a growing pile of test cores and results. The rule of thumb I found, see earlier posts on this subject, is very helpful when dealing with unknown cores. But I went searching and I was surprised at the reasonable cost of buying brand new cores from Mouser. 

For 200:50 ohm bifilar wound transformers I am narrowing down on a "standard" part. I am confident that small cores will cope with the maximum power a 12V class A amplifier can deliver across 200ohms (12^2/400 ~ 0.4Watts) so size is unlikely to be a selection criteria.  Taking into account price and performance across HF I present some results for your consideration:

Small Binocular Core ~A$0.30 each

I could only get 3 bifilar turns through this tiny toroid with the relatively large wire I was using so those bifilar windings were jammed on top of each other. Despite this the results were surprising:

3 turns 200 ohm termination and Open circuit termination

With finer wire and 4t I expect this core will result in a useful transformer. Confirmation in due course.

The next physically larger size core is substantially larger than the toroids below with a similar cost. So if 4 turns doesn't cut it then I'd put these to one side for now in this application.

EPCOS / TDK Toroid ~A$0.60 each

After 7 bifilar turns suggested more turns were needed I guessed 11 turns. Very good transformer at 30MHz but still needs a few more turns. I'll try 13 turns next.

11 turns 200 ohm termination and Open circuit termination

 

Fair-rite Toroid 43 Mix ~A$0.75 each

Results for 7 bifilar turns were pleasing. 8 turns might be better. I only had one of these cores and have put it to one side for now.

7 turns 200 ohm termination and Open circuit termination

Next steps:

The Epcos/TDK toroid is, on results to date, the leading contender. I still have a few smaller (and cheaper) ferrite toroids to test and I am interested to see how they perform. At some point I will have picked my standard for 200:50 ohm transformers and I will write it up.  Then it will be time to get back to my amplifier testing which is how I ended up down this rabbit hole.

 

73's

4:1 Transmission Line transformers - Rule of Thumb

Having unwound some of the cores I recently measured and finding some were not bifilar wound at all I started again. 

I took another, different small core, and wound a few turns on it. The exact number is unimportant at this point. It's the red curve below. I then wound a bifilar winding on another identical core with the same number of turns, that's the yellow line when the second winding is open circuit. What I noticed was the bifilar wound toroid appeared to be "stretched". When terminated in 200 ohms the blue curve eventuated.

Where either the red or yellow curve crossed the R axis, that frequency was a reasonable approximation of the frequency where the 200 ohm load was transformed to 50 ohms.

 

 

I picked up three cores, the same as each other but different from previous tests, and wound 9,10 and 11 bifilar turns on each one.

9 turns

10 turns

11 turns

Again, this relationship of the frequency where the R axis is crossed when the load is open is close to the best frequency for a 200:50 ohm transformation. 

 

I  believe I'm onto something now. Let's repeat this with 3 larger, EMI suppression cores taken from a VGA cable. Not the huge ones, but about 15mm long.

 

 

 

 

11 turns 

  

  

8 turns

7 turns

With 7 tuns I had a transformer suitable for 2 MHz (marker 1) to over 10 MHz (marker 2). I will repeat this for 6 turns and if suitable use that transformer in the test boards for general purpose transistors as RF amplifiers I am working on.

Conclusion:

You can quickly determine the frequency a toroid can be used to make a 200:50 ohm transmission line transformer with a bifilar winding. 

• If the toroid has a single winding, or adding some turns if it has no windings, adjust the turns until the R axis is crossed at the lower end of the frequency range you are interested in. Then replace the winding with a bifilar wound transmission line.
  
  
• If the toroid already has a bifilar winding you can treat it like a core with a single winding by leaving the hot end open circuit.

 

73's

Richard VK6TT

4:1 Transmission Line Transformers - Experimental results

Progress on my testing of general purpose transistors for RF amplifiers had stalled because I ran out of 22uH chokes when populating the test boards. Using a 4:1 transformer on the collector of the first stage would work but I wanted to understand, if I could not eliminate, variation between the test boards.

I had 5 different 4:1 transformers already wound in the parts bin. The pic shows these and acts as a legend for the charts to follow. There is also a 1206 sized part for scale.

 

Now to determine just how well these transformers worked. I swept each transformer over the range 1 - 100 MHZ with the nanoVNA. Firstly with the transformer un-terminated, then terminated in 200 ohms. A perfect transformer would have transformed the 200 ohm load to 50 ohms across all frequencies, showing as a dot in the center of the Smith Chart.

Core 1, the red lines below, appeared to be a ferrite core and I had expected it to work well as the frequency increased. It failed. While it presented roughly a 45ohm resistive load to the VNA across these frequencies, the reactance was significant until the frequency approached 100MHz. Expanding the swept range showed 100MHz was about as good as it got. Given this came from commercial equipment, perhaps it works better in a 75:300 ohm application.

Core 2, the yellow lines,  was a tiny twin hole balun core. I know this came from a commercial TV distribution amplifier. The open circuit trace suggested it might work up towards 100MHz, but when terminated in 200 ohms it also presented 45 ohm resistive at the top end of the swept frequencies. At 2MHz it was 34+j17 ohms.

Core 3, the green lines, was something I had wound on a ferrite bead, say 15mm x 10mm,  removed from a power cord for EMI suppression. I had no expectations. The open circuit trace showed it might have enough reactance at the low end, but it quickly presented a capacitive reactance to the VNA above marker 4, 10MHz. While the 200ohm load transformed well in the 2-4MHz range, but above 10MHz it was unsuitable.

Core 4, the light blue lines, appeared to have potential above 30MHz when inspected in the open circuit configuration. Terminated in 200 ohms it was a pass above 30Mhz. I suspect a few more turns would probably make this a suitable toroid for a HF transformer.

Core 5, the purple lines, appeared unsuitable from the open circuit trace. This proved to be the case when terminated.

Tentative Findings:

There is still more investigation to take place but some themes were noted from the work to date.

If you have a core with unknown characteristics choose one that tracks the outside of the Smith Chart when a representative number of turns is swept with a VNA. There are frequencies for which the blue trace tracks the outside of the Smith Chart. More turns would rotate the curve clockwise. Perhaps core 3 would benefit from less turns.

The useful frequency of the core will be in the range of frequencies where the open circuit response of the bifilar winding crosses the axis eg 2-4Mhz for the green trace. Getting the right number of turns could be the key to having a useful transformer for your applications. 

Being on the edge of the Smith Chart means the core losses are low and do not impress themself onto the response as a series resistance. The blue line above (LHS) is an example of this though it starts to spiral inwards above 30MHz which suggests it is not ideal above 30MHz.

The EMI suppression core would work well at 80m and perhaps 40m with further testing of turns. Above that the core losses, which a suppression core would be expected to have, appear limit it's usefulness.