09 April, 2012

Whatever happened to the 1 Volt QRP Transceivers?

Output stage of the 1 Volt DL2AVH transmitter
(from http://www.lichtnetzwerk.de/1volttxvr_dkumentation.pdf)
I am intrigued by minimalist transceiver designs like the Pixie 2 which I built some years ago. Therefore the "1 VOLT Challenge" from Dayton 2000 is also something I wish I knew more about. It had these winners:

1. Duncan Walters, G4DFV - The HAMEOBA - A 100 mW Single Cell CW Transceiver (30m)
2. (tied) Helmut Siefert, DL2AVH - A 30m 1V QRP transceiver
2. (tied) Charles Fletcher, G3DXZ - An 80m CW QSK Transceiver
4. Jim Roberts, NC9H - A 20m 1.5V Transceiver

Now, what happened to these designs? I'll try to figure it out from sources on the web.

The DL2AVH 30 m 1V transceiver is documented with a schematic and a photo of the breadboard. The key feature is an 80 mW output at 1.1 Volt supply voltage rising to 180 mW at 1.5 Volts. It used a push-pull pair of BFP196 low noise, low distortion broadband amplifier transistors in the output stage (see figure above). I consider the output stage to be the most challenging part at this low supply voltage. Other parts like audio amplifiers are frequently made for 1.5V supply voltage in for instance powered headphones, so that kind of circuity is more straightforward. The DL2AVH receiver is direct conversion and there are separate oscillators to drive the receiver and transmitter. It seems to have been corrected in 2007 and a printed circuit board seems to have been designed also. But I don't have the final documentation, it may however have been published in the German QRP-Report at the time.

I couldn't find any information on the G4DFV and G3DXZ designs, but found that the NC9H design had a superhet receiver. The rest is probably documented in the QRP-ARCI reports from July 2000 which I unfortunately don't have. 

My low voltage transmitter:
74LVC240 output stage (to the right), oscillator with 74LVC04 (lower left),
LM317-based variable power supply (upper left)
Some years ago I was inspired by these low voltage designs. I also saw several 74HC240 logic gate transmitter designs, originally coming from N7KSB in QST some time in the 90's it seems. That chip runs at up to 8 Volts, so in order to use that concept I had to use a different 74-series logic family.

I made a push-pull output stage with 4+4 gates from an 74LVC240 octal inverting buffer (supply voltage 1.2 - 3.6 Volts), see image to the right. All I was able to achieve was a mere 23 mW at 14 MHz and 10 mW at 3.5 MHz with 1.5 Volts supply voltage. This is really poor compared to the 180 mW of the DL2AVH design. I couldn't really figure out why this was all the power I was able to get.

Now, the question is if there is more information somewhere, and also more inspiration to get from the designs from the Dayton competition. Perhaps also there are newer concepts that have been developed in the mean time which are worth trying?

10 comments:

  1. Hi OM Sverre,
    came across your 1V transmitter project via
    a link on Bert/PA1B QRpp blog.
    Maybe I missunderstood your approach, but seems
    the 74LVC240 is a 3.3V device with a minimum supply specified at 2.7V ?!
    Most digital gatters employ two or more transistors in series sharing supply voltage,
    so a discrete push-pull solution might be hard to beat - anyhow, good luck for your project,
    will follow its progress on your blog, thanks a lot for sharing your results.
    72/3
    Peter/DL3PB

    ReplyDelete
  2. Hi Peter and thanks for your comment.

    What made me try the 74LVC-family was a statement on the data sheet saying "Supply voltage range of 1.2V to 3.6V". Maybe it is optimized for 3.3V, but this should indicate that it works at lower voltages also. My guess is that there is one transistor to ground and one to Vcc in the output stage, but I cannot see that that should cause any problem - or is the output stage more complex? I don't have any schematics for the inner workings of the 74LVC circuits, so I cannot tell.

    ReplyDelete
    Replies
    1. Hi Sverre,
      yes, one should read the whole datasheet before making a comment :-)
      Unfortunately the datasheet I found does not specify
      voltage for 'high' and 'low' at such low voltage supply.
      At 2.7V these voltages are given Vcc-0.5 and 0.4V with a current of 12mA, most likely worse at 1.5V.
      So your 20mW is actually more than what one would expect from a push-pull at 1.5V - I have no schematic either,
      but my guess is that these limits are the reason for the limited output power, in most discrete designs of low voltage transceivers the final transistors were choosen for lowest saturation voltage.
      OK, the 10...20mW aren't that much, but will allow QSOs
      with little help from propagation, although a 100mW or so
      make it much, much easier.
      Best
      Peter/DL3PB

      Delete
  3. Oh yes you have a point there. This should give an output swing of the order of only 0.4V - 1V which is not much. Perhaps there is a better logic family available which will have lower on-voltages?

    By the way, I enjoyed your diode-only transceiver on PA1B's blog (http://pa1b-qrp.blogspot.com/2012/02/parasaki.html), the Parasaki. Even with 2 mW you could make a contact from Germany to France, impressive!

    ReplyDelete
  4. Unfortunately I'm not an expert on digital stuff and most of the modern low voltage devices are designed for low current as well -
    just enough to cover the desired speed - maybe someone else has an idea.
    Thanks for your kind comment on my paraski adventure, actually it took me several month before I could finally make a QSO - this QSO turned out to be kind of jackpot in as much Chris/F8DZY later sent me a MP3 recording of my signal, sounds weak but 'easy' to copy with some margin left when frequency was clear.
    The power tunneldiodes (100mA peak) were offered on ebay, while all
    the other tunneldiodes are usual 4.7 or 10mA types - but cannot suggest duplicating the circuit, it requires very careful tuning
    of the parametric stage to obtain a stabile output, that can be switched between transmit and receive/mixer without failure.
    Meanwhile Mike/AA1TJ has tested an ultra simple PLL stabilized tunneldiode push-pull oscillator, but he's dealing with several projects at the same time - let's watch his blog and see...
    72!
    Peter/DL3PB

    ReplyDelete
  5. At low voltages low resistance is the KEY to get more power. For That you need a low Vce sat voltage transistor. Or an FET with a LOW Rds on. Panasonic has a transistor with .0018 ohm on resistance. It' number starts with SK86-------. It is tremendous and can be paralleled also. I have a dc-dc converter that puts out 2 Watts easily with these kind of parts. It should be easy to repeat this at RF. Remember Low Wire resistance is very important at this level.

    ReplyDelete
    Replies
    1. That's definitely an interesting FET, thanks for the tip!

      Delete
  6. I was also looking at the DL2AVH schematics. It has a push-pull output stage. It is capable of more output power than a single-ended circuit and the use of bipolar at this voltage makes sense.
    However, you'll note the extremely low power, because of the losses on the transistors and other components. You can improve the efficiency of the transistors by running slightly higher voltage. In today's technology, the target should not be a single AA NiMH (I don't like the environmental impact of alkalines!) which is rated around 1.2V/2800mAH, or a more realistic 2000mWH. At 25% key down average (contest mode!) that would be sufficient for 2-3 hours at 1W output, assuming we can get that power level.
    Instead, I'd consider a 3.7V/5000mAH Li-PO 18650-type cell. Realistic capacity is probably around 12WH, which means you can run 2 watts output at 40% efficiency and 25% key down for over 8 hours.
    If you don't like the 18650 battery then you can get a compatible (same physical dimensions) adapter for 3 AAA batteries. The power capacity drops tremendously, but it's an acceptable standby source.
    The 3.7V battery also allows you to run 3.3V microcontrollers, so it's much more user-friendly for designing a modern "pocket" QRP transceiver!

    ReplyDelete
  7. Is it allowed to cheat ? E.t. put a DC-DC voltage converter as part of the schematic 1.5V -> 5V & use that ?

    ReplyDelete
  8. No, a DC-DC converter defeats the whole purpose of the exercise. Also, I am pretty certain that it was not allowed in the original "1 Volt Challenge"

    ReplyDelete