SNAP Questions & Answers

Here's a list of Questions and Answers regarding the construction of the SNAP Atlanticon Kit.  If you have a question and it's not addressed here, please drop a note and we'll get it cleared up for you.

And remember that the SNAP QRPp Transmitter Kit schematic, layout, construction details and photos are located at http://www.njqrp.org/atlanticon/snap.html .  QRPers can build the project and bring it along (or submit it by proxy) for judging at Atlanticon on the evening of March 25.  Winner gets an RH20 transceiver kit from Red Hot Radio!


Q: Can I use a metal can 2N2222A in place of Q1?  What about a 2N2222.

A: Any 2N2222-family transistor will be allowed.  In order to obtain the most power output from your circuit, it might be a good idea to provide a socket for your transistor and swap in various other 2N222 transistors you have in your junk box in order to find the one with the greatest gain.  The gain does vary among manufactured batches, vendors and case styles.  It might be fun and a learning experience to find out about this.


Q: How "far" can you go in substituting for the 2N2222A?  Can I use any NPN transistor of my choice?  How about a TO-92 Darlington?

A: There is actually no substituting allowed out of the 2N2222 family of transistors.  If it doesn't say 2N2222, it won't be "legal".  Thus, darlingtons won't be allowed in the judging.  (Unless I'm missing the point in that there *are* 2N2222 darlingtons.  If so, this doesn't meet the intent of the rule either.)


Q: If the 51 ohm resistor is supplied with the kits, wouldn't the differences in the tolerance of these resistors effect the power out measurements, assuming a simple voltage reading will be used? Or will the actual resistence of the load resistor, R4, be measured first?

A: Per the rules stating "exact component values must be used", the load resistor *must* be a 51-ohm value.  But you're right, tolerance differences will indeed affect the power readings.  That's one of the little tricks previously mentioned that the intrepid homebrewer could use.  That is, for this load resistor (as for the bias resistors), slight legitimate tolerance variations can be used to selectively provide more power output. It's up the the homebrewer to swap in different toleranced components (perhaps after pre-calculation to learn which is better) in order to get as much power output from the circuit as possible.


Q: Is changing the squarish type 9v battery to a different form-factor (such as 3 3v batteries in a stack) allowed or not allowed?

A: Actually, it is required to use a standard, small, rectangular 9V battery.  This keeps the playinf field a bit level.


Q: Can I use an On-Off switch?

A: Sure, an on/off switch would be allowed.  It doesn't alter the functional operation of the circuit, and it's easier than attaching/detaching the battery clip all the time.  You certainly don;t want to run down your 9V batteryt by having it connected all the time, either.  A fresh battery will always provide for the greatest output power.


Q: In the interest of "beautifying", can additional connectors be added, such as xtal socket, machine pins?  How about a case, altoids can, PCB box etc?

A: Beautifying can indeed include cases, sockets, etc.  Same too for the method of construction on the pcb material.  SOme guys are building up theirs on both sides of the board! to make it smaller and to fit into a small Altoids tin!  You can add connectors, sockets, etc, as long as the electroical operation of the circuit isn't altered. But part of the "beauty equation" would be the actual quality of the Manhattan-style construction too ... how nice are the pads?  How nice are the component leads to the pads, and to the ground, etc.


Q:  How did you get the calculations for power using the RF Probe circuit shown?

A: The meter circuit diagrammed in the package will indeed read pretty close to the RMS value of the voltage.  The RF detector circuit has built-in scaling to give approximate RMS readings for sine wave signals.  An ordinary diode detector outputs the peak value of a sine wave.  To convert the reading to RMS one needs to multiply it by .707.

In the diode detector shown in the SNAP package this scaling is done using a resistive voltage divider.  If the external DMM has an input  resistance of 10 megohms (most do these days) a voltage divider is formed by the series 4.7 meg resistor in the RF probe and the DMM resistance.  This ratio is 10/(10+4.7) or .68 which is less than 5% from .707.  For a simple circuit this is probably close enough!


Q: I discovered my SNAP was going into 3rd overtone (~10.8MHz) Osc, and still putting out about 10-20mW! Reasoning that the xtal was being overdriven, I put in a 2N2222, TO-18 (metal) case (only thing I could find in my parts stash). It's gain characteristics allow the circuit to function.

A: Yes, as mentioned above, some crystals operate strangely, even in a "standard circuit". This brings to mind that we'll have a spectrum analyzer on hand to determine that the SNAP entries are operating at the fundamental frequency.  (Don't want any false readings misleading the judges! :-)


Q: What is keying this thing? Would a keyer be in series with the battery and pull everything to ground? Or is this just a smaller circuit designed to be used in bigger projects?

A: The SNAP transmitter project is based on the Atlanticon QRP Forum guys building the circuit as neatly or uniquely as possible, and having as much power as possible being delivered to the 51-ohm load resistor "dummy load".  Thus "keying the circuit" hasn't been stressed up to now.

But some of the guys have already got their SNAP transmitters on the air and making contacts, by removing that dummy load and connecting a 50--ohm antenna to the circuit output.  They are keying the circuit with a hand key that connects the +9V to the Q1 oscillator circuit, much in the same manner as the Tuna Tin 2 transmitter is keyed.

There was a guy (Jake Carter, N4UY) who described a way to connect up his TiCK keyer chip and paddles in order to key the TT2.  He used an intermediate PNP driving transistor like a 2N4403) to be the "switch" that applies the +9V voltage to the circuit.  We actually have a "TT2 App Note" article describing this technique in the coming issue of QRP Homebrewer, due in people's mailboxes early next week.  It can also be applied to the SNAP transmitter.


Q: I'm having a small problem with my SNAP. The OSC doesn't always start. If I touch the XTAL with my finger and apply power it starts fine and run _IF_ I hold my finger on the XTAL can. Removing my finger right away causes the circuit to behave as described above. If I touch the XTAL to get things started, then hold my finger in place for a few seconds before removing it, the circuit works fine.

A: Well crystals can be funny beasts.  They exhibit a variety of resonances and are usually constructed to work properly in a predefined circuit. The SNAP just may be different enough from the originally intended purpose that some crystals in a given batch may not oscillate reliably  without some special care.

I suspect that what you are seeing is an overtone response that is favored in this circuit.  Commonly higher frequency crystals (24 MHz and higher) operate in the overtone mode since "fundamental" frequency crystals get increasingly fragile as their "cut" frequency goes up.  The overtones are close to odd multiples of the lowest resonance frequency analogously to the way that a 40 meter dipole exhibits a resonance in the 15 meter band.

It is mildly unusual for a crystal to favor its overtone mode rather than the fundamental - in fact one of the usual problems is getting overtone rocks to oscillate at the overtone frequencies and not lower! But considering the fact that these are mass-produced "economy" crystals, their characteristics may not be as tighly controlled as those that get a lot of individual attention and cost as much as a good dinner!

When you touch the crystal's case you are probably adding a slight capacitance to ground that makes the circuit "want" to sing at the fundamental. A simple "no-cost" and no extra parts solution might be to simply solder a piece of small-gauge bus wire to the case and ground it!

 

Return to the Atlanticon home page



Last Modified: March 9, 2000