Why not build your own.  Buy a mini itx chassis.  Buy a mini itx mobo.  Make some mechanical mods and add one of the mini QRP rigs to a shielded area of the ITX case (you might have to do some metal work or add an internal box.  Hook the works up to the itx sound card and RS232 port.  Find a small keyboard and display.  Bundle it and run linux and fldigi.

Or... get a netbook and strap it onto a metal box with a QRP rig in the box.  Rig up some simple interconnect and voila! mobile QRP psk.

Lest you think this is a pipe dream, I have a very simple interface I made for my K2 and HP mini 1010nr netbook.  The K2 has an internal battery and SSB adapter.  I put the works in my pack with a home brew dipole and sligshot laucher.  Hiked 2 miles up a mtn. and operated PSK31 from under a tree in amoungst the snow banks.

Or you can wait for somebody to do it for you.

I've done two things to eliminate this sort of problem.

1) buy a cheap inverter and rip it out of the case.  I then repackage it in a metal box with common mode ferrites and 0.01uF caps on the DC lines.  I used a commercial EMI filter can on the AC output.  With proper RF type wiring a cheap inverter is made low RFI.

2) Take the brick power supply (or an after market one) and rip it out of the plastic case.  Again add common mode ferrites, 0.01 caps, and emi filter, put it into a metal box (with a third wire ground to the box).  Now you have a low RFI computer power supply.

If you wonder if the laptop it self is noisey, you could do some experiments running it off batteries and bring it near your antennas a see if you get noise.  If you really are concerned about cable noise, add some dummy i/o cables while on batts. and see if there is noise.

If the laptop itself is noisey, that is more difficult and probably beyond where you want to go.

Cebik analyzed the optimum length of dipole to use for multiband operation.  I think he came up with 88 ft for 80  thru 20 meters.  Check the web site Cebik.com.  Lots of good info on how the pattern changes with frequency and center impedances.  Read and learn.

I think I hear a myth busters idea here!

A few = 7AH/(.12A) = approx 6 hrs.  In direct sunlight I would guess.

The device suggested will maintain your battery but it would be a long pull to fully charge it.  You can optimisticly estimate the time by taking the amp hrs and dividing by charge current (assuming 100% discharge).

You will need a lager panel if you are going to discharge the battery fully and expect to operate during the day.  Also maybe you will not be always having sunny weather?  If I remember, the 817 has a fairly high RX current as well.

I would consider a little larger.  Also if I was car camping, I would consider a means to charge from your car in case its cloudy.  Perhaps an inverter and 120V charging circuit.

I use an old TS870 for PSK.  Unfortunately you need to monkey with some software to control it because the narrow filters are on CW and you need to xmt on USB/LSB.  The software allows split with an automatic offset.

I have used the ICOM pro series as well.

I have an IC 7000 that would work well also.

I would highly recommend the K3.  It was actually designed with Digital operation in mind and you can even send PSK via the CW paddle!

I would guess any modern DSP radio that has the ability to switch the DSP filters down to 50 to 100Hz on SSB RX would work.

The alarming strong signals are due a lot to the sun spots coming back.  Its not the "over use of high power".  I'm glad to see PSKers are starting to wake up to the use of single signal reception.

For air wound inductors there is an old time formula for inductance given dia, winding pitch and length. See the ARRL handbook. There are some old papers from the 30s that attempt to quantify the losses vs winding geometry and wire dia.  You might find these by poking around the internet.  Be aware this is complicated stuff to compute since it is actually an electromagnetic problem and any formulas or simulation programs would limitations to certain geometries etc.

For inductors with cores, its even harder, since there is a magnetic material involved that can be non-linear.  Here you need to consult the data from the core manufacturers.

For simulation, you can't do better than LT-Spice from Linear Tech.  You should visit their web site.  There are engineering books on spice such as "The SPICE book".  Search amazon etc.  There has also been a series running in QST that uses SPICE to explore simulating electronics.  Check it out.

Tapped inductors in SPICE are handled by the coupled inductor model.  A parameter K is used to define coeficient of coupling.  This is a mathmatical abstraction derived from mutual inductance.  BTW the coupled inductor model also covers transformers (tapped inductor is actually an auto transformer).

You sound like you're relatively inexperienced (at least in simulation).  Inductors can be tricky depending on what you are trying to simulate.  I would suggest you get some equipement like an impedance analyzer and make measurements on your prototype inductors and then use the measurements to extract the values for loss and series resonance etc.

You need to think closely about the aims of your simulations.  In some cases you just are looking for gross behavior.  In that case simple models with just an inductance and guess at loss resistance (Q) is good enough.  When you are pushing limits, you need more parasitics modeled.

Finally, you need to understand there are several fundamentally different domains in which simulation occurs.  Hams tend to think in the AC steady state (frequency) domain of reactances, swrs, reflection coefs. etc.  There is also a transient (time) domain where the actual time response is computed.  Each has its place.

AC steady state is linearized and is faster but does not model non-linear behavior.  Thus can't model thinks like mixers and intermod.  Transient simulation takes much longer and won't directly give you things like SWR etc.

Be prepared for a voyage of discovery.  All this seems pretty complicated I admit, but that's why electronic engineers need 4 to 5 years of training at the college level + 5 to 10 years of real world experience before they are proficient at this stuff.

That being said, have at it.  If you are willing to scale the wall, you can simulate the things you are interested in.

I've done it.  It never has seemed to cause any problems.  There is no way to tell if it really made things better.  What have you got to loose by trying it?

I have used water pipes for the ground of end fed wires before and had good results.

I would try it.

How about improving your antenna and feedline.  Its easy to compromise your station with a lossy feedline.  Antenna gain and height is very important at VHF.  Finally consider a mast mounted preamp with T/R switching.

I used to do OCSAR 10 with and SB102, a transverter and an antenna mounted preamp.  The TX feed line was a piece of Heliax.  The magic is all in the antenna feedline and preamp.

The best hidden antennas are made of thin wire and strung in trees.

End feed the wire with a tuner.  Make the wire between 1/4 and 1/2 wave on the lowest band.

You will need a wide range tuner.  Don't scrimp on it.
A balanced tuner is also nice.

If you can get a good ground you can use a single ended tuner.  If you have a balanced tuner, you can cheat and connect one of the balanced outputs to the wire and the other to whatever you are using for a ground.  The "ground" wire can be long, but it will have RF currents on it and will radiate.  The balun in the tuner will have provide the isolation to your unbalanced coax to the rig.

This will result in a relatively efficient antenna, so you will be able to run QRP effectively.  The low power operation will help keep you out of the neighbors consumer devices and maintain your low profile.

Don't waste your money on really short loaded antennas.  Especially don't waste the money on loaded verticals with very poor ground systems.  They are expensive dummy loads.  Also every non-ham knows an antenna looks like a whip or a CB antenna and they will immediately blow your cover.  Every time their TV flinches it will be your fault.

Put money into a good tuner. One that will load a variety of loads.  Then start stringing wire.

You might try varying the AGC time constant in your RX or the RF gain.  CW is on / off keying and when the off part happens the output of the RX is noise.  If you can get the on time to have more RX output than the off time, you get less false triggering.

There is not much the computer can do with a crappy fist that has poor intercharacter and interword spacing.

If I recall CW get has a display of the received waveform and the thresholds applied. You can look at that and see if you can fiddle with RF gain and AGC to get better copy.

After playing with CW decoding you understand why RTTY has both a mark and space tone sent.  It improves decoding dramatically.

Nobody seems to have mentioned the fact that the 43 foot vertical is not resonant like the usual 1/4 trap vertical.  Therefore the current distribution on the so called raidials will not follow the usual pattern.  There will be a question as to whether they decouple the feed line and provide the "ground plane".

My guess the 43 foot idea will result in a lot of common mode and a crappy radiation pattern.

If you insist on mounting a vertical on the roof, a tried and true way to make it work is with a trapped 1/4 and tuned radials.

Otherwise, hit the EZNEC and have fun screwing around with your 43 foot piece of pipe on the roof.

Consider loading up the feedline to the chimmney as a random wire.  Use a tuner and work it against a cold water pipe.  It will work better than your taktenna. I'm guessing you will run over 20ft of line to the antenna from the shack.  That length of wire fed at the end against a ground or counter poise will be a better radiator than the taktenna.  

I encourage you to try this after you try your taktenna.  You should not go by eham reviews.  They are subjective to put it kindly.

Unlike some of the posters, I have done it in the past.  I have used a 1/4 vert on 40m and end fed inverted L on 80m.  It does work and you wont electrocute the neighbors.  I have know idea if it was very good, but I did not have the time, inclination and room to put in the "proper" radials.

Some posters on eham over analyze everything.  Its not optimum.  Its better than a single crappy ground rod, worse than radials but has the virtue of quick and dirty.

The posters are quite right to point out the "I can work anything I can hear" testimonials and log books "full" of DX are BS.

Read the manual on how to use the rotate command.  Rotate the wires after entering them as a straight horiz dipole.  If I recall you select the end you want to be the rotational axis first.

The alternative is direct coordinate entry via trig. calcs or the suggested graph paper.

You're gonna have trouble if you don't remember your high school algebra and trig.