Because I drive a 1999 Ford F-150, this page will focus on my own experiences and others' Ford-related experiences that I have found useful.  However, these ideas are likely generally applicable, in varying degrees, to any make/model vehicle.

FORD VEHICLES HAVE A PROBLEM WITH THEIR FUEL PUMPS!

ALL Ford fuel-injected motors (which I believe is all motors since 1991) have an in-tank fuel pump that is a disaster on HF.  Guaranteed to radiate S5 on 10m to S9+who cares anymore on 40m and down.  Fortunately Ford recognizes this and makes an electrical filter that seems to work 90% of the time (I don't know why 10% don't, but some Usenet searching shows some failures-to-fix).  The Technical Service Bulletin (TSB) is 98-7-3.  The link is a scanned copy of the 1998 version of the TSB, which actually covers some 1999 vehicles.  Ford updates this TSB annually to account for the continuing carnage into the next model year (sigh...why not solve the problem?)   The filter install requires dropping the fuel tank, so unless you're a gearhead too take it to the dealer to have it done.  In-warranty vehicles should get the service, parts and labor, gratis.

The test for fuel pump noise:  turn on the ignition to the Accessory position BUT don't start the engine.  Listen to your radio.  If you hear noise when you turn the key to "ON" that lasts for a second or two, then quits, it's the fuel pump.  If not, LUCKY YOU!!

If you're out of warranty, or don't want to deal with the dealer (and I wouldn't blame you one bit), there are (at least) two alternatives available.  I have not personally tried these alternatives, but they seem reasonable.

1. Identify the fuel pump power harness.  On the 1999 F-150, it goes to the middle-top of the tank, runs laterally to the driver's side, then up along the left frame beam to the engine compartment.  On my vehicle it can easily be seen by looking into the rear wheel well on the passenger side (but my truck is also 4x4 with 17" wheels, so the body sits up well from the undercarriage).  Install a few snap-on ferrite chokes (see Toolkit below!).  Remember that ferrite=iron and iron+water=rust, so do what you can to shield the chokes from the elements.
2. Supposedly the primary problem with the fuel pump is that the ground lead runs the entire length of the body (it runs along the entire cable bundle to the engine compartment; the ground wire actually attaches at the very front end of the left frame beam, an easy 15' wire run)!  If you can manage to reground this wire much closer to the pump, that 15' "antenna" will be greatly diminished.  The TSB makes specific mention of this; if you do have someone do the filter install for you, make certain they follow the instructions regarding the ground wire.
3. Filter Cons:  In Jan/Feb 2000 QEX, so-and-so has an article about vehicle noise suppression that mentions a mysterious device the author calls a "Filter Con".  I don't know if this is standard terminology, but I've been in electronics for 15+ years and have never heard of one...but basically what they are is small low-pass filters built into a solder-in or bolt-in package that looks very much like a feedthru capacitor.  Most manufacturers refer to them as "coaxial EMI filters" or something like that.  The specs are impressive:

Ignition Noise and "Grounding" your Vehicle parts

An interesting thing about the 1999 F-150 is that it is a distributed low-voltage ignition system.  There is no coil/distributor setup, but rather each cylinder has its own coil pack right on top (at least on the V-8's...the V-6's have one coil pack per cylinder pair).  This has the advantage of the system being +12VDC all the way up to the spark plug (practically); there is no point in changing high-voltage spark plug wires or distributor caps/points/rotors because there are none!

One nice thing to see also is that Ford has a "Radio Interference Capacitor" for both sides of all engines (one for each row of three or four cylinders) shunting the coil packs to the engine block.  This is factory standard installed equipment.  The capacitors are located way back in the engine compartment nearly to the firewall; they are very hard to get to.

I got curious one day and took one of the Radio Interference Capacitors off the block and onto my Boonton 160-A Q-meter (with standard inductances) to measure its impedance at RF.  I'm happy to report that the Radio Interference Capacitor's  RF impedance is <2 ohms from 3-30 MHz.  I'm not happy to report the ~12" small gauge pigtail it's integrated with...

With the low-voltage distributed ignition, fancy plug wires (Magnecore, etc.) probably won't do anything beneficial.  Utilizing 1/2" shield braid around the wires may still be of benefit.

I've read on the Usenet that some have had ignition noise problems specifically with "split-fire" style spark plugs.  I don't know what mine are but I'm told that almost all recent engines use them.  Could be that swapping out for conventional resistor plugs will help.

The general advise on Ignition noise is to bond everything...hood, doors, exhaust pipe, tailgate, bumper...anything not welded to the main chassis (that is, the chassis that is the primary return path for the ignition system.  In a pickup, then, the pickup bed should be added to the list as only the cab chassis is in the ignition return path).  Wide braid is best, since it can withstand the vibration and flexure in vehicle operation.

Inspect your vehicle closely.  On the 1999 F-150, there are several attempts at factory bonding:

• Left frame member to cab chassis (approx. 10AWG rubber moulded)
• Right frame member to cab chassis (two places, no less...one ~10AWG rubber moulded, one 1/4" braid)
• Engine block to cab chassis (1/4" braid)
• Battery (-) to cab chassis (~#12AWG)
• (by default) alternator to engine block (through the metal bodies)
• Hood to cab chassis (1/4" braid, behind the hinge on the passenger side...look WAY back there and you'll see it)

Don't be afraid to try strange stuff...anything that is metal should be bonded to the chassis.  One fellow had great success bonding his (metal) fuel line to chassis.  Many report significant improvements with exhaust system bonding or hood bonding.  Grab a bag of materials and have at it!

Grounding your Equipment

From my experience, the #1 everyday diagnostic tool you can have in your mobile shack is a Peak Reading wattmeter.  I heartily endorse my Autek WM-1.  Keep it inline and on Peak setting.  Why?  Read on...

My original HF-only install (after the SG-500 amplifier was added) had a "single-point" ground, from the amplifier chassis to vehicle chassis.  My radios and amplifier are all mounted in the center of the front seat (over the "transmission hump") and there is easy access to chassis ground right in front of the transmission hump (there is a small black plastic cover that is easily removed without tampering with the dash as a whole.  Even an "official" green-screw grounding point on the passenger side).  The "single-point" ground used 1" tinned braid, about a 5" length from the amplifier chassis to the ground screw.

Then I added all the VHF and satellite equipment (70cm preamp and filter; 2m brick amplifier, 6m preamp and filter), all underneath the rear bench seat, plus the IC-290H to the front console.  Ever since then, I would observe that on certain bands, and with varying frequency, the peak power on SSB would not show 500-600W, but appear "damped" to only around 100-150W.  If I whistled or hummed into the mic, the output would go up to full power.  This happened usually on 10m, always on 12m, sometimes on 15m, more often than not on 17m, and never on 20m or lower.  This "damped" response only happened with the amplifier on; with it off, I always got >90W peak in SSB.

Curious if the wattmeter just couldn't take the "heat", I pulled out my SM-220 station monitor from the home shack and hooked it up to monitor amplifier linearity (the "trapezoid" test).  The trapezoid showed the classic overdrive envelope, but (1) the amplifier was being driven with only 50W, well within its specifications, and (2) decreasing drive did nothing to the shape of the waveform (of course decreasing drive changed the size of the trapezoid).  Reconfiguring for a simple oscillogram, I monitored a string of dashes, and saw a most interesting waveform:  instead of nice squarish pulses, there was a very exaggerated logarithmic rise time.  The signal would make it to full power output, but only during the last 10% or so of the keyed dash.

Obviously something was damping the transient response of the entire system with the amplifier active.  Since the syllabic nature of SSB is much faster than CW, the SSB output was equivalently taking place well down the logarithmic rise time as displayed on CW.  That's why the peak-reading WM-1 read low on SSB, but not CW; and the meter averaging was long enough in Average mode for it to respond to the full power part of the output waveform.

Also, particularly when working Mode J satellites, I would experience plenty of "hash" on receive when I transmitted.  This is nothing new to Mode J (see the Mobile Satellite page), so I built and installed a 70cm narrow-bandpass filter.  While this helped, it didn't clean up all the trash.  It was hard to tell if what I heard was my transponded signal horribly distorted, or just crossband desense in the receiver.  And then suddenly he RX and audio would clean up and all was well, so I don't know even to this day if there was a real problem (that is, a problem that others could sense).

Solution:

I suspected that the lack of local grounding straps on the VHF and UHF equipment in the back might be contributing to some of the problem:

• All three bands (6m, 2m, 70cm) have bandpass filters as the final element before the antenna, but filters need a good ground reference to perform their filtering function.  That is, any common-mode pickup of RF on the outside of any coax shield will pass right over the filter (chassis or body) with no attenuation.
• Two bands (6m and 70cm) are connected to my HF radio, so any common-mode pickup on HF transmit would find its way back to the HF radio (feedback loop).

Isolating Your Accessory Runs

Grounding Materials (or, the "Grounding Toolbox")

Which to use, strap or braid?

In my readings I have seen different parties state that (braid/strap) is better for grounding at RF.  Well the Q-meter needs more work so let's check it out.
Measurement samples:  13" long pieces of 1" wide braid, 4AWG wire, and 1" wide Cu strap at 28 MHz.

• 4AWG        193pF to resonate     170nH        X=30 ohms        Q=100
• strap            210pF to resonate    150nH        X=25 ohms        Q=150
• braid            273pF to resonate    120nH        X=20 ohms        Q=100

Hardware

• Always use stainless steel hardware (screws, washers, etc.) for attachments to the vehicle body.  Star washers too if available.  If the packaging doesn't say "stainless" then it isn't.  Yes, stainless is significantly more expensive.  Ask for it!
• Stainless hose clamps in all sizes are handy for attaching braid to tubing/pipes that you cannot drill (exhaust pipes, etc.)
• Plenty of braid or strap...braid is flexible, but absorbs water.  Strap has higher Q, but doesn't do nearly as well long-term under vibration etc.  Make your own strap out of flashing copper or brass--check your local yellow pages under "Metals", "Surplus", or "Salvage"...or a hobby shop or old-fashioned hardware store (Ace, True Value, etc.).  Note that new metal is significantly more expensive than salvage, and performs no better...Braid you can get yourself from old coax (although that is also very expensive, if the coax could be used instead as coax), or you can buy braid from shops selling antenna supplies (e.g. cablexperts, RFConnection, etc.).

• A small rotary tool (Dremel or equivalent) is exceptionally handy at cleaning painted metal down to the shine.
• Your hands and wrists will appreciate an electric screwdriver

Many recommend finishing off a bonding point with some sort of rubber or silicone sealant to further keep out weather.  Also, you should use a conductive anti-oxidizing compound within the joint.  There are different types, depending on if aluminum is involved (must be some chemistry there...).  RFConnection sells good copper-based goop (they also sell aluminum-based as well).  Aluminum goop is readily available at the hardware store as "NoAlOx" or similar--look in the electrical section as it's used in industry to prevent corrosion in conduit joints.

Ferrite Chokes

Know Your Ferrite Mix!  For 1-50MHz noise/interference suppression, you want to use Mix 31 toroids/beads.  It has been my experience that the majority of commonly-available surplus ferrite material is Mix 43 or 61.  Mix 43 and 61 are great for *tuned* HF circuits but are suboptimal (but usable) or useless, respectively,  for HF noise suppression.  The same applies for the cute "snap-together" beads--most of them are Mix 43.  The rectangular Radio Shack ones are Mix 61--utterly useless at HF!  If you're not sure, here's how you can tell:

• Break out your Impedance Analyzer (Autek, MFJ, AEA, whatever.  I personally endorse the Autek but it's also the only one I've used)
• Slip your BUT (bead under test) or TUT (toroid under test) onto a small length (about 6-8") of insulated wire of moderate gauge
• Solder the wire into a small loop onto an SO-239 panel socket (i.e., one end to the center pin, the other end to the flange)
• Connect to your Impedance Analyzer and set the frequency to 20MHz

If you're wrapping a toroid, be aware that distributed capacitance between windings is a factor.  Wrap no more than 3/4 of the toroid (if you wrap more, the two leads can be close enough together, and have enough shunt capacitance, that the toroid/coil is essentially out of the circuit), and maintain as much spacing between turns as possible.  Both R and X (or Z if you prefer) go up as the square of the number of turns, so it doesn't take many.  Again, use your Impedance Analyzer to assess the situation ahead of time.

 Amidon Inductive has good technical information and probably the best one-piece price for ferrite materials.  The one complaint I would have is their website has no mention/information of Mix 31, although they sell it.  Fair-Rite has extensive technical material on their website, and a very broad line of materials; I believe they are available via special order through Amidon.

Also, note that ferrite chokes are most effective in low impedance circuits.  See the Fair-Rite website above for a discussion.

Some Ways to Measure and Assess your noise issues

Instruments:

1. Good old oscilloscope.  Doesn't have to be too fast (my Tek 485 definitely not put to the test here!).  Put your radio in AM mode (no filters, widest possible selectivity) and pipe the audio out to the 'scope.  Rev the engine, etc., and observe the waveform amplitude and period.
2. Spectrum Analyzer.  Monitor the actual RF picked up by your antenna with the engine running.  This type of measurement requires a good understanding of how a spectrum analyzer works, though...it is not straightforward.  And it requires a "real" spectrum analyzer...not a $200 "kit conversion" or a service monitor!
3. A computer with sound card!  Really, there are some excellent Audio Spectrum Analyzer applications available for sound cards.  Why not pipe your AM-demodulated audio right into the sound card and look at the spectrum?  If there's an RPM dependence you can relate it to frequency.  An added plus might be if you have multiple sources you're tracking down and you can separate them in frequency (maybe one whose frequency changes with engine RPM and one that doesn't).