I have been trying to understand how to ideally handle the cable shield
on a USB device. (Full Speed USB, in this particular instance.)
As seems to be the case with many signal integrity issues,
contradictory recommendations abound, each with its own unsupported
claims. Even authoritative-sounding sources such as Texas Instruments,
Intel, FTDI, and Cypress Semiconductor seem to disagree on the correct
way to handle the cable shield on USB devices.
Contrary to my initial supposition, the purpose of the USB cable shield
is not to protect the USB data lines from outside interference, but
rather to prevent the USB device from radiating EMI.
Here are some of the options that have been recommended.
Note that (2)series capacitor to pass high frequencies onlyseems to
directly contradict (3)series ferrite bead to block high frequencies
(1) Connect shield directly to signal ground.
connector shell be tied to the ground plane.
Intel. EMI Design Guidelines for USB Components. Sec 5.4 (p. 9).
(2) Connect shield to signal ground through a capacitor.
(Possibly with high-value parallel resistor approximately 1 Mohm.)
FTDI. Debugging FT232BM and FT245BM Designs. Section 3.2 (p. 11).
Steve Kolokowsky & Trevor Davis (Cypress Semiconductor). Common USB Development Mistakes You Dont Have To Make Them All Yourself! Figure 7 (p. 7).
from the ground plane to the shield, turning the USB cable into
an antenna. To limit the USB cable antenna effect, it is
recommended to connect the shield and ground through an RC
filter. Typically, R = 1MΩ and C = 4.7nF in Figure 3-5.
Atmel AVR1017: XMEGA – USB Hardware Design Recommendations. Section 3.3.3 (p. 8).
(3) Connect shield to signal ground through a ferrite bead.
USB connector socket to keep EMI from getting onto the cable
Texas Instruments Application Report. USB 2.0 Board Design and Layout Guidelines. Sec 2.2.4 (p. 3). SPRAAR7 December 2007.
(4) Do not connect cable shield to ground on the device at all.
not connect the shield to their own ground.
Hardware Book. Universal Serial Bus: Shielding.
Whether or not the device has a metal chassis, and the handling
of chassis ground and signal grounds, (as well as how the USB cable
ground is connected to either one) is certainly important as well, but
this isn’t clearly discussed in most of the writings on USB cable shield
The device I’m developing is a bus-powered device which will likely be
in an unshielded plastic enclosure.
Thanks in advance for any bits of wisdom on this topic full of
contradictory information. I recently posted this question on si-list,
and even there I got little in the way of answers.
Previously I wrote about the bring-up of the MRF49XA Shield. Recall that instead of the standard FSK spectrum I was met with an asymmetric comb about the carrier.
Decreasing the span we see are able to measure the comb spacing, about 80kHz.
When viewed in this way, we see more of a traditional harmonic series from the carrier. I’ll assume for now that the series is not an odd function; the waveform is not square or compressed. Typically symmetric sidebands that reside xdBc from the carrier are indicative of A.M. distortion. I’ll also chalk up the asymmetry seen in at wider bandwidths is due to heterodyning.
The first place I like to look when experiencing undesired A.M. spurs is the power supply.
Setting the DMM to AC mode and turning on the frequency counter, I find a dominant 78kHz AC component on my 3V3 supply.
That’s close enough for me to be a smoking gun. A bypass capacitor is placed near the MRF49XA supply…
And the major sidebands are gone. We can still see spurs about 35dBc at 200kHz; for now I’ll assume they are Frac-N PLL spurs and move forward.
Radio Frequency Interference (RFI) can be described as any unwanted radio frequency signal, which interferes with another, desired signal. This can show up as an annoying popping or crackle on a loud-speaker, unwanted pulses on a loran, marks on a sounder paper or spots on a radar screen. The main concern is undesired noises generated by the ships electrical system which interfere with the radio equipment.
There are two ways for the noises to enter the radio. They can be CONDUCTED along the power leads and other wiring into the radio, or they can be RADIATED from the noise source and its wiring into the radio antenna system, and then into the radio. Filtering the power wiring into the radio can eliminate conducted interference. Most marine radio equipment has adequate filtering built in to eliminate conducted interference. Radiated interference is the most serious problem normally encountered on vessels. The radio antenna system and receiver have been designed for reception of very weak signals, often as low as 2 or 3 micro volts (millionth of a volt). Unfortunately the receiver cannot discriminate between these weak signals and random noise if both are on the same frequency, and if the noise has comparable strength to the required signal, then the signal is completely buried in the noise. Radiated noise is caused by any abrupt change in current flowing through a conductor. Current flowing creates a magnetic field around the conductor. When this current is changed this causes a change in the magnetic field. The changing magnetic field travels as an electro-magnetic wave for some distance from the conductor, and it is this traveling field which is radiated to the antenna.
Changes in current can be caused by opening or closing a switch which causes one sudden change in the field, and shows up as a single pop from the radio speaker. A more serious situation occurs when the current is regularly being changed as in a motor where the current through the brushes and commutator is being continuously interrupted while the motor rotates. This produces a continuous chain of pulses and is heard as a buzz or whine from the speaker.
Another source of noise is arcing caused by a build-up of static electricity. Although there is no metal conductor where the arc takes place, there is a sudden change in current through space. This change of current causes a magnetic disturbance which shows up as a noise on the loud-speaker. Some examples of this are shaft noise and rain static. This problem can also develop where there are poor contacts between various parts of the ships rigging or other metallic components.
RADIO INTERFERENCE SUPPRESSION TECHNIQUES
The first problem with noise suppression is to identify and locate the source of the interference. On a ship, noise sources can normally be traced by listening to the receiver and switching on various pieces of equipment while noting any change in noise level from the loudspeaker. Some noises may show up only with the engine running, or when the engine is running in gear. Noise from rigging and other metal parts of the ship can be identified by listening to the receiver while various parts are moved. These noises will show up as crackles or pops from the speaker when the offending parts are moved. As one noise is eliminated others may be heard which were drowned out by the first noise. Therefore noise suppression can be a time consuming task.
There are three basic methods of reducing RFI. The first is to prevent the radio interference from reaching the antenna by shielding. If the noise source is totally enclosed in a metal can, then the noise is contained and cannot reach the antenna. An example of this is the ignition system of a car which is enclosed in the metal car body. While there is some leakage of radio noise, the amount is considerably reduced.
On aluminum and steel ships the noise problems are generally less than wood and fiberglass ships because of shielding. The shielding is only totally effective if it is electrically bonded to ground, and there are no unfiltered wires passing through the shield. Wiring through the shield can act as an antenna, picking up the noise inside the shield, conducting it outside and radiating it to the radio antenna.
The second and more practical method of noise suppression is to eliminate the noise at the source. This is achieved by installing filtering or smoothing circuits across the noise producing device or contacts. A capacitor across a contact slows down the rapid change in current when the contact is opened.
The slowed change in current through the conductors results in a weaker magnetic field around the conductor and therefore less radiated interference. The above example refers to one contact being opened.
The same principle applies when current is being rapidly and continuously interrupted as in a motor commutator. In a motor the common practice is to put capacitors from each brush to ground or the frame of the motor. Here the capacitor serves the same purpose as with the switch; it slows down the rate of change of current through the brush-commutator contacts, and therefore reduces the poise radiated from the motor wiring.
In extreme cases, it may be necessary to install choke coils in the power lines to the offending device. A choke, which is a multi-turn coil of wire, has a similar effect to a capacitor, that is it slows down a rapid change in current through a conductor. Normally a choke is used with capacitors to provide the most effective filtering.
The third important part of noise suppression is bonding. Bonding or grounding should provide a low resistance path to ground for any radio frequency noise which is present. This path ensures that the noise is conducted away to ground instead of being radiated.
All motors and other electrical apparatus on a ship should be bonded to ground. The bonding conductor should be at least No. 8 copper wire. If the ground lead must be long (10 ft. or more) flat copper strapping should be used. Under no circumstances should copper braid be used for long ground leads as its radio frequency resistance increases rapidly when it becomes corroded.
All bonding should tie to the main engine or engine bed, the ships ground strap which ties the engine and zincs together, or as close as possible to the through-hull bolt for the radio ground plate. If the bonding must be connected to the radio ground, it should not be connected more than three feet from the through hull bolt. Any farther than this can cause the noise to be conducted up the ground strap and into the radio.
When the bonding wire or strap is connected to the equipment, care must be taken to ensure that a permanent low resistance connection is made. Painted surfaces must be cleaned to bare metal, and the bolt securing the ground strap should have a lock washer to ensure a permanent connection. Wherever possible the connection should be soldered. A poor connection can create more noise than no connection at all.
RADIO INTERFERENCE SOURCES AND CURES
Alternator interference can normally be recognized by its characteristic whine when the alternator is charging. The whine will vary in pitch with engine speed and vary in loudness with the charging current. If there is any doubt about the origin of the whine, the belts can be removed from the alternator and the engine run again without the alternator.
To cure alternator noise, which is pulses radiate from the output lead, it is necessary to filter the output lead as close as possible to the alternator. The most effective filter is a 0.5 microfarad coaxial capacitor. A coaxial capacitor is one which passes current through its center with the capacitor completely surrounding the current carrying lead. As the alternator current flows through the capacitor, it should be rated to handle the alternator’s maximum output current.
Capacitors may be used in parallel if necessary to increase current rating. The capacitor should be securely mounted on the alternator and a short heavy lead run from one end to the alternator output. The output lead from the alternator should be removed, and reconnected to the other end of the capacitor. Ungrounded electrical systems will require two capacitors, one on the negative and one on the positive output terminals. NEVER connect a capacitor to the field terminal of the alternator.
In severe cases of interference, it may be necessary to shield the alternator wiring. The lead from the regulator to the field terminal of the alternator, and the alternator output lead, should be enclosed in a woven shield braid. Both ends of the shielding should be grounded.
Small Electric Motors
Shipboard motor noise can normally be identified by listening to the receiver while switching various motors on. All electric motors on board should be checked including the following: fridge motor, freshwater pump, electric toilets, or toilet pumps, stove fans, ventilation fans, electric windshield wipers – in all cases where motor noise is experienced, the motor should be checked for possible problems. The wiring to the motor should be securely connected to prevent intermittent connections and arcing.
The brushes must be seated on the commutator check brush length and spring tension. The commutator should be smooth and clean—a light polishing with sandpaper will often help reduce noise.
Motor noise can generally be cured by installing capacitors on the motor. The noise is mainly generated by arcing at the brushes so the capacitors should be placed as close as possible to the brushes. On larger motors it is often possible to connect directly to the brush holder connection, and mount the capacitors inside the motor or outside adjacent to the brushes. Metal can 1.0 microfarad type suppression capacitors should be used.
On smaller motors where it is impractical to connect directly to the brushes, the input leads to the motor should be bypassed with 1.0 microfarad capacitors from each lead to ground. In addition to the capacitors, a good ground connection must be installed to provide a low resistance path to ground for the noise. This connection must be at least No. 8 wire and where the distance is greater than 10 feet, the motor should be bonded to grounded copper pipes or a one inch copper strap should be used.
Where the above methods do not cure the noise, coaxial capacitors may be required in the supply leads to the motor. In extreme cases a filter unit consisting of choke coils and capacitors may be used in the power leads.
Shaft noise is easily recognized as a slow rhythmic rushing sound when the shaft is turning slowly. It will often show up more in one direction than the other. The noise is created when static electricity is built up on the shaft due to its turning in the bearings. The static voltage arcs across the gap between the shaft and the bearing creating radio noise.
To eliminate shaft noise, it is necessary to short out the static on the shaft. A spring loaded brass or phosphor bronze grounding brush running on a clean part of the shaft as close as possible to the bearing, and grounded to the bearing should eliminate any static. The brush and bearing should be bonded to the ships ground system.
Due to the environment around the shaft, carbon brushes are not suitable for shaft grounding. The carbon soon becomes fouled with oil and water which can insulate the contact between the brush and shaft. Soft metal should be used to avoid any wear on the shaft.
If the brass to shaft contact is not adequate or becomes contaminated with oil, a “Chore Girl” can be wired onto the brass brush. The copper Chore Girl makes good contact with the brass and, due to its abrasive nature, cuts through the oil film to make good contact with the shaft. Adequate pressure should be applied to the Chore Girl to ensure a reasonable life before the Chore Girl must be turned over or replaced.
Gas Engine Ignition Systems
The most troublesome source o f interference on small gas driven vessels is the engine ignition system. The release of energy across the electrodes of a spark plug involves very sudden changes of voltage, and current in the ignition wiring. This causes a popping noise at low speeds, and a loud buzz as the speed increases.
Most of the radiated noise originates from the high voltage leads to the spark plugs. Resistance in the high voltage circuit considerably reduces the high frequency interference generated. The best place to put the resistance is as close as possible to the spark plug. Resistor spark plugs must be used as a first step in reducing ignition interference. In addition, resistance type high voltage wiring will further reduce the interference. The use of both resistor type plugs and resistive spark plug wire will suppress the noise more effectively without producing detrimental effects on engine performance or spark plug life providing the proper spark plug heat range is selected.
Any deterioration in the ignition system can completely defeat the advantages of resistor plugs and wiring. The ignition system must be periodically checked for such things as: loose connections between wiring, plugs and distributor; excessive wear on the distributor rotor; dirt or cracks in the distributor cap; dirty or fouled plugs.
In addition to the high voltage secondary circuit, the low voltage primary circuit can add to the noise problems. The main offender in the primary circuit is the breaker points. Noise generated at the points can be conducted and radiated by the ignition primary lead. As this lead generally goes through the vessel to the key switch on the dashboard, the noise must be eliminated from this wire. The most effective way is to install a 0.5 microfarad coaxial condenser as close as possible to the ignition coil primary terminal.
Besides the ignition lead going to the ignition switch, there is often a tachometer lead from the ignition system. As many tachometers depend on pulses from the engine for there operation, it is not practical to filter the pulses from this line.
Where an ignition operated tachometer is used, the wiring to the tach must be shielded. This can be done by replacing the existing wire with shielded wire or pulling a copper braid over the existing wire. The braid should be securely grounded at the engine.
The above techniques will considerably reduce ignition interference. If a further reduction is required, the ignition system must be shielded to prevent radiation of the remaining noise. This can best be accomplished by using a shielding kit on the entire ignition system. These kits are available for most engines and consist of shields for the plugs, shield braid for the high voltage leads, and metal shields for the coil and distributor. Ignition shielding kits can reduce the high voltage available at the spark plugs. In this instance it may be necessary to install a higher voltage coil.
While the shielding kit may provide good suppression when installed, the performance will deteriorate with age. When we shield and connections become corroded, there are many points where radiation may occur. The effective life of a shielding is probably no more than two years. This plus the cost of the kit ($150 -$200) makes them useable only in the most extreme cases.
Another effective method of shielding is to enclose the entire engine compartment in a metal screen cage. The screen must completely surround the engine and all leads into the engine compartment must be filtered with capacitors.
RADIO INTERFERENCE SUPPRESSION COMPONENTS
Radio Suppression Capacitor one microfarad, 100 volts min. Sprague Type AR1 or Aerovox Type 1120 OR Radio Suppression Capacitor 0.5 microfarad, 100 volts minimum
Aerovox Type 1140 (0.5 microfarad capacitor will not be as effective as a one microfarad capacitor and should only be used when space is restricted) Coaxial Capacitors
Coaxial Suppression Capacitor 0.5 microfarad, 50 volts DC 40 Amps Cornell Dubilier Type NFF 055B1 OR Sprague Type 48P18 60 Amps Cornell Dubilier Type NFT 877 OR Sprague Type 4SP100
Why cooling is important
Electronic chips generate noise and heat during use: your digital camera or smart phone gets progressively hotter the more you use it. Sensitive applications, such as astrophotography MUST eliminate this heat, as heat is noise, black sky will be grainy black sky.
This heat is also present with the RTL stick, appears as frequency drift and internal noise – radio will not stay on a frequency.
Gain applied in Configuration panel increases heat, hunt weak signals and need to increase gain = increases internal noise generated by the stick.
Improve cooling – reduce noise.
Recently introduced temperature-controlled oscillator promises to solve this issue, around 50 dollars. Regular RTL stick sells for 8 USD. Hmm.
Do you need to worry about it?
No: Leave the stick warm up for 15 minutes – half an hour, let it reach thermal equilibrium, after stick is warmed up set ppm correction. My particular stick drifts 4 ppm from cold to warmed up, then stays on frequency.
Yes: You are a perfectionist, or you want less internal noise with high gain settings.
Performance increase with modifications below will be really, really small, less noise on shortwave, maybe one or two less noisy lines on WXSat images.
The brain of the stick lives under the plastic cover, open up the stick and drill holes, especially above the black squares, or leave the board bare without its plastic cover.
Stick will run cooler, frequency drift will be reduced somewhat.
Heat-sinks are a possible solution, reliable connection is hard to achieve due to small surface area, edge of the chip can break off.
Additional airlow with PC fans, run on 5V, telephone chargers can be used. Two PC fans installed outside a PC power supply case, with stick in the middle provides a cheap air-cooled solution.
Using fans have the following drawbacks: 1) Needs external power source, 2) magnets in close proximity to the stick, 3) the awful sssssssh noise fans generate.
Electricity plus water equals disaster, insulator (hot glue or epoxy) must be used to seal electrical components from water. Tried several methods back in the Pentium 4 days, messy, expensive, hot glue won’t dry evenly, epoxy eats components: water cooling is unsuitable.
Place stick into metal can, with antenna and USB cable connected and led through lid of can, position stick middle of can, slightly above bottom.
Pour sunflower or vegetable oil into can with the RTL stick inside, completely fill up can.
Repeat: fill up can as much as you can, any interior surface will pick up condensation.
Cover with lid, plug in USB and connect antenna, fire up SDRSharp.
Oil will provide cooling, it is not conductive, see stick blue light on image below.
Place can in the fridge or freezer: sunflower oil theoretically has a very low freezing point, practically after one night at -20 degrees Celsius can will be very cold to touch.
Active heating is unnecessary, oil volume and metal can surface will be enough to dissipate heat.
Cherry on the cake: FM Noise Reduction
Metal can also acts as a Faraday cage, FM broadcast pickup will be greatly reduced.
Traditionalists love to place stick in metal enclosure for the same reason, not realizing that heat buildup will effect performance – a chip running at a very low temperature will have lower internal noise and greater performance.
Cleaning stick after oil cleaning
Stick should not be connected to computer or any other electricity.
Lukewarm water in sink, spoonful washing powder or cup of liquid detergent, the more the better, dissolve until glassy reflection and feel of glove on hand. Pour liquid detergent on stick/ place stick and USB cable in washing powder, wash off oil and detergent with lots of water.
Repeat steps above 2-3 times, wash hands, clean with towel / toilet paper.
Leave stick to dry for a day. small heat, not close to radiator, never on top of a heater.
Will be good as new.