Troubleshooting

When there is a problem it's important to proceed in this order:

• Cockpit Error - There really is no problem with the hardware.  You are expecting something unrealistic, the hardware can not do what  you expect, you have improperly configured the hardware, etc.
• Bad Connection - Cycle all controls a number of times.  Apply contact cleaner and cycle controls.  Disconnect and reconnect all connectors and mechanical joints to burnish their contacts.  This solves 90% of the problems at this stage.
• Look for common failure modes.  So far shorted capacitors, especially Tantalum types have been the most common failures I have seen.  Also look for semiconductor devices that test as a short between most if not all terminals.  This can be done using just a DMM in the Ohms and Diode modes of operation.
• In very rare cases do you need to use more advanced techniques.

What Goes Wrong

When Electronic Equipment is unused for years oxidation forms at the contacts.  This adds resistance to the joint and can cause the equipment to be DOA.  When I assembled the 2102 static RAM chips for my SWTP 6800 computer kit there was an instruction to NOT use sockets because the reliability of the socket was lower than the chip or solder joint.  By using sockets you would be creating a problem.  The more reliable sockets are Gold plated to avoid oxidation.

Below are some real examples.

HP 204B Oscillator

When received from eBay the oscillator did not function.  A check of the power supply showed that it was working fine.
After spraying the OFF-Range switch with Tuner Control Cleaner & Lubricant (Radio Shack - 64-4315) the unit started working.
The problem was oxidation on the switch contacts.

R-1004/GRC-109 Receiver

The receiver was DOA upon receipt from eBay.  A check of the power supply showed the correct voltages.
The resistance of each tube filament was checked when it was out and the tube was installed and removed from it's socket a few times.
All the filaments were about the same resistance.  The most common failure mode of tubes is that the filaments burn out and then will test as open circuits.  After assembly the receiver worked well.  The problem was oxidation between the tube pins and the socket.

Quantic Q-5200/SM GPS Timing Receiver

When received the LCD display backlight came on and the display showed 2 rows of all black 5x7 dots and two rows were not lit at all.
After removing the two main printed circuit boards and reassembling the receiver normal operation was restored.
The problem was oxidation between the one or more cable connectors and the mating contact.

URC-68 Air Crew Survival Radio

When received the radio was DOA.  After removing and reinstalling all the modules in their sockets a few times each the radio worked.
The problem was oxidation between the module pins and the sockets.

OE-254 Antenna

This antenna was installed quickly to do some tests on PRC-68 Family Radios.  When taking it down it was almost impossible to separate the mast sections.  The copper had oxidized inside the screw thread joint and much heat and force was needed to get them apart.  After polishing the male and female copper threads Radio Shack Lube Gel 64-2326   US 5037566 - Lubricating composition and method for making same. "...provides high corrosion resistance due to the absence of dissolved air and moisture.."] was applied to the joint before mating the antenna sections and Coax-Seal (Radio Shack: 278-1645) was used to seal the open part of the joint.

Lube Gel was also applied inside all outdoor RF coax connections to keep water out of the joints and prevent corrosion.
It may be that using Lube Gel in all of the above cases would prevent a future occurrence of the same problem?
Lube Gel appears to be a modern snake oil that can prevent many ailments.

PRC-77 Radio

After receiving this used but operational radio I cleaned it up using Windex, a toothbrush and a Rag-in-a-box both outside and inside.
As part of this process I found a number of screws that were either very loose or about to fall out.
The lesson is to give a close visual examination and that can best be done by cleaning.

HP 241A Audio Oscillator

The front panel had been bashed in.  This broke a plastic part that holds the radio buttons used to select frequency.  After removing and disassembling the front panel the plastic part could be removed and epoxied back together using J B Weld.  The neon power indicator was held on with one of those flat metal spring with teeth.  The plastic part of the indicator was broken by whatever hit the panel so the only thing to do was break the plastic behind the panel.  Hot melt glue attached the front of the plastic to the panel and holds the neon tube where it was.  From the front it looks original.  All of the button caps were removed and cleaned.  They had what may have been tobacco smoke obscuring the numbers.

TA-838 Mil Analog Phone

When received the phone did not work.  Opened up the case and disconnected the internal connector. Visual and olfactory inspection showed nothing amiss.  Reconnected the internal connector and reassembled.  Now phone works.  the connector must have had one or more poor joints.  Note this is a Gold plated connector.
It worked for about a year and it's dead again.

PP-8444A/U Battery Charger arrived DOA from eBay

Was sold as looking good and "as is" so no complaint.  After some troubleshooting it appears that on both charger boards C5 a Tantalum cap and U3 an MC34163 Inverting switching regulator have failed. thus causing a short (<2 Ohms) directly across the 24 VDC input to the charger board.  A shorted Tantalum cap was the problem with the SG-1144 Signal Generator.

SG-1144/U Signal Generator

The power supply had a number of shorted Tantalum caps.

SB-22 Switchboard

Before cleaning the resistance on one of side of the "D" battery holder was about 27 Ohms, way too high to pass the current needed for the talk battery or the buzzer. After cleaning it was below 0.01 Ohms.

After sitting in storage for years with the batteries removed the battery holder threads developed oxide.  When fresh "D" cells were installed the switchboard would not work.  The fix was to througherly clean both the male and the female threads of the battery holder and then apply Contax Oxide Inhibiting Compound (contains castor oil and carbon black)to all the threaded battery holder joints. I got this bottle at the OSH store in Sunnyvale, CA.

There was a time where the U.S. National Electrical Code allowed Aluminum wire in residences instead of Copper wire.  When joining Al and Cu wires a poor joint might result depending on how the joint was made.  To help solve this problem you can apply Contax.  So you'll find it in the electrical supplies section of a good hardware store.

FTS4060 Cesium Frequency Standard

This unit arrived DOA.   At the time I thought that it was shipping damage (and that is probably the case) since the other two boxed units that came on the same pallet worked fine but this one out of the box was dead.  Since it was no more than parts, I took every module and board that could be removed out and figured out all the connector and PCB pin outs.  In order to sell it as a parts unit it was reassembled and just for fun powered up.
It now locked and started working.  So cycling the connections may have  brought it back to life.

TS-1836C In Circuit Tansistor Tester

This is an in and out of circuit transistor tester.  Although the DC battery check works, the beta test appears dead.  It uses a 1 kHz signal to measure beta.  The 1 kHz signal is coupled in many places by electrolytic caps.  There are 5 caps that are all 25 uF @ 10 Volts.  The AC impedance at 1 kHz would be about 6 Ohms, but the ESR meter is showing as high as 46 Ohms, almost an 800% error!  All the electrolytics in this meter need to be replaced.  After replacing the high resistance caps it still does not work.

By starting with the schematic diagram and erasing all the switch contacts and are not being used and then anything in series with them the resulting schematic has missing connections.  Very strange.  Although it may be that lines were erased that should not have been, repeated audits have not found them.

If you have one of these that's working please let me know.

Dead CRT Computer Monitor

Symptoms: burning electronic part smell.  Failure to turn on.
A visual inspection found this burned out component.  The schematic symbol under it looks like the letter "Z".  It's located at the AC mains input to the power supply.  Right behind it is a bridge rectifier and a 330 uF cap.  The coils to the right are probably part of the line filter network.  The two blue wires go to the front panel, not sure why since the  front panel (left of photo) On-Off switch has a plastic connection to the line switch (to the right of photo).

What is the dead part?
It might be series connected (inrush current limiting) or shunt connected (overvoltage protection  MOV).
When trying to trace the PCB wiring using the DMM in continuity mode a chunk of the PCB (around the burned area) fell out.  So I've scrapped this monitor.

I'm thinking it may have been the problem or the cap behind or a nearby cap may be bad.  The ESR checker will be used, but the first order of business is to find out what this is.

Note the rectangular black plastic part just behind the split component.  The solder connection looks like a donut, i.e. it's not wetting the lead, a poor quality joint.

Static & Electrocution

Electrocution

Often static electricity is blamed for what I call electrocution.  For example, I was asked to investigate why the yield was very poor on an expensive polar discriminator assembly that used four back diode detectors on each housing.  Static discharge was the assumed cause of the problem, but the real problem was a soldering iron that had about 15 VAC on the tip.  Back diodes can be blown with only a few volts.  The work bench had maybe a half dozen soldering irons (all anti static grounded tip type) and only when the defective iron was used would the diodes be blown.

Static

Some definitions

Antistatic: Preventing or inhibiting the buildup of static electricity.

Since static electricity is generated by moving electrons from one insulating material to another the current involved is very small so a high value resistor between the two objects can easily equalize the voltages and prevent buildup.  A pure plastic bag that can generate static electricity can be treated in a number of ways so that it has a high resistance and then will not generate static.  This is an anti static bag.

Electrostatic Discharge (Wiki Talk): The sudden and momentary electric current that results when an excess of electric charge, either stored on an electrically insulated object or on an isolated conductive object, finds a path to an object at a different electrical potential (such as ground).  

Getting shocked in a warm room in the winter time when touching a metal object is because of ESD.
Lightening is an example of ESD.

It's very important to see the distinction between these two ideas.

I sit on a chair with plastic wheels in a room with wall to wall carpet.  I used to use a "chair mat" to get a hard surface so the chair wheels would roll, but plastic wheels on plastic is a static generator and often I would get a static shock.  Using antistatic spray would help but was not a cure.  Placing a sheet of plywood over the mat (could have removed the mat) solved the problem.

When working on a PC I leave the power cord plugged into the wall.  This has a danger factor in that the power switch has hot line voltage and maybe some spots on the power supply are line hot.  But it also connects the chassis to line ground so that if you touch the chassis before you touch any static sensitive device the static will be grounded.

Wrist straps are not good grounds, but rather have a 1 to 10 Meg Ohm resistor in series between the human and ground.  This way you are not grounding someone and making a human electrocution possible.  If you want to make your own ground strap by using a piece of copper wire and a ground lead, be sure to insert a 1 MegOhm resistor in series.  

Many companies that work with static sensitive components, like microwave parts or disk drives, have special flooring that is conductive and the workers wear shoes that are conductive or use heel straps that make a high resistance path from the floor to the ankle.  This way a person walking does not build up a charge.  Also the chairs and stools have grounding chains always in contact with the floor so a moving chair does not generate a charge.

When packaging a static sensitive component many people make the mistake of thinking that using an "antistatic" bag will protect the component.  THIS IS NOT THE CASE.  An "antistatic" bag will not generate static but does not protect from static. 

It is easy to destroy the component while it is in a sealed "antistatic" bag.  Just walk across a carpet on a dry day and touch the top of the bag while it is sitting on a metal work bench.

 The proper way to protect the static sensitive component is to wrap it in Aluminum foil or use one of the conducting Faraday cage type bags.

PS Magazine 2004 Issue 623 page 42 has published this information about what antistatic means. (PS I'm in Ukiah, CA, not San Diego) 

Printed Circuit Boards

A way to get a double sided board with vias is to use Express PCB.  They have a prototype service that will supply 3 boards that are 3.8 x 2.5 inches and have tin-lead reflow but have neither Silk screening, solder masks, routing nor multi-layers.  There are a limited number of drill and pad sizes.  You can put a number of smaller boards on one and separate them yourself using a Tungsten carbide tipped scriber (General Tools Model 88CM, Catalog No. GEN-31116 ).  A straight edge is needed for the first pass but all the succeeding passes are in the groove.

In the 1960s we used shoe eyelet's to connect the top and bottom PCB traces in double sided boards.  They were installed and formed using a motor powered machine then hand soldered.  This was a military approved way of PCB construction and much easier than soldering a wire and cutting off both ends.

2002 - update on cutting printed circuit boards.  I got a single speed skill saw at Wal-Mart, chosen for it's low price and that fact that it had a couple of nuts as part of the flat surface.  By drilling a hole through the 1" thick workbench top and making a short slot for the blade, then mounting the saw underneath the workbench using a couple of flat head screws, with only the blade sticking up through the slot I now have a reciprocating table saw.  By using a fine tooth metal cutting blade you can separate 25 to 100 PCBs.  This bench mounted $10 saw turns out to be quite handy.  The screws that pinch the blade turn out to be an M4 thread and have been replaced.  It takes a lot of finger strength to cut small boards since you need to hold them down very firmly or they will jump up and down with the blade.  It makes a lot of dust that you should not breathe.  It takes a little more than 50 mils for the saw kerf.

2006 Update - A small 8 inch Mini Shear Break designed for working sheet metal does a nice job of cutting 1/16 inch PCBs.  Although I think I know why the first one is not broken waiting for parts to arrive from China, I've received an email from another user that said his would not cut boards out of the box.

Replacing soldered in ICs

After having identified a bad IC that's soldered into a PCB I used the following procedure.
Cut all the IC leads using diagonal cutters.  It took a couple of passes to get them all cut on one side of the IC.  Then bend the IC up and unsolder and pull each lead one at a time.  Then press the IC down and cut and desolder each of the pins on the other side.  This destroys an IC that was bad already but greatly reduces the chances of damaging the PCB.

3 May 2006 - Next using a solder sucker clean out all the holes.  Install the new IC and solder in place.  This may not be so easy.  Sometimes neither solder wick nor solder sucker will get the solder out of the hole.  If a piece of wire that's the right size for the hole is chucked into a Dremel tool then cut off so that only a board thickness plus a little is above the chuck face.  It can be used as a drill bit on the highest speed.  Another option that has also worked for me is to use a sewing needle held in a pin vise to push a hole into the solder, sort of swaging.

Using Solder Wick

6 May 2006 - If just prior to using the solder wick you wet about 1/4" with RMA flux it will work much better.  This may be for two reasons.  First the liquid conducts the heat better than air and second, the flux cleans the solder wick of oxides allowing the solder to bond easier.  This makes the difference in being able to get solder out of small diameter holes or not getting it out and needing to drill as above.

Flux The key property of solder flux is that it's boiling point is higher than the melting temperature of the solder.  This makes for excellent heat transfer.  The shape of a soldering iron tip it typically conical or cylindrical and so touching the iron to a wire or circuit trace makes for a small contact area and poor heat transfer. If a small amount of flux is applied and the iron located where flux is present then you have much better heat transfer. I'm about to order a bunch of flux pens.

This is the construction of the A.C. Power cord for the M455-1 Power Supply.  The connector was second hand, i.e. all the cups were full of solder.  The wick was flat and a little wider than the O.D. of the pin but using tweezers it could be formed into a shape that would fit into the solder cup (after using the wick in the normal way to get down to the solder cup).  Then using a very pointed iron (the one used for SMT soldering) and trapping the wick between the iron an the protruding part of the cup the solder was sucked completly out of the cup.  The "A" pin at about 2 o'clock is the one that's been cleaned, no point in doing pins that aren't going to be used.

The black jaws are on a Panavise Jr.

Also when using shrink tubing in very tight quarters like this don't use a size that's a tight fit because when the wire insulation distorts you will not be able to pull the tubing over the joint.  A pair of tweezers was used as a heat sink to keep the shrink tube from shrinking.  The tweezers were held closed by a spring like used in the 257477BA Battery Adapter.

Here the Panaviswe Jr. is tipped on it's side and the rotation of the ball clamp has been adjusted so that there's a three point support (the base, the end of the adjustment arm, the  tip of a jaw) that's stable.

This is a great way to hold connectors while they are being assembled.

Soldering Surface Mount Devices - seperate web page on Surface Mount Technology

Soldering Iron Tip Maintenance

When finished using a soldering iron I used to clean the tip and power down the iron.  This results in the tip oxidizing and getting pitted.  Much better to clean the tip and put a big glob of solder on it just prior to power down.  Also the modern tips come with a silver colored plating, they are not raw copper and this helps maintain the tip, so these should not be filed.

Reverse Engineering

Most of the equipment that I have reverse engineered made use of printed circuit boards.  If the board has part ID information you can use the same part names, but if it does not (most of the time) they you need to name each part as well as have a way to identify each lead.  Lead identification for ICs is easy since they have standardized pin numbering, diodes have an anode and cathode, but resistors and other parts don't have lead specific markings and so you need to keep track of which end goes where.  This can be done by using a flat bed scanner to make color images of each side of the board and then adding part names, like shown in manuals.

Often it's necessary to use a continuity checker (beep Ohms mode on a DMM) to find out what's connected to what.  This is especially true for multi layer boards.  When working with military boards you will find that they often have conformal coating that is thin and transparent.  So when you place a probe tip on top of what appears to be nice shinny solder you will NOT make electrical contact.  A quick check is to put both probes on the same pad, yet not touching each other.  If there's no continuity then you know you're working on a coated PCB.  One way around this is to push and/or rotate the probe to punch through the coating and into the solder, but the same pad may need to be probed many times and this really chews up the pad.  A better way is to use a large sewing needle held in a machinist's pin vise.  The meter probe can be put into the hollow metal handle or connected with a clip lead.

When reverse engineering some thought needs to be given to possible DC shorts.  When working on the FS5000 Spy Radio I smoked a PCB trace that was on one of the interior layers.  This could have been avoided by fusing my test fixture DC power leads (both the hot and ground).

Loctite

Loctite is great for keeping screws from coming loose, BUT should not be used near plastic parts.  For the very first production lot of my 2577BA Battery Adapter I used Loctite 222 on a 6-32 screw that went into an Aluminum spacer.  BUT the 1/8" thick plastic sheet between the two showed severe cracking when the Loctite set up and the whole batch had to be scrapped.  I think what happened was some Loctite got into the space between the screw and the plastic and when it set up it expanded and destroyed the plastic.  Needless to say I don't use Loctite in this application.

Rubber Parts

Rubber (and synthetic equivalents) biodegrade.  Air and/or water will cause them to disintegrate.  Insulation will turn to dust and hoses, gaskets, O-rings, etc. will fail.  In older automotive applications this is the most common failure mode I have seen.

My 1929 Rolls Royce and the 1934 Rolls Royce used steel tubing between the vehicle frame and the engine where today a rubber hose is used.  They did this by forming 3 or 4 turns about 3 " in diameter so that the tubing would never be stressed past it's elastic limit as the engine rocked back and forth.  Of course the engine did this only on acceleration or engine braking, when idling you can balance a Nickel on top of the radiator.

The O-rings between the steel cylinder sleeves and the aluminum block of the all the RR cars with a V-8 engine is a real disaster.  It's a certanity that one of the huge number of rubber hoses carrying water or the rubber seal in the water pump that also has to act as torque stop, will fail while you are under way resulting in the engine block over heating.  Overheating those cylinder seal O-rings results in failure to seal the water jacket.  If it's the lower O-ring that failed you will get water coming out of the weep holes in the block since there are two O-rings at the bottom, the lower O-ring seals out crancase oil, so if oil is coming out the weep hole it means the bottom O-ring has failed.  Needless to say replacing the O-rings requires  major engine work.  If the aluminum bloct that's about 1/8" thick between cylinders cracks then it can NOT be welded since the aluminum is plastic impregnated.  But you migh be able to use an epoxy, didn't try that.  I forget where the water goes if the upper O-ring fails.

Cleaning

In order to make images of equipment I use a digital camera or color flat bed scanner.  Both of these methods show up every spec of dust so cleaning is needed to get good images.

Dirt

Dirt is the most common thing that needs to be removed so the first cleaning step us to use water  to wet the dirt, a tooth brush, tooth pick, small brush, pipe cleaner to scrub a little and a clean dry paper towel or tissue to wipe dry and clean.  Often knobs need to be removed to clean under them.  Engraved labels often have dirt covering the white paint and need extra scrubbing.

Very Old Masking Tape

DE-Solv-It , Contractors' solvent available at K-Mart - removed tape from a BC-611 radio that nothing else could touch.

Tape adhesive

This web page has a number of methods to remove tape adhesive as well as other stuff.
Goo Gone™ (Magic American Corp.- Cleveland OH) has received a number of recommendations on the Tek list server also works for decomposed foam.

Rust

Oxygen, moisture  and about anything can combine to make corrosion.  There are many commercial rust remover products on the market.  Many use phosphoric acid. (Wiki: Rust Remover).  Renaissance Metal De-Corroder says it's made of : Amine comples of hydro-oxycarboxilic acid.  But the only references to the key indegrient are ads or references to this product.  It has the look and feel of phosphoric acid.  Available from the Gemmary.

If paint primer is applied to bare metal it affords NO protection from rusting.  primer is by design a porous coating so the the paint overcoat will bond well.  Leaving a car outdoors with just primer is a great way to rust it (ask how I know this).  The second time I used a special paint made for ships called Corroless.   I picked this brand after reading the paint shoot out test results in Skinned Knuckles magazine since it got the highest ratings.  Rustolum and POR15 were not as good.

I've recently been doing things that involve metal working and a number of times I've come across vapor phase rust prevention (aka: Volatile Corrosion Inhibitors or Vapor Phase Inhibitors) papers.  There's a brown sheet of paper behind the plastic laminated cal certificate in the box lid of the pin gauges.  That's the special paper.  My local machine shop ships metal parts to me with some of the paper in the packaging.  It has some shelf life so is not a permanent solution, but one worth knowing about.

4604227 Vapor phase and surface contact rust preventive composition, Stauffer Chemical Co, Aug 5, 1986, 252/389.2 ; 252/389.61; 252/392; 508/399

Protective Packaging carries moisture barrier bags and Vapor Corrosion Inhibitors paper.  There web pages have a lot of good info.

When rust forms in a threaded fastener it becomes very difficult to remove the fastener.  Many people use WD40 but  that's the wrong thing.  WD stands for Water Displacement which is what it was made to do.  For getting stuck nuts, bolts or other items loose a penetrating oil like Kroil, is what works best.

Storage

Design Defects

In some cases it is sold as a "Feature". 

HP Doppler Radar Module

In some cases it kills the product or manufacturer.  Long ago HP came out with a doppler radar module.  The idea was to hook up an antenna, supply DC power and get the audio doppler signal out.  It involved a lot of specialized construction and tooling.  But there was a design defect that greatly degraded the signal to noise ratio compared to a different topology.  The product disappeared from the market.

LCD Backlight in PRC-126

In other cases is just fixed.  The first PRC-126 radios added a back light to the LCD display and it also was supposed to be night vision goggle compliant.  The designer placed a very dark green filter above the LCD and this worked great for night vision goggle use and I'm sure that's where all the effort was focused since this was a new high tech area, BUT you could not read the LCD in daylight.
Later radios fixed this by placing the dark green filter between the back light and the LCD.

Rolls Royce V8 engine

The Rolls Royce cars with V-8 engines use an aluminum block and steel cylinder liners.  The aluminum thickness between cylinders is maybe 1/8" or less.  When (not if) the rubber O-rings sealing the steel liners to the block fail (see Rubber Parts above) the engine overheats causing irreparable cracks in the thin block wall.  The last I heard Rolls was going to buy engines from BMW.  It's too bad that the company has been going down hill ever since Mr. Royce died.  The cars he designed were built to be reliable.  RR won many races, not by being fast, but by finishing.   That's why Lawrence of Arabia used the PII chassis as the bases for armored cars.  The PII was also featured in one of the Indiana Jones movies.

Fluke Test Leads

The new style test leads that come with the Fluke model 87 DMM and are made by Fluke have failed three times.  The failure mode is that ALL the strands of wire fracture in the same plane about 1/8" past the strain relief behind the shrouded banana plugs at the probe end.  This is not a stress type fracture where you can see deformation of the metal, but rather as if the wire had been cut in a shear.  This has happen to me with 3 different sets of test leads.  Each time Fluke was notified not only that there was a failure, but that it was most likely a design or manufacturing defect that caused it.

One of the symptoms of the failure is that the meter reads zero volts when in fact there is voltage in the circuit.  This is a safety issue if the meter is being used to determine if a circuit is hot prior to working on it. 

My guess is that there is some type of work hardening stress (thermal or mechanical) applied to the strands in the manufacturing process and a small amount of flexing of the wire causes a fracture.

30 Nov 2006 - It turns out that the test leads were NOT intended for DMM use.  I have purchased DMM test leads and they do have strain relief on both ends, unlike the above test leads that did not have strain relief on the probe end.  And that's where they failed.

Lap Top Li-Ion Battery

Precision Time & Frequency

GPS 1 PPS Stability

Counter Trigger Settings

Power Line Cords

Prior to the current IEC type of cord end

there was a different cord end (Belden 7A125)  that went into the equipment as shown below:
This is called a PH-104 and is made by Volex and can be ordered from Mouser.  There are two versions the standard and reversed.  In the reversed version the hot and neutral wires are reversed from the normal configuration going to the instrument.  The plug shown is marked "Belden 7A-125V" and it's cord is marked " 18-3 TYPE SVT E-3462 LL-7874"  I don't know if this one is standard or reversed?

Links

This is the 04946 hits since time this page has been accessed since since 24 Aug 2001.