Troubleshooting with simple tools
From QRPWiki
Salmoncon 2006
Troubleshooting Problematic Radio Circuitry With Simple Test Equipment
By Wayne McFee NB6M
Contents |
INTRODUCTION
Almost all of us, at one time or another, have built a piece of electronic equipment that didn’t work when we finished assembling it. How many of you have had that experience? I know I certainly have.
Once we get over the frustration and anger of realizing that our nice, new piece of gear isn’t doing what it is supposed to do, we need to figure out why. We need to troubleshoot the rig in order to locate the problem, so that we can correct whatever is wrong and get it working properly.
THE TROUBLESHOOTING PROCESS
The first thing we need to do, when a piece of gear doesn’t appear to work, is to make sure we have made the proper connections to it.
CHECK THE OBVIOUS
Is the power supply or battery pack connected properly and turned on?
Do we have the headset plugged in?
Is the on/off switch on the equipment turned on?
Is the volume on a receiver turned up sufficiently?
Is a dummy load connected to the transmitter?
Is the key plugged in?
Once we have checked to make sure we have all the proper connections, thrown all the appropriate switches, and turned up all the right controls, and the rig still doesn’t work, then we get into the troubleshooting process.
Troubleshooting makes use of the following procedures:
Visual Inspection
Voltage Measurements
Resistance Measurements
Signal Tracing
The focus of this presentation is on accomplishing these procedures by the use of simple test equipment. Let’s take a look at what those pieces of test equipment are.
A LINEUP OF SIMPLE TEST EQUIPMENT
A strong light and a magnifier for making Visual Inspections
A 50 Ohm, resistive Dummy Load
A Volt/Ohm Meter with a high input impedance
A homebrew RF Probe
Clip Leads
A .1 uF bypass capacitor
An Electrolytic Capacitor of a nominal value (2.2 uF to 47 uF)
An RF Signal Source, in this case a simple Crystal Oscillator Could also be a simple VFO or even a Transmitter
An Audio Signal Source, in this case either a simple Audio Oscillator or a simple circuit for supplying an audio signal from a stereo or radio receiver
If available, a General Coverage Receiver
The following diagrams provide the circuitry for an RMS Voltage reading RF Probe, an Audio Test Setup, an Audio Oscillator, a Crystal Controlled RF Oscillator, and two different types of Frequency Counters. None of these should tax the budget.
VISUAL EXAMINATION
If what we are looking at is a piece of gear we have just finished building, these are some of the things we look for:
Is anything burned or showing evidence of excessive current?
Are any of the connecting wires broken or loose?
Are the connecting wires installed in their correct locations?
Are there any obvious soldering problems, such as part leads not soldered, cold solder joints, or solder bridges?
Are there any traces broken or torn on the printed circuit board?
Are all the parts installed in their correct locations?
Are all the polarity-sensitive parts installed correctly?
If the piece of gear we are looking at has been working, but now is not functioning properly, it is fair to assume that the parts are in their correct places and that polarity-sensitive parts are installed properly, so we examine carefully for any of the other aforementioned problems.
If a careful visual examination turns up no obvious problems, then we turn to the use of test equipment in order to locate the problem.
TROUBLESHOOTING WITH SIMPLE TEST EQUIPMENT
Let’s assume that we have a newly built transceiver kit that doesn’t work. Where do we start, in order to determine the problem?
Ensure that all connections are made
First, make sure that the power cord, earphones, key or keyer paddles, and a 50 Ohm resistive dummy load is plugged in to the rig.
Apply DC Power
Switch the power supply or battery pack (if it has a switch) on.
Now we get to the actual use of our simple test equipment in finding the problem.
Check for DC power
The first item to use is our voltmeter, specifically to see if DC power is actually being felt in the circuit. We can do that by checking at any point on the printed circuit board that should have DC power on it, but we will probably want to check first at the main, positive DC power line (wire) where it connects to the printed circuit board.
If we actually have DC power applied to our printed circuit board, what’s next? Although we could continue checking voltages all over our printed circuit board, we will probably arrive at the trouble more quickly by tracing signals.
Signal Tracing
Usually, a problem in a rig effects only one of the functions of the rig. The transmitter may work fine, but the receiver is inoperative. Or vice versa. We need only troubleshoot the circuitry that isn’t working. If neither the transmitter or the receiver works, then we check those portions of the circuit that connect to both, such as the VFO or the keying and receiver muting circuitry.
Taking DC Voltage and resistance measurements throughout the circuitry is certainly of value, but in order to narrow down and isolate the stage in which the trouble resides, it may be quicker to trace signals through the effected circuitry. This is done with a combination of either the simple audio oscillator or crystal oscillator, clip leads and DC blocking capacitors, either .1 or electrolytic, and earphones or speaker if testing the receiver, or dummy load, RF Probe and Voltmeter if testing the transmitter.
Once the trouble has been isolated to a particular stage, then voltage and resistance measurements are made in order to determine exactly where the problem lies within that stage.
The following troubleshooting example was performed on a NorCal 40A transceiver in which the VFO and Receiver chain worked, but there was No Transmitter Output.
A NorCal 40A with no TX output
The initial portion of the troubleshooting and repair of this rig was undertaken at Deception Pass State Park, during Salmoncon 2006. The subsequent continuation of the troubleshooting process, and repair of the rig, was conducted at NB6M.
The Problem
No Transmitter Output, Receive Function Normal
Locating The Problem, Using The Simple Test Gear Presented Previously
Step 1
Referring to the NorCal 40A schematic, a series of test points were identified, which are the RF inputs and outputs of each successive stage in the transmitter chain.
These are labeled on the schematic as Test Points A through J, beginning at the antenna jack and working backwards through the transmitter chain to the transmitter mixer.
Note that that the output filter, between the Power Amplifier and Antenna Jack, was considered as a separate “stage”, as was the tuned circuit between the Transmitter Mixer and Buffer Amp.
An RF Probe and Digital Voltmeter were used to check RF voltage levels throughout the transmitter chain, beginning at antenna jack and working backwards through the identified test points.
Both the approximate normal RF Voltage levels, and the levels seen in this NorCal 40A’s transmitter chain are shown in the following list.
Normal Approximate RMS RF Voltage Levels Actual Readings
A 9.4 Volts .2 Volts
B 7.6 Volts .2 Volts
C 7.6 Volts .2 Volts
D 1.5 Volts .5 Volts
E 4.5 Volts 1.5 Volts
F .8 Volts .2 Volts
G .24 Volts .050 Volts
H .25 Volts .055 Volts
I .6 Volts .6 Volts
J .25 Volts .175 Volts
Step 2
Since all RF voltage levels after Test Point I were lower than normal, attention was focused on the parallel tuned circuit made up of C38, C39, and L6.
2a Attempts were made to tune C39 so as to peak the output across the tuned circuit, in order to arrive at the expected input to Q5 of approximately .24 Volts (240 milliVolts). These attempts failed.
2b With DC power disconnected, a resistance measurement was made between the actual pertinent leads of C37 and R10. Results - 0 Ohms.
This was done in order to be sure that a good connection was had between the two, which effectively checked both the solder joints involved and the applicable PC board traces.
2c A resistance measurement was made between the L6 end of R10 and ground, which effectively checked the solder joints on either end of L6. Results - 0 Ohms.
2d L6 was removed, and was rewound with new magnet wire, in order to eliminate the possibility of shorted turns. Results - no improvement.
At this point, the troubleshooting procedure was interrupted due to time constraints At Deception Pass State Park, and the troubleshooting of the rig was continued later at the combination shack and shop of NB6M.
Step 3
Since rewinding L6 did not appear to provide the ability to peak the signals across the parallel tuned circuit made up of C38, C39 and L6, even though 29, 28, and 27 turns were tried, in succession, the next step was to check the output of the Transmit Mixer, U4, to be sure that it was supplying the necessary 7 MHz signal to C37.
The RF voltage level on pin 4, and on the pertinent lead of C37 was .6 Volts, RMS, so even though the output of the oscillator in U4, measured on pins 6 and 7 (Test Point J) was somewhat lower than the volt readings obtained in a functioning NC 40A, it would appear that if the signal on pin 4 was actually a 7 MHz signal, then the Transmit Oscillator and Mixer were functioning properly.
This was done with the Uni-Counter, the low-cost Frequency Counter presented in the Troubleshooting Lesson at Deception Pass State Park.
The results of this test showed that the .6 Volts RMS signal coming from pin 4 of U-4, the Transmitter Mixer, was in fact a 7 MHz signal, as it should be.
Now knowing for sure that a 7 MHz signal of the proper voltage level was being supplied by the Transmitter Mixer, and realizing that tuning the tank circuit made up of C38, C39, and L6 did not seem to have any real effect in peaking the output, attention was turned back towards the output end of the transmitter chain.
Step 4
Since there should normally be approximately .8 Volts RMS of RF on the base of the driver transistor during transmit periods, and a similar level could be supplied by the simple crystal oscillator circuit previously presented, a 7.040 MHz crystal was plugged in to the simple crystal oscillator, and clip leads were used to apply its output across the top of R11 and ground.
The transmitter was keyed, and the Drive Level Pot, R13, was adjusted so that 1.5 Volts RMS was read on Test Point D, the base of the PA transistor, Q7.
This confirmed that the driver stage was working, and that it provided plenty of drive to the base of the PA when appropriate levels of RF were applied to the base of the driver transistor.
Results – No TX Output (still approx .2 Volts output, as before)
Given that the driver stage was proven to work properly, and that there was still
almost no output from the PA stage when an appropriate level of RF voltage was
applied to the base of the PA, attention was turned to the PA stage itself, and the lowpass output filter.
Step 5
A DC voltage check was done on the collector of the PA. Results – 0 Volts.
A careful visual inspection revealed that there was a small bubble on the side of RFC1, the RF Choke in the Collector lead of the PA transistor.
With DC power disconnected, a resistance check across RFC1 determined that RFC1 was open. RFC1, an 18 uH inductor, was replaced with a 22 uH inductor that was on hand.
A quick DC voltage check confirmed that now there was DC power on the collector of the PA.
Transmitter output was again checked, using the simple crystal oscillator and a 7.040 MHz crystal to supply RF to the driver.
Results – still no power output.
Step 6
The PA transistor was carefully removed, a similar, new unit was installed, and the transmitter output again checked, with the simple crystal oscillator supplying RF to the driver.
Results – Still no power output.
The original PA transistor was re-installed.
Step 7
The output filter end of C44 was unsoldered and lifted out of its pad on the PC board.
Short Clip Leads were used to connect a 50 Ohm dummy load between the lifted end of C44 and ground, and the transmitter output to the dummy load was checked again, while driving power across R11 was being supplied by the simple crystal oscillaor.
Results – Good Power output, approximately 8.5 Volts RMS read across the dummy load.
Conclusion – Since the PA stage put out an appropriate amount of RF power when its base was supplied with 1.5 Volts of driving RF and the output of the PA was lead directly to a 50 Ohm dummy load, there must be a problem in the output filter circuitry.
Step 8
A careful visual inspection revealed that the solder joint between the ground lead of C46 (820 pF) and its ground pad appeared suspect. There appeared to be some foamy-looking matter around the lead itself, in its ground pad.
Since an 820 pF, 100 Volt, C0G capacitor was on hand, C46 was removed, and the new part was soldered in place, taking particular care with the solder joint between the ground lead of C46 and its ground pad.
The output end of C44 was soldered back into its appropriate pad on the PC board, the dummy load was plugged back into the antenna jack, and the transmitter output was checked again, with the simple crystal oscillator supplying RF to the driver.
Results – Good power output, approximately 8.5 Volts read across the antenna jack.
Step 9
The output of the simple crystal oscillator was disconnected from R11, and C39 was again tuned for a peak output.
Results – now this tank circuit peaked properly, and with this tuned circuit peaked, the drive level pot was adjusted for 1.5 Volts RMS of drive to the base of the PA.
Indications are that the incorrect impedances being felt back through the circuitry may have prevented this tuned circuit from being peaked properly.
Step 10
Since 1.5 Volts of drive applied to the base of the PA only resulted in just under 1.5 Watts of power out, the turns count of L7 were reduced from 18 turns to 15 turns, in order to present a lower impedance to the collector of the PA, and while transmitting, the drive level was re-adjusted for 1.5 Volts, resulting in 2.0 Watts of output from the transmitter.
The transmitter offset frequency was adjusted (trimmer C34) for an offset amount matching that of the receiver BFO, the rig was put on the air, and shortly thereafter a contact was made with another QRP station in North Dakota.
Summary
Troubleshooting radio equipment involves nothing more than a logical, step by step approach, requiring only a few pieces of simple test gear.
In this case, although initial RF voltage measurements throughout the transmitter chain seemed to indicate a problem in the tuned circuit between the Transmitter Mixer and the Buffer Amplifier, this was shown to be incorrect.
A simple Frequency Counter, the Uni-Counter, was used to verify that the .6 Volts RMS of RF coming from the Transmitter Mixer was, indeed, a 7 MHz signal. Knowing that, a simple Crystal Oscillator was then used to supply RF to the Driver stage, which verified the operation of the Driver stage, and narrowed the problem down to the PA and Output Filter.
A DC Voltage check on the collector of the PA brought attention to RFC1, which turned out to be an open circuit and was replaced with a near-value part, which was on hand.
When it was shown that there was still no power output of the transmitter, with RFC1 replaced, and the PA transistor itself replaced, a 50 Ohm Dummy Load was substituted for the Output Filter, determining that the PA stage itself was now functioning properly, and focusing attention on the output filter.
A careful visual inspection located a possible cold solder joint at the ground lead of the center capacitor in the output filter. This capacitor was replaced, and care was taken to ensure that the ground lead had a good, clean solder joint.
These actions took care of the problems, and there was now good output from the transmitter, with the crystal oscillator still supplying RF to the Driver stage.
The crystal oscillator was disconnected, and with the transmitter’s amplifier chain now functioning properly, the tuned circuit between the Transmitter Mixer and the Buffer Amplifier peaked properly.
It would seem that the problems in the PA and Output Filter were felt all the way back through the transmitter chain, due to a marked change in loading felt from stage to stage.
Roger, K7RXV, explained the most likely cause of the failure of RFC1. He stated that the high impedance felt on the collector of the PA, due to the center capacitor in the output filter not having a good ground, caused the Zener diode on the collector of the PA to fire, which in effect shorted the supply voltage to gound, through RFC1, causing it to carry more current than designed, resulting in an open circuit. This Zener diode is in that location as a protective device to keep the PA transistor from destroying itself if there is a large impedance mismatch felt on its output.
Roger also explained that, without collector voltage, the Base-Emitter junction of the PA transistor would carry more current than normal, with the result that it loaded down the previous stages, giving a false indication of the performance of those stages.
It is always interesting, challenging, and just plain fun troubleshooting RF circuitry. And, if taken step by step, logically, the problem will be arrived at in short order, even with the simplest of test equipment on hand.
