The Wavetek SAM is a cable-TV test instrument introduced in 1980. The design has since evolved through many iterations, and early models are now cheap. This writeup covers the original SAM I and the SAM II.
SAM means signal analysis meter. The device is a calibrated, battery operated receiver that tunes 4 to 300 MHz. (One option increases coverage to 450 MHz; another adds the 470–890 MHz UHF-TV band.) Its primary purpose was to measure the amplitude of cable-TV signals. It has a wideband RF attenuator with 10-dB steps and a precision level meter calibrated in 1-dB steps over a 20-dB range. The minimum measurable signal level is −40 dBmV and the maximum is +60 dBmV (10 µV to 1 V). A small speaker and volume control help identify signals. The SAM I has analog tuning in multiple bands with a calibrated circular dial. In addition to signal level, the unit can measure residual hum (unmodulated test signal required), S/N (a two-step narrowband measurement calibrated for 4-MHz video bandwidth), and AC or DC voltage on a 100-V scale. A 150-MHz oscillator provides a +20 dBmV calibration signal with a specified accuracy of ±0.25 dB.
The most interesting secondary function is the spectrum analyzer. This mode generates horizontal and vertical signals for an external oscilloscope. The display is logarithmic with adjustable frequency span and sweep rate, or with manual scanning. It's fascinating to hook up a scope and watch the FM spectrum change in a complex way as you rotate an antenna. The span can be narrowed to a single signal or widened to cover an immense amount of spectrum. For someone without one, the spectrum analyzer may be the unit's most attractive feature. It is an option, but most units seem to include it.
I got the SAM primarily for aiming fixed rooftop antennas. It can measure signal levels even from low-gain antennas. I soon discovered that the SAM also was useful indoors for aiming a rotary antenna. It has much better resolution than an LED signal-strength indicator and won't saturate on strong signals as a tuner's analog meter will. It permits very accurate antenna aiming. I was surprised how quickly the response of a small Yagi dropped off when pointed slightly away from a station. It was easy to pick up a dB or two using the SAM.
With the SAM you can aim antennas, accurately compare antenna gains, precisely measure antenna patterns, check the ellipticity of a circularly polarized array, measure feedline, balun, or power splitter losses, and determine the gain of antenna or distribution amplifiers. You can use off-the-air signals for any of these measurements. With a signal generator and power splitter, you can measure antenna or tuner return loss and then calculate mismatch loss. Resolution to a couple tenths of a dB is easy. Absolute accuracy is rated as ±0.5 dB at 70° F, or ±1 dB from 0°–120° F when calibrated at the operating temperature.
The hand-drawn, low-frequency calibration correction curve inside the top cover is a nice touch. Another is the way the unit turns itself off. Inside the lid is a storage compartment that opens with a spring-loaded knob. When you close the lid, the knob pushes the front-panel power switch to the off position.
The IF strip is linear, not limiting, with AGC applied to the first stage. The detector responds to AM as well as FM. This makes the audio sensitive to multipath distortion, which can be useful when aiming an antenna. AM detection lets you hear NTSC video buzz and copy signals in the 108–137 MHz aircraft band. It makes the portable unit handy for chasing down power-line noise, which is easy to localize using AM. With a short wire antenna I was able to hear WWV at 5, 10, and 15 MHz as well as many shortwave broadcast stations. I was surprised that the wide IF didn't jumble several signals together, but the loudest stations came through clearly. The dial is calibrated down to 4 MHz but the unit actually tunes to DC. I could receive many stations in the 540–1700 kHz AM broadcast band at reduced signal strength.
The IF uses broadband LCR networks and a single 10.7-MHz ceramic filter. The stock filter is a 280-kHz Murata. While selectivity was adequate for strong signals, I found it difficult to measure weak signals on adjacent FM channels. I unsoldered the filter, installed a socket, tried several substitute filters, and finally settled on a 110-kHz Murata. I compensated for 2 dB of additional loss by readjusting the IF-gain pot.
This shows a ceramic filter in its socket. The IF-gain pot is at the lower right. The slide switch breaks the IF AGC loop for testing, substituting a fixed AGC voltage determined by the pot above it. I have yet to find a need for this feature.
This is the spectrum analyzer display of the FM broadcast band using the original 280-kHz filter.
This is the response using the 110-kHz filter.
Ceramic filter source and load impedances in the SAM are lower than Murata recommends. But a slightly asymmetrical passband shape may just be due to the filter being on the response slope of the other IF tuned circuits. Once you've socketed the filter and aligned the unit as described below, it's easy to check. Run the spectrum analyzer on the calibrator signal, narrow the sweep until the filter shape is apparent, and then substitute a .01-µF ceramic capacitor for the filter.
In the FM band the FINE TUNE control originally spanned 8 MHz. This wasn't much help in tuning FM signals. I replaced the wire between the control wiper and PCB with a 2.2MΩ resistor, as shown above. The PCB trace leads to a 1MΩ resistor. The total resistance of 3.2MΩ helped a lot, but I wound up installing a 10MΩ resistor for a total of 11MΩ. This provided a fine-tuning range of about 700 kHz in the FM band.
Increasing the fine-tuning series resistance offsets the dial a few MHz. If you don't intend to recalibrate all bands, use two resistors. 4.7MΩ in series with the 1MΩ, plus 1.8MΩ from their junction to ground, yielded a fine-tuning range of 900 kHz in the FM band, with negligible dial offset.
The battery pack consists of twelve sub-C NiCd cells of unmarked capacity. The charging current is about 170 mA for a partially charged battery, dropping to about 155 mA at full charge. The circuitry draws a nearly constant 160 mA; the light adds 20 mA. The manual specifies an 8-hour operating time, which suggests a battery capacity of about 1300 mAH. Charging the battery with the unit on results in a net inflow or outflow of a few milliamperes, depending on battery charge.
When the NiCds died I installed 2100-mAH NiMH AA cells to provide longer operating time and less weight. Three four-cell AA holders just fit inside the plastic sub-C enclosure that mounts on the inner case wall along with an insulating sheet. This made conversion easy. (I found cell holders with solder tabs to be intermittent. Holders with preattached wires worked fine.)
With higher-capacity batteries it would be tempting to increase the charging current to reduce recharge time. However, the small line transformer in the charging circuit already gets slightly too hot to touch. The transformer feeds a bridge rectifier and 1000-µF capacitor. The 20 V of DC developed charges the battery through a #1891 lamp to provide constant current. Circuit simulation suggests that lowering the peak winding current by disconnecting the capacitor, adding a resistor across the lamp, and charging with pulsating DC averaging up to 250 mA will not increase transformer temperature.
The battery recharge point on the meter originally corresponded to 12.9 V (1.075 V/cell). I found that the regulated voltage (11.93 V in my unit) did not drop until the input voltage had fallen to 12.1 V. To make this the indicated recharge point, I installed a 360kΩ resistor across the 24.9kΩ 1% meter resistor. 1.0 V/cell is the typical cutoff voltage recommended by NiMH battery manufacturers so this worked out nicely. In another SAM the original recharge point was 12.7 V and dropout occurred at 12.3 V. 680kΩ was required across the meter resistor, which in this unit was a series combination of 5% resistors.
If the cell capacities differ much, one or more cells may completely discharge and become reverse-biased before the battery voltage reaches the recharge point. This can shorten cell life. If you see the meter suddenly drop when checking battery voltage, often a few seconds after applying power, it's time to recharge regardless of what the meter says.
The SAM has dozens of adjustments. I experimented with all except those in the front-end module.
The discriminator is the rectangular metal can with colorful slugs on top of the main circuit board.
The red dots locate adjustments on the underside of the board. I'll identify these left to right as upper group U1–U2, middle group M1–M6, and lower group L1–L6. Here's what they do:
U1 Maximum sweep frequency U2 AFC-off frequency offset M1 Trimpot not installed, red dot only M2 Spectrum analyzer vertical position M3 Spectrum analyzer low-level linearity M4 +10 dB calibration M5 Meter calibration M6 Video position L1 Meter linearity at upper end L2 Spectrum analyzer operating point L3 IF gain L4 Hum calibration L5 AGC voltage when test switch engaged L6 Second-LO coil slug
There are two additional adjustments not visible in the photo. The slug within the calibrator module sets its frequency, while the trimpot beside it sets its output level.
It's best to align the unit at the temperature at which it will be used. Turn it on and wait until the tuned frequency stabilizes. Then adjust things in the following order.
Connect a frequency counter to CALIBRATION OUT. Adjust the slug within the calibrator module for 150 MHz. Then connect a 75Ω RF voltmeter or power meter and adjust the trimpot next to the module for +20 dBmV (−28.8 dBm). Be sure to account for the loss of the measurement cable.
Connect an FM signal generator with 1-kHz modulation deviated 25 kHz to RF IN. Turn AFC on, center FINE TUNE, and tune the main dial to peak the signal. Then simultaneously adjust L6 and FINE TUNE for maximum meter reading. Next, simultaneously adjust the blue discriminator slug and FINE TUNE for maximum. After carefully peaking FINE TUNE, move the signal generator above and below its original frequency until AFC loses lock. The upper and lower lock ranges should be approximately equal. If not, slightly rotate the blue slug, repeak FINE TUNE, and check again. When satisfied, peak FINE TUNE and then adjust the pink discriminator slug for minimum audible distortion. Remove the modulation.
Carefully peak a signal with AFC on. Then switch AFC off and adjust U2 to repeak the signal.
Switch to spectrum analyzer mode. Find the setting of L2 where the signal at VERT saturates high and the setting where it saturates low. Set L2 midway between these two points. This adjustment also affects the meter in signal level mode.
Switch to signal level mode and peak a signal with AFC on. Adjust L1 and L3 for best meter linearity. You can also adjust L2, but make sure VERT does not saturate in spectrum analyzer mode. I was able to get the center level correct within 0.2 dB when aligning the outer levels of three successive 10-dB input steps to the meter scale.
Apply a signal with AFC off and adjust M2 and M3 for best vertical linearity in spectrum analyzer mode by switching the step attenuator over a wide range. I was able to linearize the upper 40 dB or so, with a little expansion in the next 20 dB.
Jumper CALIBRATION OUT to RF IN with a cable no longer than a few inches. Tune to 150 MHz, set the attenuator so that +15 and +25 show in the meter windows, lock the CALIBRATION SET screwdriver adjustment at midpoint, set GAIN to CAL and AFC on, and fine tune to peak the meter. Then adjust M5 so that the meter is at midscale.
Turn GAIN to +10dB and adjust M4 until the meter is at the rightmost scale mark. M4 and M5 interact somewhat. Readjust each until both calibrations are correct.
In signal level mode, VERT provides the video signal (detected AM). Apply a 100%-modulated, 1-kHz, AM signal and adjust M6 so that the waveform does not clip high or low. M6 very slightly affects the meter. After adjusting it, recheck the meter midscale calibration and readjust M5 if necessary.
The vertical board provides trimpots to calibrate the dial. The six pots on the left align the low end of each band. The pots follow the order of the rotary bandswitch. The six pots on the right, somewhat deeper, align the upper ends. Center FINE TUNE and turn AFC off. Pick frequencies near the lower and upper band limits and adjust the pots using an accurate signal generator. The two adjustments interact somewhat. Check the calibration near the center of each band before moving on to the next. Dial accuracy is specified as ±1 MHz.
With the UHF option, bands A-E cover 470–890 MHz when the switch is at UHF. The vertical board contains ten additional trimpots, recessed and in the same order as the VHF pots. The little red squares around the dial edge mark UHF video-carrier frequencies, while the red numbers at the bottom identify the UHF channel range for each band. What a neat way to pack so much information onto a single small dial. I recalibrated so that the red squares mark channel centers. This seems best for ATSC signals.
The SAM II is similiar to the SAM I except for the LCD frequency readout and TV-channel keypad. My unit had a maximum frequency of 300 MHz. A single tuning knob covers the entire frequency range in ten turns; there is no fine tuning control. Frequencies are displayed to 0.1 MHz when tuning manually and 1 MHz when using the keypad. There is no way to enter an arbitrary frequency with the keypad. Both the meter and LCD have lamps.
My SAM II had a retrofitted 12-V, 2300-mAH, sealed lead-acid battery. The charging circuit was unmodified. The battery recharge point on the meter corresponded to 12.2 V. Right at this point the regulated voltage (11.98 V in this unit) began to drop. Panasonic literature indicates that a 12-V sealed lead-acid battery has 20 to 50 percent of its capacity left at 12.2 V at room temperature. Therefore the effective capacity of the 2300-mAH battery in my SAM II was somewhere between 1200 and 1800 mAH.
The vertical board in the SAM II is the frequency synthesizer/counter. It uses common CMOS ICs, all socketed. Two trimpots labeled G and O are coarse and fine VCO adjustments. When misadjusted the PLL may not lock on every keypad channel. About an inch to the left of G is a wire loop. Attach a scope probe and adjust G and O for a stable trace when punching up all channels.
With the keypad I brought up a local NTSC TV station, switched to SND for the aural subcarrier, and with AFC off adjusted second-LO coil L6 for maximum meter reading. This yielded correct keypad tuning for all TV channels. However, the indicated frequency was about 0.2 MHz low when tuning manually. Since I was more interested in tuning FM signals than accessing TV channels with the keypad, I reset the second LO for accurate manual tuning.
My SAM II had some overtuning. On the low end the frequency went through zero and came back up. On the high end the knob continued to turn after the frequency topped out. I mounted small 10kΩ trimpots to the end lugs of the tuning potentiometer. With these I was able to set the band edges at DC and 300 MHz. This brought the tuning rate down from 40 to 30 MHz per turn in the FM band.
The SAM II had a number of strong digital spurs below 2 MHz. But the only in-band spur I found was a −5 dBmV signal near 35 MHz from the second LO, plus a little leakage from the calibrator at 150 MHz.
The SAM Jr is a stripped-down version of the SAM I. It measures signal amplitude only and has no spectrum analyzer or calibrator. An IF bandwidth of 600 kHz makes it unsuitable for use on a crowded FM band. Similar units from other manufacturers use a single-conversion IF near 45 MHz. If the SAM Jr operates this way it would not be easy to modify for a narrower passband.
The SAM III is similar to the SAM II, but the synthesizer can tune any frequency, not just TV channels. The SAM IIID has an RS-232 interface.
The SAM IV is an AC-only SAM IIID with built-in CRT.
The digitally synthesized SAM 1000 and SAM 2000 have 125-kHz frequency steps. This resolution is not suitable for tuning the FM broadcast band with a narrow IF filter.
88–108 MHz