GPS-disciplined crystal channels for the Hammarlund SP-600-JX

Bringing parts-per-billion frequency stability to a 1950s spy-radio classic using a $18 programmable crystal module and a GPS receiver

The Hammarlund SP-600-JX is one of the great communications receivers of the vacuum-tube era — a precision general-coverage set built to military specifications, covering 540 kHz to 54 MHz, with a front-panel crystal switch that lets the operator select any of six pre-set fixed frequencies at the flick of a knob. For decades the JX served on intercept floors, embassy basements, and the listening posts of cold-war signals organisations, and even today a well-restored example will run rings around most modern radios on AM broadcast and HF utility work.

But the JX has an Achilles’ heel that gets worse every year: the crystals. The six fixed-channel positions on the front panel each take a third-overtone quartz crystal in an HC6 holder, and after sixty-plus years those crystals have aged, drifted, cracked, or simply gone missing. Replacements in useful frequencies are scarce, expensive when you can find them, and even brand-new quartz drifts with temperature.^[1] Worse, the factory channel frequencies almost never match what you’d actually want to listen to in 2026 — they were chosen for monitoring tasks that no longer exist.

This article shows you how to replace those six crystal positions with a tiny programmable module from QRP Labs called the ProgRock2^[2], disciplined by a GPS receiver, giving you eight fully programmable channels with frequency accuracy that would have been unimaginable to the engineers at Hammarlund — better than one part per billion. Along the way we’ll cover the broader package of improvements that turns a tired SP-600 into a receiver that performs significantly better than it ever did when new.

The modification is fully reversible if you wish, costs less than a tank of petrol, and requires no exotic tools or test equipment beyond a multimeter and a soldering iron.

A quick primer: what the crystal channels actually do

If you’re already comfortable with superheterodyne theory you can skip to the next section. For everyone else, here’s the short version.

A communications receiver doesn’t listen directly to the radio frequency you tune in. Instead, it mixes the incoming signal with a locally generated frequency — the local oscillator, or LO — to produce a fixed lower frequency called the intermediate frequency, or IF. In the SP-600-JX the IF is 455 kHz on the low bands and uses a second conversion on the higher bands.^[3] All of the receiver’s selectivity, gain, and AGC happens at this fixed IF, which is why superhets work so well.

Normally the LO is generated by a tunable circuit — the main tuning dial. But the SP-600-JX also gives you a second option: you can switch the LO over to a crystal-controlled oscillator instead, locked to one of six quartz crystals on the front panel. When you do that, the receiver is no longer tunable — it sits rock-solid on one frequency, determined by the crystal. This is enormously useful for monitoring a specific channel, because crystal-controlled stability is far better than any tunable LC oscillator can manage.

That’s the feature we’re upgrading. The ProgRock2 module replaces the quartz crystal with a programmable digital synthesiser whose frequency is locked to GPS satellite time, giving you something hundreds of times more stable than the original crystal could ever have been, and freely reprogrammable from a laptop.

The ProgRock2 module in plain language

The ProgRock2 is a small circuit board — about the size of a postage stamp — made by Hans Summers G0UPL at QRP Labs in Turkey.^[4] It costs about US$18 and arrives fully assembled. At its heart is a Silicon Labs Si5351A clock generator chip (or the equivalent MS5351M), which can produce three independent square-wave outputs anywhere from 3.5 kHz up to about 200 MHz, all on a single tiny board.^[5]

Three features make it ideal for our SP-600 modification:

• Eight programmable banks. You can store eight complete sets of frequencies in non-volatile memory and switch between them using three logic inputs. That maps directly onto the existing six-position crystal switch on the SP-600’s front panel — with two spare banks to spare.
• GPS frequency discipline. Feed the module a 1-pulse-per-second signal from any GPS receiver, and it continuously corrects its output frequency against the GPS satellite atomic clocks. The resulting accuracy is better than one part per billion — on a 10 MHz signal, that’s less than 0.01 Hz of error.^[6]
• HC6 crystal-can form factor. The board was deliberately designed so it can be installed inside an empty HC6 quartz crystal case, with the original pins re-used for power and signal. It can literally plug into the same socket the old crystal came out of.

Programming is done with a USB-C cable and any PC terminal program — no special software, drivers, or programmers required. You type in the frequencies you want, hit save, and they’re stored forever. The full operating manual is available from QRP Labs as a PDF.^[7]

A particularly good independent review and walkthrough of the ProgRock2 has been published by Breck Hutchins K4CHE^[8], which is worth reading alongside this article if you want a second perspective on the module itself before committing to a build.

Why GPS discipline is the killer feature

A factory-fresh third-overtone HC6 crystal from the 1950s might have been accurate to plus or minus 20 parts per million at the test temperature — call it 20 Hz of error per megahertz. After six decades of aging, thermal cycling, holder corrosion, and the slow migration of contaminants onto the quartz blank, even a “good” surviving crystal will typically be off by 50 to 200 Hz at HF. Worse, the error drifts as the receiver warms up: the JX crystal oven (where fitted) helps, but it’s not perfect, and on the high bands the crystal frequency is multiplied, so the error is multiplied right along with it.

In practical terms, this means that when you select crystal channel 3 expecting to land precisely on, say, WWV at 10.000 MHz^[9], you actually land somewhere between 9.9998 MHz and 10.0002 MHz, drifting slowly upward for the first hour as the radio warms up. For broadcast AM this doesn’t matter. For utility, ALE, time signals, RTTY, or any SSB work, it matters a great deal.

With GPS discipline, the ProgRock2’s output is referenced to the same atomic-clock standard that runs the global positioning system. The frequency is correct to within fractions of a Hertz the moment you power it on, and it stays correct forever — no warm-up drift, no aging, no temperature compensation needed. On a 30 MHz signal, you might see 0.03 Hz of error. The dial reads what the signal actually is.

Once you’ve experienced this on a vintage receiver, going back is genuinely painful.

Bill of materials

Item	Quantity	Notes
QRP Labs ProgRock2 module	1	Order with right-angle 2×6 headers from the QRP Labs shop
QRP Labs QLG3 GPS receiver	1	Or any GPS module with 1pps output
QRP Labs LPF kit (band of choice)	1	Or build a 5-pole Chebyshev manually
5V linear regulator (LM7805 or LM2940)	1	LM2940 preferred — low dropout
Bridge rectifier, 1A, 50V	1	For deriving DC from 6.3V lamp winding
Electrolytic capacitors, 470µF and 10µF	1 each	25V rated minimum
10kΩ resistors (pull-ups for bank select)	3	Quarter-watt fine
Coupling capacitor, 100pF silver mica	1	For injection to mixer grid
RG-174 or RG-316 coax	1 m	Internal RF wiring
Small bracket and standoffs	1 set	For chassis mounting

Total cost as of this writing: approximately US$60 to US$80 including the GPS module and shipping. The most expensive single item is the GPS receiver; if you already have one with a 1pps output you can save US$30.

Safety first — and this is not optional

The ProgRock2 itself operates at 5 V and is not dangerous, but the SP-600 chassis it lives inside very much is. Treat the radio with respect.

Implementation: the big picture

The modification has six distinct sub-tasks, and it pays to understand all of them before you start cutting wires. Take your time. The SP-600 chassis is cramped, the under-chassis wiring is dense, and the original Hammarlund harness is point-to-point with cloth-covered wire that frays if you look at it sideways. If you don’t have a schematic and service manual to hand, the SP-600-JX service documentation is available from the Boatanchor Manual Archive.^[10]

The six steps are:

1. Derive a clean 5 V DC supply from the existing dial-lamp winding.
2. Disable the original crystal oscillator stage so it doesn’t fight the ProgRock2.
3. Inject the ProgRock2 output into the original mixer through a low-pass filter and level pad.
4. Re-task the existing six-position crystal switch as a bank-select control for the ProgRock2.
5. Mount the GPS receiver and route its 1pps signal to the ProgRock2.
6. Program the eight frequency banks via USB and align the receiver.

Each is straightforward in isolation. Let’s take them in turn.

Step 1: deriving the 5 V supply

The ProgRock2 needs between 3.5 and 12 V DC at a few tens of milliamps. The cleanest and easiest source in the SP-600 is the 6.3 V AC dial-lamp winding on the power transformer. Rectify it with a small bridge, smooth it with a 470 µF capacitor, and regulate it down to 5 V with an LM2940 low-dropout regulator. Add a 10 µF capacitor on the output for stability.

A few notes:

• Do not tap the B+ line. Designing a regulator to drop 250 V to 5 V is wasteful, runs hot, and introduces a failure mode that can put high voltage on your ProgRock2 if the regulator dies.
• Do not use a switching regulator inside the chassis. The switching noise will couple into the IF and the front-end and you will spend a week wondering where the new birdies came from.
• The 6.3 V lamp winding is rated for several amps, so adding 50 mA of load is negligible. The dial lamps will not dim.
• Mount the regulator on the chassis with a small heat sink — the LM2940 will dissipate around 0.2 W which is harmless but it likes a heat path.

Verify the supply with a meter before connecting it to anything. You want 5.0 V plus or minus 0.2 V, with no audible hum riding on it when checked on an oscilloscope or audio amplifier.

Step 2: disabling the original crystal oscillator

The SP-600-JX uses a 6C4 triode as the crystal oscillator (V4 in most variants — consult the schematic for your exact model^[10]). This tube runs a Pierce oscillator circuit with the selected crystal in the feedback path between plate and grid. When the front-panel switch is set to one of the XTAL positions, this tube generates the local oscillator signal that goes to the first mixer (V5).

We want to silence this tube without ripping it out — because the same socket and surrounding wiring will become our signal injection point. The cleanest approach is to lift the cathode resistor at the chassis end and ground the cathode directly. This kills the bias, the tube no longer oscillates, but the plate circuit and coupling network remain intact and we can drive them externally.

Alternatively, and this is reversible in seconds, you can simply remove the 6C4 from its socket. This leaves the plate load disconnected, however, which can cause issues with the following mixer stage, so the cathode-grounding approach is preferred.

Whichever you choose, label the modification clearly with a small adhesive note inside the chassis. Twenty years from now, you — or whoever inherits the radio — will thank you.

Step 3: injecting the new oscillator signal

This is the most important step to get right. The ProgRock2’s output is a 3.3-volt peak-to-peak square wave. A square wave is rich in harmonics — it contains energy not just at the fundamental frequency you programmed, but also at 3, 5, 7, and 9 times that frequency, falling off slowly. If you connect this directly to the mixer of an SP-600, those harmonics will fall inside the receiver’s tuning range on other bands and produce a forest of false signals and birdies. The receiver will appear to have signals everywhere it shouldn’t, and you’ll hate the modification.

The fix is a low-pass filter between the ProgRock2 output and the mixer. The QRP Labs LPF kit is a five-pole Chebyshev filter on a tiny PCB, available for every amateur band.^[11] Choose the one whose cut-off is just above the highest frequency you intend to program. If your crystal channels span 3.5 to 30 MHz, the 30 m or 20 m LPF is too restrictive — use the QRP Labs 6 m LPF (50 MHz cut-off) so all of your programmed frequencies pass cleanly while harmonics above 50 MHz are crushed.

After the LPF, the signal is a respectable sine wave at about 1 V RMS. The original 6C4 oscillator delivered something in the order of 0.5 to 2 V at the mixer grid, so this is roughly the correct level. Couple it to the mixer grid through a 100 pF silver mica capacitor — the same component value Hammarlund used in the original injection network. If you find the receiver is overloading (compression on strong signals), add a small resistive pad: 100 Ω in series and 100 Ω to ground will drop the level by 6 dB without significantly disturbing the impedance match.

Keep the wiring short. The injection lead from the LPF to the mixer grid should be no more than four inches, and ideally shielded coax (RG-174 with the braid grounded at both ends).

Step 4: re-tasking the crystal selector switch

This is the most satisfying part of the modification, because it preserves the original front-panel feel. The SP-600 crystal switch is a multi-wafer wafer-switch with one wafer dedicated to selecting which crystal is in circuit. We’re going to keep using that switch — it just selects ProgRock2 banks instead of physical crystals.

The ProgRock2 selects banks via three logic input pins. Three bits gives eight banks (binary 000 through 111). The original crystal switch has six positions, so we map them as follows:

Switch Position	Bit 2 (BS2)	Bit 1 (BS1)	Bit 0 (BS0)	ProgRock2 Bank
XTAL 1	0	0	0	Bank 0
XTAL 2	0	0	1	Bank 1
XTAL 3	0	1	0	Bank 2
XTAL 4	0	1	1	Bank 3
XTAL 5	1	0	0	Bank 4
XTAL 6	1	0	1	Bank 5

Wire one wafer of the crystal switch so that each switch position grounds the appropriate combination of BS0, BS1, and BS2 lines. Add a 10 kΩ pull-up resistor from each of the three lines to +5 V. When a line is grounded by the switch it reads logic 0; when left floating it reads logic 1 via the pull-up. This is standard active-low logic and the ProgRock2 handles it natively.

The two unused banks (6 and 7) can be reached by reprogramming or by adding two extra switch positions on a panel-mounted toggle if you have the space and inclination. Most operators find six channels plenty.

Step 5: GPS receiver and 1pps connection

The QRP Labs QLG3 GPS receiver is a compact PCB-mounted module with an SMA connector for an external antenna and a single output pin that delivers a 1-pulse-per-second logic signal locked to GPS time.^[12] Wire this output directly to the ProgRock2’s GPS input pin with a short length of small-diameter coax (RG-174). The QLG3 runs on 5 V — the same supply you built in Step 1.

Mount the GPS antenna outside the chassis, ideally with a view of the sky. A magnetic-base puck antenna on the equipment rack or window sill works well. The cable can be several metres long without issue. Do not put the GPS antenna inside the SP-600 enclosure — the chassis is a Faraday cage and you’ll never get a lock.

Once the GPS has acquired satellites (typically 30 to 60 seconds from cold start, with at least three satellites visible), the ProgRock2 will indicate GPS lock with a brief blink of its onboard LED at each 1pps pulse. From this moment on, your crystal channels are atomic-clock accurate.

Step 6: programming the channels

With the ProgRock2 powered up, connect a USB-C cable from your laptop to the module. Open any serial terminal program — PuTTY on Windows, screen or minicom on Linux or macOS — at 9600 baud, 8 data bits, no parity, one stop bit. The ProgRock2 will respond with a menu. Programming a frequency is as simple as typing the bank number and the frequency in Hz. The complete command set is documented in the official ProgRock2 manual.

Remember: you are programming the local oscillator frequency, not the received frequency. The local oscillator is offset from the received frequency by the IF — 455 kHz on the low bands of the SP-600. Whether to add or subtract the IF depends on whether the receiver uses high-side or low-side injection on the band of interest, which varies. The simplest approach is to program a candidate frequency, switch on the receiver, and look for the signal you expected. If it isn’t there, try the same frequency plus or minus 455 kHz. Within two tries you’ll find it.

For the high bands (band 5 and 6 on most JX variants), the receiver is double-conversion: the first LO mixes the incoming signal down to a first IF (typically 3.955 MHz), and then a second crystal-controlled oscillator mixes that down to 455 kHz. The ProgRock2 normally replaces only the first LO. The second conversion oscillator usually has its own dedicated crystal at a single fixed frequency (3.5 MHz on most JX models) and can be left alone, or replaced by a second ProgRock2 if you want full atomic-clock accuracy on those bands.

Suggested channels for a modern listening post: WWV at 5, 10, and 15 MHz; CHU Canada at 7.850 and 14.670 MHz; the 40 m and 20 m ham band edges; and a couple of HF utility frequencies of personal interest.^[13] You have eight banks — use them well.

Making it perform better than ever: stability and noise reduction

GPS-disciplined crystal channels solve one specific problem, but the SP-600 has several other age-related issues that quietly rob it of performance. Address these and the receiver will outperform anything Hammarlund ever shipped from the Mahwah factory.

Migrating mica disease in the IF transformers

This is the single most important fix you can make. The silver-mica capacitors inside the SP-600’s IF transformer cans suffer from electrochemical migration: silver ions migrate across the mica dielectric over decades, gradually short-circuiting the capacitor. Symptoms include weak gain, poor selectivity, instability, regeneration, and occasional motorboating.^[14] By 2026, essentially every unrestored SP-600 has at least one IF transformer suffering from this.

The fix is to disassemble each IF transformer, remove the original silver-mica caps, and replace them with modern silvered-mica or NPO ceramic capacitors of the same value. This is fiddly work — the cans are small, the original caps are buried in beeswax, and you’ll need to retune each transformer after reassembly — but it is the single most transformative restoration step on a tired SP-600. Plan a weekend.

Paper and electrolytic capacitor replacement

The SP-600 contains dozens of paper-wax and paper-foil capacitors in the audio, AGC, and B+ filter lines. Every single one of these is now leaky to some degree. Leaky coupling caps cause distorted audio, weak AGC action, and unexplained DC drift on tube grids. Replace them all with modern polypropylene or polyester film capacitors of the same value. The electrolytic filter capacitors in the power supply have also dried out and should be replaced with modern equivalents of equal or slightly higher capacitance and at least the original voltage rating.

This step alone will drop your noise floor by 6 to 10 dB on most sets and clean up the audio dramatically.

Tube testing and matched replacement

Weak tubes are silent thieves of performance. The RF amplifier tube (V1) and first mixer (V5) particularly affect sensitivity, while the audio output stage affects fidelity. Test every tube with a proper tester (TV-7 or similar) and replace anything below 70% of nominal emission. NOS Western Electric or Telefunken examples in critical positions can be transformative. Don’t bother with the cheapest current-production tubes — the difference in noise figure between a tired 1950s 6BA6 and a fresh JAN-CRP Sylvania 6BA6 from old stock is significant.

Power supply ripple reduction

The original SP-600 power supply uses choke-input filtering with relatively modest capacitance by modern standards. Adding an additional 100 µF filter capacitor across the existing B+ filter, or replacing the choke-input with a capacitor-input filter, drops residual hum by another few dB. Be careful not to exceed the rectifier tube’s peak current rating if you go to capacitor-input — the original 5R4 or equivalent has limits. A solid-state rectifier replacement plug (which I have covered in a separate article on this site) addresses this neatly and eliminates the rectifier filament drain into the bargain.

Grounding and bonding

The SP-600 was designed with single-point grounding, but sixty years of cold solder joints, corroded chassis screws, and oxidised paint under washers degrade this. Go through the chassis and re-tighten every grounding screw. Where paint is between a grounding lug and chassis, scrape it back to bare metal. Add a heavy braid bond between the IF strip and the front-end shielding compartment. This kind of work pays off most on the upper bands where ground loops translate directly into in-band birdies and hash.

Front-end antenna matching

The SP-600 antenna input is balanced and high-impedance, designed for a doublet feeder rather than a 50-ohm coaxial line. Feeding it directly from a modern 50-ohm antenna will work, but the impedance mismatch costs you 3 to 6 dB of signal and increases susceptibility to common-mode noise on the coax shield. A simple 4:1 balun on the antenna input (or a proper antenna tuner ahead of the receiver) recovers this loss and reduces noise pickup noticeably.

RFI from modern domestic noise sources

A 2026 listening environment is vastly noisier than 1955. LED lighting, switching power supplies, plasma TVs, solar inverters, and PLC ethernet devices all radiate into the HF spectrum at levels that drown out genuine signals. Identify and eliminate the worst offenders in your own house first — turn things off one at a time and listen for the noise floor to drop. Replace LED globes near the radio shack with incandescent or RFI-quiet LED types. A common-mode choke on the receiver’s mains lead is also worthwhile.

Testing and final alignment

With the modification installed, the rebuild work complete, and the ProgRock2 programmed with your chosen channels, the final step is alignment. The good news is that you now have a GPS-disciplined frequency reference inside the receiver — so for the first time in the SP-600’s life, alignment can be done against an absolute standard rather than against a drifting signal generator.

Use a known-good channel (for example WWV at 10.000 MHz, with the ProgRock2 set so the receiver should land exactly on it) to verify dial calibration on band 4. If the dial is more than a hundred Hertz off, retrim the appropriate band-set capacitor. Repeat for each band. Then perform a standard IF alignment using a signal generator at 455 kHz, adjusting each IF transformer for maximum output while listening to a steady carrier from one of your crystal channels. The combination of a stable LO and freshly recapped IF transformers usually yields an IF passband shape much closer to the factory specification than the receiver has seen in decades.

Allow the receiver to warm up for at least thirty minutes before judging the result. A properly restored and GPS-disciplined SP-600-JX will hold a CW signal so steadily on its crystal channel that you can use it as a frequency reference for other equipment.

What you should expect — measured improvements

A representative SP-600-JX, before and after this complete package of modifications, typically shows the following changes:

Parameter	Before	After
Crystal-channel accuracy at 10 MHz	±50 to 200 Hz	<0.01 Hz
Warm-up drift over first hour	100 to 500 Hz	None
Number of usable preset channels	0 to 6 (whatever crystals are present)	8, freely chosen
Audio noise floor (typical)	-50 to -55 dB	-65 to -70 dB
IF selectivity (after recap)	Variable, often skewed	Restored to spec
SSB intelligibility on crystal channels	Poor (drifts in and out)	Excellent and stable

A few closing thoughts

This is not a modification I would have endorsed twenty years ago. At that time, original crystals could still be obtained, replacement IF transformers existed in the NOS market, and the SP-600 community was largely focused on faithful restoration. In 2026 the calculus has changed. The crystals are gone, the receivers that survive deserve to be used, and the operating environment has become so much noisier than the 1950s that a vintage receiver’s best chance of remaining relevant is to be made as accurate and as quiet as we can possibly manage.

The ProgRock2 modification is, in that light, deeply faithful to the original Hammarlund engineering intent: the JX’s crystal channels were always meant to deliver rock-solid frequency stability for serious monitoring work. We are simply using 2026 technology to do what the original Hammarlund engineers would have done if they’d had access to it.

If you carry out this modification, please document it. Photograph the install, label the chassis, keep the original crystals in a labelled bag taped to the cabinet, and write a note for the next owner. The SP-600-JX will outlive all of us, and the next steward of yours will want to know what was done and why.

Questions, corrections, or photographs of your own build are welcome — please get in touch through the contact form on this site.

Credits and acknowledgements

No piece of work like this is done alone. The following people and organisations deserve direct credit for making this modification possible:

• Hans Summers G0UPL — designer of the ProgRock2 and founder of QRP Labs. Hans’s commitment to producing thoughtfully designed, affordable, and well-documented kits for the amateur radio community is the reason this modification is even practical. His broader catalogue of synthesisers, transceivers, and balloon trackers has done more to advance accessible HF experimentation in the last decade than any other single source.
• Silicon Labs — for the Si5351A synthesiser silicon that sits at the heart of the ProgRock2 and dozens of other amateur-radio products. The Si5351 family demonstrates what becomes possible when a major semiconductor manufacturer publishes thorough datasheets and reference designs for hobby-grade use.
• Breck Hutchins K4CHE — for the excellent independent walkthrough of the ProgRock2 that complements the official documentation and shows the module in real-world use.
• The Boatanchor Manual Archive (BAMA) — for preserving and freely redistributing the SP-600 service manuals, schematics, and alignment instructions that make any serious restoration work possible.^[10]
• Chuck Rippel WA4HHG and the broader boatanchor restoration community — for decades of patient, careful documentation of the SP-600’s specific failure modes, particularly the migrating mica problem in the IF transformers, without which restorers would still be chasing ghosts.
• The original Hammarlund engineering team at Mahwah, New Jersey — who in the 1950s designed a receiver so well that it is still worth investing this much effort in seventy years later. The SP-600 stands as a monument to a generation of engineers who built things to last.

Any errors, omissions, or misinterpretations in the article above are entirely my own.

Footnotes

1. Quartz crystal aging is well documented in the time-and-frequency literature. Even high-grade reference crystals typically drift on the order of 1 to 5 parts per million during their first year, with continued slower aging thereafter. Industrial-grade HC6 crystals of 1950s manufacture, with the dielectric and electrode materials of the period, age considerably faster. See for example Vig, J. R., “Quartz Crystal Resonators and Oscillators for Frequency Control and Timing Applications,” US Army Communications-Electronics Command technical report (multiple editions). ↩
2. QRP Labs ProgRock2 product page: https://qrp-labs.com/progrock2.html. The page includes full specifications, the user manual, firmware update files, schematic, and pinout diagram. ↩
3. The exact band-by-band conversion scheme varies between SP-600 sub-models. Single conversion to 455 kHz is used on the lower bands (1 through 4 on most variants); double conversion with a first IF at 3.955 MHz is used on the higher bands (5 and 6). Consult the schematic for your specific model number. ↩
4. QRP Labs is based in Turkey and operates internationally. Hans Summers G0UPL has been producing amateur radio kits and finished products since the late 2000s. The full catalogue is at https://qrp-labs.com/. ↩
5. The Silicon Labs Si5351A is a clock generator IC capable of producing three independent output clocks from a single crystal reference, using fractional-N PLL synthesis. The MS5351M is a pin-compatible alternative used when Si5351A supply has been constrained. Datasheet available from Silicon Labs at https://www.silabs.com/. ↩
6. The 1pps output of a typical consumer GPS receiver is accurate to better than 50 nanoseconds RMS, which after integration over many cycles by the ProgRock2 firmware translates to fractional-Hz frequency accuracy at HF. The exact figure depends on satellite geometry, antenna placement, and the GPS module used. ↩
7. ProgRock2 manual (version 1.00b, February 2023): https://qrp-labs.com/images/progrock2/ProgRock2_1_00b.pdf. Covers programming syntax, GPS discipline configuration, firmware update procedure, and all hardware connections. ↩
8. Breck Hutchins K4CHE’s ProgRock2 reference page: http://k4che.com/Progrock2/progrock2.htm. Linked directly from the QRP Labs product page as a recommended external resource. ↩
9. WWV is the high-frequency time-and-frequency station operated by the US National Institute of Standards and Technology in Fort Collins, Colorado, broadcasting on 2.5, 5, 10, 15, and 20 MHz. Carrier frequencies are maintained to within 1 part in 10¹³ as referenced to the NIST-F1 atomic clock. https://www.nist.gov/pml/time-and-frequency-division/time-distribution/radio-station-wwv. ↩
10. The Boatanchor Manual Archive (BAMA) hosts service manuals, schematics, and technical documentation for the SP-600 and its many sub-variants. The community-maintained edition is at https://bama.edebris.com/. ↩
11. QRP Labs LPF kit product page: https://qrp-labs.com/lpfkit.html. Available in versions for every amateur band from 2200 m to 6 m. Five-pole Chebyshev design with toroidal inductors; harmonic suppression typically better than 40 dB. ↩
12. QRP Labs QLG3 GPS receiver product page: https://qrp-labs.com/qlg3.html. Active patch antenna included. Outputs include serial NMEA data and the 1pps timing pulse used by the ProgRock2. ↩
13. CHU Canada is operated by the National Research Council and transmits on 3.330, 7.850, and 14.670 MHz. Details at the NRC CHU page. ↩
14. The silver-migration failure mode in vintage silver-mica capacitors is well documented in the boatanchor restoration community. The problem is not unique to the SP-600 — it affects Collins R-390, R-390A, Hallicrafters SX series, and most other receivers of the period that used silvered-mica caps for IF tuning. The mechanism involves electromigration of silver from the silvered electrode through small fissures in the mica dielectric. ↩

---

Mike Peace VK6ADA / r-390a.net Administrator