Testing vintage quartz crystalsΓÇöespecially the large, low-frequency pressure-holder types used in 1940s radiosΓÇörequires a specific approach. If you simply hook a signal generator directly to one pin of the crystal and a scope probe to the other, the high input capacitance of the scope probe (typically $10\text{ to }15\,\text{pF}$) and the stray capacitance of your test leads will severely pull or damp the crystal.
Because the crystal behaves as an incredibly high-$Q$ series resonant circuit at its target frequency, you need to isolate it from your test equipment to see its true, unforced resonance point.
Here are the two best methods for mapping that mystery crystal on your bench using your OWON generator and scope.
---
## Method 1: The "Low-Impedance Resistor Pad" (Best Out-of-Circuit Test)
To prevent your test equipment from detuning the quartz slab, you must swamp out the stray capacitance using low-value resistors. This creates a simple voltage divider where the crystal acts as a near-dead-short only at its exact series resonant frequency ($f_s$).
### The Bench Setup:
1. Connect a **$50\,\Omega$ or $100\,\Omega$ resistor** from your signal generator's hot output to one pin of the crystal. 2. Connect a **$100\,\Omega$ resistor** from the other pin of the crystal to ground. 3. Connect your scope probe across that second $100\,\Omega$ resistor (the one going to ground). 4. Tie the grounds of the generator and scope together.
### The Testing Procedure:
1. Set your OWON generator to output a clean sine wave at roughly 2 V RMS.
2. Set your scope to view AC voltage, zooming in so you can see a few millivolts of deflection clearly.
3. Slowly sweep the generator's frequency from 450 KHz to 465 KHz. Move in tiny increments (like
10 Hz steps) because a high-Q crystal's resonant peak is razor-thin.
4. **Watch for the Peak:** As you hit the exact series resonant frequency, the impedance of the
crystal drops close to zero ohms. Suddenly, a massive sine wave will burst onto your scope
screen. Note this exact frequency.
5. **Watch for the Valley:** If you keep tuning slightly higher (usually just a few hundred
Hz or 1 KHz up), the signal will suddenly drop off a cliff into a deep valley. This is the
*parallel* (anti-resonant) frequency ($f_p$), where the crystal acts like an open circuit.
> **What to document for your site:** For the SX-28's crystal filter bridge, the **Series Resonant Peak ($f_s$)** is the actual number that matters. That peak is the exact frequency your IF transformers will need to be aligned to.
---
## Method 2: The "In-Circuit" Sweep (The Real-World Test)
If you drop the crystal into the radio's crystal socket, you can use the SX-28's own circuitry to map it. This has the advantage of showing you exactly how the crystal interacts with the real-world impedances of the phasing bridge.
### The Bench Setup:
1. Disable the radio's Local Oscillator (pull the oscillator tube or inject directly past it) so it doesn't create conflicting signals. 2. Inject your OWON signal generator directly into the grid of the Mixer tube or the 1st IF tube. 3. Connect your scope probe (set to $10\times$ to minimize loading) to the secondary of the very next IF transformer following the crystal filter. 4. Switch the SX-28's front panel Crystal Filter to **Position 1 or 2 (On)**, and center the Phasing control knob perfectly.
### The Testing Procedure:
1. Keep the generator amplitude low (around $100\text{--}500\,\mu\text{V}$) to avoid driving the IF tubes into distortion or activating the AVC. 2. Sweep the generator slowly across the $450\text{--}465 KHz range. 3. The scope will show a definitive, sharp peak when the generator matches the crystal.
---
## Evaluating the "453 kHz" Crystal
If your spare crystal peaks at **$453 KHz** while your current one is stuck up at **$462 KHz**, you have a fascinating choice to make for your website's ultimate alignment workflow:
* **The Math:** The factory target is 455 KHz. * Your current crystal is **$+7 KHz** high ($462 KHz). * Your spare crystal is **$-2 KHz** low ($453 KHz).
A $-2 KHz tracking error is vastly easier for the SX-28's factory inductors and trimmers to absorb than a $+7 KHz error. If you drop the $453 KHz crystal into the radio, your front-end tracking problems on Band 1 will likely vanish entirely, your S-meter will start moving normally, and your sensitivity will skyrocket because the front-end RF stages will finally align with the local oscillator's target window.
When you run the out-of-circuit resistor pad test on that spare crystal, where does its sharpest series peak actually land?
Page: /1_need_help/Gemini_26/Gemini_26.shtml
Last modified: Tuesday, 21 Jul 2026