BUILDER ALERT #9: Noise Blanker Changes
Feb 3, 2000
APPLICABILITY: Noise Blanker boards shipped prior to 1/31/2000. Noise Blanker Manual revision B or earlier.
We have discovered that the original KNB2 noise blanker impacts K2 receive performance somewhat, even when turned OFF. (Nearly all noise blankers, including ours, cause more intermodulation problems when turned ON. When turned OFF, they should have little or no effect.) Qualitatively speaking, with the noise blanker installed, the K2's receiver dynamic range goes from "superior" to "average" (as compared to other multiband commercial transceivers). For more details see the technical notes at the end of this document.
If you live in the U.S. you may not have noticed the problem, since it still takes very large signals to cause IMD. But our customers in Europe--many of whom live near short-wave broadcast stations--found that they had an increase in intermodulation on 40 meters at night. We confirmed the problem in our lab, and have now re-designed the noise blanker to eliminate it (see Technical Notes at the end). We're grateful for the help we received from our EU customers and the ARRL in testing the new circuit.
We will send the new parts free of charge to anyone who would like to make the changes. First, review the modification instructions below. Then, if you'd like to proceed, send a request for the NB Mod Kit by E-MAIL, MAIL, or FAX. We will only be sending out COMPLETE NB Mod Kits, so there's no need to ask for only a subset of the parts.
PLEASE DO NOT TELEPHONE to request the parts, since handling all of the phone
calls will just slow things down.
In the case of e-mail, please send the request ONLY to:
and use the following subject line: NB Mod Kit Request
Note: All noise blankers shipped after Jan. 31, 2000 will include the new parts, along with assembly manual revision C or an appropriate errata sheet.
Noise Blanker Modification Instructions
Note: Be very careful not to damage PC board pads or traces when removing components; if possible, use a fine-point, temperature-controlled iron. You may find it helpful to use small-size solder wick on *both* sides of the board before lifting components out. If you use a vaccuum-desoldering tool, use the largest possible unit. Smaller "solder suckers" are not nearly as effective.
1. Remove C1, C2, C3, C4, C5, C6, C7, C16, D1, L1, and L2.
2. Install the following new components at these locations.
Note: C3, C4, C6, and C7 MUST have +/- 5% tolerance. Use NPO disc, C0G monolithic, Silver Mica, Polystyrene, or other high-quality type. (We have included Digikey/Mouser part numbers for reference.)
|PCB Location||New Value||Part Number|
|C1||120-ohm resistor||Digikey 120QBK-ND|
|C2||300-ohm resistor (or 330 ohms)||Digikey 300QBK-ND|
|C3||820 pF, 5%||Digikey P4860-ND|
|C4||100 pF, 5%||Digikey P4849-ND|
|C6||1500 pF, 5%||Digikey P4863-ND|
|C7||1800 pF, 5%||Digikey P4864-ND|
|C16||39 pF (any type, 5% or 10%)||Mouser 140-100N5-390J|
|D1||1-megohm resistor (5% or 10%)||Digikey 1.0MQBK-ND|
Note: Use *only* the core type and number of turns shown below, and space the turns evenly. DO NOT adjust the turns count based on an inductance measurement unless you have complete confidence in your inductance meter--readings can vary widely for different instruments. The Autek RF-1 and MFJ 259 series, for example, can vary over 20% from the actual value, even more for different frequencies of measurement.) If you do attempt to fine-tune the inductors, the most important thing is to make them *equal* in inductance.
|L1||1.1 to 1.2 uH (16 turns #26 or #28 enamel wire, evenly spaced on 80-90% of a T37-2 [red] toroid core, nominal 1.15 uH)||T37-2 (Amidon)|
|L2||Same as L1||T37-2 (Amidon)|
3. Solder a miniature .01 uF capacitor (any type) between pins 6 and 7 of U1 (MC1350) on the bottom of the noise blanker board.
4. On the K2 RF board, remove R88, R89, and R90. Install a short jumper at R89 (0 ohms). Install a 2.7 K resistor at R88. (Digikey 2.7KQBK-ND)
We re-designed the noise blanker's band-pass filter to reduce its impact on the post-mixer amplifier. Extremely strong in-band signals coming from the 2N5109 were seeing a fairly wide impedance range coming from the NB filter, depending on how the builder adjusted the trim caps.
In redesigning the filter, we removed the original -5 dB pad from the RF board (R88-R90), and instead incorporated the 5 dB of loss into the NB filter itself (hence the 120 and 330-ohm resistors). By using these resistors as filter coupling elements, we were able to keep the same filter topology. Filter loss at 4.915 MHz is about 6 to 7 dB, which is about the same as the original noise blanker loss (1 to 2 dB) plus the original 5 dB pad. Return Loss (RL) through the filter is now around -14 dB over a wide range of frequencies and termination impedances. (This is equivalent to saying that the filter shows an SWR of 1.5:1 to the driving amplifier.)
Receive dynamic range with the modified NB installed is excellent, with a 3rd Ip of +15 to 20 dBm and a dynamic range of 97 to 99 dB (confirmed by ARRL lab tests).