38 Special Prototype Lab Tests

by Paul Harden, NA5N

About the AmQRP Club
Frequently Asked Questions
Kits
Projects
QRP Forums
Contesting
References
Links                                                     

"38 Special" Prototype Lab Tests

LAB TEST: NorCal 38 Special

QRP KIT FROM: NorCal QRP Club

A Few Comments with the lab tests following.

Doug Hendricks KI6DS sent me the prototype kit of the 38-S, which I received Dec. 24th and built Christmas Day. It consisted of the PC board and most major components. Since NorCal is awaiting the rest of the parts to arrive, I did have to scrounge some of the parts, but managed to find exact values. My PCB (the prototype) did not have the parts masking on it as will the kit units have. Be thankful for that ... with no silk screening, you start out not having a clue what holes C11 goes into -hi. Board assembly time was about 3.5 hours; there are 4 toroids to wind. When done, I counted 93 parts on the board. Only external parts required is the RF gain (1K pot), the TUNE control (100K pot), and desired connectors for the antenna, phones, key and power. All connections to the board are made to numbered holes along the side of the board; the holes are on the board for the TicK Keyer chip from Embedded Research, the 5W IRF510 mod, the RIT and the extra IF filter crystal. I was not given any info on incorporating these mods, however. I was told to build my prototype "stock" and leave it that way. Thus, the following measurements are based on the stock kit. Ori said he mailed me the info on the 5W IRF510 mod, so hopefully they'll let me add that before long.

SIGNIFICANT SPECIFICATIONS:

   Tuning Range: 10.110 to 10.133 MHz (23 KHz)
       Type VFO: VXCO; 22MHz crystal, VVC tuned (with 1N4004 diode)
            MDS: -128dBm
 IF & IF Filter: 12MHz IF; single 12MHz crystal filter
  Dynamic Range: 90dB 
    Selectivity: -6dB BW: 190 Hz  (Includes the on-board CW audio filter)
                 -12dB BW: 390 Hz
                 -30dB BW: 1.5 KHz
  LSB Rejection: -12dB (see note 1)
Receive Current: 38mA (see note 2)
   Output Power: 400mW (+26dBm) into 50 ohm resistive load
                 350mW measured on a 30M inverted vee
Harmonic reject: 2nd harmonic at -16dBm or 42dBc below fundamental

NOTE 1: There are holes on the board for adding a toroid and second IF
        filter crystal to lower the LSB rejection.  I have not added
        this to the prototype yet to measure its effect.
NOTE 2: 38-S is powered off of +8v from a TO-220 7808 (off +12v in).  The
        7808's had not come in yet when they sent me the prototype.  I
        made +8v with a zener and series-pass transistor, which likely
        elevates the receive current a few mA.

IF FILTER -6dB bandwidth is about 1.4KHz (Noise) and 1.0KHz (CW input)
          from the stock single-crystal filter.

AUDIO FILTERING/AMPLIFIER uses the NE5532 dual audio op amps. The two sections make for a surprisingly tight filter. If you're not a fan of audio CW filters, this circuit will change your mind.

There was lots of discussion with Ori about whether the 200Hz bandwidth was accidental or typical. The other three prototypes built by others yield very similar results. I therefore offer the numbers I recorded in determining the selectivity so those who build it can compare with their results. (And please post what yours does).

Bandwidth   low freq.   hi freq.   Bandwidth
---------   ---------   ---------  ---------
 -3dB         650 Hz      770 Hz     120 Hz    Resulting audio frequency
 -6dB         620         810        190       at output with RD signal
-12dB         520         910        390       generator at -100dBm.
-30dB          90        1610       1520
 

A FEW SOAPBOX ITEMS:

1. There has been some circuit/component changes to improve performance 
   that do not appear on the NorCal web page or Winter QRPp article.

2. The 38-S is kinda different in that it uses a common NE602 for both
   the Product Detector and the Transmit Mixer, with CD4066 analog
   switches doing the signal switching between XMIT and RCV.  I measured
   less than 0.2dB loss through these switches.  The scheme works nice.

3. At the upper end of the tuning range (10.130) are those RTTY stations.
   I found those to be excellent for aligning and checking the receiver.
   At midnight, those RTTY stations produced 150-200mVpp audio output
   when the tones were centered in the CW filter passband. 

4. Hardcopy plots were made of the IF and audio filtering, output
   spectrum, etc.  I'll be sending these to Ori and Doug for inclusion
   into a future QRPp or on the NorCal web page if they want.

6. I'll re-do some of the tests when I do the 5W IRF510 mod.  Then, I'll
   go on 30M early evening some night in conjunction with the FOX hunt
   or something if you want your first chance to work a 38-S.

Nice quality board and parts, worth well over the $25. For those who have already ordered the 38-S, I think you'll be pleasantly surprised at its performance and will find it a real working radio.

72,

 
Material and concepts presented on The American QRP Club (TM) website is Copyright 2003 by The American QRP Club, Inc.
These pages are designed and maintained by George Heron, N2APB
Page last updated:  April 15, 2004