This is the integrated dish unit and not the 'offical' white box. It's more of a "White Tub".

A buddy of mine, Galen (W8LNA), told me about the availability of a few White Box units from a Ham back East (Jeff, WA3ZKR). A few traded emails later, a big box was on the door step!

Some on-line White Box mod references. Why reinvent the wheel? when somebody has already gone to a great deal of trouble and then wrote excellent step-by-step instructions!

• G3PHO mods pages.
• WA6CGR article from Microwave Update 1991 .
• John's "Geekworks" Notes and improvements

The LO: Galen ordered new xstals and sent me one. (103.0794 MHz x 102)-146 MHz IF = 10368.1. The choice of 146 as the IF was to get away from the seak signal freq (144), and to give a little 'play' room for oscillator stability. What's a few MHz among friends? The xstal osc. is multiplied 17 x (1752 MHz), and then phase locked to a VCO and finally mulitplied 6x (10514 MHz) for the output. I popped off the rock heater and plugged in the new rock. Applied +20 v, let it warm up a bit, and then listened for the VCO and the xco on a separate receiver (Yaesu ft-5000). Nada...tuned out the leads on the rock were shorting out. Snip 'em and try again.

Well, the rock was there, very strong, and pretty stable. I could chase the multiples up the band. There was even a little signal at the VCO freq. wandering pretty wildly.

I put in the G3PHO LO mod to tie down the AFC (zener diode and a voltage divider, and the cheap and easy VCO 'lock' indicator - LED and resistor to ground from the VCO alarm).(NOTE, in G3PHO's LO mods, the VOC LOCK indicator 'lights up' when locked. In the WA6CGR mod, the LED goes dark when locked).

Is there output at X-band? An old x-band radar detector would have been informative, but lacking even that I need more test gear. I ran into Phil, W6HCC at a hamfest and small talk tuned to the white tub. Phil was willing to give me a hand, so the white tub went to visit. Putting the LO on a real u-wave power meter showed no output. A quick glance on his Spec Analyzer showed no VCO output.The Xco wouldn't even tune to the xstal freq...hmmm. Poking around with a voltmeter, and refering to G3PHO's LO scehamics, a problem quickly became apparent. The regulated bias for the phase locked source (the real fancy transistor transistor refore the x6 multiplier) showed rail voltage, 20V!. Well, maybe the regulator is paws up? (2n2907). A quick trip to Phil's 2n2907 bin; snip, snip, blob and daub, in goes a regulator can.

Fire the LO up again, and STILL rail voltage on the bias. Phil just happened to have an old M/A Comm white box LO sitting amungst his collection. We checked it's voltages for the PLL transistor and they were about where G3PHO's shematic indicated. Plug my rock into Phil's LO and it locked right on (so said the SA). The output was measured at about 20mW. Phil didn't like having an AFC circuit on his LO, so he yanked it out (similar to WA6CGR's LO mod, yank the varactor diode and assorted resistors and replace with a jumper wire). Phil kindly gave me his LO, in trade for my non-oscillating LO. Thank you very much!

So now I have an X-band LO! The VCO Lock indicator lights right up at 18v, but looses lock at 20v. After it warms up a bit, it'll stay locked at 20v, but it's not as stable. And, of cource the Voltage controlled Oscillator (VCO) drifts up as the voltage changes. Phil mentioned that these M/A Comm LO's are frequently tossed because of there instability (see John Ackermann's M/A Comm LO stability tests). At this point I'm just too excited to have an x-band LO to worry about a few MHz...what's a few MHz among friends?

Now on to the filters!

I surfed and surfed for SOMETHING that I could cobble together for a source. The closest thing I found was a cheezy comb generator with a diode (I don't have) connected across a BNC connector with a short loop a wire (http://www.vk2zay.net/combgen.html). I tried a similar conglomeration with a 1n34 (capacitively coupled and a pot to tweak bias and a dc return) and fed a milliwatt or 5 from my 3.4 box (third multiple to 10368). Not a lot of luck, but then I couldn't really 'tune' it to favor much of any output anywhere but 3.4. With a little more desire, time and test equipment (an x-band receiver), I could probably make it function to some extent, (mounting it in some waveguide might help) but...

I followed the mod article with the 'turn all screws a half-turn before alignment', and hoped I'd get lucky...nope no beacons. So, "Hello Phil?"... We tweeked and tweaked, and couldn't make the filter move...Phil recalled that he had not ever bothered messing with the Rx filter on previous White Box hacks (Now he tell's me! It's probably been 20 or 30 years since he saw a whole one of these...Oh, well). -30 dBm from a synthesized driven brick sitting across the room was still screaming, turn, turn, tweak, tweak. Didn't make a difference. As best as we could tell, the things a barn door. (Maybe too much signal?)

Put it all back together (after about 3 hours of fiddling), pointed it toward Mt. Thorodin (K0RZ/B 10368.044 MHz) and fired it up. Punch some numbers on the calculator to figure out where it should be and...

There is was. A few kHz off, but dang close. S3 signal. I pack up, run home and set up again...nuthin. I drive to the edge of town and set up portable...there it is! Still S3 (60 miles away). Feeling pretty excited as I drive back, I notice that neighbor Ken (kc0cou) is home and I figure I'll show off my x-band receive...set up in front of his house...nuthin! And I'm only >0.5 mi. from where I just was! Still with mostly open ground looking towards the beacon. Must be a little chunk of granite (or trees, or houses) along the path. Funny part is, Ken had borrowed a 10 Gig box for a while and could hear the beacon just fine with a little horn pointing south. There's been a great deal of construction to the south since then, but...hmmm. It makes me feel better about not hearing it from my house either.

The little antenna is very sharp. Three degrees or so if that much. I need to work on the power supply (get it out of the tub to make it lighter and facilitate Tx/Rx switching), come up with a mount (tripod like thing), and then it's on to the IF Tx attenuator and the transmit filters..."Hello Phil?..." I'll definitely need his help aligning those guys.

I got the PSU out of the tub and into a box. This lets me put in switching for the x-band tx/rx (+12v pins), and an tx/rx relay (run off of an unused PSU 12V output). The relay is a surplus motorola thing (hamfest find) that 'really' wants 23 V, but switches quite happily at 20V. I put the Rx on the NO, so if there's ever any kind of problem I'll transmit in to the transmit port and NOT the Rx port. Don't forget the diode at the relay terminals to keep the transients down. A single dpdt switch throws the thing between tx/rx (one side is the 12V for the x-box, the other is the 20V for the relay).

I'm a sucker for blinky lights. One led for pwr on indicator; one from the ACF Lock for VCO lock indication; then one each for rx/tx status (My dpdt switch is center off, so I can let the LO warm up with nothing else going on). I though about putting in another led, just to round things out, not to tell me anything (just 'cause I like blinky lights), but I ran out of diodes...Since I'm using the same +20V source (power transistor from the PSU) for both the relay and the LO, there's a little change in frequency when switching between Tx/Rx. It's not a whole lot (about 50 Hz). It's noticable, but not a big problem.

For the IF TX attenuator (the x-box only want +20 dBm - thats 100 mW if you don't have a conversion chart handy), I figured I'd have to knock 5w (+37 dBm) down to a managable level. 20-37=-17 dB (dBm for the purists). The ARRL handbook, chpt 25. is a good reference for such things. The ARRL microwave experimenters manual is another excellent reference. You might look at N5ESE's attenuator pages for some ideas if you're not sure of were to begin. I stuffed 9 x 100 ohm 1 watt resistors (rs#271-152) in a 'pi' configuration into a little box with bnc connectors on each end. For 16 dB atten, use 68.8 ohms in series and 2x 153.8 ohms parallel. My cobbled combination of 100 Ohms comes up with 50 ohms series and (2x)150 ohms in parallel (before and after the series resistor). Plug one end with a 50 ohm load, and apply a little dc on the other end. Measure the volts in, and out.

The approx attenuation is then = 20*log (Vout/Vin). Mine measures -16.9 dB. I'll measure it at 144 MHz when I go visit Phil again. And try to get a picture when I figure out how to get them to show up on pages on the server.With the FT-100d on lowest power, an FM carrier for 1 min. started to get a bit warm. That'll work!

One last trip to W6HCC's for test equipment and fiddling (gotta retune those x-band Tx filters). We measure the Tx attenuator on the power meter (-20 dB)! The added line loss; the six connectors; the relay; the dead bug style of construction, all add a little loss that really starts to add up.

We had a little trouble coupling the Spec. Analyzer to the 'correct' output. With the back cover off of the x-band module, leakage was probably out problem. Twist the resonator screws back and forth (never did touch the appature screws) and nuthin' much really happened. Finnaly Phil decided to take the SA probe and couple it directly with the waveguide output out the module. Now when the screws where turned...WooHoo! The 'undesirable' LO peak would go down (10.5 GHz) and the desired heterodyne (10.368 GHz) would peak up. (Make sure your looking at the 'desired' signal and NOT the image...the SA really helps).

Now that things were pretty close, we switch over the power meter for final tweaking of the filter. 230 mW output! Outstanding! We check for saturation with the IF drive (using a calibrated signal source) and it seems that -8 dB was about the max. Perfect!

We TX a little, listening to the output on Phil's home 3 cm receiver just to check audio quality then put it all back together in my rover set up and test it again. Super! And just in time for the up comming August 10GHz and up contest weekend. I need to coble together a little tripod, instead of just sitting the tub up in the back of the truck but I'm ready to go!

Minimum (Novice) mod final specs:

• 5w IF drive gives 230mW out to a 33 dBi dish (2ft.), or about 400W EiRP!
• NF was not measured but is probably in the 12 or more dB rage

If you've put a 'WhiteTub' or White Box on the airsend me e-mail and tell me about it!

go back to main page