AH6RP amateur radio blog

From a note by Ed Tanton, N4XY, to QRP-L 08 Apr 2003:

"OK... I took one of these [surplus SMT] inductors and measured it at 525nH (0.5mH) [...] and brought the inductor near the 'face' of the 1.5' x 3.5' x 5/16' (3.5cm x 8.9cm x 0.8cm) magnet. Interesting: the inductance changed to ~ 200nH (0.2uH). Then, I moved the tweezers/inductor to the end of the magnet. Wow. The inductance went from its usual 525nH to 43nH (0.043uH)!!!"

Merton Nellis, W0UFO, commented:

Yes, inductors with magnetic material as cores change inductance from a maximum with no biasing magnetic field to a lower value as a bias field is applied. [...] The bias field can be applied to a core with a separate winding carrying d.c. rather than with a permanent magnet. This principle is used to make saturable reactor controls and magnetic amplifiers.

Mounting a magnet near a toroid with a screw for adjustment would let you build variable inductors without having to locate a slug-tuned coil form. Also, it could be used to build an SMT variable inductor with an adjustment big enough to be useful. Attaching a man-sized knob directly to a eensy SMT inductor is begging for the leads to be torn off the board.

Re: [QRP-L] CW QRP Kits? (ATS-3 not sold out): "Yes, the ATS-3 IS sold out. Just changed the web page yesterday. Time to get busy on the next version..."

Oops. Looks like I waited too long for the ATS-3, due to work and such. I hope Steve can get the ATS-4 ready soon. Time to go and look at the Small Wonder SW+ series again. They are single band, no DDS, but a third the cost of the ATS-3.

In this note on the Energizer L522 9v lithium battery, Everready claims "the L522 at 1200 mAh has twice the capacity of alkaline batteries."

This is Ultralife's specs for the battery that Thomas sells:Ultralife Batteries - U9VL. Note that the maximum pulse discharge is rated at 400 ma... that is just adequate for the ATS-3, which draws about 420 ma at 7.8v (see AE5X's statistics for a 9v alkaline transistor radio battery).

Weight is good at 33.8 grams. Also note the discharge/capacity curve for 60 ohms (125 ma) shows a decline to 7.0 volts at 80% capacity (960 mah apparently.) Short circuit current is initally 3 amps, but falls to about 1.3 amps and then levels off until minute 3, where the current falls again, presumably when the current limiter kicks in (and that's not reversable.)

Current draw from the ATS3 would be about 110 ma in a QSO. The battery will probably last a bit over 5 hours to 7.0 volts. [How did I get that? Assume 44% key down while transmitting... that's exact for "PARIS", 50 elements, 22 with key down... and say 50% transmit time in a QSO, that's 22% transmitting and 78% receiving time. At 7.5 volts, the ATS3 has an output of 2.05w, drawing 404 ma transmitting and 27 ma receiving.]

On continuous recieve, the battery would last around 42-45 hours.

Thomas Distributing carries the 9 Volt Ultralife Lithium Battery, with a 1200 ma nominal capacity... at 10 ma drain, to 5.4 volts. At 27 ohms, it briefly supplies about 10 volts, drops quickly to about 8.3, and remains above 7.2 for 39 hours (see curve at Thomas site, note 27 ma into 300 ohms is 8.1 volts.)
That's about 1,050 mah.

Kester - Alloys: "SnBi58 Alloy

* Low melt point lead-free alternative potentially suitable for some consumer electronics. Low melt point precludes its use for applications where operating temperature is close to 138C.
* Large Bi proportion greatly reduces melting point of the solder, but alloy is more brittle. Bi improves wettability, but is somewhat offset by higher oxidation rate. In the presence of lead from HASL boards or components Bi can greatly reduce thermal cycle fatigue resistance due to the formation of Sn16Pb32Bi52 (MP=95C) which can diffuse along the grain boundaries."

C&L Finescale in England offers "145 deg C Tin/Lead/Cadmium in the form of a 1.2 mm wire," product code C2002, 0.5 kg for 25 pounds. They also have eutectic tin/bismuth as a no-clean paste, but a 35 gram syringe is 27 pounds.

Here's a small quantity of Sn/Bi wire solder: Tools (at alltronics.com): "Melcor Low Temp Bismuth Tin Wire Solder... Melcor part number 16-65-019-01... includes 2 feet of 138 deg C solder (Indalloy # 281) and the instruction sheets, spec sheet and warning sheet (don't eat it). Alltronics part number 04Z054... Unit Price : $2.95". They don't mention the diameter, but from the illustration I would guesstimate 1/30" very roughly.

Melcor themselves sell 16-65-019-01 online for $1.35, 25/$32.50, apparently without the literature that Alltronics includes.

According to the table at Indium Corporation of America » Wire Products the only indium-free wire solder that is normally available with a lower melting point than the eutectic Sn/Pb/Ag alloy is Indalloy 281, 58/42 Bi/Sn, melting at 138 C / 281 F. The shear strength is poor, 500 psi, compared to 6200 for 63/37 Sn/Pb, and the conductivity about half.

Incidentally, the name in the table for the eutectic Sn/Pb/Ag alloy is Indalloy 62/36/2...! That probably negates part of my previous post; I would imagine this implies that the Radio Shack solder is most probably 1.4% Ag and not 2.0% Ag.

Common eutectic solder, 63/37 Sn/Pb, melts at 183 C, 361 F.

Indium Corporation has a spreadsheet of solder compositions here in HTML or here in an Excel spreadsheet . It lists 62.5/36.1/1.4 Sn/Pb/Ag as melting eutectically at 179 C, 354 F. It doesn't list exactly 62/36/2 Sn/Pb/Ag (e.g., Radio Shack 64-013 0.022" wire solder, 1.5 oz for about $4) but the alloy selection section of this guide from Indium mentions that it has a melting range of 179 - 188 C, or 354 - 370 F.

That's only a 4 C or 7 F difference in favor of 62.5/36.1/1.4, and 62/36/2 is actually WORSE, since it is still slushy up to 5 C / 9 F hotter than the melting point of 63/37.

52/30/18 Bi/Pb/Sn melts at just 96 C, 205 F. That's a 87 C, 156 F difference from 63/37. That seems interesting, although the only reference to it in electronics that I have found was in an explanation of why Bi/Sn/Cu lead-free solders are on hold for the next decade or so: if there were any lead at all on the board it would dissolve into the solder to form this very 52/30/18 Bi/Pb/Sn alloy, and melting at 96 C is seen as a problem.

(There are much much lower melting points available in indium-based solders, but they form brittle alloys with copper, making them useless for electronics; see the substrate metallization section in this guide from Indium.)

I haven't been able to find much about solder availability in small quantities of anything more exotic than 62/36/2, though. Bummers.

Several cautionary tales of SMD at home are in the SMK-1 trials and tribulations thread in QRP-L. (The SMK-1 was a much smaller QRP transceiver, now out of production.) The oddest piece of advice was to stretch a nylon stocking over the nozzle of a vacuum cleaner so you could find lost SMD components by vacuuming them up. Oh. On second thought, make that ONE SMD component; if you vacuum up two different unmarked components you would have to measure the eensie things before you could solder them in.

KD7S has plans for an SMD hold-down that probably contains the absolute minimum number of parts - just two. One is a bent piece of coathanger wire and the other is a wooden slat with a groove in the bottom.

A slightly bigger version of the toolpick-and-beeswax method of handling SMD's: from the Elecraft reflector.