Micro908 Kit - Kit News & Updates

January 15 -- Round 2 kits shipped.

September 21 -- Round 1 kits shipped.

September 12 -- "Shipping: 10 days and counting!"

The title says it all ... we're awaiting the production pcbs from the fab shop and they are on a good path to provide us with the boards by the 23rd of this month. All we need to do after receiving them is drop one each into the boxes, tape them up, and drive them to the UPS place (for US customers) and to the Post Office (for foreign customers).

As mentioned before, it will take about 3-5 days for our kitting team (Tom & Nancy Feeny) to finish physically taking all 200 boxes to the shipping point, as it takes finite time and action at the counter for each box. But this will all start on the 24th and some will receive their kits before others ... just have confidence that yours is coming!

Over the last three weeks or so, during which final part and logistics issues have been handled neatly and effectively by Paul Maciel AK1P and the Feenys, I've been able to do some fine tuning on the Antenna Analyst software ... and I must tell you all that the results have been really stunning. I've been able to make two quite significant improvements in the basic operation of the instrument.

NOTE: If you don't care about the internal operation of the instrument, or if you don't want to feel confused about hearing the detail of these final improvements, just close your browser right now. I don't tell you anything that you need to know as an user of it.

Okay, the two improvements ...

1) Smoooooooth display of R & X values while tuning.

I had a chance to focus on a small problem that was nagging me all along. Those of you who have seen the Micro908 AA908 software in operation at the NJQRP and NorCal club meetings might have noticed some bouncing around of the readings as one manually dials across a band of interest with an antenna (or dummy antenna) in place. All along I've thought this was the norm, as complex impedance readings do indeed kind of jump around, even with other commercial units.

But I really dug into it and found a "mathematical discontinuity" in my computations for resistance that has previously been obscured. If you look in our technical manual at the equations for the R computation, you'll see the complexity I'm talking about. The numbers are also so large that I needed to devise a pair of 4-byte fixed-point accumulators to handle the required range, while also keeping enough digits for accuracy along the way.

Long story short, there was a point at which the denominator of a computation overflowed the 4-byte accumulator and I hadn't noticed that it. I fixed that condition and now and the improvement can be compared to driving your car on a very rough & bumpy road for a long period of time, then suddenly coming upon a stretch of freshly paved blacktop. Smoooooth operation!

2) Improved Accuracy

One of the salient-yet-important advantages of the Micro908 design is its reflectometer - the bridge arrangement of diodes, C's and R's (including the antenna load) that detect and transform the test RF signal into DC so each channel can be measured prior to computation of SWR, R and X. These signals include: Vf (forward voltage), Vr (reflected voltage), Vz (the impedance-related voltage), and Va (the total or "all" voltage). [Note: I think I previously mentioned the addition of this last fourth reflectometer channel. Va was needed to greatly simplify the Z computations.]

You'll recall from our published papers and from prior discussion that we use op amp circuits after the reflectometer to amplify its very small DC signals, and we have a diode in the feedback of the first op amp stage. The purpose of this diode is to counter the nonlinear effect of the first diode in the chain (i.e., in the reflectometer), thus providing a more linear signal at low levels.

Okay, with the stage now set, you can probably guess that unit-to-unit variability will exist among the population of kits. This variability is caused by the different voltage gains in the op amp stages (despite our use of 1% resistors), and the diodes themselves will introduce some variability in the four voltages read by the microcontroller since we don't hand-match the diodes for near-identical V-I characteristics around the "knee" of the device.

The beta units all worked just fine without any trimpots or other calibration techniques ... It doesn't matter one hoot whether a resistance comes up 100 ohms or 115 ohms for an antenna measurement, but I felt that having it read much closer to the expected value (like when a known-value resistor is used as a load) will lend much greater confidence in the instrument's value. So my special "software calibration" effort began.

The solution was rather neat in that a "Calibration" command was added to the Config menu and the user is instructed to sequentially place on the output connector: [1] an open (no load), [2] a short, [3] a 50-ohm resistor, [4] a 100-ohm resistor, and [5] a 270-ohm resistor. The Vr, Vz and Va raw data readings were all taken and compared to Vf, and a "correction factor" was computed and used to make each reading be what we know it should be with those specific load conditions. I then use the correction factor in the computations for SWR, Z, R and X to make the readings dead-on, irrespective (within reason) of the differences in the channel-to-channel gain or diode characteristics.

The Calibration command generates the correction factors, which are stored in EEPROM. This means that you only do this step once and it'll forevermore be a calibrated unit. Pretty cool!

- - - - - -

Okay, for those of you who stuck with this note to the end, I hope you found this detail interesting. If anything, it'll give you some background for when it comes time to build, calibrate and use your Micro908 when you get it. It's been a long road for everyone (including me and the kitting team!), and I'm sure you'll all be thrilled to start assembling and using your kit.

For your information, we'll soon be opening up the Project page and letting the general ham population see what we've got here, as it'll be "Open Ordering Season" and we'll be starting to take orders for the next round of kits. You guys will be on the leading edge of things with your Round 1 kits and I expect some interesting dialogue to come about on the lists for this project.

We'll post an update (and pics) when the production boards are received in about 10 days, which is also the signal for when shipping will start. Thanks again to everyone for hanging in with us on this project.

August 17 -- We've been making great progress heading down the "kit delivery runway" for every aspect of the Micro908 project ... except for the production pc boards. We ran into some snags along the way and are (once again) awaiting the final pcb's to come from the fab house. The good news is that we have all the snags worked out and when we get the boards in about 2.5 weeks we can immediately ship. So that will put the shipping date (with confidence) during the week right after Labor Day, which means that most of you will receive your kits the following week (week of Sept 13th). Sorry for the delay - we're doing the best we can during this final stage ... and it'll benefit everyone for by getting the pcb right the first time.

To give you all some solid evidence that we really are indeed ready to ship ... please visit the new Photo Album pages http://www.amqrp.org/kits/micro908/photoalbum.html to see the photos just posted for the completed, production-level product. The "Photo Album" link at the top left of the page takes you to the new pics. We photo documented the shipping box contents - all the parts bags and such - and I think you'll be as excited as I am about the kit. It looks really fabulous. Tom and Nancy Feeny have done a terrific job in packing and preparing the kit for shipping. I think you'll agree.

So it won't be too long, and I guarantee that the wait for this kit will be worthwhile. It goes together real nice, and the positive reports from our beta builders are beyond my expectations. ("Whew", actually ;-) There were no problems with the assembly of the SMT parts, in large part to the "SMT Cards" prepared by the Feenys, and also due to the clear, printed Assembly Manual that comes with the kit. And each of the beta-built units operated when power was first applied, and they are now happily measuring antennas and other resonant circuits with the instrument.

July 18 -- Just a quick update here ... nothing much new to report other than we're working at getting the the kits assembled. All parts have been received now, except for the known last-coming ones: Enclosures, Panel Overlays and the PCB itself.

Enclosures -- They are being drilled en masse right now and should be delivered to the Feeny's (kitting team) this coming week.

Panel Overlays -- These are *very* slick and quite cool. We went to aprofessional company for these and they are 4-color printed on the backside of clear acetate, with a layer of glue applied. They come pre-punched for the controls, of course, so all builders will need to do is peel off the protective backing and press it down into place. We had one done for the front panel, as shown in the proto photos, *and* we had them done for the end plate (with BNC connector, et al), *and* for the side panel (with the power jack, switch and two audio jacks). These overlays are due to the Feenys within 2 weeks.

PCB -- We're working on building up proto #2 pcb's. Once the build (and assembly doc) is qualified, I can be confident of ordering the bunches of the production version, masked and silk screened pcb's. That fab process takes only two weeks, and that's the gating item for this project ... once we get these production pcb's (and check them), we'll drop them into the box and ship it!

Software -- Joe and I are tweaking the calculations module to get the most accurate readings possible. Probably surprising to most, computing Z, R and X from raw data readings of Vfwd, Vrev and Vz is incredibly complex. Pun intended. (Complex number format.) Complicate this by doing it on an 8-bit microcontroller, and the softare gets hairy pretty quickly. I've been doing this stuff for a long time and this whole software part of the project is clearly not "falling off a log" simple. In fact, there's a good reason why you don't see many antenna analyzers on the market with this approach. But we're there.

CDROM -- All data files are sitting in the "master folder" waiting the final (first offical) version of AA-908_v1 software. Once that's ready (per above comment), the ISO master can be created and CD duper volunteer David N0YMV will be able to turn the crank to produce enough CDs for those in our first round of shipments. The CD contains manuals for all components (Micro908, HC908, DSPx), official distribution software AA-908_v1, the m908 Loader program for your PC mentioned last time, and a good terminal program (TeraTerm) for your PC. For those who just want to "build and use", you likely won't even have a need for this "Micro908 Resource CD". But if you wish to play around with this flexible kit, you'll find the files and information contained on the CD to be of immense value.

July 9 -- As I said in my note: All HC908 cards shipping now, and those included in the Micro908 Kit, have version 3.0 of the HCmon code.

But what probably makes the "Auto-Load" feature confusing for those who haven't yet studied the project carefully yet is there are are *several* software programs working at the same time in the Micro908.

CRUDE HC908 DAUGHTERCARD MEMORY MAP
=====================================
FFFF
        |
        | This space is used for HCmon debug monitor v3.0
        | and never gets updated when we change applications
        |
EF00
        |
        |
        | This space is used for "applications" like the Antenna Analyst,
        | Audio Filter AF-908, "Portable PSK", etc.
        | This area of Flash memory can be reprogrammed by the user
        | by using the m908 Loader program we talked about.
        |
        |
8000
        |
        | System level software, RAM and I/O port mappings
        |
        |
0000

So you see, the area at the top of memory for the HCmon program is what I described as needing to change to support the new loader program announced today. Those people who previously purchased the HC908 Daughtercards have v1 or v2 of the debug monitor and they'll need to get it reprogrammed by AmQRP in order to use the new, simpler m908 Loader. (The HCmon source code is freely available, but it requires a hardware programming "pod" that costs about $150. BTW, we'll be publishing a simple schematic that will allow a homebrewer to build one of these programmers, if desired.)

July 8 -- We've made some neat progress in completing a very cool PC program called "m908 Loader" that is used to conveniently "auto-update" new software into the Micro908.

Up to this point, when a new version of software is available for the Micro908 platform (e.g., an updated Antenna Analyst version, or a newer software program altogether like the Audio Filter "AF-908" coming next), the user would ...
    1) download the new program to his local PC,
    2) connect up to the Micro908 using a terminal program on the PC,
    3) put the Monitor Jumper on the pcb board (within the battery compartment)
    4) issue the HCmon commands to clear memory
    5) issue the HCmon command to initiate a load,
    6) navigate to the new file to be loaded and invoke the load.
    7) remove the Monitor jumper on the Micro908
    8) reboot the Micro908 for the new program to become active.

Whew, that's a lot of steps and it can confuse the computer novices among us!

Well, in chatting about this with Bob Hillard, WA6UFQ, we came up with a much simpler way for users to update their Micro908 instruments. Bob developed the "m908 Loader" program that conveniently sends a specified S-record file (i.e., the new program) to the Micro908. All you need to do now is ...
    1) Download the new Micro908 application program from the Micro908 website (as before),
    2) Put the Micro908 into "Update Software" mode (underneather the Config pushbutton),
    3) Start up the m908 Loader program on the PC and specify in a pull-down menu the file you wish to send to the Micro908,
    4) Click the Load button presented on the screen and wait for the load to complete! In less than 2 minutes you have the new program automatically starting up on your Micro908 platform ... no fuss, no muss!

There is a slight price to pay for this feature though ... HCmon, the built-in debug/loader monitor that comes pre-loaded on the HC908, needs to be upgraded to version 3.0. Up till now, we've only been shipping the daughtercard with HCmon v1 or v2 loaded, but HC908 daughtercards shipping from this point forward have HCmon v3.0 monitor loaded, of course. But if you are planning to use your existing HC908 Daughtercard with your Micro908, you'll need to get your HC908 card re-flashed with HCmon v3.0.

We'll do that for you at no charge. Just mail your HC908 Daughtercard to Tom and Nancy Feeny in a nice protective envelope or small box (like the one we originally used to ship it), and they will do that reprogramming and send it back to you.
    Tom & Nancy Feeny
    1480 Meadow Drive
    Walled Lake, MI 48390

This is a way to get a head start on being able to get your Micro908 Antenna Analyst Kit operational with all features right away. (Note, if you do not get your HC908 Daughtercard software updated, all other features/capabilities of the Micro908 will still work, just not this auto-update feature.)

Thanks very much to Bob Hillard for his work in developing the program ... this is going to save Micro908 owners lots of headaches downstream.

PS: Once you get your HC908 Daughtercard upgraded to include HCmon v3.0, you'll be able to try out this Auto-Update feature yourself, even *before* you have the Micro908 Kit in hand. I'll describe more about this in a few days, and will post the m908 Loader PC program on our private project web page above as well.

July 2 -- Progress is marching along and looking pretty good. Many wheels are in motion on all fronts. I've just posted to our private Micro908 website the final schematic ... this has many corrections, fine-tunings and a couple of neat improvements over what has previously been made available. (I purposely have withheld postings of the numerous interim schematic updates so as to not confuse people ... but this one should be "golden and is currently being proved out with the Proto2 pcbs.

BTW, Joe and I had a successful field trialing of the proto units over this past FD weekend. It was a windy (but gorgeous) location on the northern end of the Chesapeake Bay, right at the water's edge, and we discovered some interesting instrument behavior in the rapidly-changing SWR conditions of our wire antennas. We found that we need to slow down the response of measurment system, as the rapidly-changing SWR and impedance measurements caused rapidly-changing digital readings ... which looks pretty confusing to the operator, when you think about it. (A slight wobble of a normal needle reading is easily read, while an erratic bobble of a digital readout is not such an easy matter.) I could slow down the sample rate on the real-time nature of the measure-and-display loop, but we thought it better to put more filtering in the reflectometer to similate the mechanical "filtering" of a needle-based system. So far the approach is working out real nicely. We'll soon post some cool photos of the side-by-side measurement gathering we had going with the m908, MFJ-259B and the Autek RF Analyst.

Good news on the matter of batteries. You might recall a previous discussion wherein we had recommended use of alkaline batteries to give the higher voltage (as contrasted to the lower voltage produced by rechargable cells) that can be regulated down by the onlboard DDS Daughtercard, thus providing a constant (good) supply voltage even when battery voltage descreases. Well, we tried a lower regulation voltage (8V instead of 9V) and we still have good-enough signal output to the reflectometer to produce a constant good RF signal for measurements even when using rechargable cells! This means that we *will* be able to recommend use of eight NiMH or NiCD rechargable AA cells in the unit.

Further, we put a 100ma trickle charger onto the pc board that will recharge exhausted cells within about 24 hours ... perhaps just right for plugging the unit in overnight to get ready for the next day's antenna adventures. The only caution is that you must use NiMH or NiCD cells, or remove one internal jumper if you only have alkaline cells available. (Also, to prevent overcharging with this simple trickle charger, you will need to be careful to not leave the unit continuously powered by an external source.) Both of these are easy-to-remember situations.

The last notable change you will see in the schematic is the addition of a small memory device ("SEEPROM", or Serial EEPROM ... sort of like nonvolatile FLASH memory). This component is the one that will store all the SWR & Impedance readings for later uploading to the PC for graphical display. It will also hold the DSP program for the DSPx Daughtercard, which only has non-volatile memory at the moment, and the HC908 controller loads the DSPx board memory from the SEEPROM memory at power-on time. This feature will be available in the second revision of the Antenna Analyst software later this year, but we needed to get the 64Kx8 bit memory onto the board at rollout time.

Let us know if you have any questions about the schematic or with anything else about the Micro908 Antenna Analyst.

June 23 -- We're making good progress ... Most of the parts have been received and kitting operations have begun. The enclosures are being cut and yet another house is making the front panel overlays. The final version of the prototype pc board is in progress, and much of the manuals have been written. We've had some great help from Bob Hillard, WA6UFQ, in creating a simple one-icon software load function for software updates from the PC. Our main beta test site has been proceding extremely well and the user interface for the Micro908 "Antenna Analyst" (as the instrument is now called) has been refined to completion - it's a dream to use!

However the logistics have come along slower that expected, I've had several unexpected business trips to eastern Europe and southeast Asia, and there have been a couple of SNAFUs along the way (like a mistake I made in the pcb artwork for the almost final pc board layout). These items have conspired to slow the kit rollout team down just a bit and the expected ship date is now set for the 3rd week of July. As often happens with product releases in the real business world, the confidence in ship dates becomes better as the date approaches and I'm pretty confident of this July date now.

Please understand that the four team members and I are doing the best we can in between normal (and unexpected) interruptions, and we are actually making some marvelous progress. I do realize how eager most of you are for this kit and we're doing our best to get you a quality product as soon as we possibly can.

I've placed a draft version of the Micro908 Technical Reference Manual onto our private Project web page - I'm sure most of you will enjoy reading it, and I'll gladly entertain questions and corrections you might have. Additional material will soon start appearing on the web pages as well.

May 22 -- Been working like crazy on things here in the shack -- on the m908, on HOMEBREWER magazine, and at work (of course) -- and just wanted to update you on some progress on the Micro908 project.

1) Joe and I have been seeing wonderful progress on the "tuning up" I mentioned last time. The reflectometer is the most important part of the whole instrument and we've been giving it appropriate scrutiny. With some resistor/gain adjustments in the LM6484 buffer amps, it's now acting totally predictable.

2) I've been tuning the user interface and now have a nice interaction between the dial and control pushbuttons. The PBs get you into the appropriate mode, and you then see all available options as you rotate the rotary encoder dial. The desired option is selected by pressing the dial pushbutton. Probably like many of our cell phones these days, but pretty slick, anyway.

3) I've been experimenting with a rotary encoder that is detentless (i.e., smooth-turning), and I very much prefer the detented rotary encoder that we have in the Micro908. The detents provide for a much better defined menu selection; and frequency selection is just fine too.

4) I've also refined Manual mode operation ...this is a fun mode to use! You can selecting this mode by pressing the ENCODER pushbutton and you will see the following display (sample data here):
F: 10,000.00
S: 2.1 R: 49 X:+23
- The Freq: is the frequency at which the SWR minimum is found.
- The S: value is the SWR. In this case it is 2.1:1
- The R: value is Resistance, the real part of Impedance. In this case it is 40-ohms
- The X: value is the Reactance, or the reactive part of Impedance. In this case it is 23-ohms inductive. (Capacitive reactance would be indicated as a negative number.)

Here in Manual mode you have the ability to manual adjust the Frequency and see the SWR, Resistance and Reactance values at whatever frequency is dialed up. Frequency is changed by adjusting a single digit indicated at the point in the display where the cursor ( _ ) is. Upon initial entry, the 10 kHz digit is the adjustment point, as shown by the digit with the cursor in the display above. To move the cursor to a different digit to be adjusted, press the ENCODER pushbutton simultaneously while turning the ENCODER. The cursor will move to any of the seven available digits, allowing subsequent up/down adjustment of that digit with the ENCODER pushbutton is released. Rotating the ENCODER clockwise will increase the digit value, and correspondingly the signal generated by the DDS. Counter-clockwise rotation will decrease the digit and the generated signal. When the digit is incremented past 9, or when it is decremented past 0, the digits above the selected adjustment point are correspondingly rolled up or down. Using this frequency adjustment scheme, the user can conveniently pick an “increment” digit and manually scan frequencies with the desired granularity. Rough scans can manually be done by positioning the cursor under the 100 kHz digit or the 1 MHz digit, giving a wide and course scan of the frequencies with a quick twist of the ENCODER dial. The signal frequency can then be set to the area of interest and the cursor set to a lower granular digit (e.g., 10 kHz or 1 kHz) in order to manually perform a detailed scan while watching displayed results for SWR, Resistance and Reactance.

5) We will specify some "dummy antennas" that you might wish to build up to use in checkout of your m908 Antenna Analyzer when all built. These will simulate a different antenna resonance condition across the range of the unit. Each will consist of a cap, inductor and resistor mounted on the end of a BNC plug. We'll soon specify the values of the R/L/C in an application note.

6) Joe has expanded the section in the manual that describes how to use the m908 Antenna Analyzer to perform all sorts of other measurements, in addition to just using it as an antenna analyzer. With the m908 you can check L/C component values, phasing stubs, determine feedline lengths, and more. This will be a very interesting section for users.

7) I've been working diligently on the assembly manual. As you might guess from our other NJQRP and AmQRP projects, the assembly section of the manual will be very detailed and photo-descriptive. You shouldn't have any trouble in constructing the m908.

8) I am currently working on adding a "Custom Band Selection" feature will allow a user to specify something different than the standard 1-10, 10-20, and 20-30 MHz bands to be Scanned. Suppose, for example, that you scan the 1-10 MHz band and determine that there is a resonance of interest that you want to look more closely at. Well, you can custom scan start/end limits that focus more closely on the area of interest and get a finer granularity for that scan. These limits are stored in EEPROM so you'll have them again when you turn the power back on next time.

9) I am also currently working on a "data pass through" function that will allow the DSPx board to accept a new software load. Although not needed or used in the Antenna Analyzer program for the m908 right now, of course, this is a needed function that Bob McGwier and Lyle Johnson will need to be working on for PSK31 operation coming up in the next application to come alive on the m908.

10) SCHEDULE -- As mentioned in the last update, we're a running a little behind and expect to be shipping kits later during June, as opposed to earlier. Many of the refinements I've mentioned above have been achieved while on transatlantic plane flights -- the flight attendants really think I'm nuts with the m908 connected to my laptop during the flights, but I get lots of software mods done then. But being on the road somewhat limits my time on the hardware aspects, which wait for me to be home. Same too for Joe, who is often on the road too, more Stateside though for him. Anyway, here are some of the logistics elements we're dealing with ...

a) Final prototype pcb's (with all the final changes) weren't actually ordered until yesterday, so I'll be getting them this coming week. I and a couple others will be building them up to shake out the assembly instructions and the built/test cycle. That should happen within the week following board receipt, then I can have confidence in ordering the production boards with silk and solder mask.

b) The people cutting the enclosures for us (Electronic Precepts of Florida) are making progress now. We are a small player to them and thus are "filler work", but it's getting done. I expect them to complete their preparation of 200 enclosures and AntAnal overlays by the time the kitting is done.

c) Parts are starting to be delivered to the Feenys for initial sorting and kitting preparation. They are working on the packaging and are generating a good plan for things.

d) Lyle Johnson has been great in readying the DSPx Daughtercards and we nearly have all required for our initial run. This is one cool little daughtercard and it makes the m908 quite the computing powerhouse! Again, it's not used in the Antenna Analyzer application but it'll play a key role in the next application to go live in the fall ... Portable PSK.

e) The Feenys have prepared all DDS Daughtercards and they are sitting on the shelf waiting for inclusion in the m908 packaging.

f) Mike Huffstetler, WA6OUW (KitBuilders) is busily working on assembling and testing all the HC908 Daughtercards needed for our first m908 run. He should be complete within about three weeks.

11) KIT PAYMENT -- I have collected about 75% of the payments for those who have reserved an m908 kit. I have a letter going out this coming week to all those who haven't yet paid up, either because you are just waiting for the end-of-month deadline, or perhaps because you haven't seen all this news and stuff posted on our private project page. Anyway, if we don't have your payment by June 12 (yes, it's extended a bit), we'll advertise for someone else to join this first round of kit shipment and you'll lose out on the deal. Gotta draw the line somewhere :-)

Okay, that's about all I can think of for now. Let me know if you have any questions along the way and understand that Joe and I, the Feenys, and some others are working real diligently to bring this kit out as planned. So far so good!

April 24 -- The automated "Order Status" checking has been updated, located at http://www.amqrp.org/kits/micro908/micro908status.html . It is now current with the latest address information we have for each of you and the amount you have submitted as payment for the Micro908 Kit. If you have submitted payment ... thank you! You are enabling us to purchase the parts for this first round. If your "Payment Made" field is [zero], that means we're still awaiting your order. You'll have about another four weeks to make your payment before we have to drop you from the first round. Please let n2apb know if you have any questions about this, thanks.

April 24 -- Been busily hunkered down on the project lately but we wanted to give some quick updates to you guys ...

PARTS
Parts ordering has been in full-swing. I've already received many and now have boxes stacked high from Mouser, Digi-Key, Mini-Circuits, Motorola and Pac-Tec. A bunch more still need to come in before I get them all over to the Feenys for kitting, which should commence within another 2 weeks.

The parts availability of the 68HC908AB32 MCU that I noted before has been solved (!) By liberal application of a "sqeeky wheel" at key sources, I was able to shake loose enough of these microcontrollers to suffice for the HC908 Daugtercards being used in the Micro908. I've got these parts en route to our assembler now (Mike at KitBuilders) and he'll soon have enough of them ready for us at ship time.

Similarly for the DDS Daughtercard, which is also an integral module for us, we've worked up engineering enhancements that make it better applicable for use in the Micro908. This amounts to use of a more powerful RF amp (MAV-11) and a voltage regulator that will require only a single bias resistor in the presence of a varying supply voltage. For those of you guys who plan on using the DDS Daughtercard you already have in-hand, we almost have ready an Upgrade Kit that is simple to apply on your DDS cards. For those who have ordered a Full Kit or a DDS Daughtercard option, the Upgrade Kit will come in the parts bag. It's easy to add the parts (no major surgery) and the results give you a solid 4Vpp output signal, which will be nice for use with the internal reflectometer.

All parts are on order for this next group of DDS Daughtercards and they are fully dedicated for our first round of Micro908 shipments. (If you've been following the DDS Kit saga on the QRP-L, my stance is that we will make the kit available again for general sales later in June-July, which is *after* we ship the Micro908. See, there indeed is a rhyme and reason to our planning activities ;-)

However now the "tent pole" item (critical path) has shifted over to the display units. Seiko suddenly got out of the LCD business and my sources for the L1682 display dried up like a solder sponge on a hot desert. I did find a replacement unit that is "nearly" the same size and "almost" available. But "nearly" and "almost" are the operative words. The unit is slightly taller than the Seiko LCD and I need to have the enclosure people modify their hole-cutting plans to accommodate this new display. The "almost available" deal is that I have some, but not all, of these new displays in-hand. The good news is that they *should* all be in within 4 weeks and the vendor preparing the enclosure *should* be able to make the hole bigger. (It's harder to make a hole smaller, you see.)

FUNCTIONAL TUNE-UP
Joe has been continuing to tune the reflectometer data and computations during these final countdown weeks before availability to ensure best possible results from a product of this class and in this pice range. Recall that this is not a multi-$$$$ instrument from HP, but the results we get are certainly good enough for our use. And the stability from the use of the DDS, and the flexibility and re-use of the platform all make the Micro908 a terrific project.

PRINTED CIRCUIT BOARD
I'm about to submit for another prototype PCB which should eliminate all the cuts+adds from the prototype boards currently shown and pictured. This should be coming in next week or so and I'll give it a shakedown build and test. If it all looks good I'll submit for the production PCB run, which includes the nice solder mask and silkscreen legends, which should take two weeks and bring us right up to expected shipping time. The PCB and the Manual-on-CDROM will be the last items to drop into the box before shipment.

MANUAL-on-CDROM
As the name suggests, we'll be providing a fairly detailed Assembly and Technical Reference Manual for the Micro908. The specifics that address the Antenna Analyzer software running on this platform will be contained in an extensive appendix. We'll also be including various "sub-manuals" for the components also used in the project, such as the HC908, DDS Daughtercard and DSPx Daughtercard.

CAVEAT EMPTOR -- MEA CULPA -- E PLURBUS UNUM
This is the fine print ... The design and kitting team volunteers in the club always do the best we can, and often come out on top of the ogre during these new product wrestling matches. However things outside our immediate control sometimes occur and we need to go with the flow ... be it individual sicknesses, business travel, parts shortages/errors from vendors & suppliers, and so on. This is a huge kit introduction - the biggest one we've ever tackled - and I kindly ask that you all work with us and understand that we are doing our best to keep to this aggressive shipping schedule. But if one crisis or another happens along the way, please have the confidence that we'll also do our best to minimize the overshoot and keep you all informed of what's happening. Thanks for your understanding.

Meanwhile, please keep an eye on the Project page (www.amqrp.org/kits/micro908/project.html) as we'll be posting new information at an increasing rate in these final weeks.

April 8 -- The Micro908 reflector is established for owners who have reserved and/or paid for a Micro908. If you are not already receiving the [Micro908] email postings, please let me know (n2apb@amqrp.org) and we'll get your email address straightened out. This is a great way for owners to receive up-to-date information about the project and get fast help from others in the group. Also, the Archives are available online at http://www.applegate.org/pipermail/micro908/ .

April 6 -- The AmQRP server had some problems and the web pages were not available ... right when we introduced this important page for Micro908 owners. Most people were patient in waiting to to the ordering info - thanks very much!

April 5 -- The Micro908 Project Page (you're reading it right now) was established as the main resource location for all Micro908 owners to get the latest software, news, construction tips & errata, etc. We're also setting up a mail list for Micro908 owners to send questions, comments & dialog among all 300 other owners. This new list is not intended to be used instead of QRP-L, but is certainly a way to get the immediate attention of others in our group, including that of the designers.

April 4 -- Kit orders are now being accepted! The Micro908 Bulletin #1 was distributed to those who had previously reserved a kit. Only this group of about 300 individuals are eligible to purchase the kit and participate in the first round of kitting with us.