Clifton Laboratories 7236 Clifton Road  Clifton VA 20124 tel: (703) 830 0368 fax: (703) 830 0711

E-mail: Jack.Smith@cliftonlaboratories.com
 

 

To search within the Clifton Laboratories site, enter your search term below.
 

 

Home
Up
Updates
Current Products
Documents
Book
Software Updates
Softrock Lite 6.2
Adventures in Electronics and Radio
Elecraft K2 and K3 Transceivers
 
W5EWA Build

This page documents Stan's, W5EWA, build of a advance-release Z90.  I've edited his comments for style and added my comments in blue text.
 
Stage 1

Time: 1H 45M

Comments:
Spent 1 3/4 hrs. on stage 1. I like to have never found C501. Is
there any way you can make a searchable .pdf out of the parts layout sheet?

[Yes, done and sent to Stan this morning. You may view it by clicking here. I divided the layout into a grid, and prepared an index for each component, e.g., R124 at F3, where F3 is the intersection of the letter F grid line and the number 3 grid. This will be part of the standard documentation. I may, time permitting, integrate it into the Assembly Instructions parts list.]

During the testing phase, it turned out that the 12V wall wart I
selected to use was wired backwards and I could not figure out why the damn
voltage would not go through D501. I was about ready to put the Hakko 808 in
to service when I decided to double check the voltage from the wall wart.
Other than forgetting to solder one capacitor, everything else was good to
go. I found the capacitor that I had not soldered while I was trying to
figure out why my voltage checks weren't working.

[That was also a test of the purpose of D501--to protect against reversed voltage!]

So I had a couple of time consuming little side excursions. This
time included reading the addendums and doing a inventory of the parts box
to make sure I had all the stage bags and everything was there. 

Minor error in the manual on page 14, paragraph 3 (U501). The
paragraph says "insert U506"...should be "insert U501". [Fixed]
 
Stage 1 completed

 
Stage 2 & 2Z

Time: 1H 40M

Comments:
Stage 2 & 2Z went well, with no anomalies. I found one minor mistake
in the Stage 2Z instructions on page 23. R12 10ohm bwn bwn blk, should read
R12 10ohm bwn blk blk. [Stan had sent these comments earlier and I overlooked them when updating this page. Sorry.] [The error at Page 23 is corrected.]
Stage 2 & 2Z Completed
 
Stage 3

Time: 50 minutes

Comments:
Stage 3 was about 50min. Most of that was locating a piece of coax
to use and making the pickup loop, and then going through the testing
procedures. Fortunately I already had a short piece of coax with a PL-259 on
one end.

I was initially confused about where J701 was. Reading further on
explained that it was actually ON the DDS board.

[No photos - only one part added to PCB.]

 
Stage 4

Time: 1H 10M

Comments:
Stage 4 took 1hr 10min. No anomalies, and installation was straight
forward. Measured voltage drop across R109 (U104) was 7.22 vdc, and across
R108 (U105) was 7.24 vdc. Included pictures are of the Stage 4 area, and,
most, but not all components were included in the pictures.

[7.22V indicates the Gali-74 bias current is 72.2 mA. Both Gali-74's show similar bias current. Matched bias currents is not a design target, as there is no reason for it. However, a major difference between the two bias currents could be evidence of a problem with wiring or components.]

 
[Stage 4. Shows the two Gali-74 amplifiers]
 
[Close up of Gali-74. The large blue components are 2512 size 100 ohm resistors, rated at 1 watt dissipation. The smaller surface mount parts are all 1206 size. Tan objects are ceramic capacitors and the black and blue objects are 1/4 watt resistors. The circular holes are small diameter "vias" or pads that connect the top and bottom ground planes. The vias serve two purposes. One is to ensure that the top and bottom ground planes are connected via many low inductance paths. Secondly, the Gali-74 devices run warm, and the vias provide a thermal link to the bottom ground plane foil and improve the Gali-74's heat sink.]
 
Stage 5

Time: 1H 20M

Comments:
Total of 1hr & 20 min for Stage 5. I blew the second and final fuse
while testing on Stage 5. I don't recommend doing this, but I bypassed it
with a jumper so I could continue to work on the kit. According to Jack, the
extra filter capacitance may be causing an additional surge when the power
is applied causing the fuse to pop. Jack is sending me a different type of
fuse that resets itself after a predetermined amount of time so that this
will not happen again.

[The Z90's original design, and the first three or four prototypes, used a small filter/bypass capacitor, 82uF/63V at C503 with a 1.1A polyfuse protector. To provide faster trip action, I switched the protection component to a 1.25A fast blow picofuse.

If you are not familiar with the polyfuse, it is a device that physically resembles a disk ceramic capacitor, but has a low resistance to AC and DC current until a certain threshold is reached. At that point, it switches abruptly to a high impedance state, acting like a fuse. Unlike a fuse, when the overload is removed the polyfuse returns to its low impedance mode. Hence, the polyfuse does not require replacement and there was no need to design the Z90 with a rear panel fuse holder.

If you wish to learn more about the polyfuse, a useful source is  http://www.circuitprotection.com/06Databook/fundamentals/PSWFundamentals.pdf

In testing some typical wall-wart power supplies, I found more ripple than I wanted with the Z90's 82uF capacitor, so C503 is now an 1800 uF/25V part. A side effect of increasing C503 so dramatically is that the turn-on current can now be quite high, particularly if the power supply can deliver many amperes. This power-on surge then blows the picofuse, even though there is no circuit fault.

Stan, and also Bob, K7HBG, have seen the picofuse blow when switching on their Z90s during construction when using a power supply capable of high output current. Hence, I am returning to the original design and will supply the kits with a 1.1A polyfuse. Since the PCB is designed for the polyfuse, there will be no construction issues.

I mailed a 1.1A polyfuse to Stan today, and all kits shipped will be provided with the polyfuse.]


    Otherwise everything went well. I did have a question about the
nomenclature for U202 the TUF-3+ mixer. In the manual on page 47, at the top
of the page, it is referred to as TFM-3+. I am not familiar with this device
so not sure if this is a typo or just another designation for it.

[My error. The correct identification is a TUF-3+ and I will correct the Assembly Instructions. Minicircuits also makes a TFM-3+ mixer and I have used both the TFM-3 and TUF-3 mixers when developing the Z90. The + suffix indicates an RoHS compliant part.]

    I did change supply voltages in mid-build, because I was rapidly
reaching the outside of the envelope of the current rating of the wall wart
that I was using. So I went from a 12.5 volt supply voltage to a 14.1 volt
supply. This changed my current readings, so I took the time (not included
in the stage build) to go back over the current readings for the Gali-74
MMICs, too be sure that they still agreed with the book. Everything was
right where it should be. After stage 5 completion I am seeing around 80ma
bias on the Gali-74 MMICs. The current draw without the DDS installed is
around 359ma. The current with the DDS installed is around 469ma. This is
with the new supply voltage of 14.1 volts.

[The Z90 will actually run with a supply voltage as low as 9V. However, its input and post-mixer amplifiers will be running way below design specifications and the Z90's intermodulation performance is degraded when operating with a supply voltage below 13.5V. There's an additional diode drop inside the Z90, as the design uses a series diode for reverse polarity protection. Thus, a 13.5V supply will give about 12.9V to the Z90's internal circuits. The protection diode, D501, is a 2A Schottky diode, but at forward current levels approaching 1A, it exhibits about 0.6V drop.]
 

[At this point, Stan has 6.7 hours of construction time.]

 
[Stan provided two photos as well]

 
[The 18F4620 PIC is the large IC at the bottom of the photo. The vertical "IC" is actually a relay for crystal filter switching. The silver rectangle is the TUF-3+ mixer and the transistor with the heat sink is the 2N5109 post-mixer amplifier.]
 
[The input BNC and the attenuator switches and components are at the top of the photo. Between the BNC and mixer is the Gali-74 input amplifier.]
 
Stage 6

Time: 1H 10M

Comments:
1hr 10min on Stage 6. No problems encountered. No indicated tests to
perform on this stage. Assembly was straightforward...stuff the board,
solder the parts. With the 1/8th spacing of the Xtals from the board, the
Xtals seemed a little flimsy on the board. Once I attached the grounding
wires it stiffened them up fairly well. My spacing was probably more like
1/16th inch, but I think Jack is just trying to avoid any possible shorts
between the can and the adjacent lands and ground plane.

[Yes, Stan is correct. I've revised the Assembly Instructions to suggest the crystals be spaced 1/16" (1.5mm) above the PCB. I don't trust the solder mask as a long term insulator, so I recommend a small gap. Alternatively, if you wish to install the crystals flush with the PCB, a layer of thin Teflon pipe sealing tape as found in the plumbing department of your local hardware store could be used as an insulator between the crystal housing the the PCB. Punch the crystal leads through the tape and trim it to be about the same size as the crystal base.]

[At the end of Stage 6, Stan has 9 hours of construction time.]

 
[The two crystal filters are the most pronounced objects in the photo. This image shows all of the Stage 6 build and parts of earlier stages.]
 
[Complete PCB as of the end of Stage 6.]
 
Stage 7

Time: 1H 10M

Comments:
1hr 10min on Stage 7. No surprises, all parts present and found no
errors in the manual.
 At this point the basic board is finished, with all parts attached.
The only thing left to be installed on the board, I believe, is the CCFL
inverter and the shielding for the Log Amp. The heat sinks for the two
regulators have to be put on their respective parts at final installation.

The only checkout for this stage is the current draw. My unit is
drawing 489ma with the DDS installed.
 

[At the end of Stage 7, Stan has 10 hours and 10 minutes of construction time.]

[Photos provided by Stan:]

 
[Stage 7 construction includes the area within the wide rectangular trace and the voltage reference parts below the rectangular trace.]
 
[The log amp (bottom IC) and output buffer amp (top IC),  relay (upper left) and 15 MHz low pass filter module (vertical white component at left bottom.)]
 
[Complete PCB as of the end of Stage 7.]
 
Stage 8

Time: 1H 0M

Comments:
Finished with Stage 8, and also the initial check out and
calibration. Stage 8 took about 30 minutes, and the initial check out and
calibration was about another 30 minutes for a total of about 1 hour.

All went well. The unit calibrated as expected.

Cabinet assembly is all that is left.

[At the end of Stage 8, Stan has 11 hours and 10 minutes of construction time.]
 
[The completed PCB, with CCFL Inverter installed, heat sinks in place, DDS module mounted and ready to go into the enclosure.

The board is resting on the three-page errata sheet included with the advance build documentation. I've since incorporated almost all of the errata into the Assembly Manual.]
 

[A larger view of the completed PCB.]

 

 
Stage "9"

[There isn't a real Stage 9, as I did not assign a stage ID to the enclosure, but it's a good reference ID.]

Time: 1H 15M

I called it Stage 9, but it is Cabinet assembly...the last part
before the fun part, seeing it in action hooked up to your trusty rig.

Everything went together well. Fit and finish are really good. It
took me approximately 1hr 15min. I went slowly and made sure I understood
what screws to use where. I was initially a little confused about how the
cabinet went together, but this was in my mind from just looking at the
parts. Once I got in to the manual it was really clear, and much simpler
than I had expected.

[I am very impressed with the quality of the panels from a dimensional and mechanical prospective. The laser cut steel has perfectly smooth holes and is right on my dimensions. Unfortunately, that quality work is marred by a poor paint job. The TenTec enclosures are high quality (that is to say, expensive) units and fit together nicely. I wish that TenTec had found a way to provide electrical continuity from the upper and lower cabinet halves to the side rails and chassis pan, but they didn't.]

[I will also add that I spent a lot of time calculating where the panel holes should be, including making paper templates. When the panels arrived, I was confident that they would fit, assuming they were made according to my drawings. Nonetheless, I admit to a feeling of relief when the panels fitted into place perfectly in my first test assembly.]

I am including pictures of the front bail assembly, in case there
are any questions about how it goes together. These things can be a source
of confusion for the un-initiated. The first one I ever installed on my
Elecraft K2 had me somewhat confused.

Be sure and install the feet with the bail in the down (or deployed)
position. Much less stress on the feet when trying to install the screws.
The first foot goes on with the screws not quite tight, then slip one end of
the bail under that foot. Make sure it is pushed in all the way under the
foot. Have the bail in the deployed position and then place the other foot
on top of the other end of the bail while lining the holes up for the
screws. It should line up fairly easy and the screws should start with a
minimum of effort.

[I'll add this explanation and a few of Stan's excellent photographs to the Assembly Instructions. In the case of the manuals that are already printed, I will add it to the Errata sheet.]

One thing...these screws go in hard, so be sure you have them all
the way in the bottom cabinet. A larger than normal Phillips screwdriver
will aid in getting the screws seated all the way. They will protrude
through the bottom of the cabinet about 3 or 4 threads or so.

[One more thing. So far, all the Z90 enclosures I've shipped have the bail holes pre-punched. TenTec recently made a cost reduction change and no longer provides punched bail holes, leaving it to the purchaser to drill or punch using a TenTec-supplied template. I hope that the cabinets TenTec shipped me are all punched for the bail, but that may or may not be the case.]

[I also will mention that Stan had to do a bit of metal work to repair a bent enclosure corner. This may have happened when TenTec shipped the enclosures to me, as they were bulk packed into large boxes with plastic peanuts. He was able to repair the enclosure without difficulty.]
 
[Bottom of the enclosure with the bail installed.

I will also note that Stan keeps a very clean workbench, much different than mine in that regard.]
 

[Stan's completed Z90, bottom cover on, top cover off. The front panel looks better in a photograph than it does in person as the scratches, pin holes, orange peel and other paint problems aren't visible at this resolution.]

 
 
Stan's Total Project Time: 12H 25M