 |
A
product from the NorCal QRP Club
...
A
collection of tips & techniques for |
About the
AmQRP Club |
| Click on this Title to find the article | Contributor |
| NC20 status de K7QO | Chuck Adams, K7QO |
| Additional NorCal 20 Construction Notes | David Fifield, AD6A |
| NC20 tips | Gary L Surrency, AB7MY |
(The following are several post from Chuck as he traveled the enjoyable
trail of building this NorCal Kit. Check his page out too! ...ed)
31 January 99
Gang,
Up through section 8 on page 9 and all is well.
Something that you might try in section 8. After powering up, place a
mechanic pencil with your finger on the metal part on the center lead of
the volume control pot, VR4. With changing pot values you can hear varying
A/C hum levels.
All this with a movie and dinner in between. Just be sure to take your
time and enjoy this fine kit. You can't rush it and I have been writting the
part numbers and their values on a separate sheet of paper and checking them off
as I go. Helps me keep things straight and so far haven't missed placement
of a part. Knock on wood as they say....
Page 8 L1 assembly diagram and instructions do not agree on the orientation of
the nylon screw, but that is minor. Do not try to turn the screw too hard
as it just barely long enough to hold a light pressure.
This info just for those that haven't started yet and will later on this
upcoming week. These are minor things that newbies might struggle with and
lose some time, but you learn a lot in the process. Been there done that.
Good job Dave, Doug, Jim, and a cast of 1,000s....
Oh, 1% resistors are the blue ones for the new people.
This is definitely not a one day kit. Mileage may vary.....
FYI
--
Chuck Adams K7QO k7qo@primenet.com CP-60
http://www.qsl.net/k7qo/
1 February 99 Gang, This is fun. Up through Section 14 now and everything is working just fine. No problems. Was scared right after putting in the Tick Keyer but that scare was caused by weak battery since the only one with small connector is 0.8AHr Gel-Cell. Was it Gary in AZ or whoever that commented the other day about MODE B? MODE B was not my problem at Pacificon, it was MODE A. MODE B I can do. :-) And since it was Gary from Embedded that handed me the keyer that did that I think he owes me a Tick-3. :-) Just kidding guys and girls.... The most important thing that you can do when building this kit for the 500 in the first round (and the only round unless someone picks up the commercial rights from Dave) is to take a typing page. At the first of each section list the part numbers, one per line. Then go to the parts list in the appendix and write beside each one the value, the color codes if it is a resistor, etc. Pictures if you need them. I know this takes time, but it is time well spent. This is not a race. You want the rig to work the first time everytime you power it up. Mine has so far. But the list helped me prevent one mistake where I was about to place a 10 ohm resistor where I needed a 10K. The mind must have been daydreaming listening to Eric Clampton and some R&B. :-) The parts density is high and you just have to take your time and be careful. Consider it brain surgery. :-) OK the doc has to get back to the patient. Pictures on the web for the late night owls and owlettes. Where is that diet coke and the brownie that Phyllis made? Oh, here it is. Ciao/Chow.... dit dit -- Chuck Adams K7QO k7qo@primenet.com CP-60 http://www.qsl.net/k7qo/
1 February 99 Gang, It's all together and it is beautiful. Got it all finished about 11a.m. this morning then got a call about lunch that I had scheduled with the gang that I used to work with. So off to lunch. Get back and finish up the transmitter alignment. Just a smidge under 5.0W with Heath PSA-9 P/S the one used with the HW-9 series. OK, so off the dummy load and onto the antenna. Ah ha, atmospheric noise. That's good as it goes away when I disconnect the antenna. Looking good. Everything ready to rock and roll and then keyer starts sending dahs and I don't have a paddle hooked up like I did last night during part of the assembly proceedure. Well to make a long story short I took me a number of hours tracking down the problem. I had the Tick in a socket so I thought that was the problem. Out comes the tick and the socket, clean up the area, and then when measuring resistance note that the dit side is infinite but only 2.64K on the dah side. Also 2.64K on pin 1, so I figure they are shorted. Scrape the area between the land for the dah and pin 1. Still 2.64K. Scrape again. Still same story. Get the spray relay cleaner and clean everything. OK, after a few hours of this stuff I decide to measure the resistance between pin 1 and 6. What? 2.3K? Well, I finally find that the stereo socket has something in it. Bummer. So new socket, cleanup, alignment, and set to 0.95W with OHR WM-1A and all is well at this time. Hind site is 20/20. OK, on the air at 0205UTC and hear AA7QU calling CQ TST. Oh, ARS is tonight. So first Q is AA7QU in OR his 579 at 4 and my 559 at 0.95W. So one state down and 49 to go. I'll catch up on email and stuff later when 20 dies. Pics on the web at 11. :-) I'm on the air. Any other NC20s show up and be square in the QRP neighborhood. Review to follow in the a.m. Rig is clean, nice keying and I have the AGC disabled. Have to go back and reset it later as it triggered on me at the wrong time during all this chasing a minor problem. :-) dit dit -- Chuck Adams K7QO k7qo@primenet.com CP-60 http://www.qsl.net/k7qo/
(Chuck answers some email questions about his NC20...ed) 20:31 1 February 99 Jim et.al., Good questions. On the resitors and parts. What I do is bend the top lead over but at about 30 degrees from the body of the part. When I place the part into the two holes there is a friction fit due to the spring type arrangement. By carefully sliding the part in until it is about 1/16 inch or so and I just eyeball this and stick with whatever the first one is. Two reasons for this. One, yes you can probe the bottom near the board and the capillary action of the solder also forms a nice bead around the base so I figure double solder joint. That is why it is very very critical that when you do this you know for sure that you have the right part in the right place, but on removal of the socket I didn't have a problem even with it. The only IC that is socketed is the Keyer chip as recommended in the book if you want to consider a chip upgrade from Embedded Research. I can say this about them. If you saw there talk at Pacificon a couple of years ago when they got started, those guys know their keyer history and the know the difference between MODE A and MODE B when it came time to program it into their chip. I love their MODE B. I think I answered all the questions. I gotta get back on the air. It's what all this stuff is about. Operate and operate..... dit dit es tnx -- Chuck Adams K7QO k7qo@primenet.com CP-60 http://www.qsl.net/k7qo/
Chuck answers some more questions about his NC20 ... 2 February 99 Gang, Thanks for all the email and comments. 1. I take all my pictures with a Epson PhotoPC camera. The camera is about 3 years old and I have done all my photos with it that you see on the web page, both past, present, and future. It generates .jpg images which I download to a TI laptop and then over to the web page. This makes the process fast and clean. It uses standard 37mm lens attachments and I have a series of 3 lens that vary the focal length down to about 12cm or so for the real closeups. The only reason that they come out as good as they do is the Epson camera. It surely isn't my ability as a photographer. :-) The camera does not have a square viewing screen so sometimes I miss the focus or the area of interest is slightly off center and if I feel inclined I redo it otherwise it is what you see is what you gets.... 2. On the 10T turn pot. This is one that I got from the QRP-L group buy. Now hind site is 20/20 every time. When I installed it I played with the configuration and was more concerned on getting the front panel on at 90 degrees so I missed the cover up for the two holes for the resistor just behind it, but it came out just fine mounted underneath as Dave left room with the case to have a small family of critters living in the basement. :-) This is the first rig of all that I have to have a 10T pot for tuning. Glad I did it this way. You get 30KHz with the pot supplied but I get 84KHz with the 10T pot. And I put the AFA in auto mode so that I get the freq in morse everytime I stop and beeps every 1KHz unless I'm tuning fast. 3. The pictures will stay on the web space until I run out of space, so that will be some time and they will be on the QRP-L CDROM starting from today on. 4. I chose the 0.1" headers and Berg connectors to jumper the AFA options. That is what the two strange blue critters are up next to the 10T pot on its RHS facing the front panel. 5. I'll put front and rear facing photos on the web sometime today showing the assembled product. I assumed that these would be boring photos, but did get questions. 6. Thanks for all the email on debugging techniques. For me debugging is a series of steps to trace one or more problems. It first requires some understanding and reading of schematics. Follow the path of the signal or control lines from the source to the destination or the reverse. You know the problem has to be there. Rarely will something else cross a boundary unless it is a solder bridge and if you are following the path you'll locate that problem also. A 8X magnifier of sometime is a must for this kit and most kits anyway, no matter how young your eyes are. :-) Things are closely spaced everywhere. When you first complete the receiver it may scare you that you don't hear anything. Peak the trimmer caps before you give up and think you have a problem. For those of you that are just now getting your kits. You are gonna like this one. Take note of all the postings for hints and kinks. I guess this business of kit building and experimenting is like exploring. There are the first ones that get to fall into all the pits and holes and send back warnings or information about things to watch for, both good and bad. Then the rest of the explorers have a guidebook to help them prevent the same mistakes (well, in most cases :-) ) and maybe even do the job better and neater. Thanks to Dave Fifield, AD6A, for the design work and being the lead explorer. Thanks to Doug Hendricks and Jim Cates for all the hard work and all the hours of devotion and dedication to the workings of NorCal. And to all the beta testers for their feedback and work that made the final NC20 rig what it is today. Dave, AD6A, taught me two things. 1. How to drive on the left hand side of the road. :-) Well, I never thought that I'd have to wind toroids with one winding in one direction and then another link in the opposite direction, but it does work. I'm sure the digital photos in the manual will help a lot of folks and everyone building pay attention. The toroid section is a gem. You'll note how evenly spaced mine are in the photos. That comes from having wound a bunch in my lifetime. :-) Take your time and think about what you are doing. People complain and whine about winding toroids. I personally like it as it is about the only component in the kit that you get to hand assemble. :-) The others you just stick legs into holes and melt solder. 2. It took me two toroids to finally catch onto the trick of melting the insulation off the fancy wire used here. :-) I think OHR had the wire in one kit also, but I just used the old #11 Exacto (tm) blade and scraped the insulation off. But back to melting - I start on the inside close to the toroid and get the wire hot and a glob (TX talk for relatively large amount) of solder to cover the lower half until it melts then add solder to cover the wire and then very slowly move towards the end of the wire. The gunk (TX talk for mess) moves ahead of the glob of solder and doesn't wind up on the final windings. :-) Using a clothespin to hold things steady helps. Don't burn yourself here. OK that's it. I'll let the rest of the builders add to the thread as they come along the path. It is my sincere hope that what I did here helps in some small way with the others. I'm sure there will be lots of questions and answers. The experienced builders will really enjoy the work. The newbies will have to be sure to take lots of time as there is a lot to learn, so just don't get in a hurry. The most important thing that I think that I can and have emphasize is to take your time and enjoy the building. I know we all get anxious to get on the air with it, but if it doesn't work then you'll spend a lot more time debugging and the total time will still be up there high. Not to mention the loss of a part or two if you fry (TX talk for destroy) them. OH, one thing that I need help on and now that I think about it here is another hint. You night consider putting in X6 and X7 very early while there is room around them. Here is why. You need room to solder the grounding wire and you won't have it if the parts around them are already in. I have not figured out a neat installation technique for the grounding wires to the crystals. Mine are still ungrounded. I wanted to bend a lead wire in a kinda S-shape to solder to the base side of the crystals and then to ground to make a neat installation. But I think I will just go with soldering to the top of the case like I've always done. It works. In fact, today I'll get back on the Elmer205 series now that people have had time to work on the quiz and give the results. We'll later talk about what grounding crystal cases really really does in the physics of filters and response factors of crystals. I have waited this long before grounding the crystal cases as I wanted to experiment on a before and after scenario and see just how much difference it makes. Only other mistake in manual that I can think of is that when you got to align the transmitter offset the "dah" paddle is the one that turns off the sidetone on the Tick keyer. Lastly, my paddle problem came back. The stereo jack comes out with a vengence. If you guys and girls think back 5 or more years ago before QRP-L came along. When a new kit came out, what did people do to get information back and forth? Phone calls (expense) and snail mail (time and more expense). Now the Internet is not for free, but it sure gets the cost down and speed up on getting an exchange going. Anyone wanna go back to the olden days? Some do that can't handle the large number of emails, but I think that the overload situation is probably better than no or little information. IMNO. This group stays the way it is until Jim, N3VXI, says different. It's his system that the mail reflector is running on. See ya on the air. I'll be there around 14.060 or so and let's use the area and move the QRM away. :-) Look for the weak one, that'll be me.... dit dit es tnx -- Chuck Adams K7QO k7qo@primenet.com CP-60 http://www.qsl.net/k7qo/
David Fifield, AD6A
31 Jan 99
Folks,
Thanks to input from several of you, here are a few things to
note in addition to all the stuff in the manual:
1. The 82pF C0G capacitor (C108) for the VFO may be labeled
either '820' or '82J', these mean the same thing.
2. Some rigs will have NE602AN chips (U4 and U7) as per the circuit
diagram and parts list and some will have NE612AN chips instead.
There is no need to panic. At HF, these two devices can be
considered identical.
3. Use the drawing on page 8 to construct the VFO coil (L1) rather
than the text. The screw is easier to fit from the top of the
coil. I changed it for the drawing and then forgot to modify
the text.
4. There is just enough red 26awg wire in the kit to do all the
inductors. If you run out of the red stuff, use the green
26awg wire instead - there's plenty of that, and it'll make
the rig more colorful.
5. When you have built the VFO, audio amplifier stages and AFA
you may find that everything appears normal except the AFA
won't announce anything for you. This means the VFO is off
frequency too far - usually too low (e.g. 4.8MHz) - spread
the turns of L1 more and twiddle with TC1 until you get an
AFA response.
6. It's okay to put any/all of the IC's into sockets if you like.
Of course, the better the socket, the better performance and
longevity you are going to get out of the deal. Try to use gold
plated turned-pin sockets if you can.
7. You can modify the values of R1 and R100 to use different values
of 10 turn pots you have handy. However, only go up as far as
20K maximum - 50K pots, for instance, will have problems which
I will explain privately if you like (it's very long winded...).
That's it for now. Please let me know if you spot anything else or
have any hints for people, I'd like to keep everything compiled neatly
together.
Have fun putting it together, and don't forget to get on the air
and actually use it too!
72 de Dave Fifield, AD6A
Gang,
Nice comments by Chuck on his kit.
Chuck, on your paddle problem:
This sounds to me like the classic "piece of component lead flash, or
other trash under the socket". I have had this kind of hair-pulling
problem myself a few times. Closely spaced component holes in the PCB
will sometimes shave off a tiny bit of solder plating as you press a
transistor or other device down near the board. S&S Engineering
specifically mentions this in their kit manuals as a cautionary
statement. I had some of the debris from a part slip under one of the IF
filter crystals in my SW-40+ kit, causing it to periodically go
"deaf". I
finally fixed it by using a short blast of compressed air on the top of
the PCB. Similar intermittents can easily occur under IC's and/or their
sockets.
When I first built my prototype NC20, I had no AGC action. It was due to
a transistor lead sliver bridging across the top of the J310 AGC amp.
Since it was still partially stuck the the transistor's lead, I had to
use a tiny tweezer to fetch it out. These are just some tips for anyone
who might come up with a similar problem.
I grounded my crystals the following way:
Solder them to the PCB first, then pre-tin a small spot about two-thirds
of the way up the side of can, on the same side as the ground hole in the
PCB. A little extra flux may help wet the nickel plating. Then, solder a
cut-off resistor lead straight into the hole. I merely used a resistor
with full leads, as I knew there would still be plenty of resistor left
after I "borrowed" some for the crystal. ;-) Solder it in, straight
up-and-down.
Then, bend over the lead so that it contacts the crystal can where you
just pre-tinned the crystal case. Cut off the resistor lead to the
correct length, then hold it against the crystal with a small screwdriver
or other tool while applying the soldering iron. Use a bit of fresh
solder if needed, or pre-tin the resistor lead. Remove the soldering
iron, and hold the lead still for a second or two until the solder cools.
The whole idea is to have both the crystal and grounding wire all
prepared before actually doing the soldering, as you don't want to
overheat the crystals and change their resonant frequency - thus altering
the careful matching Doug has already done for us. You also do not want
to rupture the hermetic seal at the bottom of the crystal package, since
that is where the crystal is assembled. That is why it is better to stay
up near the top of the crystal can when soldering.
As for the 84 KHz of tuning range Chuck got:
I also had about this much band coverage with mine. Originally, we were
getting about 71 KHz. This was after the NC20 debug team decided to
dispense with a 1N914 diode in the tuning pot voltage divider, as it was
a small source of temperature-related VFO drift. That is why one extra
diode appears in the kit, as Doug had already packaged them! ;-)
Since I don't do any operating above 14.071 MHz, I felt I would reduce
the range a bit. R9 was reduced from 220k to 200k. I couldn't find a 200k
resistor in my junk box, so I used two 100k resistors in series. One end
of each resistor goes into the PCB holes, and the top leads of those
resistors are bent together and tack-soldered. Other tuning ranges are
easily accomodated this way.
Also, I wanted less RIT range than the +/- 3.5 KHz or so I was getting.
So, I experimented and found that by using a 680k resistor for R5,
instead of the original 330k, I could obtain about +/- 2 KHz RIT tuning.
It varies some from the lower tuning end to the higher end.You may adjust
this resistor to produce the RIT tuning you prefer. I found that the
center detent control was making it difficult to move the RIT frequency
just a little bit during a QSO, because the large original RIT range
required only slight movement off of the detent, and therefore it would
"snap back" to center too easily under these conditions.
If your rig seems a little bit lacking in sensistivity, here are some
suggestions:
As Dave Fifield mentioned in the manual, some J310's are better than
others. Dave Meacham and I both found we needed to replace some J310s in
our kits to produce satisfactory VFO and RX sensitivity. Now, let's not
all mass email Doug for replacement J310's un-necessarily!! Remember, we
*were* building prototypes, and in the original prototypes we had to
ground the Gate lead of RF amp Q5, since it was missed in one of the
early PCB designs.
Also, we all did a heck of a lot of work around the VFO JFET, Q2. So it
is possible that during our work, we partially damaged some of the J310's
by handling or replacing them. You all should be careful to not static
damage (ESD) any of the active devices in the kit while building. Winter
time is especially conducive to static situations, so don't automatically
dimiss the ESD precautions that should *always" take place with handling
semiconductors.
Keep in mind that many JFET's vary a lot in performance. It's nice to
have some spares for this rig and other projects, so get some for
yourself and stock up. RF amp Q5 and VFO oscillator Q2 are probably the
two parts most likely to affect performance that use the J310.
A small improvement in RX sensistivity and improved noise figure (NF) can
be achieved by replacing the 2N5179 transistor at Q7, with a Motorola
MRF904. This device is in a similar 4 lead metal package, and has a lower
NF (1.5 NF vs 4.5 NF, and ft of 4 GHz vs 1.4 GHz) than the 2N5179. If
you have some or can get them easily, you might give it a try. I saw a
small improvement in my rig this way.
But beware! It's not easy to unsolder a 4 legged transistor unless you
really know what you're doing, so try it before permanently installing
either device. Or, cut the leads off the original device before
attempting to remove it.
You can just let the unsoldered leads temporarily hold the part in place
during some comparisons. After the best part is found, you can solder it
in. This procedure is part of my "kit blueprinting mods" I often use
to
eek out that last bit of performance. :-)
Although the MRF904 has a little more gain than the 2N5179 (16db vs 15
db), it does not affect the Q7-Q8 shunt-feedback amplifier gain, since
this is set by the ratio of R36/R34, or about 25. But, the MRF904 is a
much quieter device, as indicated by its location at the top of the
Low-Noise Transistor chart in the ARRL Handbook; on page 24-18 in my 1998
copy. So it *is* a superior device, and direct comparisons by me bear
that fact out in operation. If someone has the test equipment to derive
the actual MDS figures, then the improvement can be verified and
substantiated. My ears were good enough for me. ;-)
Other devices were also tried for Q8, with little improvement in NF, as
it is less a factor in the weak signal levels coming from the crystal
filter than Q7 is.
That's all for now. More tips will be made as we go, so stay tuned. ;-)
72,
Gary Surrency AB7MY QRP-L #571 Chandler, AZ (near Phoenix)
Material
and concepts presented on
These pages are designed and maintained by George Heron, N2APB
Page last updated: January
4, 2004