SW-30+ Manhattan Style Project
Chuck Adams, K7QO http://www.qsl.net/k7qo/ is leading a group in building an SW-30+ (30m QRP Transceiver). Follow the link above to the K7QO's website for more information on the build, the schematic, parts list, instructions, etc...

This page will chronicle my progress through the build. To date I have built a few small circuits "manhattan style" and one working rig (a Pixie II). The SW-30+ is definitely my most ambitious manhattan style project to date. I have built many kits so I'm not totally in the dark here. ;)

My plan of attack is to build the radio unmodified, using Chuck's layout guidelines exactly. I plan to learn from this experience so in the future when I branch out on my own I'll have had the benefit of seeing how a real "pro" does it first.

I should also add that the SW-30+ is a Dave Benson, NN1G design and can be purchased in kit form from Dave. See http://www.smallwonderlabs.com for more details. One of my first kits was a SW-80+ (80m QRP xcvr) and it is a great little radio.

Back to my QSL.net homepage

The build...

5/12/2001
workbench.jpg - My workbench - not real pretty, but I built it myself and it works!

vfo_start.jpg - The start of the VFO along side of Chucks suggested layout

vfo_close_1.jpg - A close up of the start of the VFO and more detail of the board

vfo_close_2.jpg - Another attempt at a close up

Soapbox
Seems to be going OK here. I think I built the VFO a little low on the board but it shouldn't be much of a problem. I thought space might be an issue but it didn't turn out that way. I'm going to spread the circuit out a touch from here on out.

I learned a great deal right off the bat from Chuck's layout. I wouldn't have done it the same way and think my way would have caused some problems in the future. It's great to have the benefit of someone elses experience!

5/17/2001
vfo_finished.jpg - Finished the VFO

Soapbox
Had to pull off two caps because they needed to be NPO caps and I used regular old Monolithic. If you're wondering what the difference is, it has to do with how the capacitance of the capacitor is affected when it gets heated (or cooled for that matter). NPO caps are very stable (i.e. their capacitance doesn't change much when their operating temperature changes), while regular ceramic and monolithic are not temperature stable.

There are numerous types of capacitors available. Paul Harden, NA5N authored "The Handyman's Guide to Capacitors", which was originally published in the QRP Homebrewer (NJ QRP Club) and is now available in PDF format at the GQRP Club website: na5n.pdf I highly recommend this small, four page pamphlet.

5/27/2001
vfo-vr.jpg - VFO and Voltage Regulator

vfo-osc-c3-c2-u1pin6.jpg - Oscilloscope shot of junction at C3/C3/U1PIN6 - Volts/Div: 20mV (10x attenuator probe) so each reticle equals 200mv. Voltage p-p measured is about 900mV

vfo-osc-l1-c6.jpg - Oscilloscope shot of junction at L1/C6 - Volts/Div: .2V (10x attenuator probe) so each reticle equals 2V. Voltage p-p measured is 10V

vfo-osc-c10-c9-u5pin2.jpg - Oscilloscope shot of junction at C10/C9/U5Pin2 - Volts/Div: 10mV (10x attenuator probe) so each reticle equals 100mV. Voltage p-p measured is 400mV

I measured the period at C3 as .41uS.

Freq = 1/period = 1/.00000041 = 2439024.39Hz or 2.439MHz

Next, I powered up the VFO next to my ICOM-746. I can hear the VFO oscillating at exactly 2.436.05MHz, which confirms what I read about accurate frequency measurement with an oscilloscope - it's not.

Soapbox
This is the first time I've used my oscilloscope in an actual project. I just acquired it about a month ago at a hamfest. I'm completely self-taught when it comes to electronics (I'm a software engineer by trade). The reason I mention this is just in case I've made some major mistake here with my measurements the excuse canon is pre-loaded. ;).

5/30/01
After posting this, Cla, KA0GKC wrote to say:

"I think that your conclusion about the frequency measurement with the scope is actually false. You're measurements between the scope and your receiver agree within 1%, that's not a bad measurement. If fact it's quite good when you stop to think about accuracy. Your scope, if properly calibrated, displays the signal with great accurary, reading the scales to within 1% can be difficult."

Cla goes on to say:

"The final VFO frequency will be determined by the "real" crystal filter bandpass frequency and the VFO will need to be tweaked anyway during final tune up."

"Keep up the good work and have fun!"

Which makes a lot of sense to me. Thanks for the input Cla!

6/01/2001
I'm now building the First Receiver and First Transmit Mixer sections. I've built these in advance of Chuck posting a plan for them, but I think they are fine.

first_rcv_xmit_mixer_1.jpg - First Rcv and Xmit mixer - Photo 1 - This is taken from the left side of the board, the xmit mixer is on the left, vfo in the middle and rcvr mixer on the right.

first_rcv_xmit_mixer_2.jpg - First Rcv and Xmit mixer - Photo 2 - Taken from the top of the board, mostly shows the first xmit mixer.

first_rcv_xmit_mixer_3.jpg - First Rcv and Xmit mixer - Photo 3 - Taken from the bottom of the board looking left.

first_rcv_xmit_mixer_4.jpg - First Rcv and Xmit mixer - Photo 4 - Same as Photo 3, but with a magnifying glass.

Soapbox
What I did was pull out all the parts I was going to need for each section ahead of time to get a feel for their size. I then sketched out on paper while following the schematic and came up with a few different layouts. I picked the one that made the most sense and went with it.

For the IC's (602's in this case) I used IC sockets. I folded all but pin three out straight (pin 3 goes to ground). I then pulled pins 1, 4, 5 and 8 out at approximately a 35 degree angle and pins 2, 3, 6 and 7 at approximately 10 degrees. I dropped a small blob of solder where I thought pin 3 (the ground) would go and then soldered pin 3 on the socket to ground. This allowed me to accurately position the pads for the other pins. I plopped a small drop of super glue where the pad would go and slid it in. After placing pads for all the pins and solder the socket lead to the pad, the socket feels like it's held pretty firmly.

I am not 100% sure I have the pins of T1 Id'ed correctly. Here is a post I sent to the QRP-L yesterday. I haven't received a response yet, so I'm still up in the air about it:

The pins on T1 are ID'ed as follows; with the pins on the bottom and
side with the part id stamp facing you, tip the top away from you so you
are looking at the bottom of the can. Going clockwise from the top left
the pins are 1,2,3 then straight down to the bottom (i.e. right side), 4
and 6 (on the bottom left). I have U1 pin 1 going to T1 pin 1, U1 pin 2
going to T1 pin 2, T1 pin 3 not connected, T1 pin 4 going to a pad which
will eventually go to the RF gain and T1 pin 6 to ground and the two
tabs on the side of T1 to ground. 

Does this sound correct?

I'll post any responses here.

UPDATE: There were no responses. Chuck has posted various measurements and experiments he's doing with regard to removing the IF cans and replacing them with Toroid inductors. So I've read into his responses enough to figure out that I've got the IF cans set up correctly.

6/14/2001
I actually finished these sections last week sometime but was called away on personal business.

On Chucks web site there is some discussion about making pads for IC's. Chuck used a small miter tool and hobby knife to make some very attractive pads. I attempted to duplicate his results with some success. I used a hacksaw blade and a cutting tool from my model railroad (HO scale). I used the new IC pads for two of the IC's in the radio. The other two other two are mounted using the "moon lander" method which I described previously.

ic-pad-cut-1.jpg - This is a shot of the IC pad and the tools used to create it.

ic-pad-cut-2.jpg - Another shot of the same.

man-ic-pad-1.jpg - Shot of the pad and the radio with the crystal filter started.

man-ic-pad-2.jpg - Same thing from a different (better) angle.

pad-layout-1.jpg - I glued down the pads for the filter and the IC prior to mounting most of the parts. For the crystal filter, I used two long strips of copper pad material.

planning-preamp - I've been doing my own layout ever since the VFO. Here is a shot of the preamp being planned on paper with the radio right next to it. More on this in the "Soapbox" below.

train-1.jpg - Here is a shot of a small section of my HO train layout (figured I'd include it here since I mentioned it above).

Soapbox
I'm starting to feel pretty comfortable laying things out and building on my own. Nobody has posted a suggested layout since the VFO, so I just sort of went off on my own. I've got a bit more done then what is shown here and will take some pics and upload them soon. I've almost completed the receiver.

I've switched from using plain 'ol 4-for-a-buck at the dollar store super glue to the "Scott Super Glue Gel." The Gel makes it very easy to accurately place the pads.

6/18/2001
It lives!

These pictures span from the finishing of the receiver and then jump to the transmitter being finished and the radio getting on the air. I will fill in here with o'scope shots at various spots on the rig soon. The Soapbox below will outline what has happened here and the problems I ran into (and there were problems!).

filter-detector.jpg - This is the IF Filter and Detector (U3). I built the filter unmodified but plan to go back and install Jim Kortge, K8IQY's 500Hz butterworth filter.

if-filter-detector.jpg - This shot shows the connections from the IF filter to the detector.

final-audio-amp-1.jpg - This one side of U4 being finished. The other side was much more difficult.

final-amp.jpg - Shows Q6, the final transistor. Also has a lot of detail of the LP filter and driver.

audio-amp.jpg - Finished audio amp can be seen here. The resistor that is up over the top is R10, which I mention below (because I forgot it initially).

lp-filter.jpg - The keying section is in the middle right, the low pass filter on the top right.

lp-filter-2.jpg - A close-up (too close actually) of the lp-filter.

lp-filter-3.jpg - A bit further away shows nice detail of the LP filter.

rightside-bottom.jpg - Nice shot of the IF Filter, Detector, Audio Preamp T/R Mute and Final Audio Amp.

transmitter.jpg - This is the xmit side of the board.

itlives.jpg - I built the transmitter in a day. Here is my MH101 SW-30+ just after its first QSO with WY5P, Gary in Houston TX. Not too shabby for 2 watts from the top of VT.

Soapbox
There was a bad batch of parts in the finals that a group of us bought. That left me with an almost finished radio - no good! I have an SW-80+ kit that I put together a long time ago but hardly use. I stole the final from it for use in this radio.

After finishing the radio, I was looking forward to putting it on the air. I had visions of hooking it up to a battery and my headphones would be filled with signals... it didn't happen like that... at all.

First of all, I heard nothing. A little pop when the rig came on, but that was it. I took the radio back to the bench and traced the audio amp. I noticed that I had left off R10. After installing R10 it still didn't work.

I hooked up an RF generator and started tracing the signal backwards with my oscope. When I got around Q3, the keying switch, my probe slipped and I shorted something to ground. A pop was heard and I let the smoke out of Q3 (actually it shot its plastic guts straight down on the board). Q3 is a 2N3906 and I didn't have a spare in my junkbox. Some research on the net and viewing of some spec sheets showed me that a 2N4126 is identical except it handles less Voltage. It was still within spec for what I needed and I happened to have a few of them so I used one. I confirmed on QRP-L that is in fact a common replacement.

Still nothing in the amp though. Then I discovered that some of the IC sockets had been pulled off their pads. I reheated them and tested them with a continuity tester. OK, good to go... nope, still nothing.

I then rechecked my RF Gain and Tuning pots and realized I had them both wired in wrong.

Finally, signals in the phones!

I then touched the keying wire to ground in hopes of hearing a signal. Nope. Back to the scope.

I was getting some AC voltage off T2 and into T3 but it died just after Q4. I then realized I had R24, the 500ohm variable resistor wired in wrong. After fixing that I started to get RF out!

I peaked T2 and T3 with my scope and adjusted the aforementioned R24 to maximum gain while keeping an eye on my waveform. The waveform distorts with too much current let loose from R24. I'm now getting about 2 watts out and it looks clean.

While keying the radio, I was noticing a loud click on key up. I posted to QRP-L about and Jim Kortge replied with this:

Maybe, but most likely not.  More probably a 
wrong value of capacitor on the keyline, or 
gate of the mute FET, or perhaps a bit of unwanted 
coupling between the keyline and the audio, or 
maybe excessive voltage drop in the power 
leads/supply to the rig.  Can you tell if the 
click is on key down or key up.  It might help 
us to figure out where the problem is.

Jim was correct. The problem turned out to be a bad connection from the Gate of Q1 to C24. After reheating the solder joint the clicking stopped. The rig now works great! I should also add here that Jim has been instrumental throughout this build. He has answered many questions and designed in some mods that I plan on trying.

I think I've skipped over some other problems I had. I spent a good six, if not eight hours going over the rig with my scope. I traced the rcvr all the way though and the xmit all the way back. It has been a VERY good learning experience so far.

My next update will be after I finish an enclosure for the radio. I will also post some oscope shots that show how to tune up the radio.

"Do not be afraid of standing up in front of everybody and failing miserably. People before you have failed and people after you will fail. If you don't fail at something, you aren't trying hard enough and you aren't going to grow. Failure is not an option, but it is a very real possibility. The key to not failing is to look at others who have failed before you and don't repeat their mistakes."