Podcasting since 2005! Listen to Latest SolderSmoke

Monday, October 31, 2016

Hacking the Hackable BITX 40 Module: VFO is the Way to Go!


I am having a lot of fun with Farhan's new BITX 40 Module.  I think I'm doing exactly what Farhan intended people to do with this rig:  work on it, modify it, improve it. 

I've been working on frequency stability.  I was, I admit, skeptical from the start about the stability of a  thumb-sized,  SMD, varactor-tuned VFO with a ferrite or iron powder toroidal coil.  Don't get me wrong -- it worked.  But it drifted. It seems to me that it would be asking too much to expect a VFO like this to be drift-free. (But I may be wrong -- are there any SMD, varactor-tuned VFOs out there that DON'T drift?)

First I thought it might be the 9 uH metallic core toroid.  So I replaced that with a 10uH choke -- no ferrite or iron powder in there.  That seemed to help a bit, but SSB QSOs would still quickly drift into Donald Duck chatter.  Then I thought it might be the varactor diode.  I let it warm up.  A lot.  Still, it drifted.  Then I thought it might be the trimmer cap, so I took it off the board.  No change.  During this process I noticed that even slight pressure on the board caused the rig to shift frequency.  I began to suspect that the drift was just structural -- a consequence of the physical characteristics of the SMD parts and the board.  To get VFOs stable I've had to build them big:  10 X 10pf  NP0 caps to make one 100 pf cap, large air-core coils, and big sturdy variable caps.  I'd isolate the frequency determining elements in a box separate from the powered components. This little VFO just looked too small to be stable.   

So faced with drift, at first I asked myself,  "What would Pete do?" I took an AD9850/Arduino combination off the shelf and plugged the output into the "DDS" jack Farhan had placed on the board.  I removed the 10uH choke.  Viola! With the DDS tuned to 4.7 - 5 MHz, the receiver worked great.  I briefly tried to updated the Arduino code to take into account the 12 MHz IF (so I could get an accurate frequency readout), but ran into the old painful Arduino IDE problems:  Now it is claiming there are library problems.  Not wanting to suffer through another round of digi-agony, I left well-enough alone.   I used the DDS with the old code for one day. 

But of course, I was not satisfied.  Attaching a DDS or PLL synthesizer to the BITX 40 Module just didn't seem right. Heck, it was kind of like just hooking up my FeelTech Chinese sig gen to the DDS jack.  Farhan's rig is simple, beautiful and ANALOG.  The parts are small, but you can see them.  You can put your scope probe on the collector of Q7 and see what is going on.  DDS or PLL.  It is a REAL HARDWARE-DEFINED RIG.  So I decided to build a VFO.   Pete calls VFO's "grief machines"  but for me, the grief machines are those little Arduino beasts.  To each his own.

When I build a VFO, I start with the variable capacitor and the reduction drive.  I found a nice one (with reduction drive) in my junk box. I tunes from 40 pf to 56 pf.   I decided to use the super-simple Hartley circuit presented by Wes Hayward W7ZOI in SSDRA (page 34, fig 7). 


I went with a 4.4 uH air core coil (wound on a cardboard tube from a coat hanger).  Consultation with on-line resonant frequency calculators showed that I'd need to put about 180 pf in parallel with the variable cap.  For this, I used a bunch (maybe 10?) of small value  NP0 caps in parallel.  This really helps keep the VFO stable.

As I did with my HROish receiver, I put the coil and the caps in one box, with the MPF-102 and associated parts in an  attached Altoids tin.  Everything was glued and bolted down very solidly.


I only built the actual oscillator stage -- I decided to use the buffer amps on Farhan's board.

The oscillator started right up.  I had to add and then take away some turns on the coil to get it to run in the desired range.  Then I plugged it into the DDS jack -- the receiver was working immediately.

I noticed, however, that it seemed a bit less sensitive than it had been with the AD9850 DDS. And when I grabbed the wire going into the DDS connector, audio output jumped dramatically.  It took me a few minutes to figure that out:  I think the output from my VFO was not adequately turning on the diodes in the diode ring.  When I grabbed the wire, I was putting a lot of noise into the mixer port, probably turning the diodes more fully on (but also letting a lot of noise through).

Fixing this problem part was fun:   Looking at the BITX 40 schematic, I saw that the two 1000pf feedback caps in the original oscillator were still in the circuit.  I figured those caps would be sending a lot of my VFO energy to ground.  So I fired up my hot air rework station and deftly removed C91, the 1000 pf cap that is connected to the base of  Q9.   Instantly the receiver started inhaling as it had with the DDS VFO.  That was a very satisfying fix.

This whole VFO project was very satisfying.  It was all done in one day, and all the parts came out of my junk box. I think I ended up with an LO frequency source that matches up in a pleasing way with the analog circuitry in Farhan's rig.  And here is bonus that I think is just what Farhan had in mind:  this kind of circuit adds a definite homebrew element to the module rig.   

I found that this external VFO improved stability significantly.  I don't know if it is as stable as the DDS, but with the external VFO the receiver no longer drifts away as I listen to SSB signals. 

Saturday, October 29, 2016

DONE! Jan's AMAZING Mate for the Mighty Midget Receiver


Hi Bill,
  
Finally the Mate for the Mighty Midget is finished, just in time for the G-QRP Valve Day 12-13th of November.
Got the LO fixed for 40m by lowering the parallel capacitor from 150 pF to 100 pF.
Also the 68 pF series capacitor was lowered to 33 pF for some more band spread on 40m.
It now receives from about 7.0 to 7.4 MHz and from 3.45 to 4.0 MHz
Had to exchange C1 in the end, the one used initially quit every now and then.
I only had a large 3 section variable in the junk box covering 10-550 pF, which works fine now.
For the lower end of 80m I had to add additional 47 pF next to the 47 pF trimmer caps, so there it is about 600-650 pF max!
At the high end of 40, it is also just not too much.
The meter was used as a position indicator for C1.
Tried several ideas, but with no separate tube for AGC, I couldn’t get it to work as a S- meter
Read something about audio derived AGC, maybe this is worth a try.
 
The BFO can be switched off for AM reception.
Simultaneously the input on the mixer side of the crystals is disconnected but still coupled by some capacitance of the switch wires.
AM reception is possible, but not very good.
Need to find a better solution which doesn’t degrade the crystal filter properties to much.
(By the way, the detector regen. control ads about 4 dB to the AM sensitivity)
The receiver will mainly be used for CW/SSB reception, so maybe it stays this way for a while J
 
I hooked it up to the W1TS transmitter, which was very loud.
Didn’t foresee a T/R relay (learned a lot from this project ;-)  ), so added this one between the front plates next to the RF and audio gain control.
The quit down everything a little, the RF gain pot is lifted of ground as suggested by James, N2EY at QRZ.com.
It helped a lot, but was still too loud if tuned exactly in the bandpass of the crystals.
The T/R relay now also switches an adjustable potentiometer at the input of the audio pre-amplifier.
 
The dial cord has no lag, and works very well for fine tuning.
Unfortunately the reduction drive went from 1:19 to 1:9…, the tuning capacitor only has a 180 deg. span.
Something to remember for the next receiver.
 
It’s a nice little receiver and quite stable after warm-up.
The only extra luxury a next receiver will have, is AGC.
But with no AGC it’s easier to tune the antenna tuner by ear J
There’s now a complete homebrew station here, antenna, feeder, tuner, receiver, transmitter, power supply, al home made J
 
About the Mystery Hole….
If you haven’t guessed it by now, or Pete hasn’t told you, it is revealed in one of the pictures below.
 
I also made a little video:
 









Friday, October 28, 2016

Another Great DSB rig from New Zealand


So many great Double Sideband projects come from Down Under.  There are the various versions of the famed ZL2BMI rig.   And Peter Parker VK3YE has long been the acknowledged guru of DSB. In fact, Peter sent me an enthusiastic e-mail about the new ZL DSB rig pictured above -- his e-mail arrived before the message (below) from the intrepid builder.  I detect a bit of the "Tucker Tin" influence in this rig. (But perhaps this one is more Tupper than Tucker!)  Charlie's work has graced out blog posts before: http://soldersmoke.blogspot.com/search?q=zl2ctm
Be sure to check out his video:   https://youtu.be/JsAuWGkyvmE   (and below).


Hi Bill.

You asked recently on the podcast for listeners to let you know what they had on their work bench. Well I’ve been working on a tramping (hiking) radio, which is now complete. It’s a DSB 5W rig designed for 80, 40 and 20m, as well as our New Zealand mountain safety radio system. I designed everything in LTSpice as was suggested by Pete, N6QW. That was great, as I could ‘desolder’ components with the mouse and instantly see what impact it had on the output. An amazing tool that’s free! I highly recommend it.

Once again I’ve used upside down strip board for each stage, which are tacked down onto an un-etched copper board (earth plane). That seems to work really well for me.

The rig uses an Arduino mini driving a small OLED screen and a Si5351 DDS. The Si5351 is going straight into a SBL-1, which seems to work fine too. The AF strip is a 2N3904 before a LM386, which has enough drive to run a speaker. The TX amplifier is a three stage one with shielding between each stage. It’s made up of two 2N2222A stages followed by a BD139. That in turn is followed by three simple filters, one each for 80, 40 and 20m.

All-in-all it works really well. I’ve uploaded a quick video at https://youtu.be/JsAuWGkyvmE

The next project will be a proper SSB rig using a crystal filter salvaged from an old Codan 7727. Like this one, it will use an Arduino and a Si5351.

Finally, I am certainly no expert in homebrew, but I hope my ‘dabbling’ will help inspire others to pick up the soldering iron and give it a go. If I can do it, then anybody can! There is certainly a great sense of achievement to operate a rig you built yourself.

Regards, and thanks to you and Pete for all your inspiration.

73’s
Charlie
ZL2CTM


Thursday, October 27, 2016

I.G.Y., The Nightfly, Donald Fagen, Jean Shepherd and SolderSmoke


OK, so from time-to-time we  talk about IGY, the International Geophysical Year. I was born during that scientifically momentous period.  A lot of cool stuff happened.  Amazing propagation conditions too.  So for a while (around SolderSmoke 149) I was using the opening bars of Donald Fagen's song I.G.Y. as the intro for the podcast.  That song comes from Fagen's album Nightfly.  The album cover appears above.

This morning I got two e-mails from Steve N8NM about another connection between SolderSmoke and IGY.  At first I thought he was pulling my leg.  But  before I show you the e-mails, let me show you another picture:


That's Jean Shepherd.  Can you dig it?

Steve writes:

Hey Bill! 

I'm Listening to #149, where you introduced Donald Fagen's "I.G.Y" as the new theme song.  Have you heard that the protagonist on that album is based on none-other than Shep, K2ORS.  Don't know if that's necessarily true, but the album title (The Nightfly), Fagen being from NY, and the era depicted certainly make that plausible...

I guess it's factual - This is from an interview Fagen did with New York Magazine:
Your first solo album, The Nightfly, was inspired by fifties jazz disc jockeys. Which ones were your favorites?
Symphony Sid was very popular. Mort Fega was probably the best all-around jazz D.J. Ed Beach on WRVR would do this very scholarly afternoon show, and I’d listen to that when I came home from school. But the figure of the Nightfly was based more on a guy who didn’t play jazz records, Jean Shepherd. He was a monologist who used to just talk and tell stories and say funny things. He was a social satirist.
Speaking of Shep:  You've done him proud, not only in your HB radio efforts, but in proving yourself to be a very capable monologist for the several years between Mike's passing and Pete's arrival.  Three Cheers for Bill!  Give that man a brass figlagee with gold leaf palm!
73!  Steve N8NM
By the authority vested in me by having once spoken to Jean Shepherd, I award Steve Murphy N8NM, the coveted Brass Figlagee with Gold Leaf Palm.  

EXCELSIOR!  

TRGHS

Wednesday, October 26, 2016

Farhan on What's New in the BITX 40 Module



Writing to the BITX e-mail group, Farhan provides some very interesting information on the philosophy behind his new BITX 40 module, and on how it differs from earlier BITX designs:

The new builders are often caught in a catch-22 : to get on air they need to build a rig from scratch. but to build a rig, they need lots of experience. A way out was to provide working boards where we can get on air quickly, and then start improving and modding the circuit. this is the spirit behind the new boards. Consider them like you would consider a raspberry pi or an arduino : simple, working circuits around which you can grow your own radio.

In the new bitx boards, I have tried to keep as close to the original bitx as I could. however, there are a few departures that I thought the bitx builders here would like to know about.

i have to admit though, strangely, i am less familiar here with bitx than many others on this form. arv, leonard, dan,  andy and others have build far many more version than I did. I just happen to be the first one to build a bitx. this as much an acknowledgement of their inputs. without all you folks, bitx would not have had the kind of traction that it now enjoys. I suspect that it is the most built transceiver in the world.

So, here are the changes from the original bitx.

1. SMD
The SMD components make for virtually error free boards assembly. We used the biggest sized SMD components. In fact, the resistors and capacitors are about the same size as a quarter watt resistor that is soldered standing up. They are very easy to desolder without messing around with the desoldering wick and solder pumps. All you do is to lay the soldering iron's bit on the component such that the flat end touches both sides at once and after a few seconds just drag the component away. I soldered the sample boards with my regular, 2 dollar, 25 watts iron without using a magnifier (I wear reading glasses). 

1. 40 Meters
It is just that with the sunspots fading away, 20 meters in the tropics is far less active than before. Many of the us South Indian hams hang out on the lower end of 40 meters every morning and evening. Hence, the choice. That doesn't mean that i can't be converted to 20m! There are several ways to change to 20 meters. Keeping the VFO same, change the crystals (and hence the IF) to 8.833 MHz and rework the band pass and the low pass filters. I will work out the details in a few weeks and post them here.

2. A new bandpass filter
The original bandpass filter was quite lossy. I didn't know how to use any CAD tools when i sketched it. I was actually on a long haul flight when I designed that filter. The new filter configuration is very interesting one. I saw it on PA3AKE's site. This is a triple tuned circuit with very good out of band attenuation while maintaining very low loss. 
In the last then years, ecomm has made it possible for us to globally access good quality toroids anywhere in the world. Hence, we have used T30-6 toroids with excellent low loss. I measured it at just 2 db, the original had more than 6db loss.

3. VCO
The original oscillator on the BITX used a variable capacitor. These were noisy, and often of inferior quality. In any case, they are no longer available. Instead, we have used a varactor diode for tuning. The greater benefit of using a varactor to tune the oscillator is that the tuning control only carries a DC voltage. You can install it anywhere. If you need finer tuning control, you can add a second lower value tuning pot in series with the main tuning pot. It is easier to add FLL to a VCO.

4. Audio muting
The original BITX used just a switch to move from receive to transmit. The receive voltage charged receiver's audio preamp's decoupling capacitor and it took time to discharge. this kept the audio preamp active even on transmit and caused a very sharp audio noise on the transmit change over. Now, the other section of the T/R relay is used to cut the audio off to the LM386 as soon as transmit line is energized.

5. A better T/R system
The original bitx didn't have a PTT. this one has two relays to switch the linear amplification chain in and out of the circuit, mute the audio and change over the antenna. These changes lead to a very stable linear amplifier and smoother change over. 

6. Mic amplifier
The original mic amplifier easily saturated. The new design, thanks to dan tayloe, has a better head room and provides very clean modulation.

7. The fixed BFO
Though the PCB has the provision for a trimmer and an inductor to pull the crystal frequency. I discovered that with five matched crystals, if you used 4 in the ladder filter, the fifth's frequency fell right into the perfect sweet spot for LSB work. You might need to add the trimmer and an inductor back for USB work.

8. DDS connector
To use the DDS, you will have to remove L4 (the VFO inductor) and inject the DDS/PLL output into the connector provided.

There are some smaller mods that people can try out:
* The current in the receive amplifiers can be reduced if you don't have any radio hams in your neighbourhood who run kilowatt amps. 
* The capacitor between pins 1 and 8 of the LM386 can be removed if you prefer headphones to speakers. 

Tuesday, October 25, 2016

VK3HN: SOTA, HB, SSB, and QRP FB!



Peter VK3YE sent me the link to this amazing site.  Wow, Paul VK3HN does great work, both with the homebrew rigs and in describing his work on them.  Check it out: 

https://vk3hn.wordpress.com/2016/10/25/summit-prowler-one-a-homebrew-7mhz-ssb-qrp-transceiver-for-sota/

Great stuff.  Thanks Peter!  Thanks Paul!


Designer: Douglas Bowman | Dimodifikasi oleh Abdul Munir Original Posting Rounders 3 Column