Thursday, September 21, 2017

A New Toy

One of the first things I found when I started building projects that rely on parts from the Far-East is that it is better to have more than one project going at any time.  Still waiting on a couple of parts for the new 3D printer that I forgot to order when I started the project. And also some J310 Fets for the Simpleceiver project. I thought I had some, but when I checked they were the wrong type.  I have managed to get the Arduino VFO-BFO board working with a modified version of Pete's sketch.
Checking an e-mail add from the local MicroCenter store that I usually buy my 3D filament from, I noticed a small CNC engraver on sale for under $200.  It looked like it would be the right size PCB mill drill for the size boards I usually do.  I mostly do toner transfer with great success, but there is always the problem of drilling holes.  I have gone to SMD or 'Muppet' style construction for most of my projects to reduce the number of holes that have to be drilled.  This looked like it might solve that problem, so when I went up to pick up some 3D filament I also picked up one of the engravers.
Size is just about perfect, less than a foot square and high, just the right size to fit in front of my small 3D printer. 
It came  assembled except for putting the motor in the holder and installing a cutter. There was no software with it, but the included documentation gave a link to a program that would send g-code, and manually control the printer.  After playing with this for a while, I went looking at a way to convert pcb files to g-code.  I use the free version of Eagle for most of my boards, so I did a search on Eagle to g-code conversion.  I turned up several YouTube videos on a add-in to Eagle that did that from within the Eagle GUI.  I tried a simple board layout as a test, and looking at the output with the CNC control program everything looked like it would work well.  Tried routing that on a scrap piece of PCB material.  Most of it came out fine, but there is one area where it did not go all the way through the copper layer.  A little more looking around I found an additional software piece that will probe the material, and generate a surface map to auto level the engraving surface.  I just need to make up a simple cable with a couple alligator clips that connect to the CNC controller board and give that a try.  If that works out I might be able to switch from etching boards to routing them, and drilling without having to get out my small drill press. 

Thursday, September 14, 2017

A SI5351 VFO-BFO UPDATE 9/15 9/16 9/21




Working on Pete's Simpleceiver + I need to build another VFO BFO.  The SI5351 based boards have been my favorite for several years now.  I have several different versions of boards that I have tried, but prefer the small module from Adafruit.  It is very small, and comes built for around $8 plus shipping.  One of the things I like about it is that the clock outputs are brought out to a header along with the SMA connectors.  I have used this board in several projects and have designed different boards for each one.  With this I plan on designing a board that can be used with many different projects.
  
The design considerations are, board size around 2" x 2",
Arduino Nano, rotary encoder, TFT display, a connector for the Adafruit SI5351 board. After an initial design and exchanging several e-mails with Pete I have expanded it to also provide a connector for a small  I2C OLED display.  Because there are several pinouts for these displays, some will plug in directly to the connector others will have to use a connector cable to the display. This connector could also be used to connect to a I2C version of a 1x16 or 2x16  line LCD display. With the correct software drivers you could use one of these LCD displays, a small OLED display, a 128x128 or 128x160 TFT display, or even a monochrome Nokia display.    I also added a connector that brings the rotary encoder connections out to a separate connector so you can  use either the bare encoder mounted on board, or one of the pre-wired encoder boards. I have not tried but you could probably use an optical encoder through this connector. 
One thing I added was provisions for a voltage divider that can be used to measure input voltage if the board is used in a battery powered project. Finally I decided to bring some of the unused Arduino pins out to an auxiliary connector for use in other projects.

Since the board layout is similar to several others I have done, I was able to modify one of them fairly quickly.  I did the board as a single sided board that can easily be made using the "toner transfer" method.  Although I usually etch my own boards, in this case I decided to order some from one of the inexpensive Chinese board houses.  If they turn out well, I will follow Pete's suggestion and make some available at a reasonable cost for those who are working on the Simpleceiver or other project that needs a VFO-BFO.  I think I will start with one of these boards as the basis of a simple signal generator for routine use.  This should be a easy way to test the different video drivers needed for the different display options available.

I ordered the boards on Sept 9 with DHL express delivery, and was pleasantly surprised when I received them on the 13th.  I quickly wired one of them up to see how they will work.  Changing some of my other sketches to reflect the correct pin configuration I tried the board with a small OLED display and a 128x128 TFT display. After getting a working display with each, I was sure that the board worked correctly. 
I installed a right angle female header strip so I could mount an external rotary encoder board, and took some pictures of possible display configurations. There are several more that I could try, but this is what I had handy.
I also took a picture of the Adafruit SI5351 board mounted on the back.  Depending on how the header pins are installed in the board, it can also be flipped to have the edge with the SMA connectors extending past the board edge.  I planned on the Nano and SI5351 board to be soldered to the board, but they could be socketed if you do not need to keep everything really compact.  As it is with everything soldered the package with the OLED or 128x128 TFT is about 2"x 2"x 1".  This should be small enough to fit in most project packages.  Now to write some software to fully test each of these configurations
9/15/17

I had some time today, so I took Pete's Simpleceiver Plus DCR sketch and modified it to reflect the changes in the pin assignments on the new board.  And after I had it working with the 160x128 TFT display I modified that to reflect the 128x128 display.  There are several different versions of the 128x128 display that use different driver chips, so I had to also change the driver library and initialization code. With a little tweaking on position on the display I was able to get a very similar looking display screen. I had taken all of the additional code in the sketch that reflects U/L side band selection and several other things that Pete has in his code to make it more modular in design.  When I get around to going to a superhet and then transceiver, these can be added in as separate .ino files in the main sketch directory.  That way it should be possible that no other changes in the additional .ino files should be necessary whatever display you are using.  Here are a couple pictures of the two displays.  Hope to get around to writing some code for a couple other display types this weekend.

128 x 128 display  Rotary encoder on board

160 x 128 display  External rotary encoder



I spent a little over an hour this evening modifying one of the earlier sketches to support a small 0.96" OLED display.  Most of that time was trying to fit everything on the screen, and make it look nice.  It is much easier to work with a screen that has more real estate to work with. This is listed as a two color display, but all pixels are either white or black, the color comes from  two different colored parts of the screen.  In an earlier project, I used the smaller color area to put current settings.  

One thing to be very careful about when using these small OLED displays is that looking at some of the pinouts for ones for sale on ebay, is that some have the VCC and GND pins reversed.  
SO BE VERY CAREFUL and check before just plugging it in the socket.  I still have one of the monochrome Nokia displays around, but with the problem I had trying to fit everything on the screen, I don't know if I will bother porting the code for that configuration.

Now that I know the boards work for multiple display types, I have decided that I will offer them for sale.  Because of Pete's urging, I ordered 50 of them to start with.  For now I only plan on shipping to US locations, but am working on getting the board layout as a shared project at the Chinese board house I used. This should make it easier for DX locations to order them directly.  I had checked at oshpark.com, but the board size made them rather expensive.  The board is single sided, so fairly easy to make with the "toner transfer method" if you want to make your own.  I will have the "toner transfer" image along with other documentation at 

https://www.dropbox.com/sh/xjfh93dfcu5hr9y/AADaHSciYgwUcPb8_gQgb_h1a?dl=0

This dropbox will also have the Arduino sketches to use as a guide in using the board with different types of display.  They are very rough right now, I just did enough work on them to get the display and rotary encoder to work correctly.  I will be doing a lot of work as I progress with the Simpleceiver project and see what Pete has in store for us.

I am offering them at $5 each or $7.50 for two including shipping. If you are interested email me at duwayne@kv4qb.us and I will give put you on the list.  

9/17/17
I have just uploaded the files for this board to the supplier I used to their shared projects area.  You can download the gerber files from there or order directly.  I have placed several orders with this supplier, and have been very happy with the results.  Their pricing is very reasonable, and they offer several different shipping options that can be very reasonable. This will be the most economical way for DX builders to purchase boards. The project is at
https://www.pcbway.com/project/shareproject/W42368ASJ5_W42368ASJ4_si5351_vfo_bfo.html

And yes I do get a commission from boards sold. So if two people buy  sets of  5 boards I save enough on my next order that I can stop at McDonalds and get a cup of coffee when it is on sale.


9/21/17
Finished up the modifications to Pete's Simpleceiver sketch for use with my PCB.  I combined the direct conversion and super-het versions into one. There is a single flag variable dcr_mode that you can change to go from one to the other. I have it just before the start of the setup area in the sketch to make it easy to find.  I have it tested as much as I can without having the complete hardware.  Pete plans on building up one of the boards I sent him, and giving the software a test run.  It is located in the folder Simpleceiver_Plus_DCR_SSB_160x128  at the dropbox link

https://www.dropbox.com/sh/xjfh93dfcu5hr9y/AADaHSciYgwUcPb8_gQgb_h1a?dl=0


Saturday, September 2, 2017

Simpleceiver Circuit boards

I have received several e-mails with questions about the boards I am making for Pete's Simpleceiver project.  I use a layout method known as "Muppet" style.  This is well documented by a series of YouTube videos by Chuck K7QO.  The first of his 12 videos covering design, layout, etching, and building is  at

https://www.youtube.com/watch?v=O6AmT1trO60&list=PLaxN7ey82-BnPIbAzacv_PMvACpAd6ZT7

The boards are laid out using the free "expressPCB" software, this is about the easiest to learn PCB software I have found.  Since it was made to be layout software for a single board house, it does have some limitations compared to other programs.  But, for making simple "toner transfer" boards it is more than adequate.
I have made a few changes to Chuck's procedure for board layout. The main change has been to modify several of the component foot prints that comes with the software, to provide large mounting pads instead of the through hole pads provided.  
I find it much easier to grab the correct Muppet component from the custom list and place them where I want. These  have a outline of the component over the pads, which makes it much easier to follow the circuit diagram.  I have pads for the common resistors, capacitors, diodes, transistors, and IC sockets I use most of the time.  Any other can be built from standard pads as required.  This also makes it easy to mix through hole components along with standard size SMD components if I am laying out a SMD board.

Since the "Muppet" board is etched on the top layer of the PCB, it is necessary to flip the image right to left for proper transfer. Chuck prints the top layer image to a PDF and then uses a linux program to flip it.  I found another program "Copper Connection" that will import expressPCB files and then print both sides in the correct orientation for "toner transfer" or photographic method.  It also has some nice features that I sometime use to help with the board layout.  One of them is to flip a board over when doing double sided boards.  Unfortunately it is now hard to find a copy of this software after they were bought out by expressPCB.

Another change is the "toner transfer" method I use.  Chuck goes through a great explanation of the "Hot" method using a laminator. I have found that the results of this method is very dependent on the brand of printer used and if you use OEM or generic replacement toner.  After seeing another method on a internet site, I switched over to the "Cold" method.  This uses a chemical solution to soften the toner, and then the image is pressed onto the blank circuit board.  I found that I can get much more consistent results using this method, than I could with the "Hot" method.  I have slightly changed this by passing the board and image through my laminator without first waiting for it to warm up.  This seems to give more even adhesion of the toner than just pressure alone.  I usually pass the board and image through the laminator from several directions to give even adhesion.  This works much better for larger and double sided boards.  I have a blog entry that describes my method and another later one that covers how I make double sided boards.

https://kv4qb.blogspot.com/2016/01/cold-toner-transfer-circuit-baords.html

After completing the layout for the LM380 audio amplifier and the product detector, I decided to combine them into a single board.  I found that I could open one of the files in expressPCB software, highlight the area I wanted to copy and copy to the clipbaord.  Then open the second file and paste the first into it.  Then I used the software to reposition components or 
add/delete traces as necessary. 


Last evening I also finished up 
the SMD version layout in two different size formats.
I have a SMD board ready to etch, and hopefully I will be able to get that done tomorrow.






Update 9/4/17
Well I etched the board, and of course I found an error I had caused when I rotated a section of the board around.  I  made some quick changes, and etched another one. Then built the AF amplifier half of the board, so I can test it without having to worry about any noise or issues introduced by the product detector circuit. Very pleased with the way it turned out.  Since I had coated the board with a thin coat of lacquer after I etched it, the only problem I had was trying to get a good picture without a lot of glare off of the board. 



I hope to get time tomorrow to do some testing on the amplifier and check what the frequency response is, and how much gain I get out of it.


Thursday, August 24, 2017

N6QW Simpleceiver revisited updated 8/26

Over a year ago I was working on a version of Pete N6QW's Simpleceiver and my version of a SMD Bitx.  Then Farhan came out with his amazing Bitx40 board.  I dropped my original projects and went to work on the Bitx40 with a homebrew SI5351 VFO.
Now Pete has returned with the Simpleceiver Plus and has plans of eventually converting it to a transceiver.  From the response Pete has received, it looks like there is quite a bit of interest in this project.

http://n6qw.blogspot.com/2017/08/a-new-line-of-transceivers-difx_19.html

I think I will revisit the Simpleceiver and restart on what I had been doing.  Pete is doing this with simple building blocks that can be reused for several functions in his design, and also other projects.
Taking the product detector and audio amplifier circuit published on his blog, I started doing simple circuit boards for each of these.
Because the boards can be combined into a larger assembly, I am using expressPCB software to design the boards.  They will be laid out for etching using the toner transfer method.  I also decided to do two versions of each board, one using leaded components and another for size 1206 SMD .  Checking my available parts, some of the parts used will be have to be leaded, so I have made provisions for using them in both versions.


The first circuit Pete  published was for a dual J310 product detector, this basic circuit will be used several times in the final system.  The leaded version is laid out similar to K7QO's "Muppet" style with fairly large pads for mounting components.  I also placed large ground pads to help align components, and to make soldering to the ground plane a little easier. The SMD version is laid out larger than it could be, but I wanted to make it easier for construction without a solder mask.

Pete has several different audio amplifier circuits, a LM386 version and a higher power LM380 circuit. I will probably go with the LM380 version, but will build both and test to see which one I like best.  I have the LM380 version layout finished, and hope to get the LM 386 version done in a day or two. 

8/26/17
Finished the LM386 version of the audio amplifier board and generated toner transfer image files for both versions.   I am putting all these files in a dropbox folder at

https://www.dropbox.com/sh/4ne31pdib5o6vbr/AABDnxczmmhOQvlfB156GvWYa?dl=0

 


In the next couple of days I plan on etching and building up some boards.  Then I can test them and see which one I want to use for the final project.










Wednesday, August 2, 2017

New 3D Printer Project

|My poor little 3D printer has been busy for the last week or so printing parts for its big brother.  The AM8 design is based on a bunch of mods to the Anat A8 copy of the Prusa I3.  There have been several different versions of parts created to build the AM8 by different people.  Just looking at the files, I was not able to determine which would work the best for my build.  I ended up printing several different versions of several parts and then will decide which version I wanted to use for my project.  As I stated before, the original AM8 design used 2040 Aluminum extrusion for its frame. I did not have any of that profile, but did have some 2020 and decided to go with that.  Because I did not want to redesign a lot of the parts, I will sister two pieces of 2020 to get the desired profile.  Unfortunately I did not have enough Aluminum extrustion, so I printed some 2020 profile out of ABS.
  
After 4 days of 18 hour prints, I had enough 2020 profile finished to build the pieces I need.  After Acetone welding them together, I was pleased with how they looked and also how strong they were.
After cutting them and the Aluminum extrusion to length, I started the assembly.  The combination of Aluminum and ABS profiles went together very easily, and fit with the other printed parts very well.  After trying some of the different versions of parts I printed, I came up with what will work best for my project.  Because of some of these choices, I had to design and print several additional parts.  


After assembling the frame and gantry, I am very pleased with how it looks and how strong it is.  The size is just right to sit on the file cabinet next to my computer desk.
Now I am waiting on several other parts to arrive to complete the assembly of the X axis and Extruder carriage.  After that I can start on the motors, and electronics.  So while I am waiting on them, I can work on the boards for the SNA Jr 2.8 that came in last week.

Tuesday, July 18, 2017

Building a new 3D Printer update 7/20/17

While I am waiting on the boards for the SNA 2.8 I decided I want to build a larger 3D printer.  When I designed the case for the new SNA, I was limited by the size of my 3D printer.  As I mentioned in one of my earlier posts, the main problem with getting a 3D printer, is that I now have another new hobby.  The largest things I can build with my printer is abut 6" x 6", and I would like to go up another coupe inches.  I looked around and found a clone of the Prusa I3 called the Anet A8 that would be the right size.  There are several kits on EBay that can be bought for under $175 including shipping.  I checked several of the user groups on Google and Facebook to see what users thought of these kits.  It looks like they are  basic machines, but with some  work and modifications they can perform quite well.  One of the 3D printing sites I frequent is Tom3d.org, he recently did several You Tube videos on building a I3 clone from scratch.  After watching them I decided to build my own instead of getting a kit.  Looking at some of these kits and mods, I found a mod called the AM8. This uses many of the Anet A8 parts, but replaces the acrylic frame with one made from 4020 aluminum extrusion.  I have many of the parts, and some 2020 extrusion that should work.  I was able to download the  files for all printed parts and several mods from thingiverse.com , now to print the parts.

Some of the parts can be printed in PLA, but most need to withstand more heat and need to be printed in ABS or PTEG.  My printer works well with PLA but I have had problems working with ABS.  Originally this printer only used cartridges, but a recent software hack of the firmware enables the use of bulk filament.  I modified several used cartridges to feed filament directly from bulk reels, and tried several things I had found on-line to help when printing ABS.  My printer does not have a heated bed, and that was the main limiting factor for using ABS.  I added a thin PEI sheet to the build plate, and found that if I put several heavy coats of hair spray on the bed I could get the print to stick.  
I also designed some simple 20 x 1 mm. discs that when added near the corners of the object being printed, helped hold the part down to the bed.   

Without the heated bed, there was still some curling of the part when it cools.  From what I found on-line it looks like it might help to increase the temperature in the build area.  I found that by simply placing a plastic trash bag over the printer, the temperature was increased enough to prevent most of the curling, and also seems to help with layer adhesion.

Now that the printer is working better, I am in the process of printing the parts I need for a  printer based onthe Anet A8 and several mods I want to incorporate in my design. 
I have finished some of the larger pieces, and find very little curling, and what is there is not on any of the mounting surfaces. The pieces are very solid, and show no signs of layer separation. Wonders what a trash bag can do.

Update 7/20/17


Looking at the instructions for the modified printer now being called the AM8, the aluminum extrusion being used is 20 x 40 size. I have some 20 x 20 extrusion so will be using that instead. I do not want to have to re-design all of the parts, so I will mount two pieces together to get the profile I need.  I do not have enough aluminum extrusion, so I looked on thingiverse.com and found some files for printing some of the desired  profile.  My printer only has a print height of 150 mm. so I have to print in sections.  I printed several sections along with some extension support  pieces, and Acetone welded them together.  The resulting piece was stronger than I had thought it would be.  I guess now I just have to print up a few more batches and put them together.  Then I can cut them to length and mount to the Aluminum extrusion.

Friday, July 14, 2017

Lets call it SNA Jr version 2.8 for now

I have been trying to get a 3.2" TFT 8 bit parallel interface display working with the STM-32 board, but have had several problems.  So I decided to temporarily go with the Nano version, and a 2.8"  display with a SPI interface.  I had much of this working previously, so that part should go along quickly.  The previous board had a lot of additional circuitry on board that had been giving me problems.  I did a simple layout with just the Nano , display with simple voltage divider level translation, Ad8307 log detector, and a Adafruit SI351 board.  I usually make my own boards, but saw an offer from a Chinese board house that I wanted to try.  The board layout was under the 10 x 10 cm. size they had on special, 10 boards for $5 plus about $14 for postage.  At less than $2 per board, I decided to place the order.   Generated a set of Gerber files, uploaded them to their web site and placed the order.  Three days later I got an e-mail notifying me that the boards had been completed and were being mailed that day.  It has been about a week now, so expect them to show up some time late this week.  

While waiting on them, I decided to work on the packaging I wanted to use.  I had designed several 2 piece clam-shell type boxes that 3D printed quite well.  I changed the overall size ,and added cutouts for the display and joystick.  I was also able to design in mounting posts for these components, and added a bezel for the display.  It took a couple minor design changes and test prints to get everything exactly as I wanted it. 

It is still a lot easier than making a design on paper, transferring to  a box, then drilling and cutting out all the holes. Then you hope you have everything right, or you might have to start over with another box.  



I like being able to just do the design, send over to the printer, and in a few hours grab it off the printer sitting next to my desk.  The wife also likes the fact that I do not have metal filings all over the place anymore.