UPDATE 1/6/2009 – For K3 Owners
To use the MOASU with K3’s, a simple mod will need to be added. The MOASU was designed to work off a PC parallel port which is active high. A K3 uses open collector devices so it is active low. By rewiring the input to the PC847 optical coupler, the K3 will control the MOASU just fine. The one limitation is the K3 can only control 9 transverter bands so only control through 10GHz is available.
To make the mod, let’s first review how the PC847 works. When the LED’s are biased on and generate light, the output transistors conduct. I tied pins 2, 4, 6 and 8 low and the parallel port pulled pins 1, 3, 5 and 7 high through current limiting resistors R1-4. To begin the mod, remove the four 100 ohm resistors – R1, 2, 3 and 4 that connect pins 1, 3, 5 and 7 of the PC847 to the DB9 connector. Ideally we would remove the ground connections from pins 2, 4, 6 and 8 and connect them to the DB9 and on to the K3. However, since this is a 4 layer board and one of the inner layers is ground, there is no easy way to isolate the pins. I am going to come up with a daughter board that replaces the PC847 that does this with no mod but for now, carefully bend pins 2, 4, 6 and 8 out 90 degrees so they clear the socket. Carefully solder a piece of hookup wire to pin 2 of the PC847 and solder the other end to the hole vacated by R1 that goes to pin 2 of the DB9. Next do the same for pin 6 of the PC847 to the hole for R2 that went to the DB9 pin 3. Continue with pin 4 of the PC847 to R3 that connected to pin 4 of the DB9 and last pin 8 of the PC847 to R4 hole that goes to pin 5 of the DB9. In the open holes of R1-4 that went to pins 1, 3, 5 and 7 of the PC847, solder one end of a 2.2K resistor in each. The other ends of all four resistors are tied together and connected to 12v.
I will get some pictures up on the website to show these mods ASAP. While this is an easy mod to do, if you are uneasy doing it, return your MOASU to me and I WILL PERFORM THE MOD AT NO COST, you just cover the shipping both ways!
UPDATE 12/1/2008 – New order for parts being placed!
There has been enough interest in a second parts buy. The total is still the same, $140 for a board and a complete kit of parts less cabinet. Email me at w8zn@comcast.net if interested.
12/1/2008 – Using the MOASU with the new Elecraft K3.
The BCD code that the new Elecraft K3 outputs from the high density DB15 plug is compatable with the MOASU. Since the K3 has 6m native, the first code for transverter #1 corresponds to the 2m position on the MOASU and the 9th transverter position, the highest a K3 can do, is 10 GHz.
This page contains information about the Mother Of All Switching Units
As seen in QST – MOASU
The World Above 50 MHz column in the April 2008 edition of QST magazine features a description of the Mother of All Switching Units (MOASU). The ARRL.org link to the MOASU schematic cited in the column has been delayed and the information is being temporarily posted here on the K8GP website. Please follow this link below to retrieve the schematic. Remember that the schematic is repetitive and only the connections to one of the transverters are shown in the schematic. The others are identical for each additional band.
July 10th update: All parts for kits and PC boards have arrived and are being sorted and added to individual kits. I should start shipping in the next couple of days.
Mother Of All Switching Units Schematic
Station Automation
Most of us who operate the VHF and above bands have built our stations over time. Adding a band or amplifier or preamp as the need arises or the money is available. Whether you are a DXer, contester, or a casual operator, station efficiency and ergonomics play an important role in how effective your station is. Some band openings only last a few minutes and being able to QSY quickly can make the difference between working someone or missing the “big†one. Contesters are always looking for ways to improve their rates. Over the years, I’ve had many opportunities to see how others have laid out their stations. They range from a Yaesu FT736 where one button changes from 50 MHz to 432 MHz and all of the brick amps use RF sensing to station layouts that resemble the wizard behind the curtain in the Wizard of Oz where a band change should be an Olympic event! Then, there was my dear father’s method of band changing, unplug this BNC, move it to here, move this BNC over there and 5 minutes later he would show up on the wrong band! My preference is being able to run the bands with a station as fast as possible. When N4UK (now W0MZ) lived in South Carolina, we ran from 50 MHz to 2304 MHz in just over a minute and we were several hundred miles apart.
Today’s radios and most all computer logging programs offer the capability to control external equipment. N3FTI had an excellent band data decoder used by many of the Pack Rats. There are commercially available band data decoders available from Top Ten and W9XT to name a few. Older rigs like the FT736/726 had individual PTT lines per band. Most newer radios have “band data†outputs ranging from a step voltage output to BCD (binary coded decimal) outputs. Anyone that is lucky enough to have a Elecraft K3 has many transverter functions built it. The K3 can control 10 external transverters and since it has 6m native to the rig, it would allow 50 MHz through 24 GHz very easily. I use Roverlog for contest logging and it can output BCD through the parallel port on your computer. Add to this, many rotors now have RS232 interfaces and with a program like Roverlog, imagine working a station on 6 or 2 and just from entering his 6 digit grid, the sharper UHF and microwave antennas automatically turning to the correct beam heading. All this without ever taking your fingers off the keyboard!
Given all this available technology, how does one put it all together? There are several devices available that will decode BCD and allow you to control external devices. What I was looking for was a integrated approach to controlling all my transverters, amplifiers, remote preamps and only have to look at a single wattmeter (at least through 2304). Through the wonders of eBay®, you can find just about every part you need to build the ultimate band switching interface. As with any project, you first need to define what it is you are going to accomplish. This interface had to be able to control 11 bands of transverters, 50 MHz through 24 GHz, have full sequencing to protect relays and preamps, deal with the fact that 432 and below were split transmit and receive 28MHz I.F.s and 903 and up were combined 144 MHz I.F.s, and switch different Bird® line sections in for different bands. Most importantly, it must perform these functions automatically from a single program and have a manual override when the computer crashes! I call the resulting unit the Mother of All Switching Units (MOASU).
Circuit Description
I began with a CMOS 4 line to 16 line decoder, a CD4514. Most other band decoders I have seen use CD4028’s which are 4 line to decimal decoders and extra circuitry is needed for 24 GHz control. The extra decoded lines also allow for expansion to 47 GHz. The active high outputs from the CD4514 drive NPN switch transistors through LED’s which control the matrix relays. For protection the BCD inputs are isolated from the computer via opto-isolators as most other designs do. To accomplish everything I set out to do would require a lot of bread boarding and lots of hand wiring, so I decided to do a PC board layout and get a custom board made by ExpressPCB. This would allow me to have separate PC mounted triple RCA jacks for various PTT outputs. I found some nice units from DigiKey that are colored red, white and yellow, obviously for some audio/video application. I also incorporated separate IDC connectors for controlling I.F. relays. Because of the high density of control lines, the PC board needed to be a four layer board using the inner two layers as ground and +28 VDC. While pricey in small quantities, I knew many rovers and multi-band VHF ops that would want these and purchasing most of the necessary items on eBay saved enough money so the PC boards didn’t kill the budget! In addition, I wasn’t looking forward to wiring 36 RCA jacks on the back panel.
Figure 1 – Block diagram shows a complete block diagram of the controller and its externally controlled items. The I.F. relays are connected to the main PCB via two14 pin IDC connectors with ribbon cables. Separate +12 VDC and +28 VDC lines are available on both of the IDC connectors for separate 50 MHz – 432 MHz and 903 MHz – 24 GHz relay control. An onboard LT1083 supplies plenty of current for several 12 volt relays. This allows a mix of both 12 V and 24 V relays in the system as long as isolation diodes are used.
The DIP relays for the switch matrix are purchased on eBay. They are 24 volt units but the standard Omron G5V-12’s used on the DEM sequencers can also be used by selecting the 12 volt relay source jumper instead of the 24 volt one. There are two DPDT relays per band in the matrix. Three sets of contacts switch ports 1, 2 and 3 from the sequencer and the fourth set of contacts sink the I.F. relays. Bands 903 MHz and higher also activate the 144 MHz transverter PTT and I.F. relay but not the power amp or remote preamp. As you change bands, a different set of DIP relays are selected routing the first three channels of the sequencer to the selected bands’ RCA connectors. An external 12 position, 2 pole rotary switch is used to manually change bands if the computer dies or you don’t have the computer running. The 12th position is for “remote†operation.
A male DB9 is the computer interface to the board. I could have used a DB25 to match the parallel interface on a computer but that takes more board space and drives up the cost. I made this by cutting the Centronics end off a cable and replacing it with a female DB9. I used a PC847X 4 channel optical coupler in a DIP package for isolation. The output goes to the CD4514b 4 line to 16 line converter to drive each leg of the matrix. Each decoded output goes through a LED and a current limiting resistor to drive the base of a NPN switch transistor. I had 2N3903’s available from my junk box but anything like 2N2222 or 2N3904’s will work just fine.
The LED’s make troubleshooting easier. Two DPDT DIP relays are used per band to switch three channels of the sequencer and the I.F. relays. Since the MOASU sinks voltage from the I.F. relays, any voltage relay can be used. If different voltage relays are used and sinked through the same contacts, you should install isolation diodes so the lower voltage relays do not provide a path for the higher voltage relays to key. There are two 14 pin IDC connectors at each end of the board, one for 50 MHz – 432 MHz I.F. relays and the second is for 903 MHz through 47 GHz I.F. relays. The third IDC connector is for a BCD rotary encoder to use if the computer dies. I have omitted a lot of minor construction details; most of us are familiar with standard building procedures. Some models I have built I’ve installed sockets for all the DIP relays, others I’ve soldered the relays in, builders choice.
Some of the extra features I added to the board allow the 2m transverter to be used as the I.F. transverter for 903 and above. There is circuitry added that will only key the transverters’ PTT and I.F. relay when 903 and above is selected, keeping the preamp and power amp connections for 2m inactive. There is an onboard 5A fuse and a place for a LT1083 7.5A regulator if you used 24v DIP relays. Since this is a 4 layer board, the “power plane†is connected to the onboard fuse. There is a separate 12V bus to run the logic; if you use 12v DIP relays, the LT1083 is omitted and a jumper is placed from the input to output pads.
Mounting the board in an enclosure can be a bit daunting, given that you need to drill 39 holes for the RCA jacks and a DB9 cutout, which is always fun. In one application, I just cut slots for the RCA’s to stick through, functional but not pretty. Connecting the MOASU to your station is fairly straight forward. RCA block J1 is the input from the sequencer. The next RCA block J2 is the 50 MHz. Outputs and J3 is 144 MHz. etc. The red jack is timing position one for preamps or antenna relays, the white jacks are timing position two for power amps and the yellow jacks are timing position three for transverters. Remember, timing position four is always the I.F. radio.
PC boards are very expensive in small quantities, about $48 for silk-screened ones. While this may be high, I gauged this outlay against my time to hand wire 39 RCA jacks and relays, no contest! If you are interested in a board, email me at k8isk@comcast.net and if we can get a group order together we can save everyone money. Or, I can send you the ExpressPCB file if you want to make your own or modify my design. All the components used were either junk box, eBay or procured from DigiKey.
Building the MOASU
First, the basics –
This entire project, which has ultimately taken a couple of years to complete, is designed to allow your computer running a logging program to control switching transverters, wattmeters, PLL’s, and just about anything else necessary to run a VHF+ station. I use Roverlog, which while not having the best of GUI interfaces, is by far the best when you compare all the features needed for VHF ops, whether single or multiops.
Most of us have a variety of switches in our station to change bands. And most of us have multiple wattmeters. The MOASU can control just about any 12V or 24/28V device that needs to change when you change bands. I have tried to make the design as universal as possible to allow for the unique properties of everyone’s stations. At my station, the MOSAU is now controlling all transverter I.F.’s, Bird line sections on 50 through 2304 and PLL’s for 1296 through 10G. Basically, all the MOASU does is takes a standard DEMI four channel sequencer and move it from band to band and also provide a ground path for external relays. Many high quality SMA multi-position relays are available on eBay for reasonable prices. Most if not all are 28V units from some military project. These relays work very well for switching 10M or 2M I.F.s between transverters. The MOSAU provides separate connectors for 50 through 432 transverters which commonly use 10M I.F.s and a connector for 903 through 47G which typically use 2M I.F.s. These connectors have both 12 and 28V available to operate whatever relays you have. Cables with RCA connectors are used for transverter, amplifier and remote preamp switching. If you are going to use a combination of 12 and 28V switches, you’ll need to add a diode in each leg of both relays in parallel so that one relay doesn’t provide a ground path for the other and lock it into an “on†state. Since you wire these relays with either 12 or 24/28V to the common pin, the diodes are placed with the anode towards the coil and the cathodes towards the MOASU.
In the original presentation, I discussed the ability to use either 28 or 12V relays for the control. While this is possible, since 90% of the multi position relays are 28V, it just makes sense to use 28v units on the board. If you have all 12v coax relays, you can use
G5V-2-H-DC12’s and just supply 12V to the fuse instead of 28V and install a jumper in place of the LT1083-12. The parts kit will have G5V-2-H-DC24’s in it. If you want 12v units, let me know and I will swap with you as I have plenty of the 12v units on hand.
One of the other options is the ability to use the 2m transverter as the I.F. transverter for 903 and above. 1N4004 diodes are placed at each band relay position you want the 2M transverter to provide the I.F. for. So if you are using 903 or 1296 transverters with a 28MHz I.F., omit these two diodes. Or, if you use a completely separate I.F. radio, omit the diodes all together. It is up to the individual builder as to how to handle reducing the power out from their 2M transverter to drive the microwave transverters. Some transverters have options for high drive, such as DEMI’s but DBNT’s will only take a watt or two.
Construction
I assume everyone tackling this project knows which end of a soldering iron to hold and knows basic construction and component layout. The only thing different about this board over most other PC boards is that it has 4 layers. There are the traditional top and bottom foil layers but the ground and +24/28V layers are sandwiched in the middle. The only precautions that need to be observed are if you have to remove a component: make sure it is desoldered completely. And if you have to drill a hole for any reason, remember there is a layer that has 24/28v on it.
Start assembly by soldering all the sockets to the PC board. The sockets for the relays will need several pins clipped off to fit the board. Use one of the DIP relays as your guide as to which pins to remove. Insert all the sockets in the PC board, take a piece of cardboard slightly larger than the MOASU and lay it on top of the sockets. Holding the cardboard tightly against the sockets, turn the board over on your work surface and solder all the pins.
After the sockets are complete, install the passive components, then the diodes, then the transistors. Last should be the triple RCA jacks. While the RCA jack assemblies snap onto the PC board, they may need to be “convinced†to stand up straight for soldering. It is a tight fit but with careful alignment, they all fit. Getting them straight will allow whatever enclosure you choose to put it in much easier to mark and drill. NOTE: When the boards arrived, I found the holes for the locking pins on the RCA jacks weren’t large enough so I have drilled them to size. The jacks will be a tight fit; you may need to wiggle them back and fourth to get them to seat properly.
After the board is complete, give it a once over to make sure transistors, diodes and LED’s are in properly, we don’t want to let the smoke out of them! If the LED’s are mounted incorrectly, a particular band won’t work since the LED’s are in series with the decoded signal from the 4 to 16 line converter to the base of the transistor controlling the relays. One possible modification to the original design would be to remote all the LED’s to the front panel so you could see which band is active. While writing this paper, I actually looked at the PC layout and thought about adding another connector to facilitate this but the board is just too busy. If you want to remote the LED’s, use 2 pin, .1 inch headers and 2 pin connector shells available from DigiKey. I added the LED’s to the PC board more to improve troubleshooting than for operator convenience.
Once the board is done, install all the relays and a fuse. Do not install the PC847X or the CD4514 yet. Apply 24~28V to the input and measure the 12V output to verify the LT1083 is working. No LED’s should be lit at this point except the 12V and 24/28V power good LED’s. Attach an alligator clip to a resistor and connect the other end of the clip lead to 12V. Take the lead of the resistor that has the clip lead on it and push it in to pin 11 of the CD4514 socket. The 50 MHz LED should light and the relays should snap in. Check the other bands using the attached chart.
Band CD4514 pin
50 —- 11
144 —- 9
222 —- 10
432 —- 8
903 —- 7
1.2 —- 6
2.3 —- 5
3.4 —- 4
5.7 —- 18
10 —- 17
24 —- 20
47 —- 19
After the checks are complete, you can plug in the CD4514. If you plan on using an external BCD encoding switch, you can wire and test it now. I use a thumbwheel switch I salvaged from an old SCSI external drive unit. There are even rotary BCD encoding switches you could mount to the front panel and label. If using only computer control, a jumper must be placed between pin 3 and pin 11 of the BCD switch connector. This applies 12v to the outputs of the opto isolator PC847X. If both a manual switch and a computer are used, a Local/Remote toggle switch should be installed. 12V from pin 3 or 4 should go to the common of the SPDT switch. For local operation, apply 12V from the switch to the common of the BCD encoding switch. For remote operation, the remaining switch terminal should be connected to pin 11 of the header.
You will need to make an interface cable between your computer and the MOASU’s DB9. This cable will be a female DB9 to male DB25. I used an old printer cable, removed the Centronics end and soldered on a female DB9. Refer to your software manual to see which pins are used in the parallel port. Pin 1 of the DB9 is ground; pin 2 is the 1 or “A†bit. Pin 3 is 2 or the “B†bit; pin 4 is the 4 or “C†bit and pin 5 is the 8 or the “D†bit. A word of caution here. I have seen one computer, a Toshiba, that didn’t pull the parallel port high all the time. It would pulse it high and not latch it. This would cause all the relays to “chatter†as the port would be pulsed. All the Dell systems and clones I’ve tried work just fine.
INTERCONNECTIONS
Now that the board is done and tested, it’s time to interconnect it to the sequencer. Since everyone’s individual setup is different, yours may vary slightly but for the most part, it’s all done in the same way. The sequencer has four ports. The first port normally supplies voltage directly to a remote preamp or to your antenna relays if they are wired so that the receiver is not connected to an antenna until that band is selected. I use the yellow jacks for sequence one. Next is position two which is used to key amps; this is normally a ground on transmit connection. Use the next RCA connector up for that. The third is to key transverters which is usually also a ground on transmit connection. The fourth sequencer position, which is also a ground in transmit, should go directly to your I.F. radio bypassing the MOASU altogether. I’ve included a block diagram to illustrate a typical set up.
To switch wattmeter line sections, I use another SMA multi-port relay to select which sensor gets connected to the meter on the common port. This relay is in parallel with the I.F. relay. If you are using a JWM multi-frequency PLL that can cover 1296 through 10G, use another multi-port SMA relay wired in parallel with the appropriate I.F. relay to switch the RF and a switch matrix to program the correct PLL frequency. Since the MOSAU uses 2A relay contacts to switch the RF relays instead of solid state switching, it is essentially bullet proof and will easily handle 6 or 7 coils.
That’s about it. If you have any questions, comments or suggestions, please let me know at k8isk@comcast.net. We are also including an update section on the K8GP website where I will post updates, hints, kinks, ideas from other users about the MOASU. Remember, everyone’s station is unique. What may be a priority for you may be secondary to someone else or visa-versa!
73 and see you on the air.
Terry
To download the complete article from above, click this link.