Per LA9DTA 11-band modification
Here's a nice modification by Per LA9DTA, whose Ultimate3 now has 11 low pass filters plugged in and can cycle through them all. The modification makes use of the Aux parameter in each of the mode-setting screens on the U3 (requires firmware v3.05 and above). The value on the Aux setting (may be 0 to F) appears on the LCD signals D4-D7. Note that the 4-bit value is reversed; so in order to control signal D4, you use the '8' bit of the Aux setting, i.e. use Aux=8 to set D4 on, or Aux=0 to set it Off.
The LCD signals D4-D7 are of course used for writing to the display, and also some are used for communication with the DDS module. During these brief interruptions the relay needs to be held and this is accomplished in Per's circuit by the 10K and 10uF capacitor combination at the transistor base (see circuit diagram below).
So: the relay-switched LPF kit comes with 6 relays. 5 are installed on the relay board, one is installed on the main U3 board. So for an 11-band U3, two of the relay-switched LPF kits are needed. That means 1 relay is left over. The spare relay is controlled with the transistor circuit shown in the diagram. The LPF 10-way connectors are wired in parallel; a few mm must be cut off the tall 5 x 2 header of the second relay board, so it can be soldered in place on the component side connections of the first relay board (as Per's photos below show). The +5V connection to the relay boards must be cut and come from the 12'th relay in Per's circuit diagram. In use: use Aux 0 to select one relay board, and Aux 8 to select the other. Simple!
For the moment, Per uses a 4m wide-band vertical with special balun; SWR is less than 1.8. Other modificaitons Per made: integrated GPS into the same case; added 2x 470nF power supply decoupling at the LCD module' added a heatsink on the four BS170 MOSFETs; added 470uF and 2x 470nF on the 5V power to the main U3 board, and again 470uF and 2x 470nF to the 12V PA power; removed R12 (see DDS mod page) and connected 2 x 470nF at the 5V power signal on the DDS; drilled a small oval hole in the LPF board for fixing the antenna cable to the PCB with strips; built a small 12V and 5V regulator board; added a power diode at the power jack to prevent accidental reverse connection.
Michael DK7MF update
Several people had reported trouble with Per's 11-band modification described above, which doesn't quite work. Evidently two relays are switched on at any one time. There was an attempt to fix this in firmware in v3.08, by switching the switched-off relay I/O pins to high impedance rather than +5V. But apparently that did not work either.
Now Michael DK7MF has provided a solution, that involves cutting some tracks and fitting additional diodes.
"Hello Hans, your U3S works great and now I made my Ultimate3S 11-band! First I tried the modification by Per LA9DTA as described on your website (dk7mf-pic1.png), but always both relays on LPF”1” and LPF”2” were activated. Why? The relais-circuit worked all right, but to the other LPF trough the resistances of the relais-coils came about 3V to the not activated LPF and activated the relais. Finally I found a practicable solution; I tried it on my LPF-boards Rev.4 (dk7mf-pic2.png): 10 diodes 1N4148, five on each board, for decoupling the relais, here drawn in blue. I just cut the 5 connections from Band (1to 5) to the relais and soldered a diode in each connection on both boards. Now only 1 relais is activated and everything works well.
"It's quite easy: you cut ON THE SOLDERING SIDE the connection between BAND5 and the Anode of D5. Then you solder a 1N4148 as seen on dk7mf-pic3.png just over the cut or with a bit of wire between the contacts, again on the soldering side. This has to be repeated 4 more times for the other BANDx and Dx. That's all."
Peter DM2PE confirmed it works well
Peter writes: "Thank you Michael for this great workaround on top of the modification from Per LA9DTA. The days before you published this entry I worked quite some time on the issue that most always both relais got activated and tried various solutions. Yours is the best solution and I have implemented it now and can confirm it is working perfectly. I have also added an 11th diode on the U3 mainboard between the relay and PD7 / Band 0 of the MCU for complete decoupling."