Mark N7EKU has posted a nice clean schematic and a picture for the audio pop fix he has just completed on his uBITX V3. It works great and saves the ears a lot.
If you check the top part of the photo, you can see that I have done Allison’s fix for harmonics by re-arranging the filters a bit and moving the original relays to the bottom of the board. This worked great too and cleaned up harmonics a lot.
In considering adding a new LPF module to replace teh original on board LPF matrix, it is helpful to remove the audio switching from K3. This reduces RF signal routing on the uBITX board. Many have found the signal routing via K3 causes RF to get into the audio so this represents good practice.
It is suggested that Audio M1 and M2 are tied together at R70, and does not involve Tx/Rx switching them with K3.
Power is Tx/Rx switched to Q70 on v4 boards, and the audio pop modification does the same thing on the v3 board.
The procedure is:
Bridge the front pad to the adjacent track
Cut the adjacent track and short the relay sides of both tracks to ground.
With the release of the v4 board from HF Signals, we finally have a design for an audio pop fix that can be applied to the v3 µBITx board with few additional parts.
Mike ZL1AXG has completed this mod and can vouch for it have removed all unwanted pops (on both transitions from RX to TX and TX to RX). It doesn’t kill the CW sidetone. His application of the mod is described below.
The mod can be placed in the same position on the v3 board as in the v4 design, using a simple Dupont header as shown below, but there are other ways of achieving the same result.
This mod only involves 5 parts. It is now the uBITx.net recommended mod to fix the audio pop. The audio pop fix summary will be updated shortly. All other fixes are now effectively redundant because they are more complex.
NB – This is an extract from the circuit diagram on hfsignals.com. There is an error with the numbering of the leads on Q74. Check first, before wiring up!
Parts required are:
2n7000 MOSFET or similar (Q74)
1 µF ceramic capacitor (C79)
1N4148 or similar silicon signal diode (D14)
100K resistor (R78)
1K resistor (R70)
Complete the following steps:
Locate R70 (100 ohm) resistor on the right hand side of the board when looking from the front panel and remove this resistor.
Drill a small hole through the board roughly in line with the two solder pads for R70 in front of the relay 7/10 of an inch to the right.
Install a standard dupont female header with 7 sockets (spacing 0.1″ per pin) on to the board (see first photo above). The first two pins are bent over at 90 degrees and solder to the pads for R70. Pins 3-6 are removed. Cut them off underneath and then pull them out with a pair of pliers. Solder the final pin 7 underneath the board
Install parts on the plug in board as per the circuit diagram below, and use a male dupont pin to connect to the T-R line. Plug the board in to its socket, connect the T-R line, and you should have a nice quiet transition going to TX.
Allison KB1GMX has come up with an improved pop fix based on the one in the v4 board design, originally submitted to the BITX20 list by Joe VE1BWV.
Allison finally got annoyed enough by the pop to fix it.
Parts count 5:
2x 1n4148/914 diodes
10K resistor (any value from 10 to 100K really)
0.1 µF capacitor
Allison has added a second diode. Why? The TX line is relay switched and relays take milliseconds to physically move contacts. So the second diode to the T/R line from the Raduino is the fast acting “audio kill”. The second diode and parallel resistor is the hold until the relay returns to RX position.
Allison assures us that this fully mutes the rig with no pop, no thump in either transition (from RX to TX or TX to RX).
Michael VE3WMB points out that connecting to VOL-H will kill the sidetone output. He notes that Ashhar Farhan VU2ESE has his V4 pop circuit connected at M2(R70) with the value of R70 increased to 1K ohms in order to hear the CW sidetone.
Mike N6CMY enjoyed his first build of the µBITx so much he built another one! This one avoided all of the mistakes of the first build.
1. Choke in series with R86.
2. All emitter resistors in buffer, predriver and driver bypassed.
3. Feedback resistor in PA increased to 600 Ohms.
4. Output Xformer replace with 2:4 on BN43-202.
5. Adafruit audio compressor in mic line with chokes to eliminate RF feedback.
6. PA powered by buck boost at 13.8V the rest of the board by 12V battery.
30W on 80 and 40, 20W on 20. RV1 adjusted to reduce output to 15W.
1. ND6T AGC installed (underneath main board) at K3 powered by 5V regulator.
2. Additional stage of audio gain installed between Q70 and U1 to make up for loss due to AGC.
Plenty of audio!! and good AGC action.
3. VE1BWV audio pop mod (similar to the one standard on the new ubitx) installed underneath the main board.
4. To further enhance the pop suppression PIN 1 at K3 is grounded.
RESULTS: Sounds fine to me.
A red/green LED TX/RX LED is installed on front and a TX/RX relay to key an external PA is installed on the back.
Nick VK4PLN has been working on a new board that will give a few extra features to a stock uBitx and plugs into the audio loom socket.
Its an Audio board providing easy access to Audio I/O pins. (add in your own AGC board, External amplifier…) It includes an area for adding an SSM2167 Mic Pre-amp module (with filtering for feedback and shutdown on TX). It also includes the simple 4 component PTT POP fix. (BS170) and a switchable 200hz CW filter. (LM324) that Nick already produces as a board for purchase.
The board has a bonus “snap off” section with a 3.5MHZ BCI filter for the RX chain.
Here is a pre-view, NOTE this is a WORK-IN-PROGRESS.
G3EJS has used a PIC processor to sequence the mute of the RX and change to TX. The PIC responds to the PTT going low by muting the audio amp, and 100ms later, passing the PTT low state to the raduino. When the PTT is released, it stops the PTT to the raduino, then 100ms later un-mutes the AF amp. He then made a breadboard version, which resulted in a silent transition.
G3EJS then made some boards and finally integrated a number of mods onto a single circuit board. Check out his pop mod circuit below:
Glenn VK3PE has built up one of the 9 component anti thump circuits described by ND6T (but it is in fact a design by VA7AT ). He has yet to actually try it in his rig.
The PCB is about 26 x 10mm in size. Only difference to ND6T’s version is he placed the 10uF’s on the PCB also. ie remove from uBITX board and fit to this board. Otherwise it follows ND6T’s web page for installation. It is made from SMD 0805 parts.
John KK5VH has been working on understanding the audio pop problem for a while. He doesn’t have a fix yet, but he has identified that:
The relays have a max actuation time of 7ms.
K1 drives K3 via the TX voltage – that means you have a 7ms (max) time from when the TX turns on and K3 disconnects the audio.
Changing the power (turning RX/TX on and off) on these one transistor amplifiers will cause a large spike to be coupled via C63 and C51 to the audio IC before K3 can cut them off.
Both the amplifiers draw tiny amounts of current, Q6 about 1.3ma and Q70 about 2.2ma. They do not have enough current draw to bring their respective power sources down quickly. Secondly, they have 47µF capacitors that hold the power up within that circuit.
John has been simulating the circuit via LTspice with some results.
He increased C64 to 517µF by paralleling a 470 µF cap across it and changed C52 from 47µF to 0.1 µF. That made a timing difference that cured the turn on pop but left a gigantic pop on turning off the rig or moving from TX back to RX.
After all of this playing around he still don’t have a good hardware answer to the problem!
However, John suggests that if the Arduino Nano controlled K3, this could solve the problem using timing delays. A simple sequence would in moving from RX to TX, first turn on K3 (disconnecting the audio chain) then switch on K1 into TX mode. When finished with TX mode, hold K3 on for a number of milliseconds to all the RX circuits to stabilize before switching back to RX on K1. Hopefully, this would solve the problem. John welcomes comments!