New AGC system

Don ND6T has been tinkering with AGC mods, and has come up with a new solution that gives much high levels of RF compression.

The µBITX presents some challenges with RF AGC system design: There is no RF preamplifier to use as a voltage-controlled attenuator, it is broadband and includes no tuned circuits in the receive path, and the intermediate frequency amplifiers are not configured for variable gain.

If we try to use a PIN diode in the RF, the high insertion loss raises the receiver over-all noise figure appreciably and the driver circuitry begins to become complex. The diode array also necessitates a fair additional current drain.

Original solution

Don’s original solution, designed for the BITX40, was to use a 2N7002 MOSFET as a shunt from RF to ground. The advantages were the simple circuit, the low current drain, and no insertion loss. The disadvantage was the limited dynamic control range (about 20 dB ). This was primarily due to the finite series resistance of the MOSFET when it was driven to full conduction, about 5 ohms. You can see this project by clicking here.

Putting another FET in series

Numerous experiments revealed that adding another 2N7002 in series with the receive RF path made it possible to add another 20 dB (or so) dynamic range across the HF spectrum. This was controlled with the same control bias as the original but the configuration requires driving the source connection of the transistor with the control and biasing the gate at 2.5 volts so that, at idle, the series transistor is driven to full conduction. This control topography requires that the new series transistor to be DC blocked from the RF line.

Pre-requisite Mod

This project assumes that you have already installed the manual RF Gain Control modification since it requires that the trace between relay K1 and relay K3 be cut. I designed mine to simply be glued on the back of the control potentiometer and cut the single-sided un-etched printed circuit board stock to be about 20 mm by 13 mm. The thickness of the completed circuit is only 3 mm and so fits easily.


Detected audio from the receive audio pre-amplifier is sampled by bringing a twisted pair of wires from the volume control. Just attach the pair across the outside terminals of the control, attaching the wire from the “hot” terminal (which also attaches to “audio1” plug on the main board, pin 4) to the vacant side of C1 on the new board. The other wire attached to the “cold” terminal (which also attaches to pin 3 of the “audio1 plug) connects the the “GND” area of the new board.

This audio sample is then amplified by Q1 and fed through C2 to be rectified by diodes D1 and D2 and filtered by C4 to become the AGC control bias. Signals below 30 millivolts RMS leaves the circuit idle, with Q2 (the series element) biased to full conduction and only adds about 0.6 dB of loss. Q3 (the shunt element) is biased to cutoff and has no effect on the receive signal. Louder signals (S9 and above) create higher bias voltages until maximum is reached (about 1 volt RMS audio) at about 3 volts DC control bias. At this point attenuation is about 50 dB at 3.5 MHz, 34 dB at 30 MHz.

The time constant of C4 and R6 sets the “AGC release” rate, several seconds from full to idle. Charging time is fast, milliseconds, so that sudden strong signals are handled without discomfort. This ratio prevents oscillation and diminishes “pumping” on strong signals. For faster recovery rates, decrease the resistance value of R6. Values down to 100 K work satisfactorily but 1 megaohm worked the best for me.


A hobby knife was used to isolate the connection pads as shown in the sketches. The remaining copper is then lightly tinned. Resistance checks should then be made to insure that there are no copper scraps or solder bridges between the lands.

I often find it convenient to simply begin in the top right corner of a board and work my way down and left so that I avoid working over previously completed circuit areas. Just place a component, hold it down, and touch the soldering iron to the board near one of the component leads. The thin solder will make a weak attachment to the component (called “tacking”). The builder can then move to another connection on that part and properly solder it using a hand to hold the solder. Then the rest of the component is soldered properly and the value checked with metering to assure a good connection and that the part wasn’t damaged in the process.







A spot of fast-setting glue holds the board to the back of the manual RF gain control. Remove the coax center conductor from the potentiometer wiper and connect it to the junction of Q3 drain and C5. Run a jumper from the wiper lug of the gain control and connect to the vacant pad of C3. Leaving the manual RF gain control in place will give you extra control should you wish it. Very short jumpers keep things solid and have purer control. Power can be supplied from any nearby 5 volt source (like pin 3 up on the Raduino, usually green wire). It just needs to be from 4 to 5 volts and only draws less than 2 milliamps.

If you want to add a switch to shut off the AGC, do so at the audio input. I would suggest a single-pole double-throw switch with the connection to C1 at the center, ground to one side, and the other to the hot side of the volume control. If power is removed from this circuit Q2 will not be biased “on” and will reduce signal levels significantly.

A good, solid ground is most important. I recommend a ground lug on the potentiometer shaft or on any other nearby ground point. A poor RF ground will yield poor attenuation. 40 dB is a ratio of 1 to 10,000! Hundredths of an ohm count.

If you don’t use an RF gain control

If you don’t have room for a manual RF gain control or a switch there is no reason to worry. This circuit can be left active and placed near the RF path connector without a problem. You could easily build this as a module that simply plugs into the connector. The most critical junction would again be a nearby ground but a third pin, this time through the board to the ground plane, would help. There should be little need to operate without it. If you need to run any tests that need it disabled, just remove the plug to the RF receive path and replace it with a shorting plug across the two RF pins.

If you adverse to surface mount construction, there is little problem in building it with leaded components. Use 2N7000 or BS170 MOSFETs and a 2N3904 or 2N2222 type NPN transistor. Try to keep wiring short and compact in the RF attenuator portion but the rest is very non-critical. If you use an electrolytic or tantalum capacitor for C4, observe polarity.

Parts List

  • 1: NPN switching transistor (1B, 2N3904, 2N2222) [Q1]
  • 2: N-channel enhancement mode MOSFET (2N7002, 2N7000, BS170) [Q2,3]
  • 2: Silicon switching diodes (1N4148, 1N914) [D1,2]
  • 2: 0.01 uF, 6 volt or higher ceramic capacitor [C3,5]
  • 2: 0.1 uF, 5 volt or higher ceramic capacitor [C1,2]
  • 1: 10 uF, 6 volt or higher ceramic (preferred) electrolytic, or tantalum capacitor [C4]
  • 1: 1 Kohm resistor [R2]
  • 4: 100 Kohm resistor [R1,3,4,5 ]
  • 1: 1 Mohm resistor [R6} (or other value, depending upon desired AGC recovery rate)


Signals below 50 microvolts (S9) are unaffected. Above that level, the AGC begins to engage. Q2, the pass transistor, begins to turn off a bit while Q3, the shunt transistor, begins to turn on. Signals are attenuated to where the audio output is held to reasonable levels and the rest of the receiver does not experience distortion. If you are considering adding an S meter this circuit may be to advantage due to the amplifier stage. The control bias is available at the junction of D2,C4, Q2,R5 and R6. This is the spot that you would use for a metering source.


AGC makes for a much more pleasurable experience. Don seldom needs to reach for the RF gain control or the volume when an especially strong signal suddenly appears. During nets I can now wander about the shack and still hear all of the check ins. Round tables don’t require a hand on the controls. Things are getting easier.

Recommended by others

Tim AB0WR says, “I am using this new AGC circuit. It works well. You still get a slight pop when an extremely strong signal first comes on but it is very short and, to me at least, not annoying at all. You may want to lower the 1 M Ohm resistor to something smaller to decrease the pop i.e. decrease the attack time. The next time I open the case I’m going to try 500K.”

Jim W0EB says “Thanks ND6T for a very workable AGC circuit!”.

It is clear that this has become the new recommended AGC mod.


Pesky volume pot

Geoff, G8BMI is cross-fertilising ideas across hobbies…

He has been assembling a uBITX, and found that the volume control has a 5mm shaft rather than the usual 1/4 inch. This means that none of his spare knobs fit.   We’ve all been there.  Most of us threw the supplied pot into the junk box and moved on with another standard potentiometer.

However, a waking inspiration suggested a very short length of 1/4 inch OD copper pipe, as used in model steam engines, could be used to ‘sleeve down’ a 1/4 inch bore knob. And it does!.

Geoff turned down the copper tube in the lathe, using a file just to ease the fit [It could be mounted in a power drill chuck to get the same result].  It only needs a skim. His tubing was 0.210 inches bore, which is just over 5 mm (5.18)


Connecting your rig to a computer

Connecting your µBITx can be done in a number of ways.   Some people will want to guarantee that their computer is fully isolated from their rig and will make or buy interfaces (Signalink, EasyDigi, etc.).  Others are quite happy to wire up a cable for USB control and another for the audio in/out.    In reality, transformers and other forms of isolation don’t necessarily always work.  RF can get into cables or the computer directly or indirectly.

Gordon KX4Z provides some useful links for reading up on rig/computer connections:


Cheaper alternatives to signalink

With relay:

Dealing with RFI


Updated manual for KD8CEC firmware

Rod KM6SN has updated his manual for the KD8CEC firmware for the µBITx that is pretty popular because it requires no hardware mods.

The updates to the manual:

1. incorporate feedback received on the manual, with some new
operating tips

2. enhance the Memory Manager section a bit

3. improve readability.

As before, please send any feedback to the email address shown in the manual.   A new manual for KD8CEC firmware rev 1.07 is forthcoming, as is a manual on the Memory Manager.


TDA2822 now shipping with socket

The TDA2822 audio amplifier chip (now infamous because of the self-combusting WX versions) now comes installed in a socket for easy replacement from HF Signals (the µBITx manufacturer).

There are still regular reports of spectacular end points for the WX chips on the list.   Most are finding that TDA replacements work reliably with no need for a regulator to feed them with a lower voltage.  To be safe and secure, however, you can easily feed them with 5-9v via  a regulator.

In related news, some constructors are substituting the readily available NTE7155 chip which is 100% pin compatible with the TDA2822.


HF Signals catching up with backlog

HF Signals seems to be catching up with their huge backlog of µBITx orders.  The latest posts from members of the BITX20 List shows that orders up to 17 March are being shipped at present.  This suggests that the manufacturer is just a month behind on shipments now, compared to two months behind in March.  This will be good news for constructors eagerly awaiting their kits.  Note that shipping time is additional (around a week for DHL shipping worldwide and 2-5 weeks for IndiaPost depending on destination).


SSB Crystal filter response

Michael VE1LEB has been scratch-building a uBITx.   This is the first time he has built a crystal filter.  The photo below shows the response curve of the 12MHz SSB filter output by his PHSNA:

To select the crystals, Michael used the K8IQY crystal test fixture and chose the closest ones to 12MHz from a bag of 50 inexpensive computer crystals.

He is not sure whether his PHSNA is accurately calibrated or how much error there is in the frequency readings. However, the -20dBm passband is less than 2.5khz.   He received a new bag of 100 crystals, and was wondering if he should take the time to  

Rod KS6SM commented,

“In the ubitx, the crystal filter is driven/terminated in a 200 ohm impedance. It is likely your PHSNA is 50 ohms.  The passband ripple you are seeing can be severely influenced by driving/terminating impedances.

“Are you measuring the filter with the transformers at each end, or are you going direct to the crystal connections?

“It is normal for the passband to be below 12 MHz. On my ubitx, the BFO is at 11,997,117 Hz, so that will give you a sense of how far below 12 MHz the passband is.”

It turned out that Michael was measuring the filter ‘naked’, without the transformers.  He measured it again but through the transformers, and the result is  much “softer”:

Satish VU2SNK said, “In both the curves without and with transformer your filters loss appears to be around -7 db.   In my opinion this is bit too high -3db is just right up to -5db is acceptable according to the experts, what type of capacitors are used in this filter? The multi layer ceramic capacitors
are really of low loss. If you use the modern SMD ones in my opinion that will reduce the loss in the filter.”

Allison KB1GMX also suggested, “One warning when measuring narrow filter use a very slow sweep. At narrow resolutions if the sweep is too fast the curve will not match the actual. It is as much detector response time as its filter delays. When too fast the it will appear tilted, and when you slow down you will see more accurate result.

The filter insertion loss in the 4 to 7 db range sounds correct. Though I might be better if higher Q capacitors are used we are not talking a 3DB difference. Also dId you calibrate out the transformers first? That can add a DB of loss sometimes more.”

In responding to the suggestions from Allison and Satish,  Michael said, “I’m using chinesium SMD 1206 capacitors.

“I realize now that I had not accounted for loss in my cables, and perhaps I’m not appreciating fully the implications of my test equipment. I’m a dopey artist, not an engineer! When shorted, the cables out-of and into the PHSNA present a 3.5 dBm loss. Each transformer adds 1.9 dBm loss, so double that and add the loss in the cables will represent a total loss of 7.3 dBm. When I bias the PHSNA for the 7.3 dBm loss and attach the filter (including the transformers on each end), I get this result, which puts me well inside the 4-7 dBm loss that you mention.”


Baofeng Speaker/Mics and uBITx

There has been a bit of discussion on the BITX20 Groups IO list about Baofeng mics, mods required to work with the µBITx, and what to do about a mic jack. This article tries to sum up thoughts on using the Baofeng mic.


The Baofeng Speaker/Mic can be obtained very cheaply, either direct from China (e.g. on or, or or via third parties in your country.   Be warned that almost none of these will be a genuine Baofeng mic.  There are many different varieties of “knock-off”, and they are of very varying quality.   However, most are adequate for µBITx use.

The wiring in the microphone is not standardised (i.e the colour of the wires varies).  In some cases the wiring simply doesn’t work.  In this instance, you should throw away the Speaker-Mic as the wiring is non-repairable.   Buy several to safeguard against the odd one that has wiring issues.

The microphone element is so-so, you can replace the one in the Speaker/Mic with the element that comes with the kit as required.  You may also need to drill out the tiny hole to be a bit bigger to make a reasonable air passage to the element.   The speaker is not high quality and will not give much volume.   When using the speaker-mic in the µBITx, most of us don’t use the speaker at all.

Standard wiring will work on the BITx40, and the LED in the Speaker/Mic will even light up.  It won’t on the µBITx, because the PTT line works differently.  This is connected to +5v from an arduino line (messing up the bias on the microphone).

Disassemble the mic, and rewire (using a multimeter to test connections) so that the PTT switch is wired separately from the microphone element.  You can use a common ground return for the mic, speaker and PTT.  See the original circuit diagram for the speaker/mic here.

Panel jack mounting

A key question that those with a Baofeng Speaker/Mic, is whether you should retain the 3.5mm and 2.5mm plug, or cut it off.

John KG5WJQ observes that they sell a combined jack on Alibaba and the price is fine.  The problem is that it is a PCB mount jack so that can result in some difficulties with mounting the jack in a case.

Craig KM4YEC uses panel mount mono jacks, one 3.5″ and one 2.5″, sourced from Radio Shack in store stock (but they could probably be ordered online).  He says,  “If you turn these back to back, and butt them against each other, with the ears turned out, it is a perfect fit for spacing.”   

Craig uses only the PTT, and the Mic element in the enclosure.  He enlarges the hole in front of the element, removes the speaker, and makes sure the only two circuits are PTT and Mic.  He went as far as removing all the SMD components and unnecessary trace runs using exacto knife cuts.

Glenn VK3PE fitted standard 3.5 and 2.5mm jacks at the rear of his uBITX build. NOTE the 3.5mm jack needs to be insulated from the chassis to work as the PTT is connected to what is normally ‘ground’ on a stereo socket. Glenn used some plastic washers to insulate it.

The challenge with this approach is that the spacing between the two jacks needs to be reasonably precise.

There is nothing to stop you from cutting off the plug and wiring it directly into the circuit.  This is a cheap option (no plug and socket required), but is a bit inconvenient when it comes to moving the rig, as the microphone is permanently attached.

The other option is to remove the plug altogether and use different connectors.   Many constructors like to use a standard 4 pin mic jack or similar style 8 pin jack used by the big three Japanese amateur rig manufacturers as illustrated below:

Mike ZL1AXG uses Kenwood wiring on a standard 8 pin mic connector as shown below.

Is there a better choice of microphone?

Dave K8WPE felt his Baofeng speaker mike was of such poor quality that he went ahead and bought an almost identical microphone from Btech.  This is a QHM22, a much better product, for US$23.00.  The speaker is top notch and reports of his voice quality are also very good. So the Btech mike might be a better choice.


RTL-SDR is mated to uBITx

Ian KD8CEC is working on a new release of his CEC firmware for the uBITx that will be more modular in approach (picking up on John VK2ETA’s mods to Ian’s firmware), interface with a range of displays (I2C versions of the 1602 and 2004 displays as well as the display that comes with the Raduino).  More importantly, however, this new version will enable the µBITx front-end of the receiver to be connected to an RTL-SDR USB device.  The RTL-SDR device will draw on the  HF receive sensitivity of the µBITx  along with the front end bandpass filter and first IF roofing filter and associated IF stage amplifier (normally at 12MHz) to produce a quality SDR receiver (for just a few $s invested in the RTL-SDR tuner).

This version is available to download (1.071 Beta) from Ian’s website now.