<HOME
<OH2GVB

Baofeng UV-5R notes

A radio transceiver for my bicycle

Modifying the charger and using the 2.5 mm charging port

Adding a 1750 Hz access tone for the speaker mic

Bicycle mount and antenna

Since these Chinese rigs are so cheap, and actually quite good, in fact really good for the price, I bought a Baofeng UV-5R to go on my bicycle. With a dedicated bicycle radio, I could adapt it to that use without sacrificing my other two handhelds. When the UV-5R arrived by mail, the Customs Declaration on the package had it listed as "Women Accessory". Ho hum...

Of my three cheap Chinese handhelds (this one, the Wouxun and the Baofeng UV-3R), this one is the nastiest—the decorative frills on it look and feel plasticky (though otherwise the build seems solid). It's trying too hard to look technical and robotic. The picture here doesn't do it justice. I think it looks tacky, and I'd feel self-conscious using it, say, on a bus. The Wouxun is more composed, not to mention the UV-3R, which is tiny enough to hide in your hand. But a bicycle equipped with an antenna and a radio transceiver is weird enough anyway, so I guess the UV-5R fits right in...

First I had to charge the radio's battery. I was too chicken to plug in the universal 100–240 V wall wart with US plug into a European power outlet... Although many el-cheapo power supplies are just fine, with the incorrect plug this would have been impractical in the long run anyway. So I used my lab power supply instead, set to 10 V. Rumour has it, a 13.8 V station supply cannot be used—apparently it will kill the charger, or at least the charger will shut down, unless it is modified first.

Once charged, the first thing I did was to set the RX/TX frequency limits. Instead of having to mess around with crappy Windows software under Wine, I was pleased to find that the current version of CHIRP allows setting these frequency limits! Like on the KG-UVD1P, you need to set 144–145 and 432–437 MHz limits in order to get the 144–146 and 432–438 MHz ranges allocated for amateur use in Finland. (Actually you get 144.000–145.975 and 432.000–437.975 MHz.) Being a bit paranoid, I first ensured all memories and VFOs were set to frequencies within these ranges. All went swimmingly well, despite CHIRP's warning that UV-5R support is still experimental.

Remember to check that menu item #32, "AL-MOD", is set to "SITE"! Accidentally pressing and holding the button right next to the PTT activates an annoying Alarm function, which may transmit its alarm noise on the air if set to anything else! When set to SITE, it only makes audio noise locally. Annoying, yes, but at least disabling alarm transmission doesn't require a hardware mod, like on the UV-3R.

The radio, with its modified speaker mic, was attached to a vibration isolating mount and thus mounted onto my bicycle's handlebars (see below). Velcro straps hold it securely and make for a workable quick-release as well. A twinlead J-pole antenna, attached to a fiberglass rod (from a children's bike safety flag) and covered with heat-shrink tubing, is attached to the front fork of the bike.

I ordered the big 3800 mAh battery to go on the UV-5R. This gives long operating time, plus it has a 2.5 mm charging jack (the standard battery doesn't). Thus it will be possible to charge the radio, without having to dismantle it from the bike mount first for dropping into the desktop charger. See below about the charger—the cheap "charger cables" available from eBay are dangerous and will explode the battery!

So what sucks on this radio? The dual watch feature. That's what. Sure, none of my El Cheapo HTs has true dual receive, but that's not my beef with the UV-5R. What sucks is how the dual watch feature is turned on and off. On my Wouxun, it's just a single press of the <TDR> key (or two keys if the keylock is on). On the UV-5R you have to go through the menu: <MENU> - <7> - <MENU> - <▲> - <MENU> - <EXIT>, what's that... six keypresses?? (Seven, if the keylock was on.) That's quite a performance, if you just want to switch off dual watch during a QSO... (The UV-3R is no better in that regard, but in that tiny, super cute, ultra miniature rig I was prepared to accept the inconvenience.) This shortcoming alone makes the Wouxun way better.

Also, unlike the Wouxun and the UV-3R, there's no way to make this radio display its battery voltage. (Actually, some software versions on this radio might have this feature, but mine does not. I'm told that holding down the "0" key displays the battery voltage.) However, just like the UV-3R, it has a fairly well behaved battery status icon with 0–3 bars. You can set the power-on display to read whatever you like, for example your call sign, but I'd actually prefer it to display the battery voltage, like on the other two HTs I have.

Reports abound on the Internet about low TX modulation on this radio. Many modifications exist to improve it. Audio reports I've received have not indicated any major problem, maybe just a little on the quiet side. This may, of course, vary from unit to unit, or it may depend on the revision of the radio. The version and whatnot info on my unit are BFB297 (holding down <3> when powering up) and 130903N (holding down <6>, also reported by CHIRP).

The Baofeng speaker mic apparently has slightly better modulation than the internal mic, but a slightly "tin can" quality. So when I modified it to add the 1750 Hz repeater access tone (see below), I filled up the empty space with open cell foam. The modification itself also improved the sound quality somewhat. Now the modulation is, in fact, a bit on the strong side. I'll be opening up the mic again to install a resistor attenuator some time soon. For the time being I'll try to talk more moderately and at a distance.

My UV-5R did not exhibit the display "oddities" observed by Gordon N5GOR on his website. Apparently these have been fixed since then.

Charger modification—Using the 2.5 mm charging port

The big 3800 mAh battery has a 2.5 mm DC jack on its side near the top. This is for charging the battery, and it is wired directly parallel to the normal + and charging contacts at the back of the battery. You can either drop the radio, or the battery alone, into the standard desktop charger that comes with the radio, or you can get a dedicated charger with a 2.5 mm connector. Such AC-powered chargers are available from eBay, and possibly they have a higher charging current than the desktop charger, as they are designed for the big battery.

However, if you search for "Baofeng charger 2.5mm", you mostly get car charger cables. DO NOT USE THESE TO CHARGE THE BATTERY! I bought one of those and looked inside—it is a straight cable, not a charger! It contains no charging electronics at all. Plug that into a car power outlet, and IT WILL EXPLODE THE BATTERY! (Well, the battery might contain a safety protection circuit to prevent it from exploding, but certainly the battery will not be correctly charged by the car cable.)

However, the coiled cable and molded 2.5 mm plug were nice, so I decided to use the cable but to connect it to the UV-5R's original desktop charger, which is a proper charger. This way I could charge the 3800 mAh battery through its 2.5 mm DC jack (to drop it into the desktop charger would require dismantling the radio from the bicycle mount), albeit with the desktop charger's standard 400 mA charging current.

So I dismantled the cigarette lighter plug from the cable, preserving its strain relief piece, and ground a suitable hole into the desktop charger's case. With a bit of hot glue, I got a nice snug fit.

With the cable in place, I simply parallelled its wires (the red and black wires coming from the right in the photo, here's an enlargement) with the charger's + and contacts. The T contact is not used by the charger—it may or may not do something inside the battery, but in the charger it's not even connected!

I also modified the charger to accept 13.8 V input voltage. I simply placed a 7810 voltage regulator (with 0.1 μF filter caps between input and GND, and between output and GND) on the charger's input. The charger draws relatively low current, so you might get away without a heat sink, but I bolted a piece of scrap aluminum onto it to be sure. There's plenty of room for the regulator and sink inside the charger chassis. In the photo you see it on the left (enlargement). I connected its input and ground (the red and black wires coming from the left in the photo) directly to the input DC jack. I milled a break into the pcb trace coming from the DC jack to the charger electronics (you can see the break underneath the brown wire in the photo), and soldered the regulator's output (the brown wire) to the other side of the break (to solder the wire onto the trace, you must first scratch away the green solder resist).

Now the desktop charger can be used from my 13.8 V station power supply, or it could be used from an automobile lighter jack as well. However, it will not work with its original 10 V wall wart any more, because the 7810 regulator wants an input voltage typically 2 V higher than its output, i.e. 12 volts. Higher input voltages, on the other hand, will only cause it to heat up more. At 13.8 V, the outer surface of the charger becomes perceptibly warm while charging. As the battery fills up and the charging current falls, the surface grows steadily less warm.

Also the 2.5 mm cable works as planned (although charging the big battery is admittedly slow—fully charging a depleted battery can take almost 11 hours—but what's the hurry), and especially with the strain relief piece still in place, it looks neat enough to be an original feature of the charger! :)

Just don't try to charge a battery in the drop-in slot and with the 2.5 mm cable at the same time. If the two batteries are initially at different voltages, harmfully huge currents may flow when they are first connected in parallel.

Speaker mic modification—Adding a 1750 Hz tone

More classy microphones have DTMF keypads and other frills on them, but the only thing I missed on Baofeng's cheap speaker mic was a 1750 Hz repeater access tone (or "tone burst"). When using a speaker mic for extra convenience, it's stupid if you have to reach for the radio itself just to get a repeater to open. And tone deaf as I am, I can't just whistle 1750 Hz at will... So I decided to add the desired function to my speaker mic. (The photo shows the modified mic. The original had an LED on its front face, the modified version has a pushbutton switch in its place.)

Interestingly, the Baofeng mic has a silicone seal between the two halves of its shell, and at the PTT switch. This looks like it's trying to be waterproof! But then the microphone hole, speaker grille and LED have no sealing whatsoever—so don't go swimming with it. Just so you know that any modifications you do won't actually degrade its resistance to rain.

You can find the complete (though maybe not completely accurate) schematic of the UV-5R on the Internet. Here is a simplified diagram of the microphone audio input stage, and the modification being considered. The mic line in the radio is biased to +3.3 V through a 2.9k resistor. I wanted to add switch S1 and some kind of "Black box" which would load the mic line with a sinusoidally varying current, thus superimposing a sinusoidal voltage signal onto the line (albeit with a small DC offset, but the DC blocking capacitor C1 takes care of that). Note that the pushbutton switch completely isolates the "Black box" from the circuit when it is not in use, so it has absolutely no effect on the microphone's performance or audio quality.
I opened up the speaker mic, and drew a schematic of what I found inside. The modifications are also shown: Firstly, I removed the LED which really serves no purpose here. This had the positive side-effect of slightly increasing modulation, and maybe improving it overall a tad. Also, I ended up installing switch S1 into its hole. Secondly, I installed the "Black box" (see below) 1750 Hz tone generator and wired up its pushbutton switch S1.

The precise contents of the "Black box" was the big question. There's almost 200 thousand ways to make an audio oscillator (174341 to be exact), but I needed one that would be reasonably stable, would produce a sine wave output, and could be powered by the radio's +3.3 V mic bias voltage. (I did not like the idea of adding a button cell battery inside the microphone just to power the oscillator, even though the battery's lifetime would be practically infinite in this use.) Others have used Twin-T oscillators for the same purpose, but I'm not entirely certain how stable they really are with temperature and voltage variations etc. At least my initial tests didn't look too promising, so I kept looking for another solution.

I began thinking about a PIC12F508 microcontroller. It has a suitable operating voltage range of 2.0–5.5 V, a relatively low operating current, and it has a precise and stable 4 MHz oscillator built in. It requires no additional components to work, and I happened to have some in shelf, and a programmer for them as well. With its output pins wired to ground through a set of resistors, as shown in this schematic, I could program it to load its supply line with a 5-bit approximation of a sine wave. This is the "black box" shown above, and the microcontroller's positive and negative supply pins (Vdd and Vss) are connected to the positive and negative mic lines of the radio (ring and sleeve connections on the 3.5 mm connector), with the pushbutton switch in between.

Capacitor C2 bypasses the PIC's power supply line (a good idea in general, and especially in an RF environment), and together with R1 in the above schematic (inside the radio), forms a low-pass filter. This further smooths out the sine wave. Also, by selecting a suitably sized capacitor, you can move the knee of the low-pass filter to reduce the tone amplitude, and thus adjust it to the same level as the radio's internal 1750 Hz tone.

So I made a prototype on solderless breadboard, calculated suitable time delays to produce as nice a sine wave as possible with the non-linear D/A-converter formed of available resistors, and started testing. Naturally I erased the PIC while programming it, thus losing the factory-set oscillator calibration value. :( So I tweaked the OSCCAL setting by trial and error, until I got as close to 1750 Hz as possible (I actually got to within 1 Hz). Here is the PIC firmware as gpasm source code and as a compiled hex-file. Since the oscillator calibration varies from device to device, you really should compile from source, and change the OSCCAL value until you get it right, like I did. Or first read the correct OSCCAL value from the PIC, before you erase the factory calibration, like I did...

Then I etched a tiny circuit board for the surface-mount PIC and the five 0805 size resistors and capacitor. Larger components are just fine too, there's plenty of room inside the microphone. (I actually used one smaller 0603-size resistor, because I didn't have an 0805 820k resistor.)

The tiny PCB fits nicely inside the speaker mic. It is hot glued to the microphone's own circuit board. The pushbutton switch is a normal microswitch with a suitably long shaft, so that the pushbutton just protrudes from the microphone. The hole originally occupied by the microphone's TX LED was drilled large enough to let the shaft through. The microswitch is also hot glued in place, and wired up to the mic's PCB and the added modification board.

When closing up the microphone, I filled up most of the empty space inside with open cell foam, hoping to improve its audio and decrease the "tin can" quality. This did not seem to do anything, however. I was also told my audio had a "dark" tone to is. I swapped the capacitor in the mic (I don't know its value) for a 1 nF one, big enough to help keep out RF but certainly small enough not to filter the audio at all. That also did nothing to improve the audio quality. So it's not broadcast quality, oh well. Maybe it's good enough for portable rag chewing.

The 1750 Hz access tone works nicely, as it should. There's no reason why a square wave access tone wouldn't work as well, so this solution is overly hi-fi... But the PIC does provide a stable time base, and since it can do 5-bit DDS, why settle for 1-bit...? :)

Bicycle mount and antenna

For mounting the radio onto the bicycle handlebars, I made a vibration isolating mount out of Finnfoam (Styrofoam). It has a hollowed out seat for the radio, where it is held in place by Velcro straps. The underside of the foam block is molded to sit nicely on the handlebars. Velcro straps also hold the mount in place. Holes are made to provide access to the charging jack on the high-capacity battery, and to enable adjustment of the volume knob.

The left-hand side of the mount has space for keeping the speaker mic, and I may also add a clip for a pad of paper and pen.

All the hollows and through holes in the foam mount were made using a makeshift hot wire. I used ordinary iron wire, mounted on the edge of a piece of perfboard (pre-drilled non-metallized circuit board), wired up to a low-voltage high-current transformer. I adjusted the heat with a variac, increasing voltage until the wire cut through the foam at a reasonable rate. (I have used the same kind of hot wire in shaping core material when building my 16-inch carbon fiber telescope.)
The Velcro straps are sold as computer or A/V supplies, and are intended for keeping bunches of cables neatly together. About half of a strap's length is of the "hooks" or "rough" gender, the other half is "loops" or "soft". In one strap I made two holes with a leather punch. The strap was attached where the radio's belt clip would go, with washers to reduce strain on the strap. This is one of the straps to hold the radio in place. The other strap simply goes around the big battery, at the bottom end of the radio.
Here the radio sits in its hollow on the foam mount. Two Velcro straps hold it in place perfectly firmly. The lower strap is visible across the battery, the upper strap is hidden under the radio. (Otherwise the upper strap would obscure the display or some of the keys.)

The radio is on the right-hand half of the mount block. On the left side there is space for hanging the speaker mic, and I might add a clip for a small pad of paper for a log and other notes.

A dual-band J-pole antenna (along the lines of a DBJ-1 or DBJ-2) is mounted on the front wheel fork of the bicycle. The J-pole is made of traditional 300-ohm TV twinlead and hook-up wire, with an RG-174 decoupling stub for 70 cm. The feedline is also RG-174, with a BNC connector. An SMA-female to BNC-female adapter sits permanently on the radio. The wire antenna is supported by the mast from a children's bicycle safety flag and covered in heat-shrink tubing. The fluorescent yellow mast is designed to attach at at the wheel axle (front or back). Here you also see the choke balun on the feedline, right at the bottom of the black heat-shrink. The choke makes three turns through eight FT50-43 ferrite beads. No part of the antenna is grounded to the bicycle frame.

I won't give any dimensions for the antenna, because (a) they depend on the building materials used, and (b) I forgot to write down the measurements while building it... :) Note that the antenna's resonant frequency will change when it is attached to the mast and covered by heat shrink, and it will change a bit more when the heat shrink is finally shrunk on the completed antenna—so tune accordingly!

Warning! Low bridges or tree branches may violently yank the antenna backwards, hitting you on the head! Unless you enjoy cuts, bruises or concussion, wear a bicycle helmet!

Here the UV-5R is fitted onto the handlebars of my bicycle. The speaker mic was temporarily installed with bits of Velcro, but it felt like it would fall of at any time. Now the speaker mic is held in place by its own clip. The radio itself is mounted solidly, and the velcro straps that hold the mount on the handlebars turned out to be perfectly strong as well. They serve as a quick-release mount, so the whole installation is easy to attach and remove.

Since the antenna is more efficient, and better placed, signals are much stronger with this set-up, compared to carrying an HT on the belt. Also the speaker mic is more convenient to use. And when resting in its mount, it points toward my head, instead of blasting loud RX audio at innocent passers-by. :)

I even added a bright red "OH2GVB" flag at the top end of the antenna...

References

Dale H. Chidester, Speaker-mike modifications to generate a 1750-Hz tone burst for European repeater operation, QST, 2/1997, 74–75.
Retrieved online at
http://techdoc.kvindesland.no/radio/audiforradio/spkrmic1750hz.pdf on 11/2013.

Edison Fong, The DBJ-1: A VHF-UHF dual-band J-pole, QST, 2/2003.
Retrieved online at http://www.arrl.org/files/file/Technology/tis/info/pdf/0302038.pdf on 07/2014.

Edison Fong, The DBJ-2: A portable VHF-UHF roll-up J-pole antenna for public service, QST, 3/2007.
Retrieved online at http://harriscountyares.org/training/KNW/KNW-123.pdf on 07/2014.


Antti J. Niskanen <uuki@iki.fi>