Tecsun PL-380 Modifications
Article : Graham Maynard, UK
Email  : graham.maynard1@virgin.net
Web   : www.GMweb2.net

The Tecsun PL-380

WOW ! what a potentially capable miniature receiver we have here,
whilst to my ears - what a truly obnoxious sounding little radio it seemed !

Click Images to Zoom
Tecsun PL380 next to Sangean ATS803

The PL-380 might embody amazing reception flexibility via its digital signal processing circuitry, but out of the box its Longwave sensitivity is very poor, its Medium Wave sensitivity poor below about 1200kHz, both being aspects which render it inferior to most of the basic track tuned solid state pocket portables made during the last 45 years. Additionally it makes no attempt to offer comfortably balanced reproduction of original broadcast program material due to it beaming sharp, spitchy and sibilant sound (especially on FM) via a tiny internal loudspeaker, which for me, made this Tecsun quite *impossible* to listen to. I still have smaller 1960s AM radios which are both more sensitive and more acceptable in their own right; they might not have DSP, but they are more competent when it comes to general listening!

Whether a radio is analogue or digital there can never be any excuse for bad audio, and designers really should consider the nature of human hearing with its inate need for balanced low and high frequency sound pressure level comfort with respect to our listening norm, as is set by the mid range frequencies we must hear in order to comprehend communications. This means that when there is little bass output - as via a tiny loudspeaker - then the treble output must also be reduced in level, and not be left to propagate tweeter like with unnatural loudness via untamed small driver high frequency efficiency.

After reading the comparative review by Gary DeBock (USA) as published in Medium Wave News, other DX magazines, and on the Internet, I bought this receiver through *e-bay* with no problems via International Air-mail from Hong Kong;-

Link to Tecsun on Ebay

Gary went on to overcome AM insensitivity by mounting a 7.5" long ferrite rod antenna on a support externally affixed to the top of the radio. This antenna is two inches longer than the radio itself, thus making the set larger than my pockets. See-


However his comparative reviews with other small radios, did not fully express how poor the digital PL-380's daytime sensitivity is when compared to more normal sized AM portables, nor how atrocious the sound is should this receiver be used as a portable radio. I found myself needing to use the AM passband selector as a treble cut, but that simply muffles reproduction and degrades clarity. As a direct result not only did I find myself very disappointed with this purchase, but my set holds the record here for speed of removing the cabinet back on a brand new piece of equipment; ie. after only five minutes of listening time.


Three back screws allowed rear removal, and then I saw with my own eyes the miniscule size of ferrite antenna bar as commented on by Gary - it is only 3.5mm by 7.5mm in cross section, and 79mm in length, so no way could this be capable of providing any radio with adequate daytime MW sensitivity. Not only that, but the antenna coil turns are wound directly over the ferrite bar without any insulating former, as is normally the case with high impedance AM frequency overwinds in order to reduce transduced voltage field losses capacitively coupled between the wire at the live end of the coil and the bar surface itself.

Click to Zoom

I desoldered the Tecsun antenna coil; temporarily connected up a 3.5"x 3/8" diameter ferrite antenna/ coil from another radio; immediately noticed a significant improvement; and observed good varactor diode tracking across the entire MW band. Then by comparing the new coil inductance with that of the original I measured more than a factor of two difference between readings; the Tecsun coil being lower at circa 250uH. Thus I cannot do other than conclude that either someone at Tecsun, or Silicon Laboratories (the si7434 chip manufacturer), has not done their calculations correctly, because the antenna fitted during manufacture does not come near to optimising the si7434 IC circuit capabilities. Indeed, Gary deBock independently found his PL-380 performed best when its antenna coil was closer to 600uH, this value being outside the range specified by Silicon Laboratories!

With adequate coil inductance the PL-380 performed well across the entire MW band, but it was still poor at Longwave frequencies unless the antenna coil was moved to the centre of of the ferrite rod. There was almost 20dB sensitivity difference at 198kHz for that same antenna coil between the centre of the rod and the end position as was necessary for good MW tracking. Thus not only was the original ferrite bar itself too small and the overwind not sufficiently insulated, its inductance was too low for efficient MW transduction, and the LW tuning capacitance too low to be using the same antenna coil. Actually Silicon Laboratories specify a separate and greater inductance coil for LW tuning, so really the Tecsun PL-380 is not properly Longwave capable at all, especially when compared to what we have long taken for granted in this part of the World.

Now I was not seeking supercharged DXing capabilities from such a small set, just a half decent portable capable of receiving regional transmissions from around the British Isles during daylight. Given that the internal ferrite bar really did need to be replaced, I had to get my head round the necessity for some internal surgery in order to fit a larger one. Here the planning was more scary than implementation itself, for removing the original ferrite bar and cutting the plastic 'chassis' proved relatively easy.

The existing ferrite antenna appeared to be held in place by some plasticy impact adhesive. Without damaging the original antenna bar or its wire I was able to remove it by very slowly pushing the blade of a terminal screwdriver between it and its mounting bed at the loudspeaker end of the plastic 'chassis'. Over a period of about ten minutes the glue stretched enough to either give way or be undercut with miniature wire cutters.

Click to Zoom

Then the plastic chassis was nibble cut with the same miniature wire cutters until there was a long slot between where the radio back would be, and the internal metal LCD screen plus the light switch pcb. See photograph showing a slot opening large enough to hold 4" of 3/8" diameter ferrite rod with antenna coil.

Click to Zoom

The bottom of this slot was then lined with fine neoprene foam to keep the new antenna coil from pressing closely against the metal screen.

Next I cut down to 4" a longer length of 3/8" AM antenna rod (make a hacksaw blade notch, face this upwards at the edge of a table, then sharply tap the unwanted overhanging end downwards to snap it off) and fitted the MW antenna coil from a scrap transistor radio. The coil was from a damaged £5:00 Mikomi CT-2388, and this winding is not likely to be bettered by any separately obtained specialist product.

Click to Zoom

Actually, for those who do not have a ferrite rod either, the 3.5" long Mikomi rod and MW coil work fine, thus that radio represents an inexpensive transplant source. The Mikomi coil also has the higher inductance necessary for optimising PL-380 input sensitivity, whereas coils salvaged from older transistor radios might be less suitable due to these having been manufactured with lower inductance for use with higher value variable tuning capacitors.


When this coil was measured on a rod away from the radio, its inductance was between approx 500uH for end, and 750uH for mid positions. However as soon as located within the radio slot, the proximity of the metal LCD screen and loudspeaker frame reduced the antenna coil inductance. Thus there was a need to move the coil down the rod in order to re-establish track tuning alignment when compared to the coil's optimum position as found with the rod being held away from the radio. I loosely set the coil plus rod into the slot previously cut to house it, *with the earthly end of the antenna coil towards the strobe wiring at the light switch pcb*, and then moved the coil up and down the rod whilst watching the S meter in a mirror to establish an optimum position for the coil on the rod.

Unfortunately, not only was antenna sensitivity reduced when the ferrite was close to the internal metal screen and loudspeaker, but background noise levels were increased due to direct digital pick-up from on-board circuitry. When just two inches away from the radio pcb, that same 4" antenna is about 3dB better for both sensitivity and S/N on MW, so clearly Gary's externally mounted antenna represents a much better arrangement for those wishing to optimise stand-alone DX capabilities.


Overall I estimated the MW reception improvement due to antenna replacement as a useful average 10dB above the original, thus providing this little radio with respectable daytime sensitivity for MW listening, even though it remains poor on LW when compared to most other radios. The PL-380 is however good for indicating the presence of night-time split-channel DX reception, even if not being the best to resolve it, because resolution is also impaired by having the ferrite antenna so close to digital circuitry, as revealed by the on-board S/N-ratio meter. Nor is there any substitute for synchronous demodulation when it comes to weak signal AM reception.

The highest signal strength reading indicated here on the Tecsun meter is 63dB in the 10 to 15mV/m fields from MW locals and the 100kW regional transmitter, this appearing to be the maximum reading possible with any powerful signal. The best S/N ratio indicated is 25dB, though in reality it might be better and more like the typical 40+dB with solid state portables (whilst 60+dB has been measured with good tube receivers). However, unless signal strengths are good, completely steady, and not close to 100% modulated, the S/N reading can fall from 25dB to 15dB whilst listening to say a 10mV/m R5-Live broadcast, this with an accompanying low level digital mush just observable on accompanying audio. At night any fading or fluttering signals, or momentary adjacent channel splatter seemed to induce noise which destabilised and affected reproduction when compared to straight analogue or non DSP digital receivers, though this might have been due to digital feedback into the closely sited ferrite antenna. (See below.)

Click to Zoom

Another aspect relates to AGC speeds. This is not a dedicated communications type receiver,so whilst the speedy receiver gain degeneration is useful, I feel that gain recovery is too fast for AM listening of fast fading and fluttering carriers, such that dissonance from momentarily unmodulated sidebands could become seriously exacerbated by the radio's tweeter like reproduction, even when a narrow DSP passband is selected. This same 'mush' was apparent and hard to tolerate on SW where the receiver might additionally oscillate if the whip antenna was not fully extended. Strange'?'

Not being one to give up I determined it was time to take the back off again to see if anything could be done about the strident reproduction. I removed one screw beside the volume control and two in the battery holder ridge to release the plastic 'chassis' from the radio's front casing.

Click to Zoom

It then became possible to click the USB+phone socket insert away from the front case moulding, and lift that end of the chassis by about 2.5cm (1"). Then I see the loudspeaker, and 'Oh' what a tiny magnet, how shallow it is, and with a plastic cone as well!

Click to Zoom

Being 16 ohm made me think of inserting a choke in series with the loudspeaker voice coil, plus a capacitor connected directly across it in order to empirically implement a high cut filter. This worked, and sounded most acceptable with values of 1mH (Toko 8RB) and a sub-min 4.7uF electrolytic, but it made the digital noise worse on weak signal reception'?' I imagined the possibility of field coupling between the tiny choke and the ferrite antenna on MW, yet the SW oscillation was worse too'?'


Checking on the Internet I read that the CD1622CB audio IC has input pins of 1 and 9, these being positioned towards the centre of the radio. So it was 'out' with the L+C and 'in' with one 0.01uF (10nF) miniature capacitor between each input pin and the central common ground track.

Click to Zoom

This simple mod immediately made the PL-380 most acceptable to listen to.

The previously raised hiss/ noise levels during AM fades were no longer so objectionable, and !'YeeHa'! the SW oscillation was completely cured. Clearly there had been an original problem due to inadequate si7434 'line out' filtering between the on-board DACs and the separate audio IC, as if transient potentials superimposed on AF around the audio IC were coupling directly with the thin antenna wire running across the cabinet back to the telescopic antenna. I had already tried positioning this wire to the very top of the cabinet back prior to fitting capacitors to the pcb, but this had not cured the SW oscillation either - it took the capacitors to do that.

It is possible that younger ears than mine might not like the same degree of treble cut I prefer, thus fitting 0.0047uF (4.7nF) capacitors might be more appropriate, and ordering two of each value at outset might be worthwhile.

So now, with a new ferrite antenna fitted to the Tecsun PL-380 pocket portable it can hold its own against other radios, and I can enjoy its useful MW performance. It also has reasonable SW plus FM sensitivities for such a short telescopic whip, whilst those two easily added capacitors have made FM so much more acceptable to listen to, it being better still if placed standing with its left side end on to a book, frame or wall. Average modified Tecsun PL-380 reception is equivalent to being about one 'S' point below that on my Sangean ATS-803A, and whilst its DSP works very well, it is still not as good as synchronous demodulation on my modified but so much larger Sangean.

Mission accomplished!

Hi Tecsun, if you are reading this, do please check out the AM antenna specifications for this pocket radio, also its LW track tuning components, because something is very wrong here. The simple modifications described here so much improve this little receiver that you might have an even more popular product - if - it could come straight off your production line with them installed. A Mk2 version with slower AGC recovery would be worth paying more for.

Personally I would prefer ergonomic Up-Down tuning buttons to reduce the need for knob spinning after selecting a SW band (yes, I do realise spot memory frequencies could be stored). Also separate Memory Store and Recall buttons are easier instead of needing to knob spin through all the 'EasyTune' memories via the tuning knob carousel in a manner which then denies subsequent fine tuning without first needing to press yet another button.

Nor have I managed to figure out why my radio keeps flipping into channel change mode whilst I am already fine tuning to search for DX between channels.


The antenna and capacitor modifications described here make the Tecsun PL-380 into a most satisfactory and thus daily used MW, SW and FM band pocket portable.

Just a little additional effort has turned this little radio into a bargain buy, and I have become very pleased to own it.

Contact;- graham.maynard1@virgin.net. Website;- www.GMweb2.net