Every now and then I read reviews of Bluetooth speakers, and I’m baffled how a same device can have both eulogistic and vilifying reviews. I know that any other kind of technical product gets 5-star and 1-star reviews on online stores, and I also know that some of the enthusiastic reviews, as well as many of the deprecatory ones, come from people who don’t know very much about the respective matter–being it audio devices, smartphones, watches, and so on. Some people are too stupid to be able to use a technical product, and some others are expecting too much from a relatively cheap one. That’s why I often encounter difficulties in making up my mind about a gizmo. This, and the fact that I am a cheapskate.

But I wanted a Bluetooth speaker, to use it with both my smartphone and my laptop. And I was amazed on how poorly seemed to sound so many good-looking speakers in the range €20-50; cheaper devices also had the flaw of having a too short battery life: how could anyone use a Bluetooth speaker that dies after only 2-3 hours?

As I happened to be in a brick-and-mortar store which displayed a number of Bluetooth speakers, I inspected the displayed ones. To my surprise, among the very few devices whose batteries were not discharged was one featuring an old-school rotary volume knob!

Now, this is extremely rare. I’m sick of all that +/- crapola. So I gave it a quick try, by pairing my smartphone with this device… and I was impressed by the quality of the sound!

The device? PHILIPS (for audio electronics, that would be Gibson Innovations Limited) BT2600B/00 (black),  BT2600W/00 (white), BT2600R/00 (red), or BT2600A/00 (blue). It can connect to two devices simultaneously, it includes a microphone (so it can be used as a speakerphone), its 750 mAh lithium battery promises up to 8 hours, and its rated output is 3.5 W RMS (the web page says 4 W, but the PDF manual is more accurate).

Too good to be true for less than €29? It depends on your expectations. Some people complained that BT2600 is “worse than the previous model” BT2500, but the previous model only promised 5 hours of autonomy for the same size of the battery. In my case, I very much prefer the devices that “just work” (provided that they’re not iPhones), and this one “just worked.”

Of course they must have used some tricks. Of course it can’t sound like a 400 W (and 40 kg!) professional loudspeaker. Of course they probably boost somewhat the bass while cutting a bit the treble, but the resulting signature is very warm and pleasant. And it can be very loud, much louder than my needs! Maybe this is why the first time it seemed to last more than 8 hours at my listening levels (after the LED starts blinking red, it was still able to play some more 45 minutes). Too bad it really needs at least 2h30 to charge, because it wouldn’t suck more than 0.3 A (more like 0.27 A), no matter the charger.

And the previous version, BT2500, had a separate on/off switch, whereas with BT2600 you need to turn the knob down below the minimum level to turn it off. That makes it even more retro, but also more prone to wear. And yet… BT2600 has some improvements in the audio amplifier:

  • BT2500B: Rated Output Power 3 W RMS, Signal to Noise Ratio > 60 dBA
  • BT2600B: Rated Output Power 3.5 W RMS, Signal to Noise Ratio > 72 dBA

So I’m happy that such an insignificant gizmo can give me joy–as much as I can experience enjoyment, because I’m rather gloomy by habit.

This is not to say that this device doesn’t have a few design flaws–it does, as every single Philips product I’ve ever met had! And I’m thinking at the Philips portable radio receivers from the 1960s, when they were actually made by Philips, and in The Netherlands!

In the case of BT2600, it’s designed to be used face up, but I never keep a loudspeaker face up, unless I want it to collect all the dust that the law of gravity brings to it. Funny thing, while they have put a rubber pad on the back, to cushion the device while placed face up, in many pictures it’s shown this way:

The problem is that in this position it has no plastic feet or anything to recommend this position. It might not be stable enough when placed on a hard surface, and the paint on the metal grille might wear away on the contact surface.

If anything, the raised plastic border that surrounds the connectors suggests this position:


Face-up is just not that practical when it comes to adjusting the volume!

I am quite familiar with this kind of oversight from Philips. For instance, I have used for many year a Philips pocket radio AE1595/00 (now discontinued and replaced with the smaller and dumber AE1500 and AE1530). Not only it was extremely robust, but that Philips pocket radio was much better than more the expensive Sony equivalents of the time. For one, Philips AE1595 was based on the excellent one-chip receiver SONY CXA1619, with mono output on the speaker and stereo output in the headphones, while the equivalent SONY receiver was using a less performant TDA chip made by… Philips! Not to mention that SONY required mono headphones.

I can’t find an official picture, but it was that ugly:
The bad design choices included the relative instability while placed in the normal position, because of the dumb plastic feet that were not helping much, and which weren’t even perfectly aligned in their long and short parts (in the following picture, the silver parts and the black parts):

And this wasn’t all there was. If you go back to the first picture (the all-black radio), you’ll notice the unnecessary and ugly bottom-left part which seems to have pimples. Well, technically it wasn’t necessary, so it’s just a stupid design. But the worse thing about it that it was a black plastic covered with a shade of anthracite paint, whereas the rest of the black housing was… a different kind of black plastic covered with a reddish shade of anthracite! Why on earth couldn’t they leave the black plastic as is… beats me. Now imagine how both plastic parts get scratched in time only to reveal different shades of black-ish plastic and paint!

Otherwise, a great device. And I could tell of various design flaws of other Philips-branded devices, including another radio receiver, AE2160.

Notwithstanding the above facts, I always had a propensity towards Philips. One can always tell it’s designed by Philips just by the way it looks, even with the branding removed. Oh, and most SONY devices have worse design flaws with regards to ergonomics! (Read “human factors” if you’re an American.) Think of a device with asymmetrical plastic feet (longer in the front, because the bottom wasn’t quite horizontal): when the feet wear through usage, the device simply topples face down! (In the affordable price range, SONY managed to make design errors even in the electronics!)

On the other hand, it’s true that Philips (or the companies that are authorized to use this name: Koninklijke Philips Electronics N.V.; Gibson Innovations Limited; WOOX Innovations Netherlands B.V.; Philips Consumer Lifestyle B.V.; China Electronics Corporation; Philips Lighting Holding B.V.) are making products that are most often worse with each new redesign! (So maybe BT2500 were really better than BT2600!)

Let’s take my preferred earbuds that are not in-canal (I hate having things inserted in my ear canal): SHE3000. I’ve used for years the white-cum-blue SHE3000BL/10, now I’m using the black SHE3000BK/10, and I have an unopened blister with another BL set! (Other colors exist too, including an all-white SHE3000WT/10, but I couldn’t find any.)

Why am I so frantic about these €9.99 earbuds? Because I hate the in-canal types (although I also own something like SHE3590BK/10, only in a different color). Because these SHE3000 sound better than many €30 earbuds. Because they’re covered in a soft rubber that helps a lot. And because, believe it or not, the bass is good and clear.

What happens is that SHE3000 is mostly sold in Asia today. The successor, SHE3010BK/27 (and seven other colors) is also priced at €9.99 (MSRP), and they’re some of the crappiest headphones out there!

Here’s the technological regression:

  • SHE3000: speaker diameter 15 mm; diaphragm: Mylar; max. power: 50 mW; sensitivity: 115 dB; freq. response 12-22,000 Hz
  • SHE3010: speaker diameter 14.8 mm; diaphragm: PET; max. power: 25 mW; sensitivity: 107 dB; freq. response 9-22,000 Hz

The improved frequency response (the only improvement) is pure bullshit. Without a chart of the exact frequency response, it’s utterly irrelevant. Besides, every pair of headphones has a specific audio signature because of the different way it reacts to intermodulation, the different way it creates harmonics, the different way a powerful bass “suffocates” or distorts other frequencies, and so on (one could measure the response to a square wave, the impulse response, etc. etc.). What matters is that SHE3000 sound great (for the price), whereas SHE3010 are not so great. (There are purists–or idiots–who, while reviewing professional on-ear or over-ear headphones, insist that one model is more suited for a specific style of music, then a different model is better used to listen another style of music, etc. Ideally, the headphones should be neutral, linear in response, which never happens. But instead of looking for headphones that favor certain frequencies, one should rather compare the so-called audio signature, which indeed might not go with all kinds of music, but it’s what makes you choose one model over another.)

Bottom line: I’m happy with cheap devices from Philips such as BT2600 and SHE3000. I’ve seen much worse for much more money.

ADDENDUM: I discovered more things about BT2600 as compared to BT2500. First, a German guy complained in a 1-star review posted on Amazon.de and on Philips.de, that BT2600 works fine when connected to a smartphone, but when connected to his laptop (Asus Zenbook UX501), the Bluetooth mouse causes severe audio distortions (stutters and cuts). He claimed that the old BT2500 didn’t have this issue, and that he even tried a different Bluetooth mouse to no avail.

I was skeptical at first, but then I tried and, to my amazement, the same thing happens to me if I use a Bluetooth mouse instead of a normal wireless one! On the other hand, if the speaker is paired to the smartphone, the Bluetooth mouse doesn’t cause any interference. Somehow, I believe that the problem doesn’t reside with Philips.

The reviewer didn’t state that he tried again the old BT2500 with the current setup, but only that he only tried two different mice with BT2600. I’m pretty sure that the interference is caused by some stupid design error in the Bluetooth circuitry in some laptops, or maybe there’s a bug in Windows 10’s drivers. The audio is broken if and only if both the BT speaker and the BT mouse are connected to the same laptop, not to two different devices–e.g. the BT speaker to the smartphone and the BT mouse to the laptop, or the BT speaker to the laptop and the BT mouse to a tablet. There is no convincing evidence that BT2600 has a design flaw, unless a BT2500 can be proven faultless on his laptop (or mine).

I could find that BT2500 is using the FW3717-85 Bluetooth module, based on the IS1685S single-chip RF and baseband IC, but I couldn’t find anything about BT2600. Since I found the service manuals and the schematics for BT2500 and BT3500 (a stereo version of BT2500), I discovered some funny facts:

  • BT2500’s battery playback time is not 5h as in the official specs, but 8h (min. 6h) in the following conditions: at 1/8 of the rated power (3 W), so at 375 mW.
  • Its rated output audio power is given at 10% THD. The audio amp is TDA7491P.
  • BT3500’s battery playback time is 5h (min. 4h) in the following conditions: at 60% of the rated power, so at 2x3W.
  • This might sound surprising, but it has a battery larger than twice: 2200 mAh (3 times BT2500’s 750 mAh!), so the rated output is not 2x3W, but 2x5W (min. 2x4W) when the battery has 4.2 V, or 2x3W (min. 2x2W) when the battery has 3.7 V. The THD at -6dB from the rated power is 0.5% (max. 1%). The audio amp is NTP8250.
  • The TEST SOUND for the battery discharge is… Hotel California Live, and for other measurements (THD, etc.) Hotel California Live + Let It Will Be + Do It Well + Night + Superwoman on REPEAT ALL! (No other details are given. So this is the music the Chinese factories like?! This must sound like torture.)

What I can suspect is that BT2600’s “up to 8h of battery time” means “at 1/8 of the rated power”: 437 mW considering 3.5 W, or 0.5 W considering 4 W (as I couldn’t find a service manual, I can’t tell which value is right; maybe 4 W with the battery at 4.2 V, and 3.5 W with the battery at 3.7 V).

BT2500’s battery (nominally 750 mW, just like with BT2600) has a standard charge current of 250 mA and a max. charge current of 400 mA; the charging time is 3h (typ.) to 4h (max.).

More interesting is that I found two French tests: TEST / Philips BT2500B/00, la nomade ne minaude pas, and TEST / Enceinte ultra-portable Philips BT2600, toujours un bon rapport qualité/prix.

For BT2500, the reviewer says that the lower midrange (“les bas-médiums”) have the tendency to resonate at times, but the midrange (including the voice) and higher frequencies are clear, and the voice has no sibilance. This means there is no excess in the range called “presence” (4-6 kHz), as the sibilance occurs in the 5-8 kHz range. Also, lower midrange can mean 250-500 Hz, 200-600 Hz, or whatever is the definition someone has read in a book, but based on the frequency response chart given in the article, I’d say that the range 300-500 Hz is responsible for the occasional resonance of the case.

Reviewing the new BT2600, the same reviewer states that the “équilibre fréquentiel” is excellent and surprisingly so for a speaker of this size, but the treble is too accented, and the bass, although still clean,  is less impressive and it drops earlier than on BT2500. Explanation: “that’s logical, as the speaker is no more of 2 inch, but of 1.5 inch”.  I failed to notice that in the specs, but she’s right!

The distortion is said to be slighter higher at the rated power–4 W instead of 3 W (but I still stick to the 3.5 W stated in the user manual); a higher power with a smaller speaker!

There’s a frequency response chart here too; I can see that there’s a peak at about 11-12 kHz, and this was noticed by our sound engineer. Still, as I said, I don’t hear too much treble and like the overall frequency signature.

Let’s compare them:

Many people would say that BT2500 has a better frequency response, and technically they might be right. But I see more linearity in BT2600 for the range 200 Hz – 10 kHz.

The comparison is made difficult by the following queerness: while for BT2500 it makes sense to look for a -2.5 dB/+2.5 dB (why the fuck didn’t the fucking software label the chart -3 dB/+3 dB?) around 1 kHz (which is at 0 dB), for BT2600 there is a local peak at 1 kHz, so that almost everything is slightly below 0 dB, except for the frequencies in the range 8 kHz-16 kHz, which can go up to +3 dB!

But still:

  • BT2500 has the 6 dB range -3 dB/+3 dB roughly 230 Hz to 3 kHz
  • BT2600 has the 6 dB range -6 dB/0 dB roughly 180 Hz to 9 Khz, which is much better than BT2500’s response!

However…

  • BT2500 has an abrupt drop of 15 dB between about 200 Hz (at -5 dB) and 100 Hz (at -20 dB)
  • BT2600 has an abrupt drop of 15 dB between about 200 Hz (at -1 dB) and 120 Hz (at -16 dB), which is not a significant loss.

This being said, despite the reviewer having tremendous professional credentials–Conservatoire de Saint-Cloud (Solfège, Harpe); Université de Versailles Saint-Quentin-en-Yvelines (Licence en Mathématiques); SAE Institute Paris (Sound Engineer Diploma, Ingénierie du Son)–I still prefer BT2600’s frequency signature (the “équilibre fréquentiel”). Yes, I’ve heard myself the abrupt drop after 11-12 kHz, but if we have -3 dB at about 14 kHz, that’s fine, because my ear can’t hear anymore over 14 kHz!

This doesn’t mean BT2500 didn’t have a better design than BT2600. Mechanically, aesthetically and ergonomically–but I don’t like its frequency response. True, BT2500 has a better “presence” (4-6 kHz), which makes it suitable for listening interviews with Sean Connery (why, you might even need a spectrum up to 8 kHz to understand him!), but I’m using this speaker for music, and it sounds quite retro to my uneducated ears. Also, remember that the reviewer said it tends to resonate because of the peaks at 300-500 Hz, despite having a larger speaker.  What I’d like to find about BT2600 is whether they changed the BT chip or not, but in the 21st century there’s no way to find a schematic diagram if the manufacturer doesn’t want you to have it!