Vintage Hi-Fi Info
All contents of this Website are Copyright. Original research, photos of our hifi & all unquoted text is ©2011-2022 by select45rpm. This is all published freely on the internet by us to further the scene, not to give any seller or forum 'expert' undeserved credibility. We Do Not Authorise any Copying, Republishing or Quoting, even as rewriting Our Research In Your Words, of using or linking to any of our Sections on ebay, any sales sites or anywhere else. No-one else has formed these opinions, so don't steal them as yours. Please do not link to our site on ebay sales trying to use our unique info to play buyers for a sale suggesting we are authorising their sale, as we are not.
| *See our NEW Hi-Fi Blog page
lots of New Sections since Jan 2017 that add a wide range of Hi-Fi & Tech related subjects plus opinion on Hi-Fi News 1970-1980 as we read through.
Includes New Articles on this page's subject. This page has been updated & read through with 2017 ideas & with an Index to navigate. (working on..)
Power Ratings are Misleading
Only Peak Clean Voltage Output Should Matter.
Transistor amplifiers use a strange way of working out wattage, beyond useless Peak Power ratings, see below. The majority of speakers used are 8 ohms though Car ones are as low as 1, 2 or 4 ohms to boost wattage ratings again, they play no louder. If things could be BS-hype free you'd be reading: Output 30v peak. Can be used on 4 or 8ohm speakers. Simple. You know exactly what you're getting then. It doesn't sound anywhere as Glamorous as 95w into 8 ohms, 160w into 4 ohms & 250w peak does it? Wattage is Current x Voltage & only really the Voltage matters, current mattered if driving insensitive speakers, but who actually does use such speakers?
Wow. A 300w amplifier. Not Really...?
Not so exciting & not worth having either. The 300w amplifier can deliver a high current to drive those awful low sensitivity 82db speakers. Most big speakers like Tannoy Gold 15"s are 95db. So your 300w amp like the Marantz 2600 is, was about £1200 new in 1979 though there are very few Speakers capable of handling 300w, it runs on ±85v HT. The 85w Pioneer SX-950 works off ±50v (output 37v = 74% of HT) and the 32w Rotel RA-610 works off ±29v (output 23v = 79% of HT) just to show smaller amps are more efficient as the table below reveals. Having pushed one amp beyond clean sine, 105w NAD 300 output 44v it could go to 53v before the relays clicked in to save it. It works on ±53.2v HT so theoretically you can go hard into distortion to get the full HT, but risk trashing the amp & speakers. So if the 300w Marantz 2500 has ±85v HT it should do about 70% of that as clean sine output meaning 59.5v. If the SX-950 at 85w does 37v, the Marantz 59.5v should mean 136w in terms of comparable volume. But it's rated 300w meaning it delivers more than twice the current of the SX-950. Power (Watts) = Current (Amps) x Voltage (Volts). The laws of Decibels means 85w to 136w isn't that much of a gain. 85w by an online calculator is 49.3dBm and 136w is 51.3dBm. A true 300w that this actually isn't would rate 54.77dBm. Extra current delivery doesn't mean it plays any louder. So on normal speakers, there is no need to have an amp over 100w as 85w to 136w, 51w extra only gives 2dB gain. A low sensitivity speaker of 82db compared to the Golds at 95db means changing speakers gives 13db gain. A 3dB difference is barely noticeable, though a critical ear can hear 0.5dB difference. An amp rated 300w is going to be designed more for Current delivery & be overdesigned, meaning Hifi quality will never be as good as the 40w-100w amps.
Power Ratings in the 1965/66 Hifi Yearbook
now get a clarification as things are getting confusing with the USA way of measurement. IHFM which is a Standards Institute of HiFi. Continuous Power means the Sine Wave Power the amp can deliver for 30 seconds without exceeding rated THD. Music Power Output means a tiny burst of a power that would trash the amp in a full second if not exceeding THD as it's so fast that is absolutely useless to know. The amp is not using it's own internal power supply, but a lab type power supply matching the voltages. It's total BS. Continuous Power at the amp's best power output spot frequency at rated THD in RMS. They quote P power = E squared / R resistance, where E= rms voltage across the load.
A Batch Of 45w Amplifier-receivers Reveal
We test all Transistor amps for their max clean 1kHz sinewave reading, we use a Scope & it reads Peak To Peak, a Multimeter would be different as it reads RMS which is 0.707 of Peak To Peak. One batch we needed to do ratings on at the same time are all 45w rated: the Sansui 3000A 45w (28v), Pioneer SX-1500TF 45w (31v) & the Sansui 4000 45w (23v) it is simply a current issue to get the 45w as low 23v and high 31v were read . These play a similar sort of volume, but only by playing them loud do you get into clipping. The Pioneer with 31v reading could cope well with loud bass notes & delivered them well but the Sansui 4000 with 23v reading clipped out & at volumes even taking the fact the amp is all-original, it just flattened out quite obviously when the Pioneer kept it clean & undistorted. The Clean Voltage Output certainly does matter more than the Wattage.
Do You Really Need More Than 45 Watts?
Assuming you are using efficient speakers or headphones, you will never really need more than 45w to power them even for high volumes in Domestic use. 100w may appear to cope better with transient peaks, but in reality we found a 100w amp doesn't really play much louder than an efficient 45w one. You read that "difficult loads" are the reasoning behind wanting 150w-300w amps, more like people foolishly buy these thinking the quality is better, but instead get overdesigned & over damped restricted sound unlike a 45w amp which has to be the optimal power rating for volume & still have a fresh musical quality. Like having a 4.5 litre car engine, the power is there, you know you have it & it massages your ego, but you'll rarely drive 200mph on any road as 70mph is the UK limit & your 200w amp will usually not put more than 30w on a loud session into efficient speakers. Both amp & car will be good for pro use, PA amp & Speedway, but you'll never be using it like that. And they will be expensive to run. Seeing quite a few of these sort of amps on ebay & they don't appear to sell shows owners are tired of them & buyers feel them too much in many ways. Both will be expensive to run. Flash Motors & Monster Power Amps are pure ego things. Are you really that shallow?
USA Hyped Ratings On Early Hifi Actually Became Illegal.
Yes a law was passed to stop dishonest Peak Power ratings being quoted to fool the unwary buyer. Peak Power is utterly worthless as it means the maximum power available for a fraction of a second, likely so totally distorted. Also if you ran your 20w RMS into 8 ohms at that Full Music Power it'd self destruct. USA amplifiers known to be 30-40w including some early Pioneer such as the SX-1000TW that we've had & review elsewhere, cheekily has stickers on saying they were "FET IC 150w" when it's a 50w RMS 8 ohm rated amp. USA often added wattage & peak power to fool buyers: 25w+25w = a 50w amp, right? NO.. Same amp delivers 55w peak, so 55w + 55w = 110w then? NO... All dishonest. Also you see 3000w Car amps.
Trio-Kenwood KR-4140 is an 18w receiver from 1970-71, the USA booklet cynically states 95w 'Power Output' meaning the 1 second peak power before destroying & also 33w or 24w per channel into 4 ohms or 8 ohms. But who buys a Stereo amp to use just one channel. The true 2-channel used rating is 21w into 4 ohms & 18w into 8 ohms. So much for 95w.
Beware of Wrong Additions
You also see foolish quotes by amateur ebay sellers of the Total Power that the unit will use at it's maximum as on the back sticker, a 95w amp may only be 20w RMS. A 580w one only 80w RMS & similar. The 580w more fairly 580VA, may be drawn by the amp at absolute full peak power for one second before exploding. Older adverts & even amplifiers added L+R wattage together, ie the Leak Delta 70 is only 35w RMS per channel, but it is 35w + 35w. Some amp makers like B+O used 4 ohm speaker ratings as their own speakers were 4 ohm, so not dishonest, but today watts into 8 ohm is the usual way so a Beomaster 4400 rated 75w is 75w into 4 ohm & only 50w into 8 ohm. Many amateur sellers quote the transformer VA (Watts) rating the amp can deliver just before destruction, the Yamaha CA-1010 is 100w but the back says 900w & the foolish seller was impressed at that.
We're saying nothing new here...
we later read that in December 1976 & April 1977, the one we read & where the quote below is from, in Hifi News the author was complaining about measuring amps in terms of heating power of resistive elements. The quote in Green below is from this article. He too used the peak voltages route. Also the Slew rate-rise time of an amp, a thing that certainly makes the amp sound "fast" & not blurring like slower ones do. Not a thing we can test beyond knowing how to listen to record crackles being served up separately from the music as only the best amps can do.
"It is surely high time that the stupidity of rating amplifiers in terms of heating watts was killed for once and all"
If you want more loudness you buy more efficient speakers. This is how a 4w 1920s valve type amp can sound very loud with 100db+ speakers, not to mention how loud an acoustic Horn loaded system can play is another thing, we heard a guy played a 1w signal played through a horn loaded highly efficient speaker could be heard miles away. Having had a Wartime Tannoy Siren type Signal speaker which was sort of horn loaded, this is no mere claim, ours worked a good wide range, though we only tried test tones through it, exceedingly loud it was.
Sine RMS Ratings with Different Watts Ratings.
The Sine Output doesn't always match the Wattage as the below shows. A 24v rated amp can be 30w or 40w. The 40w is an amp with better design with extra current available so can control speakers better . The 30w amp is on the weak side on design, current limiting is the term though that's a design thing & suggests it can upgrade to be a 40w amp, but that takes a lot to get that far & often isn't worth it. A 90w amp with 45v output is a weaker design than a 110w amp with 44v output.
Rating Amplifier Power as "dB" is Ridiculous.
Reading the Jan 1980 Hifi News, the very tedious Martin Colloms, still active in Hifi it seems, with his way of filling a page with text & after having read it you are none the wiser, if perhaps now comatose, rather foolishly in 1980 now decides to rate Amplifier Power as dB because it works out mathematically. Bless. This is foolish twice over for firstly confusing an issue that was only tidied by manufacturers finally having to put RMS not 'Peak Music Power' & then by pretentiously saying a 100w amp is 20dB/W. The facts in our table below show that Wattage can rely on higher or lower current & even amps rated low wattage can play very loud, surely making a mockery of the silly dB/W easily. One JVC amp we played as loud as we could take it on headphones with test tones & it didn't distort at all, yet it only rates about 15w. The very boring Sony TA-F550-ES rated 90w & the Marantz 1152DC rated 76w both played very low volume & could be clipped out far too easily. An amplifier is about Voltage Gain but the Master Volume of an amp relies on how the circuit is designed, the berst amps are not designed 'averagely' or with general purpose ICS either. That JVC 15w amp played good and loud but a 1971 Sony 16w receiver we could play full volume in full distortion & it still wasn't loud as the design was aimed at aged buyers surely. No way is it possible to justify dB/W & hopefully this was quietly forgotten as we don't remember it in our Hifi mag years. The basic idea of dB/W leads to this below...
Why Has No-One Tried This To Rate Amplifiers?
People buy bigger power amps as they want it louder, so why not get one good PA speaker system, to be sure it's going to survive, and then plug in an amp playing Test Tones at 50Hz, 1kHz & 12kHz to see how loud they go in the same given speaker just before distortion shows on the oscilloscope. This would remove any need for the crazy dB/W as noted above & the Heating Qualities of Watts would be dismissed too. This would help matching speakers to amps better, if generally domestic Hifi is not made as loud as it could be for domestic living reasons. This would truly rate the amp in dB gain, removing the variables of current that make a 32v clean sine amp in the table below rate at 48w or 60w, or a 25v sine rate at 34w or 50w.
THD-Harmonic Distortion: In Your Face.
To put it bluntly, the reliance on THD in maker's specs is utterly worthless BS. It means Nothing in terms of how the amp will sound. Harmonic Distortion is worked out using a Sine Wave & reading how it distorts on it's Harmonics, read elsewhere about good & bad harmonics. So the 1kHz test tone gets 0.01% THD, how fun for those who listen to test tones. Music is a multitude of frequencies all at once, it relies on the speed & reserve of the design to reproduce it. Square Wave distortion is another one quoted in reviews, slew rate & ringing. All amps rely on a design based on mix of a large quantity of transistors, resistors, capacitors & diodes. The designer may know how to fully bring the best out of a circuit, but then comes the brutal dumbing down so you don't get the Good Stuff as you'd never buy another amp. The only true test for Distortion is how the Source compares to the Output, it will usually be tamed down to not be so dynamic, deep bass will be reduced, treble can be blurry or smeary. With today's technology, to sample 5 seconds of music at the Source & at the Amplifier Output, scale them to the same visual size & see where it goes wrong. That peak got reduced by 10db, that transient didn't resolve correctly, the Sub Bass has gone missing. The "Distortion" could be of several Hundred Percent where it fails to resolve complex signals of Music, not easy Sine Waves. Go on Hifi "experts" put that idea into your sales hype & tell the truth how mangled your crappy amps are. Then stop dumbing them down so heavily like 120+ changes were needed to better, if not possibly perfect, the 2007 Marantz PM6002. Hifi should be much better than it is.
Now we are sure most think this will protect your speakers & transistors if a fault occurs. "No, Sir" is the clean version of the hard truth. All that a Protection circuit does on closer looking at the circuit, is sample the output signal through a resistor, so it doesn't affect the sound. Once the Triac or more modern Relay circuit detects "too much" DC offset it kicks in & silences the output or cuts power off to the Preamp on earlier amps to mute, in effect protecting the speaker from pure DC that would burn the voice coil out. It won't protect the amp from damaging if a fault has occured, only a fuse or fusible resistors will help limit the amplifier damage. Only the repair tech will nurse your frazzled amp back to health. Do valve amps have protection circuits? No, they have transformers that block DC but a signal played too loud will have much DC in the clipped waveform so you aren't safe either. Also the early Capacitor coupled amplifiers usually pre 1974 block DC with the capacitors but will have the similar issue with heavily clipped waveforms. Only with the Complimentary or Semi Complimentary designs do protection circuits become required. Live Band & guitar amps use valves, do you hear of speakers being wrecked? We read some people take out the Protection circits as they've been told by some amateur that it affects the sound. No it doesn't, read the circuit. Leave the circuit be, a future owner of the amp may not be pleased on finding there is no protection when the design should. Beware of amateur fiddling.
Many later amps use an excess of NFB in amp designs with way too many transistors & other excessive IC based circuitry. The more actual amplifying stages the louder the signal gets & it needs NFB to bring the level back down. The lowest Transistor counts found in an Amplifier to give a high quality sound are actually as few as Phono x2, Tone-Pre x2 & Power Amp x7. Some may add one extra in the Phono-Tone-Pre & a Power Amp with 10 transistors including Protection circuits is still a great design. But when you go to excessive overdesign like the overrated Pioneer SA-9800 from 1980 has 19 transistors in the Power Amp be sure the quality of sound even after upgrading will never be as good as the SA-9500 Mk I that has just 10 transistors. NFB as History proves is very useful in Hifi if used sparingly to reduce distortion & flatten ragged response curves. But usually hack Hifi designers add in way too much NFB just to get the pointless 0.001% THD in tests but unaware the musical value is heavily reduced. This is why Hifi 1982 onwards usually sounds so awful & boring as it may have high specs but at serious expense to the music. As with anything, the master designer can know how much NFB to add to improve but not spoil the sound. From all the amps we've heard, no such amplifier exists as if it sounded the Perfection that is possible to get with some amps, no-one would ever buy another amp.
WATTAGE vs VOLTAGE
Some examples we've found using 1kHz test tone, Peak Oscilloscope voltages read just before clipping of the wavetop.
Watts are RMS rated into 8 ohms from the manufacturer. We don't care about testing watts ratings as watts can be read high into distortion. We test each channel but play the signal through both, so none of that Stereo Amp higher power if just using One Channel nonsense you see in earlier ratings. We use 1kHz sine wave at 0db & read off the voltages on one channel at the speaker connections with an oscilloscope RMS to see the point of clipping & then check the other. Read about RMS, Peak & Music power elsewhere. We don't use any loading as no speaker is an flat 8 ohm resistor in it's response, so only this gives a reliable comparision. With loading, the readings would differ & amps have different Damping NFB & Impedance, so we simplify it in our own way with just what clean sine wave maximum is available at the outputs. It proves useful in seeing the difference in rated power around 30-32v. Also the 'shame' of three 50w amps only putting out 25v that one 34w amp can match, a tiny bit more current is less useful unless you use 4 ohm speakers than overall clean volume which again depends on design. Amps differ in frequency responses as some are heavily rolled off after 50Hz & some can't make the full 20kHz. To do these extremes in testing would give very different values & some specs are rated eg 50w RMS at 1kHz but only 40w RMS at 20Hz to 20kHz. Some amps quote a Peak Power, this will not be a clean Sinewave & having pushed one amp noted above that had a relay to protect it & it could read 25% higher before the relay clicked in, ie at 'destruction' & at almost the full HT voltage. Not one to want to test for either. Our tests appear to be reliable though some amps clearly use extra current capabilities to boost wattage as the list shows, if at the expense of clean AC output. An amplifier when Serviced with the adjustments done correctly can give a slightly higher voltage reading, ie AC bias settings. The maximum clean Sine Wave voltage ends as soon as clipping-flat topping of the sine wave becomes noticeable, rounded to the nearest whole digit.
As an example: Voltage or Wattage vs dB gain calculators show the 18w Coral Amp puts out a clean 14v RMS, the STR-6120 puts out 31v & a 110w Yamaha CR-2020 puts out 43v. Calculating simply voltage to dB shows the Coral has 23dB gain, Sony has 30dB gain & the Yamaha has 33dB gain. So 18w to 50w gives you 7dB extra gain but 50w to 110w only gives 3dB extra. With your smaller speaker losing 7dB to 10dB compared to a bigger one, the extra sound volume is easy to explain. Not so much the amplifier watts but the Loudspeaker dB gain. For PA amps in gigs & stadiums they use 95dB to 110dB speakers. The big Marshall stacks are 110db so using the 18w Coral amp you could realistically fill out a huge hall. So if you want more volume, get higher dB sensitivity speakers. You don't really need more than 50w. (as from the Loudspeakers page).
The Below are RMS voltage readings. 31v RMS = 86v pk-pk. RMS is half pk-pk divided by 1.404