I pretty much agree. McLaughlin has amazing technique and speed, but horrid tone, and all that widdly-widdly shredding leaves me longing for something with a little more melody, and a little more emotion.
Vinnie Colaiuta (the drummer) blew my socks off. I wasn't aware of him until I saw that DVD. Amazing drummer.
McLaughlin's bass player had fantastic tone in that set - a real contrast to the shrill shrieking noises that McLaughlin made with his digital amp emulation pedal.
Jeff Beck impresses me so much - a senior-citizen musician who just constantly gets even better. He plays almost everything with his right-hand thumb now, with an occasional flick of the index finger. Middle and ring fingers are almost permanently dedicated to the most sophisticated use of the 'Strat vibrato arm I have ever heard from any guitarist. It is a completely unique playing technique, and I've never seen any other guitarist use it. Nobody else can make a 'Strat sound like whale-song, either. Incredible!
Tal Wilkenfeld was another player who was new to me when I first heard that DVD. Her bass playing blew me away. Wow, so incredibly good, so incredibly young!
Recently she (Wilkenfeld) seems to have started singing and playing guitar. Unfortunately, I have rarely heard a singing voice that puts me off as strongly as hers - to me, it's quite horrid. But who knows, someone else might love it.
-Gnobuddy
Ok, I dragged the power amp over to the bench and played the preamp through the bookshelf speaker.
Our mortal enemy, Mr. Fizz, is back. Not a big surprise, as the bookshelf speaker has a tweeter.
The problem is that even in the "clean" setting, you can still hear it.
When I play an open G arpeggio, you can hear the fizz on the pick attack of the g and b strings, in particular. It's almost a crackle, or a combination of the two, a frackle.
What I am trying to determine right now:
Is the preamp generating more fizz than what 12AX7s would do?
I think a certain amount of fizz is inherent to 12AX7 preamps, but if there's too much fizz, there's something wrong.
If there's one thing I've learned diddling with DIY valve amps, it's "star ground everything."
Sure, Leo didn't star ground, but we only see the end result of the R&D, we don't see the pile of prototypes that oscillated or hummed.
Also, there is the time tested technique of pushing the leads around with a chopstick to get the correct "lead dress".
Well, with this little breadboard, I'm kind of stuck with "bus" grounding and I'm not going to push anything with a chopstick!
Not to mention all the parasitic capacitances inherent in the breadboard itself.
I know these breadboards are used up to the megahertz range, but that's with low voltage low impedance stuff.
A few picofarads of stray capacitance here and there can cause weird problems with high voltage high impedance circuits.
So, it looks like the time has come to move this thing off the solderless breadboard and on to a more suitable platform.
I have a chassis from a small solid state guitar amp that I can use, it's got all the holes drilled for the pots and jacks and stuff. I'll use a combination of good old terminal strips, plain perfboard and eyelet board. I don't like the eyelet board material Fender uses, I make my own out of perfboard.
I think I'm going to need a 24dB/octave low pass filter.
This means active filter. And the easiest way to do that is opamps.
So I'm going to need a low voltage supply.
Once I get the high voltage stuff off the breadboard, I can use it for what it was intended for, low voltage stuff.
Our mortal enemy, Mr. Fizz, is back. Not a big surprise, as the bookshelf speaker has a tweeter.
The problem is that even in the "clean" setting, you can still hear it.
When I play an open G arpeggio, you can hear the fizz on the pick attack of the g and b strings, in particular. It's almost a crackle, or a combination of the two, a frackle.
Yes, the graphic EQ cannot entirely dial out the fizz.
What I am trying to determine right now:
Is the preamp generating more fizz than what 12AX7s would do?
I think a certain amount of fizz is inherent to 12AX7 preamps, but if there's too much fizz, there's something wrong.
If there's one thing I've learned diddling with DIY valve amps, it's "star ground everything."
Sure, Leo didn't star ground, but we only see the end result of the R&D, we don't see the pile of prototypes that oscillated or hummed.
Also, there is the time tested technique of pushing the leads around with a chopstick to get the correct "lead dress".
Well, with this little breadboard, I'm kind of stuck with "bus" grounding and I'm not going to push anything with a chopstick!
Not to mention all the parasitic capacitances inherent in the breadboard itself.
I know these breadboards are used up to the megahertz range, but that's with low voltage low impedance stuff.
A few picofarads of stray capacitance here and there can cause weird problems with high voltage high impedance circuits.
So, it looks like the time has come to move this thing off the solderless breadboard and on to a more suitable platform.
I have a chassis from a small solid state guitar amp that I can use, it's got all the holes drilled for the pots and jacks and stuff. I'll use a combination of good old terminal strips, plain perfboard and eyelet board. I don't like the eyelet board material Fender uses, I make my own out of perfboard.
I think I'm going to need a 24dB/octave low pass filter.
This means active filter. And the easiest way to do that is opamps.
So I'm going to need a low voltage supply.
Once I get the high voltage stuff off the breadboard, I can use it for what it was intended for, low voltage stuff.
And these days my weapon of choice is the nastiest of the nasty, the Telecaster.
Actually I wouldn't mind having reverb. I have a delay pedal that can do reverb, but I'd prefer to have the reverb in the preamp. That issue will be dealt with later.
Concerning the de-nasty-fighter
I think we all agree that on-axis sound of guitar amps in most cases sounds too sharp, off axis is more agreeable.
So I propose to combine the forth-order filter for speaker-on-axis-roll-off with another (2nd-order?) filter of lower roll-of frequency (maybe 2.5kHz). At the end both filter responses are multiplied that way.
@ MJD-tech
Never ever play an electric guitar through a speaker box including a tweeter.
Disconnect the tweeter and the fizz is gone. Or block it with some gaffa tape at least.
I think we all agree that on-axis sound of guitar amps in most cases sounds too sharp, off axis is more agreeable.
So I propose to combine the forth-order filter for speaker-on-axis-roll-off with another (2nd-order?) filter of lower roll-of frequency (maybe 2.5kHz). At the end both filter responses are multiplied that way.
@ MJD-tech
Never ever play an electric guitar through a speaker box including a tweeter.
Disconnect the tweeter and the fizz is gone. Or block it with some gaffa tape at least.
Last edited:
I have an album "Love, Devotion, Surrender" with McLaughlin and Carlos Santana. It is one the biggest wankfests ever recorded.
Ol' Carlos was a little out of his league trying to keep up with John, but he made a valiant effort.
He can shred on the Dorian mode and bluesy type riffs, but John whips out all his patented altered scale runs and weird bends, and you can just tell Carlos was thinking "what in the world was that?"
The guitar tones were pretty high gain, possibly Mesa Boogie mark 1 amps.
I consider this more of a comedy album, but it's supposed to be, like, spiritual and stuff...
I don't like Mr. Fizz at all.
So far, the nefarious Mr. Fizz has come along for the ride every single time I've put semiconductor diodes into a guitar audio chain. When I saw KMG's circuit, I wondered how KMG could slather six diodes (!) onto each MOSFET without fizzage.
One of the things about KMGs work that impressed me was the demo clip of (mostly) clean tone from his Bogner Ecstacy MOSFET emulator. I didn't hear much harsh fizz in that demo, but there is heavy reverb and delay use, which might be there to reduce harshness.
Lately, I've played my semi-hollow electric guitar into an entirely solid-state chain a number of times. The guitar goes into a Digitech Trio+. A Boss Blues Driver, Biyang reverb, and Mooer Reecho delay pedal are in the Trio+ FX loop. Finally the Trio's "Mixer" output (which has some sort of de-nastifying filter) goes to my Acoustic AG30 acoustic guitar amp acting as a powered speaker.
And Mr. Fizz is nowhere to be heard.
So we can get fizz-free, very slightly "dirty" cleans out of a Blues Driver with a little delay and reverb helping out. I expect at least as much from KMG's circuit.
If KMG's circuit requires vast amounts of fizz-fighting, in addition to six diodes per MOSFET, two power supply rails, and a fistful of passive components per stage, it loses a lot of its appeal to me. It starts to look like too much work for too few returns.
That's disappointing. I hope you find a way to improve the situation.The problem is that even in the "clean" setting, you can still hear it.
When I play an open G arpeggio, you can hear the fizz on the pick attack of the g and b strings, in particular. It's almost a crackle, or a combination of the two, a frackle.
Production Fenders seem to be on the verge of instability, too. Bill M. had to stop offering an improved output transformer for the Super Champ XD because some of them would oscillate if you swapped the transformer.
I once opened up a dead SCXD and converted it to all-valve by adding a 12AX7 and Blackface preamp circuitry. From the perspective of audio signal flow, it had the worst designed PCB I've ever seen. Power and input ground were the same point. IIRC there was also a long trace from speaker ground to power supply ground, so speaker current would corrupt the ground reference voltage. No wonder it was on the hairy edge of outright oscillation.
I also briefly had a Blues Junior. It would oscillate if the ribbon cables from the EL84 output valves weren't dressed just exactly right along the metal chassis. Another marginally stable product.
My Princeton Reverb reissue will oscillate if you put some pedals - gainy ones - on the floor too near it. Leo cheaped out on grid stoppers and padder capacitors to manage HF bandwidth, so it seems many of his amp designs are barely stable.
Oh, man. This is one of the reasons why I didn't use a breadboard much. In the days when I was still tinkering with Hi-Fi, I've seen a high-gain audio circuit - particularly anything that used an integrated circuit - go nuts on a breadboard too many times. Discrete circuits rarely have that much gain, so behave better.
Since it's finally become clear to me that we really don't need to worry about anything above 5 kHz for electric guitar, the breadboard is an option once again. We can limit the bandwidth of our guitar electronics for stability, without worries about affecting tone.
A single good BJT (like the BC550C I used) is surprisingly good at this job. Sallen-Key filters work with a unity-gain emitter follower, but you have to vary the two cap or resistance values to set the filter Q where you want it.
With an opamp you can keep all R's and C's the same, and just tweak voltage gain to set filter Q (so one resistor value sets Q.) But you have to provide dual supply rails and better HF supply decoupling to keep the opamps stable and happy.
Long before I ever held a guitar in my hands, the sound of the Rolling Stones on AM radio permanently turned me off whatever the heck guitar it was they were using (I didn't know a 'Tele from a Stradivarius at that age.)And these days my weapon of choice is the nastiest of the nasty, the Telecaster.
All I knew was that Rolling Stones guitar sounded horridly out of tune, all the time.
I was an adult before I first heard the word "intonation" and discovered what Leo's three-barrel bridge did to the 'Tele.
The Rolling Stone's substance-abuse problems probably didn't help, either. You can't tell how bad your tuning is when you're as high as a kite.
-Gnobuddy
I found the entire album on You Tube. The wankery started mere seconds into the first track. I don't think I lasted a full minute before clicking the stop icon.
Love, Devotion, Surrender? You have to read between the words: (I) Love (my own ego). (I demand your dog-like) Devotion (to my guitar-godliness.) (You must) Surrender (to my awesomeness).
-Gnobuddy
I tinkered a little with the "2BJTE" ( 2BJTE - 2 BJT triode emulator ) circuit in LTSpice during my lunch-break today. (A copy of the original circuit is also attached to this post.)
I started my tinkering by sketching in a JFET common-source stage at the input for the guitar to connect to, and buffering its output with a BC550C emitter follower.
Then I got rid of some of the obvious mistakes in the BJTE circuit, starting with the useless 33 ohm emitter resistor and even more useless 4.7nF(?) bypass cap (R2, C2).
I also removed the useless shunt negative feedback from collector to base of the first BC108, the one that creates the interesting distortion. This was done by splitting the 470k resistor (R6) into two 220k resistors in series, and adding a 1uF cap to their junction to filter out the fed-back signal.
Removing the shunt negative feedback in turn means the circuit no longer has to be fed from zero impedance(!) to work. In LTSpice, I stuck a 10k in series with the input with little change to the output waveform.
After some tinkering in LTSpice, it seems increasingly clear that the interesting distortion characteristics really do come from the collector bootstrapping. Removing that capacitor drastically changes the output waveform, which reverts to the usual distortion from the exponential transfer function of a BJT.
I also raised the 1k emitter resistor (R7) to 15k, which should do wonders for current consumption (and battery life.)
Increasing the 500 ohm bootstrap resistor (R4) to 1k didn't seem to cause any big change in the circuit behaviour. Similarly, I was able to reduce the huge 100 uF bootstrap capacitor (C3) to one-tenth it's value.
The resulting circuit is attached (LTSpice screenshot).
Now, that output waveform...our eyes are notoriously bad at doing mental Fourier transforms, but that waveform looks quite a bit like the shape KMG was going for, doesn't it? The slightly overdriven 12AX7 stage, with negative output peaks flattened off due to grid current flow at the input clamping incoming positive signal peaks?
Also, there is no evidence of harsh corners in the simulated waveform, so one can hope for little to no "fizz".
I'm going to breadboard this up when I get the chance, and see what happens. Maybe it will make some nice distortion without fizz, who knows (fingers crossed).
-Gnobuddy
I started my tinkering by sketching in a JFET common-source stage at the input for the guitar to connect to, and buffering its output with a BC550C emitter follower.
Then I got rid of some of the obvious mistakes in the BJTE circuit, starting with the useless 33 ohm emitter resistor and even more useless 4.7nF(?) bypass cap (R2, C2).
I also removed the useless shunt negative feedback from collector to base of the first BC108, the one that creates the interesting distortion. This was done by splitting the 470k resistor (R6) into two 220k resistors in series, and adding a 1uF cap to their junction to filter out the fed-back signal.
Removing the shunt negative feedback in turn means the circuit no longer has to be fed from zero impedance(!) to work. In LTSpice, I stuck a 10k in series with the input with little change to the output waveform.
After some tinkering in LTSpice, it seems increasingly clear that the interesting distortion characteristics really do come from the collector bootstrapping. Removing that capacitor drastically changes the output waveform, which reverts to the usual distortion from the exponential transfer function of a BJT.
I also raised the 1k emitter resistor (R7) to 15k, which should do wonders for current consumption (and battery life.)
Increasing the 500 ohm bootstrap resistor (R4) to 1k didn't seem to cause any big change in the circuit behaviour. Similarly, I was able to reduce the huge 100 uF bootstrap capacitor (C3) to one-tenth it's value.
The resulting circuit is attached (LTSpice screenshot).
Now, that output waveform...our eyes are notoriously bad at doing mental Fourier transforms, but that waveform looks quite a bit like the shape KMG was going for, doesn't it? The slightly overdriven 12AX7 stage, with negative output peaks flattened off due to grid current flow at the input clamping incoming positive signal peaks?
Also, there is no evidence of harsh corners in the simulated waveform, so one can hope for little to no "fizz".
I'm going to breadboard this up when I get the chance, and see what happens. Maybe it will make some nice distortion without fizz, who knows (fingers crossed).
-Gnobuddy
Attachments
Well, here's the deal.
- It reacts and "feels" darn near exactly how a tube Fender preamp does. The transition from clean to overdrive is spot on.
I've been focused on what's wrong and have failed to emphasize what's right. - With a signal generator feeding in a sine wave, I'm getting the exact waveforms that KMG gets here where he compares to an actual 12AX7.
LND150 tube emulator attempt.
Every little nuance of the waveforms, I'm getting those too. - If I use a guitar speaker, and stand a few feet away, off axis, I barely notice the fizz.
- I've had fizz/crackling problems with pure valve guitar amps too. Even where everything looks ok with a static sine wave, there can be some unpleasant overtones with an actual guitar signal. This usually indicates some kind of "transient parasitic instability" (or something) due to improper grounding and/or lead dress, even corroded contacts on the tube socket.
- Also, I am kind of cheating. KMG uses individually buffered negative supplies for each stage. I'm using fixed bias, derived from a mere resistive voltage divider, common to 3 stages. As far as the signal is concerned, the fixed bias rail is at ground, but because of the breadboard, the rail is not "star grounded". This could be a problem.
- More cheating, I'm running 3 gain stages and a cathode follower from a single B+ node. Granted, it is voltage regulated, and should be low impedance and all, but still kind of cheating. This is actually a fairly high gain circuit. Also that B+ node is a bus, not a star.
- And finally, the breadboard itself: no star grounding, questionable contact integrity, parasitic capacitance.
That's why I need to rebuild the circuit properly to give it a fair chance.
Jury is still out. Can't blame KMG's design concept just yet.
The kind of fizz I'm getting is different than the "tube screamer" or "rat" diode clipping fizz. It really does seem like a burst of oscillation, possibly at RF, so I can't see it on my "high dollar" scope.
If you'll recall my adventures with the IRF820 regulator, it's quite possible that the IRF820 was reacting to some oscillation happening elsewhere in the circuit, rather than causing the oscillation itself. The TIP50 is too "slow" to be bothered by the oscillation.
I'm no expert in RF theory, all I know is "get the grounding and lead dress right".
Well, this circuit is simulating a Blackface reverb channel preamp without the reverb.
I doubt you could build this preamp with real 12AX7s and not need any further signal processing.
I've looked around and have found most guitar speaker simulator circuits use a 4th order lowpass and there's a peak somewhere 2 - 4kHz and another one 100-200hz
Seems reasonable to need a speaker sim with this kind of preamp.
True, and a lot of other amps as well: Mesa, looking at you.
I've never before used this kind of breadboard with a valve circuit or a high voltage solid state circuit, and I'll never do it again.
I've used the breadboard successfully to fiddle with TLO72 opamps, 555 timers and simple stuff like that.
Mine has a 6 saddle bridge, and 22 frets. It's a "parts caster". Gets the job done, nearly bulletproof, stays in tune, I don't care if it falls off the stand and gets dinged up.
Ah. It seems I completely misunderstood the entire situation and generally haven't got a clue. That probably means I should run for President.The kind of fizz I'm getting is different than the "tube screamer" or "rat" diode clipping fizz. It really does seem like a burst of oscillation, possibly at RF, so I can't see it on my "high dollar" scope.
What about a 100pF cap from each MOSFET drain to ground? With your 100k drain resistors (and ignoring loading, stray, and input capacitances), that will restrict bandwidth to about 16 kHz, three or four times more than we need for electric guitar, but hopefully enough to ensure RF stability.
I can imagine the need for the 100Hz high-pass if you're plugged into a powerful P.A. system that really does go down to 30 or 40 Hz, but most of the smaller P.A. and powered speakers I've encountered rarely have a response that goes below 60 or 70 Hz (this probably applies to your bookshelf speakers, too.)
So I have a suspicion that guitar speaker emulation doesn't really need to include additional bass rolloff for most typical situations. If we get the high frequency end right, it will probably do the job just fine.
Guitars are very personal things, aren't they? Probably my favourite is a Korean-made semi-hollow similar to an ES-335. Tonally, its almost at the other end of the electric guitar spectrum from a stinging 'Tele. A Blackface Fender seems to be a match made in heaven for this guitar - it sounds really good even with my hybrid Super Champ XD set to the clean channel.
But I also have a Les Paulish thingy I love, and I'm starting to miss having a 'Strat. But the narrow necks on Fenders are a headache for me.
I've grown weary of hauling an acoustic guitar, an electric guitar, and a bass guitar to jams, so from time to time I dream about a guitar that can do both passable acoustic sounds for chords and general rhythm playing, and also electric guitar sounds for solos. There are some commercial solutions, all beyond my financial reach. If I could find a way to put a piezo on my semi-hollow without ruining its appearance, that would probably do it.
But that's another project for another time. Now, I use my Super Champ XD and play both chordal and solo parts through the same semi-hollow guitar. If the KMG preamp sounds as good as the SCXD, I'll build one of those, and leave behind the SCXD. One less delicate box to haul around to jams.
-Gnobuddy
That seems to have the same slope at both ends of the frequency response. Probably second-order.
-Gnobuddy
I was reminded of Voltwide's comment last night.
Some days ago I stopped by at a local thrift store. A small speaker caught my eye - two roughly 4" aluminium-cone speakers and a tweeter, mounted in a D'Appolito (MTM) layout in a small enclosure with a roughly triangular cross section. Probably intended as a TV centre channel speaker, or for wall mounting to provide background music in a coffee shop.
The brand was Digital Research (do an Internet search for "white van loudspeaker scam" to learn more about this brand!), but the asking price was only $4.99. Not much to lose there, even if the speaker turns out to be dead or not to sound great.
I had a couple of possible uses in mind (as a mic-stand mounted foldback vocal monitor when my wife and I sing at jams, or in a battery-powered portable guitar amp.) For only $5, I decided to buy it.
So last night I connected the $5 speaker to a $15 class D audio amplifier board powered by a $4 switching power supply.
I made up a 1/4" jack to RCA plug adaptor cable, and connected a dirt-cheap ($25 USD) Danelectro Fab chorus pedal to act as a preamp. (It was within arm's reach; the graphic EQ pedal I would normally have used was in the other room. I was lazy.)
Finally I plugged in my beater Epiphone Les Paul SP-II, bought about 25 years ago for $150 USD (a new one costs $250 Canadian today: Epiphone Les Paul Special II - Ebony - Long & McQuade Musical Instruments )
No tone control (except on the guitar.) A full-range speaker with a tweeter. No valves. No JFETs. No de-nastifying filter.
Yikes - how nasty is this going to sound? Or will the chorus pedal suppress the harshness and save the day?
Well, after making a few slight adjustments to my right-hand playing technique, I got some really nice clean tones out of this $199 guitar and amp combination. Nice enough that my wife commented on how good it sounded from across the living room.
That's when I remembered Voltwide's comment, about how the gear doesn't matter anywhere near as much as most musicians think it does.
The Danelectro chorus pedal definitely played a role here, taking off the harsh edge that you usually get with a solid-body electric guitar plugged into an entirely solid-state guitar amplifier. Like other cheap Danelectro pedals I've tried, this Fab chorus pedal sounds really good, but looks really ugly in its cheap plastic housing.
I had anticipated a severe shortage of bass, but, at least at the low SPL I used last night, this was not really a problem. Open "cowboy chords" sounded full. There was no evident shortage of bass for single note runs all the way down to the low open "E" string.
The combination of Danelectro pedal, class D amp, and white-van-scam Digital Research speaker actually sounded good enough that I might put this together with a 20-volt cordless tool battery to make a small portable go-anywhere guitar amp. Spring is almost here, and this part of BC will soon be bursting with green things and beautiful wildflowers soon. It would be nice to be able to go to a park, sit in the middle of all that natural beauty, and play some electric guitar.
Given its shady origins and even more shady distribution/ marketing method, it proved rather hard to find an online photo of my $5 Digital Research speaker to show you guys. I finally found one - from a scam artist trying to sell it as a Polk system he claimed to have inherited from his dad!
So I stole the scam artist's "Polk" speaker photo, and attached it to this post. Now you too can enjoy the beauty of this white-van speaker special!
(It looks pretty decent. That - and the tempting backstory - are key ingredients in hooking gullible people into buying "white-van-speakers".)
-Gnobuddy
Attachments
I'm highly amused at the brand name! The owner of the "original" company of that name wrote CP/M, the first microcomputer operating system (of any popularity), and infamously blew off (I forget which, maybe both - these computer history stories are so incredible but true) Bill Gates and IBM. By the time the brand name was used on speakers, the original company had surely folded.
I've seen some model of acoustic guitar with a speaker in it to give it a little more volume or bass or whatever, and I was thinking that kind of thing is the perfect place to use one of these modern low-power (2 watt? 20 watt??) Class D power amps.
I've seen some model of acoustic guitar with a speaker in it to give it a little more volume or bass or whatever, and I was thinking that kind of thing is the perfect place to use one of these modern low-power (2 watt? 20 watt??) Class D power amps.
The version I heard was that Kildall blew off IBM, who then went to a young charlatan named Gates, who in turn bought Bill Paterson's "Quick n Dirty Operating System" for a song, sold it to IBM for millions, and so was launched on his long career of stepping on a thousand throats to get whatever he wanted by any means.
The first personal computer I ever used ran DOS, so I missed the era of CP/M.
After killing Seattle Computer Products and Digital Research, Gates went on to kill many other good companies, including Dr. DOS and Be, Inc, makers of the BeOS, the most fascinating PC operating system I've ever had the pleasure of using. By this time Microsoft had enough clout to simply tell computer retailers that if they sold BeOS, they would lose the contract to sell Microsoft products. No retailer could afford that, so BeOS died a sorry and undeserved death.
If you believe the Amazon or Ebay ads, every class D amp starts at about 100 watts (even if it runs on two AAA cells), and power output only goes up from there.
The one I'm using is currently running on a 24V, 2A power supply. Like most of them, it's based on a chip containing four drive circuits, wired up into two bridge-mode power amplifiers.
I figure at least 2 volts lost across each output device, so a maximum peak voltage of 20V across the speaker. That translates to 50 W RMS into a 4 ohm load, or 25 W RMS into an 8 ohm load. (My power supply is marginal with a 4 ohm load, but I don't think I've gone beyond a few watts output power yet.)
The $5 "Digital Research" speaker measured 3.4 ohms DC resistance with my ohmmeter after subtracting its zero offset, so it's nominally a 4 ohm speaker.
-Gnobuddy
Yes, for that app class-d is the way to go - TPA3118 from TI for instance. That is the one I use with my mobile 12V-amps.
I have been staring at the same writing on the wall.
Mass-produced Hi-Fi is dead, and class D audio power amplifiers allow manufacturers to make products that are smaller, lighter, cheaper, and more powerful. That's a pretty potent combination of positives. Nobody cares if they are not (yet) up to true Hi-Fi performance levels.
There was a recent thread on diyAudio about a 600 W (!!) TI class-D audio power amp evaluation board that was selling (during a half-price sale) for something like $80 (USD). How can you possibly compete with that with class AB? The heatsink for a 600 watt class AB amp alone would probably cost more than $80.
-Gnobuddy
agreed. The amp you are referring to is TPA3255 EVM-board. I could test this extensively here as a friend led it to me for several months. I would not really go for 600W with more than 40W power dissipation on that tiny chip. But it is a very nice choice to design a compact mono block for around 100W applications - that may be powered by a 36V pedelec battery.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.