Good point about the Class-D clipping.Honestly, I'm not a fan of diode clipping, but I think that is a far, far better alternative to the horrid sounds I've heard from class-D chips clipping!
Are you okay with the sound of the thin plastic box the Roland comes in? I have only heard it with the stock speaker, but it sounds like a little plastic bucket to me. Thwack it lightly with your hand, you'll hear the same unpleasant sound. I'm wondering if that sound will come through even with a better speaker.
Actually, the little Roland isn't plastic at all; it's 1/2" MDF and appears to be soundly constructed. If you "thwack" it, it sounds like any small wooden box would, to my ears. Would a bigger box sound better? Undoubtedly, but it would be just that---bigger. And heavier. And more expensive. Posters on other forums have reported a HUGE increase in sound quality from replacing the stock speaker with a Weber. Coupled with a power boost to avoid any clipping, and it should be good to go. If I increase the MAXIMUM power to 16 watts rms, that should give me 12db of rms-to-peak headroom before clipping whilst playing at 2 watts, which I think will be more than loud enough, and should draw only somewhat marginally more current than the stock Roland playing at 2 watts.
Eh?
Amazon says:
so 6 AA´s give you 19.2V @ 700mAH
A 15W amp fed from them will eat at least 16/18W or so, even if Class D , so continuous use will give you some 30 minutes battery life, twice as much considering playing is intermittent, has dynamics, etc.
And you need a 9.6V or so tap to feed the original Roland board .
You don´t need any DC DC converter which to boot will reduce your battery time.
Personally I´d look for way higher capacity batteries , even if heavier.
Will not fit the original battery holder but yes somewhere inside the speaker cabinet.
As of speakers: I have repaired MicroCubes, and have put my own 6" speakers inside.
Speaker frame, at least mine, *barely* fits inside, in fact "it should not" but thanks God (or chance) amp chassis does not reach the frontal speaker panel but stops about 1/2" short, letting speaker edge slid in that unused space.
That said, the original 5" Roland speaker sounds very good, considering it´s 5", driven by paltry 2W Class D and is fit inside *half* a shoebox space.
People hate it because of UNREAL expectations, such as setting COSM to "Metallica playing 8 MBDR heads into 16 4x12", full blast, at Castle Donnington" or whatever and getting angry because house walls do not come down in a cloud of dust. ..... gimme a break
That's certainly not why I hate it.
I think it sounds worst set to clean tone, which is where my friend sets it 99% of the time.
Plenty of people seem to like the little 'Cube, so there you go, as the old saying goes, one man's meat is another man's poison.
-Gnobuddy
LiFePO4 has lower energy density than Lipo/Li-ion types, so you get less mAh in the same-size cell.
You can get better capacity from larger LiFePO4 cells - there seem to be plenty of 18650 sized LiFePO4 cells advertised. (These will NOT fit AA battery holders. I do see some 18650-size battery holders advertised on Amazon and Sparkfun.)
You are facing the same Li-ion quandary as me. On the plus side I already have a suitable power tool pack I could use...on the minus side, I am concerned about the safety aspects of this chemistry. At this point in my life, burning down my home would be a disaster from which I would never recover financially, assuming my wife and I both escaped with our lives.
I saw rechargeable NiMH cells in "C" and "D" sizes at Home Depot today. But those have their downsides too - only 1.2 V per cell, so you need a lot of cells - sixteen of them to equal the voltage of a 6-cell LiFePO4 pack! And there is the very annoying self-discharge problem, the cells are always dead just when you need to use them.
If I had the budget for it, I would go LiFePO4 for sure.
-Gnobuddy
Out of curiosity, I looked at rechargeable NiMH cells on Amazon.com. A pair of these 8-cell bundles would work: Amazon.com: EBL 2800mAh AA Rechargeable Batteries Ni-MH (8 Pack) with 8 Bay AA AAA Battery Charger: Home Audio & Theater
Sixteen 1.2V, 2.8 Ah AA cells, plus two 8-cell chargers, for $50 (USD). They have to all go in series to power the portable Princeton, and come apart to go in the charger.
It is possible to solder directly to NiMH cells without damaging them, but this requires good soldering skills and the right soldering iron - and requires a different charger.
A NiMh pack is a lot heavier than an equivalent-capacity lithium pack, but in such a small capacity pack, and for this particular application, the weight difference may be negligible in practical terms.
-Gnobuddy
Sixteen 1.2V, 2.8 Ah AA cells, plus two 8-cell chargers, for $50 (USD). They have to all go in series to power the portable Princeton, and come apart to go in the charger.
It is possible to solder directly to NiMH cells without damaging them, but this requires good soldering skills and the right soldering iron - and requires a different charger.
A NiMh pack is a lot heavier than an equivalent-capacity lithium pack, but in such a small capacity pack, and for this particular application, the weight difference may be negligible in practical terms.
-Gnobuddy
On this subject, my friend wrote me this:
"I've always thought that the sound of the distortion when an amplifier approaches its limit is more important than the amount of distortion. I also think current limiting might sound better than voltage clipping"
'Nothing sounds like a Plexi driven into clipping going into a 4x12.' Just to generalize on output clipping. But then if you take into consideration that at clipping the NFB is lost and the amp is really hard clipping rather than those pretty pictures people show where the wave shape looks rounded it is understandable people are confused.
Speaking for myself, I've always thought the best sounds I've ever heard from an electric guitar came from David Gilmour. And he is also famous for using enormously powerful Hiwatt amps, designed like Hi-Fi amps with very little distortion, and set to run clean, with little or no clipping.
Instead, he got soft-clipping from Big Muff and similar pedals, subtle low-order harmonic distortion for cleans from Tube Drivers and similar pedals, and cleaned up the harsh edges from the Muff with delay, Leslie speakers, flangers, and reverb.
I wonder what Gilmour would sound like playing through his normal pedalboard, followed by a high-wattage class D power amp with essentially no audible distortion of its own?
-Gnobuddy
Instead, he got soft-clipping from Big Muff and similar pedals, subtle low-order harmonic distortion for cleans from Tube Drivers and similar pedals, and cleaned up the harsh edges from the Muff with delay, Leslie speakers, flangers, and reverb.
I wonder what Gilmour would sound like playing through his normal pedalboard, followed by a high-wattage class D power amp with essentially no audible distortion of its own?
-Gnobuddy
For a long time I was in* The Cult of Gilmour
The amps are (were) 100% clean, "tone" happens very early in the effects chain before all the short and long delay effects. The volume on the final amps were always just about at breakup (if beyond, you need more amps). See here for more
Regardless, Dave Gilmour will always sound like Dave Gilmour. (thwack)
*currently the twelve step program has me chicken pick'n on a thinline and an SE valve amp, cranked. Occasionally I'm allowed to add some reverb
Last edited:
Then not exactly 100% clean (thanks God
) , tube amps "magic" happens specially at that point, not by chance called "the sweet spot" .Around 6 or 7 volume on a classic Fender amp, YMMV on Marshall or Hiwatt.
If you scope the signal, it´s already clipping, but with smooth peak rounding and lots of natural compression thanks to power tube grid rectification and bias shifting, yummy!!!!
NO Class D amp can achieve that, ever, they hit the brick wall at full speed (rail voltages) and crash in an ugly way, plus SMPS derived rails are *stiff* .
Even worse: some SMPS, overcompensated trying to keep voltage stiff, can create some audible artifacts.
A conventional Class AB SS amp, fed from a conventional (iron) supply will also hit the rails, but just that, no supply artifacts, and rails will drop on demand , usually by 15%, so that produces *some* pumping effect, not exactly compression but variable level clipping following signal envelope, better than nothing, gives it a little "life".
Mr Quilter achieves stunning results, very natural and musical from his brick sized amplifiers, which use Class D amps and SMPS but here I am certain he worked A LOT to achieve that.
That is what worries me going down the SS-Class D path. I am putting it aside to do a tube front end with Class D amp at the moment. It is about the last thing I should be doing now but curiosity has the best of me.
Agree 100%. I suspect, though, that Gilmour would come closer to his normal tone (playing through a 100% clean power amp) than most other famous electric guitar players.
I think it may be possible to come up with either a valve or FET final preamp stage that provides just a little of that soft distortion, just before the signal goes to the powerful, 100% clean power amp. Even an off-the-shelf overdrive pedal with a nearly transparent setting might do the trick. Maybe something as everyday as a Boss BD2 Blues Driver.
I'm not suggesting we all try to become low-quality copies of Gilmour, of course - while I admire his playing and sound, I don't try to imitate it. But is interesting to see that he gets his fantastic sound without doing what most famous guitar players (Mark Knopfler excepted) do.
I have only once played through a Quilter amp. It was small, light, loud, beautifully constructed, and extremely expensive (over $1200 USD). It had none of the harshness I usually hear in solid-state guitar amps - but that particular amp also sounded muffled and dull. There wasn't a trace of the lovely singing clean tones and shimmering treble you can get from many good valve amps. To me it was a very "meh" amp - nothing really bad, nothing really good.
The music store owner told me he sold most of his Quilters to wealthy middle-aged acoustic guitarists, who valued the small size and light weight of the Quilters more than anything else, and didn't care about the exorbitant price.
Quilter probably had to do a lot more of his own R&D to design and build his power amps. Now anyone with a VISA card can buy a TPA3255-based board and a 48V SMPS and have monstrous amounts of clean power available. All we have to figure out is how to make a good-sounding preamp, whether valve or solid-state or hybrid.
-Gnobuddy
A couple of pics of my project. This started with a $4 pair of thrift-store refugee Sony speakers (first pic) and an Amazon.ca class-D circuit board.
Since the power amp has two channels, I decided to use both speakers. To minimise unwanted interference problems from two nearby full-range speakers, I had two choices: either stack them in a vertical array, or put them back-to-back, one pointed at the audience, one pointed back at the performer for monitoring.
I eventually decided to go with the vertical array, and the next problem was how to attach the two speakers to each other securely. It finally dawned on me to use a biscuit-joiner to do the job.
So far the electronics is extremely simple: a home-made adaptor connects a 1/4" jack to a pair of RCA plugs that mate with the class-D board. A Joyo British Sound (or Joyo American Sound) pedal provides Sansamp-like tube amp emulation, and accepts input from one of my guitars.
All testing so far has been done with another thrift-store refugee, this time a 24 volt switching power supply that says "Kodak" on it.
In keeping with the topic of this thread, that power supply will eventually be supplemented with a battery pack for off-grid use. The choice of battery chemistry has caused much head-scratching, and I still haven't made a decision.
This particular class D board has an onboard blue LED that lights up when power is applied. The Joyo has the usual LED that comes on when not in bypass mode. If an additional power indicator is necessary, I have some 6000 mcd green LEDs that light up brightly at 80uA (yup, 0.08 mA), so one of them would draw negligible additional power.
I expected these nominally full-range speakers would sound harsh and nasty, and I would have to design and build a "de-nastifying filter" as in a recent previous project that also used boombox speakers. To my surprise, that turned out not to be the case: for whatever reason, these speakers don't sound harsh. Most likely because the "midrange" and "tweeter" that Sony used are both actually cheap piezo devices that produce little or no output at any frequency where the guitar itself has significant signal output.
As far as playing guitar goes, this is everything actually needed. All that remains is to mount the electronics to the back of the loudspeakers and tidy up the wiring. It would be nice to have a little reverb as well, which could be done with any guitar reverb pedal.
However, in addition to guitar, I would like to add one or two XLR microphone inputs for singers, and that would require building some custom electronics, at the least, some sort of simple mixer to combine the mic and guitar signals. I may have to build the mic preamps too, if my little ART USB dual pre turns out not to have quite enough gain.
-Gnobuddy
Since the power amp has two channels, I decided to use both speakers. To minimise unwanted interference problems from two nearby full-range speakers, I had two choices: either stack them in a vertical array, or put them back-to-back, one pointed at the audience, one pointed back at the performer for monitoring.
I eventually decided to go with the vertical array, and the next problem was how to attach the two speakers to each other securely. It finally dawned on me to use a biscuit-joiner to do the job.
So far the electronics is extremely simple: a home-made adaptor connects a 1/4" jack to a pair of RCA plugs that mate with the class-D board. A Joyo British Sound (or Joyo American Sound) pedal provides Sansamp-like tube amp emulation, and accepts input from one of my guitars.
All testing so far has been done with another thrift-store refugee, this time a 24 volt switching power supply that says "Kodak" on it.
In keeping with the topic of this thread, that power supply will eventually be supplemented with a battery pack for off-grid use. The choice of battery chemistry has caused much head-scratching, and I still haven't made a decision.
This particular class D board has an onboard blue LED that lights up when power is applied. The Joyo has the usual LED that comes on when not in bypass mode. If an additional power indicator is necessary, I have some 6000 mcd green LEDs that light up brightly at 80uA (yup, 0.08 mA), so one of them would draw negligible additional power.
I expected these nominally full-range speakers would sound harsh and nasty, and I would have to design and build a "de-nastifying filter" as in a recent previous project that also used boombox speakers. To my surprise, that turned out not to be the case: for whatever reason, these speakers don't sound harsh. Most likely because the "midrange" and "tweeter" that Sony used are both actually cheap piezo devices that produce little or no output at any frequency where the guitar itself has significant signal output.
As far as playing guitar goes, this is everything actually needed. All that remains is to mount the electronics to the back of the loudspeakers and tidy up the wiring. It would be nice to have a little reverb as well, which could be done with any guitar reverb pedal.
However, in addition to guitar, I would like to add one or two XLR microphone inputs for singers, and that would require building some custom electronics, at the least, some sort of simple mixer to combine the mic and guitar signals. I may have to build the mic preamps too, if my little ART USB dual pre turns out not to have quite enough gain.
-Gnobuddy
Attachments
Class-D board: Yeeco 2X100W Dual Channel Digital Stereo Power Amp Module DC 24V Class D Audio Amplifier Board for 4Ω 6Ω 8Ω Subwoofer Sound System Speaker Car Vehicle Home Theater: Amazon.ca: Electronics
Actual power on a 20V battery would be:
Vpeak (approx) = 20-3 = 17v (assuming 3V losses)
Vpeak * Vpeak = 289
Vrms*Vrms = Vpeak*Vpeak/2 = 144.5
Prms (watts), 6 ohm speaker = 144.5/6 = 24 watts
That's per channel, and there are two channels, one driving each of the two speakers. Call it around 40 - 50 watts RMS total for both channels.
The speakers have a model number on the back, and Google turned up the fact that they originally came with a Sony mini-component system. I found the manual, but there are no useful technical specs regarding the speakers: https://docs.sony.com/release//specs/CMTHPX9_mksp.pdf
I take Sony's claims of "60 watts per channel @ 10% THD" with a grain of salt, but I'm hoping these speakers will hold up under half that power or less, which is all they'll get from the power amp I'm using.
I'm sure these speakers are far less sensitive than a real guitar speaker. I'm hoping they will be loud enough to use in a small song circle with a few friends at a quiet park. So far, I've only tested in my apartment, and kept the volume well down not to annoy my neighbours.
Sony claims the woofer in each of these speakers is 4.75" diameter, so the Microcube's 5" is actually bigger! However, having the two woofers much less than a (bass) wavelength apart should make them behave like a single 6.7" speaker, and improve sensitivity by about 3 dB at guitar frequencies. It will probably increase from absolutely abysmal to merely pathetic.
-Gnobuddy
Actual power on a 20V battery would be:
Vpeak (approx) = 20-3 = 17v (assuming 3V losses)
Vpeak * Vpeak = 289
Vrms*Vrms = Vpeak*Vpeak/2 = 144.5
Prms (watts), 6 ohm speaker = 144.5/6 = 24 watts
That's per channel, and there are two channels, one driving each of the two speakers. Call it around 40 - 50 watts RMS total for both channels.
The speakers have a model number on the back, and Google turned up the fact that they originally came with a Sony mini-component system. I found the manual, but there are no useful technical specs regarding the speakers: https://docs.sony.com/release//specs/CMTHPX9_mksp.pdf
I take Sony's claims of "60 watts per channel @ 10% THD" with a grain of salt, but I'm hoping these speakers will hold up under half that power or less, which is all they'll get from the power amp I'm using.
I'm sure these speakers are far less sensitive than a real guitar speaker. I'm hoping they will be loud enough to use in a small song circle with a few friends at a quiet park. So far, I've only tested in my apartment, and kept the volume well down not to annoy my neighbours.
Sony claims the woofer in each of these speakers is 4.75" diameter, so the Microcube's 5" is actually bigger! However, having the two woofers much less than a (bass) wavelength apart should make them behave like a single 6.7" speaker, and improve sensitivity by about 3 dB at guitar frequencies. It will probably increase from absolutely abysmal to merely pathetic.
-Gnobuddy
Ah, the old TDA7498......quite efficient little mite; your estimations are definitely on the conservative side; ST is posting 44 watts into 6Ω @ 10% THD on the TDA7498 datasheet. On this note, I agree with Fahey that power quotes at 10% are useless (especially for Class-D). I believe you'll have plenty o' power, though. Looking at the Micro Cube speaker, I'd say their claims of a 5" speaker are stretching things a bit; it's 4 1/2" at MOST; definitely the weak spot in this amp.
I find that comparing the datasheet THD vs power curve, and the datasheet maximum output power specification, can be quite illuminating. The attached images are from the TDA7498 datasheet for a 36 volt power supply, and 8 and 6 ohm loads.
Distortion is down at 0.01% - 0.02% at 100 Hz just before the amp starts to clip. When clipping starts, distortion spikes steeply upwards, and soars all the way up to 10%. The distortion increases by a factor of 500 - 1000 times (!), in other words.
Looking at the graphs, while the distortion increases a thousand-fold, the power output only increases by 2x - 3x times.
So what is the real output power? We all agree the 10% number is ridiculous. Should we go with 1%? That is still a hundred times above the distortion floor at slightly lower power; it is a hundred times more distortion than the chip is actually capable of!
To me, the honest power rating is the point at which you see the THD curve abruptly kick up, and transition from being nearly horizontal to nearly vertical. This sharp change of direction very clearly signals the onset of clipping.
For the TDA7498, and most similar chips, this point seems to occur at somewhere in the range of 1/3rd to 1/2 of the "max power at 10% THD" stated on the datasheet. Suddenly, the "100 watt" chip turns into a 30 watt or 50 watt chip. It's only a matter of 3 - 5 dB, but less technical buyers might not understand this, and prefer the "100 watt" chip to the honestly advertised 30 watt one.
I can see why the marketing department, whose stock in trade is lies, wants to claim 100 watts instead of 30 watts. Meantime, I bet the engineering department is frustrated by this, since their hard work in achieving a 0.01% THD floor has been completely hidden behind the "10% THD" spec.
The back-of-the envelope math I use to estimate output power may be a decibel or two on the conservative side, but it is quick and easy, and if it errs, it errs on the side of honesty. I think it's reasonably real-world accurate: I'm assuming 1.5 volts lost at each end of the signal swing to device saturation and power supply droop. How much better can we really do? 1 volt? 0.5 volt? Can we really keep our guitar signal peaks accurate to within 0.5 volts to avoid hard clipping, even if the amp can do it?
One of the nice thing about these class D chips is that they have brought the watts/dollar cost of audio power amplification so low that we can ignore the false advertising, and buy as much power as we need, within reason. So we have the luxury of pretty much ignoring the advertised power rating entirely. All we need to worry about, is whether the board can handle the power supply voltage and speaker impedance that we plan to use.
-Gnobuddy
Distortion is down at 0.01% - 0.02% at 100 Hz just before the amp starts to clip. When clipping starts, distortion spikes steeply upwards, and soars all the way up to 10%. The distortion increases by a factor of 500 - 1000 times (!), in other words.
Looking at the graphs, while the distortion increases a thousand-fold, the power output only increases by 2x - 3x times.
So what is the real output power? We all agree the 10% number is ridiculous. Should we go with 1%? That is still a hundred times above the distortion floor at slightly lower power; it is a hundred times more distortion than the chip is actually capable of!
To me, the honest power rating is the point at which you see the THD curve abruptly kick up, and transition from being nearly horizontal to nearly vertical. This sharp change of direction very clearly signals the onset of clipping.
For the TDA7498, and most similar chips, this point seems to occur at somewhere in the range of 1/3rd to 1/2 of the "max power at 10% THD" stated on the datasheet. Suddenly, the "100 watt" chip turns into a 30 watt or 50 watt chip. It's only a matter of 3 - 5 dB, but less technical buyers might not understand this, and prefer the "100 watt" chip to the honestly advertised 30 watt one.
I can see why the marketing department, whose stock in trade is lies, wants to claim 100 watts instead of 30 watts. Meantime, I bet the engineering department is frustrated by this, since their hard work in achieving a 0.01% THD floor has been completely hidden behind the "10% THD" spec.
The back-of-the envelope math I use to estimate output power may be a decibel or two on the conservative side, but it is quick and easy, and if it errs, it errs on the side of honesty. I think it's reasonably real-world accurate: I'm assuming 1.5 volts lost at each end of the signal swing to device saturation and power supply droop. How much better can we really do? 1 volt? 0.5 volt? Can we really keep our guitar signal peaks accurate to within 0.5 volts to avoid hard clipping, even if the amp can do it?
One of the nice thing about these class D chips is that they have brought the watts/dollar cost of audio power amplification so low that we can ignore the false advertising, and buy as much power as we need, within reason. So we have the luxury of pretty much ignoring the advertised power rating entirely. All we need to worry about, is whether the board can handle the power supply voltage and speaker impedance that we plan to use.
-Gnobuddy
Attachments
I've always wondered why vocalists use amplification when it really isn't necessary. I attended a local play where there was a guitarist/vocalist who played acoustically inside the small theater (~ 75 people) during the show. She sounded GREAT; then there was an intermission where everyone went outside during the break, and she continued to entertain us outside in the open spaces. Maybe because she felt it was outdoors that she needed amplification, so she plugged a microphone and her guitar into a Fender amp----it was dramatically WORSE---especially her voice sounded unnatural, and it really wasn't any louder than inside. Unless a vocalist has an extremely soft voice, or is playing with a full, drummer-driven band, I think "P.A." systems make singers sound MUCH worse unless they are of extremely high quality with tons of headroom. Maybe there is some advantage to being able to electronically alter the sound---adding EQ and reverb, say, but in general I think it's a step in the wrong direction.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.