Tube Emulation & EQ

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Excellent, will try and run that through my model when I get home tonight.

Thanks for the good wishes, I do feel a bit better today.

I just use a jigsaw followed by bit of sanding.
I've done that in the past, but now I have access to the little hand-held router, so I want to see if I can do a neater job with it.

Thank you for that, that was an excellent link. And there was some stuff there I didn't know.

I think we may now know one of the reasons why adding an Lpad or two-resistor attenuator to a valve guitar amp often makes it sound worse. The Lpad or attenuator lowers the output impedance of the amp a lot!

As an example, you can make a -6 dB attenuator for an 8 ohm speaker using a 4 ohm resistor in series with an 8 ohm resistor, with an 8 ohm speaker wired in parallel with the 8 ohm resistor. Because of Thevenin's theorem, even if the guitar amp itself is acting like a 50-ohm source impedance, the guitar speaker will now see less than an 8 ohm source impedance.

And now we know that's going to effectively turn down both the bass and the treble, at the same time as it turns down the volume. And the volume reduction itself also makes your ears turn down their bass and treble response (Fletcher-Munson contours). So there's a double whammy, no wonder adding an attenuator can take a good amp and make it sound "blah"!

But there is good news too - now that we know the source of the problem, we can design our attenuators differently. We can always start with a resistor in series with the loudspeaker to product the desired amount of attenuation, and maintain at least a somewhat high source impedance. And then add an additional load resistor across the amp's output (not across the speaker) to make the total impedance work out right from the amp's point of view.

I can see why you are putting so much effort into these boombox speakers.
Yeah, "stupidly expensive" is a good description! But I've always wondered what, exactly, was so special about guitar speakers that made them sound good for guitar.

After all, the prescription for building classic guitar speakers is actually a lot like the one for building very primitive Hi-Fi speakers from fifty or sixty or seventy years ago - use large, thin, floppy, flimsy cones with ribs and seams, inadequate magnets, high bass resonance frequencies, and try to cram the entire audio frequency range through one driver.

Maybe "cone cry" and other Hi-Fi evils really are good for guitar. But maybe a big part of it is just EQ - and if that will let me use $5 boom-box speakers rather than $150 guitar speakers (which are entirely beyond my current budget), then I'll be a happy camper.

I wonder if there's anything different with class D amps.
They are evolving so fast, I don't know diddly squat about the current state of the technology. But I think they have slightly higher output impedance than older class B solid-state amps, mainly because they have two output inductors in series with the speaker (one from each half-bridge output stage) to help filter out the class D switching signal.

They are small inductors, though, only 10 - 20 uH according to a Texas Instruments white paper I read, so it is certainly possible to wind them with fat enough wire to have very low DC resistance.

Some of the audiophile guys say class D has a triode-like presentation, or something.
Some of the audiophile guys also say things like "putting spikes under my power amplifier improved pacing, rhythm, and expressiveness". Which, I'm sorry to say, is utter nonsense. :D

I do have a cheap "Lepai" class D amp from Parts Express, connected to a TV, driving some cheap bookshelf speakers, also from Parts Express.
I put together the exact same setup for sound from my wife's PC. As you say, it works well enough, and the audible flaws are all from the speakers, not the Lepai.

Which is not to say that the Lepai is perfect, but either it's flaws are inaudibly small, or they are buried under the much bigger audible flaws from the little bookshelf speakers.

The little class D module I'm using for the current project came with extravagant claims of "2x 100 W stereo" or something like that. I tracked down the chip datasheet, the chip manufacturer rates it at 50 watts RMS at maximum supply voltage (32V DC) and 10% THD. :eek: If you look at the datasheet, 10% THD is deep into clipping - the amp is actually below 0.1% THD until clipping starts in earnest.

So we've got nonsense ratings from both the chip manufacturer and the module vendor. Toss both those into the virtual trashcan.

Basic amp theory to the rescue. It's a bridge-mode output stage, assume 2 volts saturation loss across each output device, peak voltage across the speaker will be about four volts less than the DC supply voltage. So maximum clean output power is about {(Vs-4)^2}/(2xRspeaker), with a well regulated power supply. Maybe just a wee bit more.

That works out to maybe 13 - 15 watts RMS per channel with my 24V power supply and (I think) 15 or 16 ohm speakers. Thirty-odd watts is a long way from the advertised two hundred, but I'm sure it will be more than loud enough for my friend's needs.

-Gnobuddy
 
<snip>
The rising impedance of the speaker also meant that a tube amp will try to put more power into a speaker at these frequencies than a SS amp will.
This is the part I hadn't twigged to. So we might actually end up needing to put a little gentle treble boost into SS guitar amp designs, depending on the speaker used.

That doesn't seem to be the case with the pair of 6.5" boom-box woofers I'm using for this project, though, or for the stock speaker and cab from my Princeton Reverb. In both those cases, my ears were happiest with an overall gentle downward tilt across the frequency spectrum from 100 Hz to 5 kHz or so, and a bit of a notch at 800 Hz.

But I'll reserve judgment until the preamp is finalized, and I've come as close to extracting what mjd_tech calls "gling" from it as I can. Mebbe I'll end up wanting more treble again at that point, who knows.

-Gnobuddy
 
Maybe "cone cry" and other Hi-Fi evils really are good for guitar.
The catch is: which ones and when, and for what sound. I have a full ranger with a "wizzer" which is pretty bloody ugly for guitar once that one "lets go".

I need to go through my library but (not surprisingly) the speaker manufacturers themselves have done a decent job of working out a range of causes of speaker misbehaviour, and which ones are instantly audible (even if 40db down). Even if they go and ignore their own results in the pursuit of fashion (vale the Vifa P13WH-00-08).

Wolfgang Klippel (00's), Steve Temme (90's) and Ragnar Lian (80's) are names that come to mind. There's a thread called "beyond the ariel" in these forums which has a lot of information in it (Olson, Geddes etc) but is mostly focused on n'est plus ultra level hifi and room interaction issues. Klippel has, if I recall correctly, a bunch of presentations on available on line and is where I'd start to come up to speed.

Some of the audiophile guys also say things like "putting spikes under my power amplifier improved pacing, rhythm, and expressiveness". Which, I'm sorry to say, is utter nonsense. :D

Pass my wine cork! Unfortunately marketing has long since trumped everything else in audio, be it guitar or hifi. And when some iconoclast comes along and points out that the emperor has no clothes (again) it usually gets attributed to some "woo" by someone who wasn't there. So we get a new dogma.

As opposed to being something well designed sounding better than the shiny, cost-engineered crap being pushed by the magazines. (My axiom: most people's stuff sounds so bad that the first decent anything they hear becomes their religion)

Fortunately there's a bunch of people here abouts still working away at the fundamentals and trying to work out why A consistently sounds different to B.

Which is why I like this thread. Thanks!
 
Last edited:
Here's some small guitar speakers that are not stupidly expensive.
Eminence 6-1/2" Guitar Speaker 20W 4 Ohm
They're 6 1/2 inch, so not going to produce window rattling bass.
Might be worth a try if the boombox drivers don't work out.
the price is right.
Some of the customers have done some creative things with these..
Someone even measured the T/S parameters:

Re: 4.01 ohms
Fs: 108 Hz
Qts: .966
Qes: 1.04
Qms: 14.2
Le: .48 mH
Mms: 5.59 g
Vas: .378 cuft
SPL: 93 dB
 
I have a full ranger with a "wizzer" which is pretty bloody ugly for guitar once that one "lets go".
Curiously enough, I am using two different 8" speakers with whizzer cones in my little 2-watt 6AK6 amp.

One is a dirt-cheap P.A. speaker intended to be ceiling-mounted and used to pipe announcements or background Muzak into large occupied areas. It has a 130 Hz (!) resonance frequency, a very light and flimsy cone, and some 93 dB@1W sensitivity.

And it turns out, if I add some treble boost to the preamp above 1 kHz, this speaker works really well for clean tones, giving them an almost "vocal" characteristic, a bit like some very expensive carved-top archtop jazz guitars. But bass is severely lacking - not necessarily a problem for single-note lead guitar playing, but not much good for rhythm work unless there is a bass guitar also in the mix.

Because of the lack of bass, I then moved to another slightly more expensive (but still cheap) 8" full-range with a whizzer cone. This one has a (measured by me) 55 Hz fundamental resonance, lower than your typical guitar speaker. It has better bass than the dirt-cheap speaker, but none of that "vocal" character. It works decently well for clean tones, nothing special, but gets the job done.

Klippel has, if I recall correctly, a bunch of presentations available on line and is where I'd start to come up to speed.
Thanks! I'll check them out one of these days.

I'm boringly staid about "Hi-Fi" now - I do all my listening with a pair of Alesis Monitor One Mk II near-field monitors, supported by a small Velodyne subwoofer. Not perfect, of course, but these monitors were designed for the flattest and most neutral frequency response possible within the budget, and a lot of effort went into both driver design, and crossover design (I knew the designer personally).

It would take a lot of money to really improve on this setup, so it will never happen...if the Alesis monitors die, I'll be listening to thrift-store boombox refugee speakers for the rest of my days. :eek:

(My axiom: most people's stuff sounds so bad that the first decent anything they hear becomes their religion)
I remember that moment for me: I grew up listening to mono AM radio. I was maybe twelve years old, I was in a local electronics store which sold parts, and also repaired electronics. While I was at the counter paying for a couple of resistors and transistors, the repair guy brought out a boombox from the back room that he'd just fixed.

He put it on the countertop, pressed play, and an ABBA song came pouring from the cassette tape inside. I heard my first stereo audio, my first introduction to Dolby B, my first electronics with at least some bass and the power to make it audible. And I could not believe how good it sounded!

That boombox was my benchmark for great sound for a number of years after that. Eventually, on the tiniest of tiny budgets, I even managed to build my own tape deck, including designing and and populating my own Dolby B circuit board using a schematic from a library book.

Fortunately, I much prefer information to religious dogma, so I didn't stay stuck with the belief that stereo cassette tapes running at 1 7/8 inches per second were the ultimate audio source. :D That cassette boombox would sound pretty mid-fi to me today!

Fortunately there's a bunch of people here abouts still working away at the fundamentals and trying to work out why A consistently sounds different to B.
I think 99% of the real research into Hi-Fi has already been done, decades ago. IMO there is not much left to be learned there, everything in the contemporary audio chain is virtually perfect (defects are too low to be audible), except for the biggest thorn in the side of everyone with good ears: the darn loudspeakers, and the darn listening-room!

But guitar amps? If there is real research out there, most of it must be locked in the vaults of the big music-electronics companies. I bet Yamaha knows a lot about how to make 4" speakers sound like a real guitar amp - but they're not about to tell their competition what went into their THR amps.

So when it comes to guitar amps, I think there is still undiscovered, or at least, un-disseminated, stuff for us to dig up, one nugget at a time...

Which is why I like this thread. Thanks!
Thank you! This thread was a lonely place for a while, and I was beginning to get tired of reading my own posts...and then Printer2, PRR, mjd_tech, you, a few others, all jumped in.

I am very grateful for all of your contributions here, whether you're sharing ideas and information for us all to ponder, or asking me probing questions that keep me honest as this project progresses.

Hopefully we will all end up learning a few things we didn't know, and my friend will end up with an amp he likes. (Fingers crossed!)

-Gnobuddy
 
Here's some small guitar speakers that are not stupidly expensive.
Eminence 6-1/2" Guitar Speaker 20W 4 Ohm
They're 6 1/2 inch, so not going to produce window rattling bass.
Might be worth a try if the boombox drivers don't work out.
the price is right.
Thank you, they do look interesting! Mebbe for a later project, time is starting to run out rapidly, and sometimes there can be a week or two delay getting things from the USA across the two-inch-wide white painted line at the border with Canada. Particularly as Christmas gets closer, and lots of people buy lots of stuff they don't need, won't use, and can't afford...

-Gnobuddy
 
I made a simple little plywood jig tonight that lets me cut circles with a little palm-size router and a 1/4" two-flute router bit. After that there was just enough time to make a test cutout in a piece of scrap ply before I had to clean up and vacate the wood-shop.

I'm happy with the test cutout - it fits the (front-mounting) speaker well, and the router produced a pretty clean cut in the ply. I couldn't do as good a job with a jig-saw, at least, not without lots of sanding after the fact.

I figured it's about time this thread had a few more pictures in it, so here we go. The woofer (6.5" outer diameter), complete with a field-canceling magnet for use near a CRT TV or computer monitor, the test cutout, the speaker placed into the test cutout.

-Gnobuddy
 

Attachments

  • woofer_1000px.jpg
    woofer_1000px.jpg
    440.2 KB · Views: 165
  • woofer_side_1000px.jpg
    woofer_side_1000px.jpg
    353.3 KB · Views: 160
  • test_cutout_1000px.jpg
    test_cutout_1000px.jpg
    229.2 KB · Views: 136
  • speaker_in_test_cutout_1000px.jpg
    speaker_in_test_cutout_1000px.jpg
    390.7 KB · Views: 138
I spent three hours in the wood-shop after I got home from work tonight, and have the plywood baffle mostly done, though I still have to cut two 3" diameter holes for the port tubes. I also bought a few bits and pieces (screws, a drawer-pull to serve as a handle, stuff like that).

I still haven't had time to test the circuit idea I came up with a few days ago. I'm slow at wood-working, and building this cab is taking a lot of time. Hopefully I will start gluing stuff together this weekend.

It's been a long day, and I'm beat, so it's off to bed.

-Gnobuddy
 
I spent a few more hours in the wood-shop today, and am still not done. But at least I finally got to the point where the first glue joints are setting as I type this. By tomorrow morning, I should have something that's beginning to resemble a cab, rather than a pile of lumber.

On the electronics front - I've spent all my free time today in the wood-shop, and it's now well afte 1 AM, so once again, no time to build and test my mystery circuit. But, as a teaser, I'm going to attach an LTSpice simulated set of output waveforms.

I have high hopes for this circuit, assuming it behaves at least somewhat like the simulation. Notice how it is quite linear (symmetrical output waveform) at low input levels (dark blue, dark green), begins to gently round off negative half cycles beyond that (orange, cyan, magenta), and finally starts to soft-clip the negative cycles (grey, dark green) when the input is turned up even more.

But even when the negative half cycles are clipped flat on their peaks, the corners of the waveform remain rounded. This should significantly reduce the presence of harsh-sounding very high order harmonics. If the real circuit behaves like the simulation, there shouldn't be much audible harshness to deal with during overdrive.

FFTs of the simulated waveforms are interesting, too - only the second and third harmonics make it to within 30 dB of the fundamental, i.e. only these are likely to be audible with guitar.

Will this circuit turn out to have what mjd_tech calls "gling" and most of us call "tubeyness" or "valveyness"? I'm hoping. Mebbe I'll find out tomorrow, if the curse of the wood-shop doesn't strike again!

This isn't KMG's diode-clipped MOSFET circuit, by the way, but something different that I came up with.

-Gnobuddy
 
Time has been tight of late, but finally managed to get the 3" circular holes for the port tubes cut, using a borrowed hole-saw and the drill press in our apartment complex's wood-shop.

It looks like there is no longer much interest in this thread, so I probably won't post too frequently here for the immediate future, though I will certainly update things any time I manage to make significant progress.

-Gnobuddy
 
Still interested, free time consumed with holiday season activities, shoveling snow, etc.
Glad to hear you're still interested, and commiserate with the seasonal busyness.

Things are getting busy here too, though we haven't had much snow to deal with so far (except for one freakishly early snowfall in the first few days of November).

There will be some progress on the project this weekend, if at all possible; I breadboarded the mystery circuit last night, so the next step is to find a little time to hook up power, scope, sig-gen/guitar, and try it out.

Meantime, here is an interesting item I found at a local Home Depot (pic attached): a satin-finished rectangle of aluminum sheet, sold as a door kick-panel, but ideal for making front panels for home electronics projects.

These come in 6" and 8" widths, and are 32" long, so one can expect to get a number of front panels out of them unless you're building enormously large electronic projects. That might make the $25 (Canadian dollars) price more palatable, particularly when you consider that these are already satin-finished.

Years ago I used to sandpaper sheet aluminum by hand, and then etch it with lye (sodium hydroxide, NaOH), to make my own satin-finished front and back panels. That was before we had poisoned the planet to the point where we finally began to realize that using chemicals like lye without the proper toxic waste disposal facilities wasn't a great idea.

-Gnobuddy
 

Attachments

  • satin_finish_aluminium_panel_Home_Depot.png
    satin_finish_aluminium_panel_Home_Depot.png
    72.4 KB · Views: 128

PRR

Member
Joined 2003
Paid Member
...an interesting item I found at a local Home Depot (pic attached): a satin-finished rectangle of aluminum sheet, sold as a door kick-panel...

You can also get them in "brass tone".

Real Stainless Steel is widely used for commercial kick-plates. Not sold at (my) H-D (tho they probably use them on the restroom doors), cost a lot more than Aluminum, and the devil to drill.
 
Last edited:
no longer much interest in this thread........season activities, shoveling snow, etc.

I have read every post, just don't have much to contribute, yet. I have this stupid idea, and the key elements to make a STUPIDLY powerful class D guitar amp. Maybe you guys will have the de-suckification aspects figured out before I work through the half finished projects on my work bench.....

Of course it is the holiday season, and with 4 grandkids to shop for ...... Then there is the matter of the white stuff falling from the sky.

There is a TI 600 watt class D amp, the necessary 48 volt power supply to run it, and a TI guitar preamp EVB all waiting for me to hook it up and make some LOUD noise. Yes, TI made a guitar preamp EVB that plugs onto several of their high powered class D EVB's. No word on how it sounds yet. It's still in the box.

Details are buried somewhere in this thread.

TI Class D EVM Board 50% Promotion

TI's page:

SIDEGIG-GUITAREVM Guitar SideGig Plug-In Module | TI.com
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.