It was another feverish day in bed for me, but I began to feel a bit better by 8 PM, so in the last couple of hours I cooked dinner, then got the 2nd speaker test-fitted in the cab, pushed the back panel into place, and dragged the new cab over to the workbench.
The "workbench" is actually one third of the MDF top of an old pool table, slapped onto a frame I built from cast-off 2x4 lumber I found at a construction site a few years ago. But that's another story!
I hooked up the new cab to the class D power amp module, and tried various combinations of electronics between guitar and power amp - nothing at all, then just the single JFET preamp stage (with treble boost above 500 Hz), then the same JFET preamp with my "de-nastifying filter". I only drove one of the two 6.5" speakers from the power amp, because I was too lazy to wire up the second one. (The second channel of the class D module is feeding a dummy load for the moment.)
Conclusions:
1) The new cab (ported, and designed to internal dimensions close to the Golden Ratio) does sound better than the old Sears enclosure did, with less "nasty" to the sound. Bass is still a bit thin, though. Overall, it does not sound great with a solid-body guitar driving it.
2) Adding the single JFET pre does make the thing sound a little better.
3) Adding the "bright cap" across the source resistor for treble boost above 500 Hz, adds some nice glassy edge / bite to the high treble, certainly usable for some styles of playing, but perhaps a bit too much.
I may dial back the bright cap to half it's current value to only boost treble above 1 kHz.
4) Adding the "de-nastifying filter" made a very nice improvement, filling out the bass, removing some harshness, and still allowing plenty of treble "bite" when the bright cap was in-circuit. Now it actually sounds good with a solid-body guitar.
My overall impression is one of relief - this is going to work! All the work building the new cab did improve the sound, and also substantially improved the appearance. The pine I used also made the new cab surprisingly light for its size, considerably lighter than the old MDF Sear's cab I had been using.
But most importantly, my ears didn't lie to me the first time around - the "de-nastifying filter" and JFET preamp really do improve the sound.
Don't misunderstand me, this is definitely not the best guitar amp I've ever heard. But it is, IMO, a decent guitar amp, and much better than the one my friend is currently using.
There is still a fair bit to do - remove the speakers and finish the cab, build a tone control, make a control panel, mount all the electronics in the back of the cab, make a cosmetic front grille, make a cosmetic outer back panel, add a handle and feet.
But I'm going to feel a lot better about doing all the remaining work now. I no longer feel like this is an unfinished R&D project with an unknown probability of success. It's going to work!
I almost forgot, I measured the free-air resonance of one of the 6.5" drivers a few days ago: 103 Hz. Which is about where a lot of "real" electric guitar speakers have their fundamental resonance. It's 2017, it is no longer necessary to have a 12" speaker just to get usable bass response down to a guitar's low "E" at 83 Hz!
-Gnobuddy
The "workbench" is actually one third of the MDF top of an old pool table, slapped onto a frame I built from cast-off 2x4 lumber I found at a construction site a few years ago. But that's another story!
I hooked up the new cab to the class D power amp module, and tried various combinations of electronics between guitar and power amp - nothing at all, then just the single JFET preamp stage (with treble boost above 500 Hz), then the same JFET preamp with my "de-nastifying filter". I only drove one of the two 6.5" speakers from the power amp, because I was too lazy to wire up the second one. (The second channel of the class D module is feeding a dummy load for the moment.)
Conclusions:
1) The new cab (ported, and designed to internal dimensions close to the Golden Ratio) does sound better than the old Sears enclosure did, with less "nasty" to the sound. Bass is still a bit thin, though. Overall, it does not sound great with a solid-body guitar driving it.
2) Adding the single JFET pre does make the thing sound a little better.
3) Adding the "bright cap" across the source resistor for treble boost above 500 Hz, adds some nice glassy edge / bite to the high treble, certainly usable for some styles of playing, but perhaps a bit too much.
I may dial back the bright cap to half it's current value to only boost treble above 1 kHz.
4) Adding the "de-nastifying filter" made a very nice improvement, filling out the bass, removing some harshness, and still allowing plenty of treble "bite" when the bright cap was in-circuit. Now it actually sounds good with a solid-body guitar.
My overall impression is one of relief - this is going to work! All the work building the new cab did improve the sound, and also substantially improved the appearance. The pine I used also made the new cab surprisingly light for its size, considerably lighter than the old MDF Sear's cab I had been using.
But most importantly, my ears didn't lie to me the first time around - the "de-nastifying filter" and JFET preamp really do improve the sound.
Don't misunderstand me, this is definitely not the best guitar amp I've ever heard. But it is, IMO, a decent guitar amp, and much better than the one my friend is currently using.
There is still a fair bit to do - remove the speakers and finish the cab, build a tone control, make a control panel, mount all the electronics in the back of the cab, make a cosmetic front grille, make a cosmetic outer back panel, add a handle and feet.
But I'm going to feel a lot better about doing all the remaining work now. I no longer feel like this is an unfinished R&D project with an unknown probability of success. It's going to work!
I almost forgot, I measured the free-air resonance of one of the 6.5" drivers a few days ago: 103 Hz. Which is about where a lot of "real" electric guitar speakers have their fundamental resonance. It's 2017, it is no longer necessary to have a 12" speaker just to get usable bass response down to a guitar's low "E" at 83 Hz!
-Gnobuddy
... It's 2017, it is no longer necessary to have a 12" speaker just to get usable bass response down to a guitar's low "E" at 83 Hz!
The laws of physics have not been repealed. In acoustics, they really work against us.
A Twelve can give 4X the efficiency of a Six (or 2X the bass extension, which we do not need here).
If you do not need large acoustic output, this may not matter. Of course we could always get response below 100CPS even with a 2-inch driver, just very weak.
I also think a "guitar" voice should be about as large as a guitar. You can get the same light energy with a small point source or a large diffuser, but they give very different light quality on the subject.
Here's something I got out of reading through this thread.
A number of distinctive tube amplifier characteristics are discussed as candidates for why we like the sound of tube amplifiers. I think all of these are potential techniques for good tube emulation.
1. 3/2 power law (or 2 power law) and unique small signal harmonic distortion
2. graceful limiting of large peak input voltages and grid current load on pickups (clean tone)
3. soft/asymmetric limiting behavior on overdrive (distortion tone)
4. ability to handle large output peak levels gracefully and with pleasant cues (loud tone)
5. sag and other dynamic effects that act as a sort of compressor
6. interaction between output resistance and loudspeaker system, damping factor
Not sure whether to include line frequency IMD products...
PS obligatory 5E3 background:
I bought a 5E3 when I was in 3rd year of high school, for $50. I then got the Jack Darr book and after a few weeks of study, I hacked into my Tweed Deluxe and cascaded the 12AY7 into 2 gain stages, added a TMB tone stack and master volume (I think I used the 3 holes that remained after removing the mic input stuff...) which resulted in the best tone I have heard in a small amp. I may have replaced the 12AY7 with a 12AX7 at some point but I don't recall it making a huge difference. When I left town I left the amp with my band buddy and always missed it.
A number of distinctive tube amplifier characteristics are discussed as candidates for why we like the sound of tube amplifiers. I think all of these are potential techniques for good tube emulation.
1. 3/2 power law (or 2 power law) and unique small signal harmonic distortion
2. graceful limiting of large peak input voltages and grid current load on pickups (clean tone)
3. soft/asymmetric limiting behavior on overdrive (distortion tone)
4. ability to handle large output peak levels gracefully and with pleasant cues (loud tone)
5. sag and other dynamic effects that act as a sort of compressor
6. interaction between output resistance and loudspeaker system, damping factor
Not sure whether to include line frequency IMD products...
PS obligatory 5E3 background:
I bought a 5E3 when I was in 3rd year of high school, for $50. I then got the Jack Darr book and after a few weeks of study, I hacked into my Tweed Deluxe and cascaded the 12AY7 into 2 gain stages, added a TMB tone stack and master volume (I think I used the 3 holes that remained after removing the mic input stuff...) which resulted in the best tone I have heard in a small amp. I may have replaced the 12AY7 with a 12AX7 at some point but I don't recall it making a huge difference. When I left town I left the amp with my band buddy and always missed it.
Last edited:
This is the point I was making. In Hi-Fi, for a long time, we've had bookshelf speakers that produce all the SPL that is needed for listening to music in a living room. Meanwhile, in the world of electric guitar, there is a widespread belief that anything less than a twelve inch speaker in a cabinet the size of a dorm-room 'fridge is incapable of handling a guitar's bandwidth.
There is a lot of ground between 2" and 12" speakers, particularly when you consider that cone area depends on the square of the diameter.
For the millions of guitar players and would-be guitar players living in apartments, MO the noise level of a household vacuum cleaner is a good benchmark for the acceptable upper SPL limit. Louder than that, and one is being a bad neighbour.
Vacuum-cleaner SPL is somewhere in the neighbourhood of 70 dB, and a pair of 6.5" speakers will very easily exceed that by a considerable margin over the full guitar bandwidth, in a reasonable sized living-room.
Closer to the 2" you mentioned, Yamaha went with 3" speakers in their THR amps. And those amps were quite a sales success, tapping into a market that other guitar amp manufacturers had not really entered.
Now we have interesting new products like the Boss Katana Mini, which uses a 4" speaker. Here is a rather unusual review by the talented Mary Spender: YouTube
Big speakers can have big impact, sure. But are not appropriate or usable in every situation. The world has changed since 1960, and opportunities for live guitar at 120 dB SPL from refrigerator- sized speakers are now few and far between (and lead to tinnitis, hearing loss, and a lonely, deaf, old age.)
It's a different topic for a different day, but I often think our lives have shrunk in many ways over the last few decades.
-Gnobuddy
Thanks for summarizing stuff in one place, great idea!
All this reminds me of the food industries attempts to make "fake butter". There are over a dozen different kinds of "fake butter" on the typical supermarket's shelves, none of which manage to taste very much like the real thing - even though people with white lab coats probably tried hard to figure out what makes real butter "buttery", and then to inject those qualities into their fake butter product.
The amp I'm building is in the "fake butter" category. It definitely has some butteriness to it, but will not be mistaken for the real thing by anyone with normally functioning taste buds!
I think KMG's fake butter may be the best in the supermarket. It comes very close to the real thing.
-Gnobuddy
When I lived in the Los Angeles area, I had a friend who had fallen on hard economic times after the financial collapse of 2007. He lived in a truly scary part of L.A. - South Central, the area made infamous by the riots that started after the police brutality inflicted on Rodney King. Not surprisingly, South Central L.A. is the cheapest place in Los Angeles when it comes to renting a place to live.
Violent police were not the scariest thing in my friend's neighbourhood. The scariest thing was seeing a little group of young boys coming by on their bicycles. If you saw that, you ran and hid.
The boys were sentinels and watchdogs for adult gang members and drug dealers. Too young for the police to arrest, shoot, or beat up, the children were sent ahead to report back if the coast was clear for the real criminals to show up and do their thing.
-Gnobuddy
Back on topic, I just spent a couple of hours in the wood shop, first sanding, and then applying clear polyurethane varnish to the speaker cab.
I also applied the clear poly to a test scrap of wood. Tomorrow cab and test scrap will both get a second coat of clear. After that I'll try applying the tinted (black) poly varnish to the test scrap to see how it turns out.
If I could, I would now be cutting aluminium sheet for the control panel. But I have to get to the supermarket, bring back some groceries, and cook us some dinner.
-Gnobuddy
I also applied the clear poly to a test scrap of wood. Tomorrow cab and test scrap will both get a second coat of clear. After that I'll try applying the tinted (black) poly varnish to the test scrap to see how it turns out.
If I could, I would now be cutting aluminium sheet for the control panel. But I have to get to the supermarket, bring back some groceries, and cook us some dinner.
-Gnobuddy
This may be an appropriate time to show a circuit that was originally posted by Nigel Wright a few times here on diyAudio (for example, post #2 of this thread: Design by ear! )
I drew up the circuit in LTSpice and tinkered with it a little, in particular, replacing the hard-to-obtain 2.2 Meg pot with a 1 Meg pot.
The circuit diagram and simulated waveforms are in the attached image. The circuit produces symmetrical clipping (like a long-tailed-pair phase splitter or push-pull output stage), but soft and progressive, depending on the setting of the drive pot and the input signal level.
I have not built or heard the circuit in action, but these waveforms look much more interesting to me than the buzz-saw nastiness generated by the widely used pair of back-to-back diodes in an opamp feedback loop.
-Gnobuddy
Attachments
I had a brief lightbulb moment while responding to a different thread, and I think it connects to something we have discussed here a few posts ago.
Piezo undersaddle pickups on (electro-)acoustic guitars have a reputation for producing a harsh sound, sometimes called "piezo quack". Piezoceramic elements also produce startlingly high brief peak voltages in response to mechanical transients.
(Many years ago, my mom had a piezoelectric lighter for her kitchen stove; the piezo was struck by an internal spring-loaded weight when you pulled the trigger, and in response, it generated sufficient voltage to make a big fat spark at the time, lighting the gas stove. Keep in mind the dielectric breakdown strength of dry air is between 20,000 and 33,000 volts per centimeter, depending on electrode shape!)
Back on guitar piezo pickups, I have one guitar which has an actual 12AU7 vacuum triode in the piezo preamp (it's a Takamine Cool Tube preamp). The manufacturer claims the tube was used because it reduces piezo harshness in the guitars sound. Most listeners with good ears agree.
I've wondered for a couple of years what that 12AU7 actually does to reduce the piezo harshness. I think I just figured it out - it is exactly the thing we talked about earlier on this thread, when we discussed how grid current flow and/or nonlinearity in valves can squash harsh transients in the electric guitar signal.
The Cool Tube preamp runs the 12AU7 on 9 volts DC. IIRC, grid current flow is higher at low anode voltages. Ideal for softening brief transient voltage peaks from the piezo element.
-Gnobuddy
Piezo undersaddle pickups on (electro-)acoustic guitars have a reputation for producing a harsh sound, sometimes called "piezo quack". Piezoceramic elements also produce startlingly high brief peak voltages in response to mechanical transients.
(Many years ago, my mom had a piezoelectric lighter for her kitchen stove; the piezo was struck by an internal spring-loaded weight when you pulled the trigger, and in response, it generated sufficient voltage to make a big fat spark at the time, lighting the gas stove. Keep in mind the dielectric breakdown strength of dry air is between 20,000 and 33,000 volts per centimeter, depending on electrode shape!)
Back on guitar piezo pickups, I have one guitar which has an actual 12AU7 vacuum triode in the piezo preamp (it's a Takamine Cool Tube preamp). The manufacturer claims the tube was used because it reduces piezo harshness in the guitars sound. Most listeners with good ears agree.
I've wondered for a couple of years what that 12AU7 actually does to reduce the piezo harshness. I think I just figured it out - it is exactly the thing we talked about earlier on this thread, when we discussed how grid current flow and/or nonlinearity in valves can squash harsh transients in the electric guitar signal.
The Cool Tube preamp runs the 12AU7 on 9 volts DC. IIRC, grid current flow is higher at low anode voltages. Ideal for softening brief transient voltage peaks from the piezo element.
-Gnobuddy
In that case a 15-inch driver may be appropriate
I don't have those kids on bikes.
The near neighbor hammers on Hudsons and grinds roof sheets. George keeps setting fires in junk cars and the fire dept sirens come. Then the neighbors a *mile* down through the woods throw LOUD parties.
I've thought of, not only the Fifteen with a PV 800 amp, but a 4-foot directional horn, to blare-back at the party animals.
Yes, between excitements, it is so quiet here I could use a Two with a milliWatt amp to amuse myself.
Elektor magazine once published a circuit that was supposed to replace a starter's pistol (for sporting events). It used a heavy-duty woofer to make the sound, most likely by charging up a capacitor and discharging it through the speaker.In that case a 15-inch driver may be appropriate
-Gnobuddy
Wowsers.I did not expect that much from humbuckers.
This was using more-or-less conventional playing techniques?
If humbuckers can produce 10 volts peak-to-peak, we can probably extrapolate that single coils can put out maximum peak to peak voltages in the neighbourhood of 5 volts, probably a bit more, as most single coils have more winds than half a humbucker.
If true, this means transients from both humbuckers and single-coil pickups are capable of swinging a 12AX7 triode all the way from full cutoff to deep into grid current flow.
I once saw an oscilloscope screen capture on the 'Net that claimed to show 6 volts peak to peak from a single-coil pickup. At the time I figured that was too high to be correct, and that the person who took that measurement had probably used a 1x oscilloscope probe but set the 'scope for a 10x one. Or simply mis-read the volts/div knob.
-Gnobuddy
I measured some Fender Prodigy single coils as putting out slightly over 3 volts peak to peak when the strings were strummed hard all at once.
A young kid sitting on the transformer box at the entrance of the neighborhood with a boom box was the sentinel for the gang that was burglarizing houses. He would have a walkie talkie to alert the thieves when the cops were coming. I pointed this out to the cop who lived down the block after his house got robbed. The kid's older brother was one of the thieves.....all were underage so none got jail time and went back into business in a different hood.
In the mid 70's I worked the evening shift and got off work somewhere between midnight and 4 AM. I slept until 10 or 11 AM, except for when the disco queen that lived in the apartment next door started blasting disco at 9 AM. Polite diplomacy failed, so I told her that I had nuclear audio capability. DIY guitar amp (solid state and nasty when pushed), two 4 X 12 inch guitar cabinets. One afternoon before leaving for work I pushed both cabinets up against my living room wall facing her bedroom. Than night at about 2 AM the entire complex got to hear Smoke on the Water at earthquake volume levels......truce negotiated, never broken!
During the Hundred Buck Amp Challenge I wired a scope across every guitar I own and wailed on them pretty hard (humbuckers and single coils). The most I saw was about 3 volts peak to peak. The loudest guitar I have is an old Guild with DiMarzio humbuchers. I captured the initial peak of a single note here, about 2.8 volts P-P.
I am looking at tiny sub miniature tubes to go inside the guitar for GraphTech piezo saddles.
Attachments
The humbucking pickup measurement was done playing a late model LP Custom with pretty hot (~8K-12K) pickups into a 10M scope impedance, captured with a stored trace.
The piezo measurement is from a single graph-tech "FAAS" bass pickup on a nylon harp string about .050 diameter and 24 inches length. This was also measured with 10M load and I'd estimate under 10pF capacitance.
I would characterize 10V as the not-ever-exceeded- but repeatable- value. At the same time, I would agree that designing for 3V p-p would be a reasonable assumption for most cases.
It was surprising to me at the time, but since then I have studied more about the dynamic range of music; I don't think it's so unusual.
I have seen some large swings out of vintage tube microphones also, which explains the physical pad on the U67, and why the overall gain structure of these mics might be a good approach.
The design decisions this informs for me is that virtual ground summing amp topologies are a good choice for gear that handles "low level" signals.
I think it also explains why soft limiting and asymmetric limiting may result in a more pleasing sound.
The piezo measurement is from a single graph-tech "FAAS" bass pickup on a nylon harp string about .050 diameter and 24 inches length. This was also measured with 10M load and I'd estimate under 10pF capacitance.
I would characterize 10V as the not-ever-exceeded- but repeatable- value. At the same time, I would agree that designing for 3V p-p would be a reasonable assumption for most cases.
It was surprising to me at the time, but since then I have studied more about the dynamic range of music; I don't think it's so unusual.
I have seen some large swings out of vintage tube microphones also, which explains the physical pad on the U67, and why the overall gain structure of these mics might be a good approach.
The design decisions this informs for me is that virtual ground summing amp topologies are a good choice for gear that handles "low level" signals.
I think it also explains why soft limiting and asymmetric limiting may result in a more pleasing sound.
Scale seems to show 0.14V peak to peak. A 10:1 probe would be wise and make that 1.4Vpp. Did you have a 20:1 probe or other attenuation?
I'm 90% sure that guitar very-very rarely deliver more than 0.5V-1.0V peak (1V-2V peak-peak), so they fit in the grid of a 12AX7 without distress. (And that players who do make more "want" distressed sound.)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.