Kind of makes you wonder if this small output amp could be used as a "lead guitar waveform generator" with its OPT driving a second set of bigger tubes, with bigger power supply and bigger OPT. For bedroom use, you plug the speaker into this amp sitting on top of the cab as a head. For club use, you turn on the chassis in the bottom of the cabinet, plug the speaker into that.
One would think for the additional oomph, a more simple class D downstairs would do. I dunno... Seems like it wouldnt be quite the same with all the speaker OPT tube interaction going on. The class D being more like a bipolar power supply that happens to be able to track an audio signal.
I wonder what you’d come up with if you worked the tubes from the other end - used a pair of big-ish 20-24 watt sweeps instead of wussy 7 or 9-pins, and run them at similar voltages and currents to make 40 or so watts (or 60 to 80 as square waves). You set the controls for the heart of the sun and nothing bad happens.
Back in the mid 90's I did something very similar. I made a little amp with a pair of 6AQ5's that made about 5 watts. I ran that amp into a small speaker box with a 5 inch or so speaker. It did a good job of making the sound that I wanted at the time. I then fed its output into an early 70's style totem pole transistor (2 X 2N3055) amp that used a DIY driver transformer. Tapping the output of the tube amp WITH the speaker cabinet connected did a good job of making a louder version of the tube amp. There was not enough gain in the solid state booster to cause feedback unless I set its speaker cabinet on top of the 4 X 12 inch cabinet that was connected to the booster output.
Replacing the speaker on the tube amp with a resistor definitely changed the character of the sound output. It did not suck though until you pushed the booster itself into clipping.
In a different experiment I used as cheap car stereo speaker on the tube amp and ran it into the booster as before. Again, there was a noticeable difference between the real guitar speaker and the car stereo speaker at the booster output. A couple of good guitar players were present for some of these experiments. One of them could tell whether the real guitar speaker (in a pine wood cabinet) was placed on the ground face down or face up. Clearly there is some interaction between the speaker and the amp that gets to a booster wired directly across the speaker.
In the last experiment, I cut the cone away from the voice coil on the car speaker, so the amp was driving a voice coil that was free to move without rubbing. Everyone involved agreed that the plain resistor sounded better. Without the cone impeding the movement that voice coil was exceeding x-max and it did not live long.
"I wonder what you’d come up with if you worked the tubes from the other end - used a pair of big-ish 20-24 watt sweeps instead of wussy 7 or 9-pins"
Been there, made that, used TWO pair....the easy way. It was a simple build that took an hour and lasted for a couple years. The little 7 and 9 pin amps were made for minimum cost. The guitar amp I used to annoy my neighbors the most.......Pete Milletts Engineers Amp, Tubelab modified 6HJ5 version that made 125 WPC. Connect a 4 X 12 cabinet on each channel parallel the inputs and drive it all with an ADA MP1 MIDI controlled guitar preamp. I had several presets stored in that thing that ranged from mellow to full metal racket. My favorite for neighbor busting was simply called Jimi. That old MP1 is still around here somewhere. It got too microphonic and intermittent to be used reliably. Near the end it would lock up and require a good whack with the butt end of a screwdriver to make it come back to life. Think two 12AX7's controlled by a primitive microprocessor and far too many parts. Somewhere on my computer is the schematics for it too.
The MP1 can be seen here (rack mount unit on left workbench) driving an 833A tube into the red glow zone on 1500 volts. This was another very loud amp. Thint a 200 watt Fender Champ feeding the same pair of 99 db EVM-10s that I am using now with 40 watts.
Attachments
Thanks for sparking up an old memory. I now have the MP-1 opened up on my bench. Thie relic dates from the early 90's since the youngest chip in the thing wears an early 1992 date code. When it worked it was quite versatile and could do anything from mellow to metal and change personalities at the stomp of a pedal. There are a pair of 12AX7's as well as one each NE570 and NE572 chips. Both are "compandor" chips that can be configured for dynamic range compression or expansion. The whole right side of the unit is an 8 bit "computer" that runs a Zilog Z-80 CPU chip.
I'll know more about this thing after I study the schematics a bit and maybe plug it in to see what happens. I like the idea of a digitally controlled tube amp, especially for saving settings and recalling them. This setup is far too primitive though.
I'll know more about this thing after I study the schematics a bit and maybe plug it in to see what happens. I like the idea of a digitally controlled tube amp, especially for saving settings and recalling them. This setup is far too primitive though.
Attachments
At least it's through hole. That's a nice blast shield you have on your bench; has it saved you from flying bits of fire?
Part blast shield, part stupid guitar player death prevention shield. Fortunately, it was never actually needed. That picture was taken in my Florida lab about 20 years ago. The lab / listening space / computer room / recording studio was a small 10 foot by 10 foot kid's bedroom. Two walls had 30 inch workbenches from the visible corner to a 19 inch equipment rack at each end. Another wall was completely covered with records, CDs and those Akro Mills blue parts boxes full of parts. That left 40 to 50 square foot of floor space in the middle.
The totally unsafe "don't try this at home kids" test amp on the right side bench at the time the picture was taken was assembled from a 1500 Volt 1/2 Amp power supply from an old tube type RF transmitter (large rack mount unit stood on its end on the left behind the yellow meter), my 845SE HiFi amp being used to drive the 833A tube which is the glowing object on the right in its a custom socket created with Vice Grips and hose clamps, a big capacitor taken from a cardiac defibrillator, an analog current meter, and an OPT, all live with lethal amounts of electricity. The plate voltage / current flows from the power supply to the real "death cap" through the yellow clip lead. There are ample opportunities for stuff to go really wrong here, so whenever this thing was live there was 1/4 inch of Lexan between me and it. My real fear was that I could accidentally touch the head of the guitar to something live on the bench while my mind was occupied with playing the guitar or poking buttons on the MP-1. This amp was built for testing a custom Transcendar OPT that Gerry made just for it. The amp only existed for about a week back in 2003. Note the modern TV set and computer screen.
There was a guy that lived down the street who had a pirate radio station. He broadcast music, talk, and other programming from his garage. Two nights a week he had live guitar lessons taught by a real music teacher with the sheet music and tabs posted on their web site. For both "lesson" nights that this creation existed, I opened the window behind the amp and placed the speaker on the shelf facing out. They could hear me playing along inside their garage studio.
The totally unsafe "don't try this at home kids" test amp on the right side bench at the time the picture was taken was assembled from a 1500 Volt 1/2 Amp power supply from an old tube type RF transmitter (large rack mount unit stood on its end on the left behind the yellow meter), my 845SE HiFi amp being used to drive the 833A tube which is the glowing object on the right in its a custom socket created with Vice Grips and hose clamps, a big capacitor taken from a cardiac defibrillator, an analog current meter, and an OPT, all live with lethal amounts of electricity. The plate voltage / current flows from the power supply to the real "death cap" through the yellow clip lead. There are ample opportunities for stuff to go really wrong here, so whenever this thing was live there was 1/4 inch of Lexan between me and it. My real fear was that I could accidentally touch the head of the guitar to something live on the bench while my mind was occupied with playing the guitar or poking buttons on the MP-1. This amp was built for testing a custom Transcendar OPT that Gerry made just for it. The amp only existed for about a week back in 2003. Note the modern TV set and computer screen.
There was a guy that lived down the street who had a pirate radio station. He broadcast music, talk, and other programming from his garage. Two nights a week he had live guitar lessons taught by a real music teacher with the sheet music and tabs posted on their web site. For both "lesson" nights that this creation existed, I opened the window behind the amp and placed the speaker on the shelf facing out. They could hear me playing along inside their garage studio.
I have started digging into how the MP-1 works, mostly via Google at the moment. I have not yet plugged it in. I got mine used on Ebay pretty cheap sometime in the 1990s. Now they seem to be going for between $450 and $750 in working condition, so I don't want to destroy it. My random collection of schematics seems contradictory and doesn't make sense sometimes. It appears that this machine kept evolving as time passed and semiconductors came and went. My unit is from 1992 according to the dates on the chips, but the PC board's copyright date is February 1988. There are at least two distinct versions of the MP-1 and a MP-2.
The ADA Signal Processors company has been gone for a long time. There are lots of conflicting stories on the web about when and why they ceased operations. The most credible story from a former employee has the shut down date in late 1997. The large lot of guitar amp OPT's that I have were to be delivered to "ADA Signal Processors" at their Oakland address. The shipping boxes have the number 9016 on them which I assume is the date code. At least 1000 of these 6600 ohm transformers hit the surplus market in the 90's, and there were also some 4300 ohm versions surplussed at the same time. There was a fire that destroyed the ADA warehouse. The actual date of the fire and extent of damage varies from report to report.
This is what I have learned about this thing so far:
The guitar signal first passes through a solid state J-fet and opamp based gain and "overdrive" stage before making its way to the vacuum tube board. There are 4 12AX7 gain stages on the board with j-fet switches on the cathode bypass caps in the first two stages. A tap after these two stages goes off board to some solid state circuitry that includes one of the compander chips and half of a stereo digtal volume control chip The other two gain stages (4 total) go directly to the other half of the volume control chip. Both outputs from the stereo chip are summed to mono where they are routed to the EQ section.
The EQ section has four independent opamp filters with digipots in the feedback loop. Hand written notes on the schematics put the filter frequencies at 160 Hz, 600 Hz, 1600 Hz, and 3000 Hz. All filter outputs are summed together. There is a path through an external FX loop, and an internal chorus circuit that can be switched, summed, or bypassed to one or both (stereo) outputs. The chorus circuit has the second compander chip wrapped around a BBD chip which is clocked by the VCO in a 4046 chip which gets its control voltage from an 8 bit DAC made from resistors and a latch chip. Hand written notes put the clock frequency from 30 KHz to 270 KHz.
Since I have no intentions of cloning any part of this, I'm not digging much deeper into it, except for what it may take to fix it. There is plenty more information here if interested:
https://adadepot.com/index.php?topic=148.0
The ADA Signal Processors company has been gone for a long time. There are lots of conflicting stories on the web about when and why they ceased operations. The most credible story from a former employee has the shut down date in late 1997. The large lot of guitar amp OPT's that I have were to be delivered to "ADA Signal Processors" at their Oakland address. The shipping boxes have the number 9016 on them which I assume is the date code. At least 1000 of these 6600 ohm transformers hit the surplus market in the 90's, and there were also some 4300 ohm versions surplussed at the same time. There was a fire that destroyed the ADA warehouse. The actual date of the fire and extent of damage varies from report to report.
This is what I have learned about this thing so far:
The guitar signal first passes through a solid state J-fet and opamp based gain and "overdrive" stage before making its way to the vacuum tube board. There are 4 12AX7 gain stages on the board with j-fet switches on the cathode bypass caps in the first two stages. A tap after these two stages goes off board to some solid state circuitry that includes one of the compander chips and half of a stereo digtal volume control chip The other two gain stages (4 total) go directly to the other half of the volume control chip. Both outputs from the stereo chip are summed to mono where they are routed to the EQ section.
The EQ section has four independent opamp filters with digipots in the feedback loop. Hand written notes on the schematics put the filter frequencies at 160 Hz, 600 Hz, 1600 Hz, and 3000 Hz. All filter outputs are summed together. There is a path through an external FX loop, and an internal chorus circuit that can be switched, summed, or bypassed to one or both (stereo) outputs. The chorus circuit has the second compander chip wrapped around a BBD chip which is clocked by the VCO in a 4046 chip which gets its control voltage from an 8 bit DAC made from resistors and a latch chip. Hand written notes put the clock frequency from 30 KHz to 270 KHz.
Since I have no intentions of cloning any part of this, I'm not digging much deeper into it, except for what it may take to fix it. There is plenty more information here if interested:
https://adadepot.com/index.php?topic=148.0
I guess that too many different things have been discussed in a short period of time here. The little 4 tube amp did start out as a self spilt output stage with an unbypassed shared cathode resistor. This was to satisfy the "no sand in a tube amp" requirement imposed by the guy who took my "wanna bet" bait that led to the whole HBAC in the first place. The 4 tube wonder had to be the cheapest amp built during that whole long thread, coming in at about $45 back then. All 4 tubes are still on the dollar menu at vacuumtubes.net but the transformers have gone up a few bucks since then. The OPT that was $4 is now $6 and the power transformer that was $15 is now $19 so it could still be built for well under $100.
The amp was refreshed a few years ago and is now a real screamer. The guy who started all the commotion hasn't been seen here since 2012, so I decided to turn the 4 tube wonder into a 4 tube screamer. I added two LND150 mosfets, a few caps, and resistors to make two topology changes. A REAL phase INVERTER made with one of the LND150's went in between the driver and the output tubes. The output stage was no longer self split, and would work with or without a bypass cap across the shared cathode resistor, but the sound quality when driven into hard clipping was definitely better with a bypass cap.
The amp was refreshed a few years ago and is now a real screamer. The guy who started all the commotion hasn't been seen here since 2012, so I decided to turn the 4 tube wonder into a 4 tube screamer. I added two LND150 mosfets, a few caps, and resistors to make two topology changes. A REAL phase INVERTER made with one of the LND150's went in between the driver and the output tubes. The output stage was no longer self split, and would work with or without a bypass cap across the shared cathode resistor, but the sound quality when driven into hard clipping was definitely better with a bypass cap.
The MP-1 has been sleeping for at least 10 years, so I decided to connect it up to the paralleled SPP board instead of my DIY preamp and plug it in to see what happens. No smoke, no fire, no sound. I remembered how to get sound, whack it! I found a suitable screwdriver and whacked the front of the unit right over the input jack like I used to. Sparks flew and the speaker made LOUD pops and crackles. The sparks were not inside the MP-1, but inside one of the old Magnavox branded 6BQ5 tubes. A few more whacks brought sound, AND all of my old presets still reside in memory. I played it for about 10 minutes without issue. I guess this means a teardown and rebuild is in order. The need for whacking seems to be associated with the level control on the front panel right where I whack it, but all the plastic jacks are bad too.
Attachments
Woke up too early this morning, got to thinking. When "power soaks" were first marketed, it was just resistors, allowing a tube amp to be appropriately loaded while delivering a line out or headphone out. Since then, they've perhaps evolved into speaker emulators, which take into account that resonance of the cone / cabinet, to get things - as a load - closer to real.
One would think, with the proper feedback, one could use a second amplifier - even solid state - as a load. In other words, you push current into its output, with the input shorted or at 0. Naturally, it will try to hold its output terminal at zero volts and not be much of a useful load, until you modify its feedback using a sense resistor, to make its output "compliant" to the same degree as a resistor would be.
Putting a notch in its feedback at some typical speaker resonant frequency would make it more compliant at that frequency, so its effective loading would be higher there, compared to other frequencies. Putting a DSP in its feedback would allow one to have, perhaps, multiple notches, resulting in multiple virtual resonances in its loading behavior. Unless time delay got in the way with the DSP doing all those calculations.
One would think it'd be easy enough to try; if you happen to have multiple amps and analog filters like in a synth hanging about. Using a regular SS amp and a 8 Ohm load resistor, feed that with the guitar's sound but through a bandpass so that amps output only moves along with any guitar sound within the bandpass. Connect that to the tube amp as a load. At all frequencies outside the bandpass, the load should be about 8. Within the bandpass, because hopefully with the SS output is in phase with the tube amp output it's not 8 Ohms to ground anymore.
Kinda like a shaker table with a force sensor, the right feedback - one should be able to hop on it a bounce like a trampoline.
The SS amp “power soak” would basically be strapped similarly to what they do to emulate a tube’s high output impedance - as a transconductance amp. The output impedance/conductance can be dialed in with the sense resistor, and can be adjusted over a fairly wide range. Instead of a resistor, it would need to be some RLC network to force the amp to follow a typical loudspeaker impedance curve.
There are some things it might not do well, though. If the load impedance is high enough, you may not be able to absorb enough voltage to deal with those high peaks that can be generated by the OPT. Clamp diodes will save the amp, but it won’t be emulating properly.
There are some things it might not do well, though. If the load impedance is high enough, you may not be able to absorb enough voltage to deal with those high peaks that can be generated by the OPT. Clamp diodes will save the amp, but it won’t be emulating properly.
Things have temporarily come to a halt while I rebuild my breadboard and workbench setup so that I can explore some not so common guitar amp ideas. I have been working on a vacuum tube based music synthesizer off and on for a couple of years and many of the concepts I want to explore would be useful to both the synth and guitar amp projects.
I made lots of vacuum tube, and a few solid state guitar amps in the 70's and 80's. Too bad I didn't write anything down. Something I read brought back a DUMM BLONDE experiment that I did back then. Many popular guitar amps return a feedback signal from the speaker output to the non-driven grid in the LTP phase inverter. There is often a "presence" control associated with this connection. NEVER replace this setup with a full tone stack. If you do you will be able to get some neat sounds like the amp is on the verge of exploding. When you turn the bass knob a bit further the amp WILL actually EXPLODE......or at least your speakers and eardrums will think it exploded, because you just unleashed a full power mangled square wave in the low bass region and the master volume control will not control it. Well, I thought it was a good idea at the time!
I made lots of vacuum tube, and a few solid state guitar amps in the 70's and 80's. Too bad I didn't write anything down. Something I read brought back a DUMM BLONDE experiment that I did back then. Many popular guitar amps return a feedback signal from the speaker output to the non-driven grid in the LTP phase inverter. There is often a "presence" control associated with this connection. NEVER replace this setup with a full tone stack. If you do you will be able to get some neat sounds like the amp is on the verge of exploding. When you turn the bass knob a bit further the amp WILL actually EXPLODE......or at least your speakers and eardrums will think it exploded, because you just unleashed a full power mangled square wave in the low bass region and the master volume control will not control it. Well, I thought it was a good idea at the time!
Oh those lovely square waves. High power ones at that. Having a voice coil that was rubbing and being before the internet and a replacement cone was not a thought, what are you going to do with a somewhat higher power speaker that is unlistenable? Well you plug it into a wall socket, right? Well, at least the outlet on the side of an industrial building (we plug cars in here in the cold of winter). BLAAAAAA......
Yes it was loud. For about two seconds.
Yes it was loud. For about two seconds.
We had these Chroma AC sources at work, for power supply testing. I was responsible for programming them over GPIB. They were capable of making any number of AC cycles you'd ask for. There were lots of MRB bins on the floors and I finally found a discarded PC external speaker. Surprised myself that the little speaker inside could take 1 cycle of 120 VAC w/o obvious damage.
That's only 119 cycles less than your higher power speaker took. I thought I once read you could plug in a Bose 901 and it would survive - but they didnt say continuously, so I guess it depends (a lot) on how long you subject it to that.
Had a “junk” 15” coaxial PA speaker (with cabinet!) dumped on me by a former boss in college. Tweeter didn’t work at all and sounded horrible. It got 60-Hertzed and pieces of cone went flying. The cabinet got holes drilled in it, and placed in the dorm trash pile (under the boss’ window) with a big “Philips” symbol painted on it.
I don't need or want a lot of power. Since I still have lots of those rusty transformers in non-rusty condition still packed in their original Schumaker 8 packs, they are a given. I have also been collecting metal octal vacuum tubes for many years and now have a big box full, so there will be at least one all metal tube amp. I have several metal 6L6 and 6V6 tubes that are known good. That amp will be something that Leo could have built in the 1940's using only parts that existed before I did, pre 1952.
I have a 4 tube 4 watt amp that I designed for the Hundred Buck Amp Challenge that's plenty loud enough inside the basement, but falls a bit short outside. I'm guessing 30 watts or so would be plenty. I just fixed up the 5 tube 20 watt amp that I made for that challenge, but I have not tested it outside yet. It's really loud in a concrete basement.
Most of the time I just plug my guitar into the Focusrite interface, load up an amp sim, and play through the studio monitors or headphones. My playing skills and song memories have degraded a lot in the last 40 to 50 years, so I find myself back in the 60's relearning the old surf music songs I played as a teenager with 4 X metal 6L6 tubes stuffed into an old Stromberg Carlson PA amp. One of the must haves that I don't have yet is a good reverb with real springs.
The purpose of these amp building experiments is to learn some new tricks, and try out some ideas, like UNSET for guitar amp duty. Anyone can clone a Fender or a Marshall. Been there, built that. I want something different, but I'm not exactly sure what that is yet. I'll know it when I make it.
How about the UNnecessary interstage transformer, driving the output tubes. Nobody's going to build that in a commercially available all tube guitar amp. Unnecessary cost adder, compared to a couple of cheap capacitors. But, having the bits to do anything you want...
Who would take bets that would add some, hopefully perceptibly and pleasant, character to the amp?
Already being investigated. Finding a suitable transformer is likely the biggest issue. I have a configuration that I call UNSET when it's used in SE. It involves standing a p-channel mosfet on its head and using it as a follower to drive the cathode of an output tube while "triodization" or Schade feedback is applied to the control grid via a simple resistive divider from plate to ground. I have seen 500 watts flow from 4 X 26HU5 tubes in this manner.
A couple of experiments revealed that a step down transformer in the cathode circuit of the output stage works too. Simulation and some crude measurements show that the impedance of a pair of Chinese 6L6GC's driven to about 10 watts is in the 300 to 600 ohm range. The impedance drops as the power level is increased.
So, if the all metal tube "MetallicAmp" was to use a pair of 6V6 tubes driving the cathodes of a pair of 6L6 tubes, I would need an 8000 to 500 ohm transformer capable of passing up to 5 watts, and handling up to 100 mA per leg of current. I might have a suitable power toroid around here somewhere.