matchbox tube preamp pedal info-kit

many of you have contacted me for more information about my matchless hotbox build, unfortunately i did not document this build well and was not able to even provide a schematic.

in the past few months, i have been documenting my builds more carefully and after finding a few improvements to the old hotbox build, i decided to build a new one...

i present to you the matchbox tube preamp pedal...

the matchbox fits into a smaller chassis and is purely point to point wired with no turret board. the component layout has also been improved. testing of the matchbox against my previous build shows no increase in noise and an identical tone to its predecessor.

more importantly, the build was fully documented with a parts list, schematic, wiring layouts, drill templates, assembly instructions and 3d pdf.

i have decided to make this documentation available as an "information kit." with little time to put into my amp projects and a tight budget limiting how many builds i can complete, i am hoping that offering these information kits will greatly help others to learn about building amplifiers, enable them to build a higher quality amplifier as well as helping me develop more infokits at a faster pace.

here is what you'll get in the infokit:

- complete schematic
- parts list with schematic references and weblinks to suggested parts and vendors
- drill templates
- three stage wiring layouts for ease of assembly
- step by step assembly instructions
- 3d pdf for detailed reference during assembly

the introductory cost of the kit will be $10 usd and will be available to residents of the US and Canada and other countries with mains voltages of 110-120 volts ac. payment will be accepted by paypal. delivery will be to your email address.

if you would like to order a matchbox tube preamp pedal info kit, please send me an email! (sorry, i cannot respond to comments)


coming soon...

coming soon are updates to projects which seem to generate the most interest.

new amp: vox ac4 'heritage' - 'ef86' and top boost with active in-phase channel mixing - working out the kinks in this one, right now....


100w ampeg b-15n (1968 circuit)


after building a successful ampeg b-15n preamp, a customer wanted me to build him a 100 watt version of it. originally, the b-15 was a 20 to 25 watt amp. 6l6 tubes can put out 30 watts each, so four of them would certainly be capable of producing 100 watts. simply doubling the 6l6 count would get me into the 40 to 50 watt territory- to reach 80 to 100 watts i was going to have to implement a fixed, adjustable bias and use a solid state rectifier.

i wanted to build a quality amp that wasn't going to cost $1000. The first issue to tackle was the transformers. b-15 transformers (individually) were in the $250 range, but they certainly wouldn't get me the 100 watts i was looking for. so what other 100 watt amplifiers are out there? fender had some 4 x 6l6 amps that were know to produce 80 watts. i located a schematic for a twin reverb and found that its power transformer was capable of providing the needed power, and at a relatively low cost.

the output transformer was a little tricky. for use as a bass amplifier the output transformer would need to be capable of a wide frequency response. i looked to a reputable source of custom transformers for the answer: edcor. they had a 100 watt model with a 3.3k primary (half of the primary impedance of the original b-15 output transformer) but with a frequency response of 20kz to 20khz. impressive! an order was placed and the transformer arrived 4-6 weeks later at a total cost of around $120. that puts the total cost of the transformers around $250- not bad at all.

mechanical design

the next issue is what i consider the most difficult issue in custom amp building: the chassis. undrilled (blank) chassis are readily available for marshall style amplifiers, but would all the components fit? some quick measurements said that it might be possible, but i put o use my new 3d modeling software to answer this question.

the first thing i needed to do was create the models of all the parts. this takes some time, but makes it much easier to do this in the future since the models can be reused. once i had the models for the tubes, sockets and power and output transformers i populated the chassis model with them and the answer was: yes! everything will fit... barely! (notice the tight clearance between the power tubes and the transformers)

here are some more shots of the model:

notice the tight clearances between the power and standby switches and the transformer. the original 3d model indicated an interference between these components. i had to buy switches with smaller lugs and move the power transformer nearer to the tubes to make it all fit. this and other mistakes would have been terrible to correct if i had already punched the chassis and then discovered the problem.

electrical design

the power supply needed a few changes to get the amp in the 80-100 watt range. a full wave, solid state rectifier was used. a standby switch was added for tube life and convenience, a marshall-style adjustable fix bias was installed, and fuse values were derived from the fender twin reverb.

the board layout was done next. here is what it looked like:

some of the custom features in the amp (as requested by the customer) came from the b-15 preamp that i built as a testbed for this amp- preamp-out/poweramp-in jack, output mute, master volume. another addition was a buffered tuner out jack. however, on this amp the inputs were modified to be a "normal" and -10dB channel. a bright switch was added to the inputs instead of a separate bright channel. output protection circuitry was also added to the amplifier so that it would remain on standby if it was run with a speaker cable plugged into the output jack.

the image below shows most of the features of the amp:


with the mechanical and electrical design nailed down, orders were placed and i handed an undrilled chassis, a pile of parts, drill templates and wiring templates to my co-worker so that he could build his brother an amp. the most difficulty that he had was drilling the octal sockets and doing the cutout for the lay-down style power transformer.

my friend did a great job keeping the wiring neat, escpecially for someone with no previous experience wiring something so complicated. most of the electrical troubles during the construction were to blame on a poor soldering iron. there were some cold solder joints or joint with insufficient solder that were causing intermittents.

here are some pictures of the completed amp:


this amp had some really big issues. the first was hum. 100 ohm resistors across the heater reduced it a bit. after a few weeks of troubleshooting, i located a poor ground path. replacing the ground wire, and an excessively noisy tube fixed the problem.

another difficult problem was the presence of a sputtery type of distortion. this distortion was being introduced at the phase inverter. i had a theory that perhaps one of the components in the
phase inverter was damaged by the malfunctioning soldering iron. i replaced all the components just in case as well as ordering parts for a twin reverb phase inverter.

after replacing the b-15 phase inverter the distortion was still there. i disconnected the negative feedback and the problem was still there. i replaced the phase inverter with the twin reverb phase inverter and now the amp had a beautiful sustained distortion (just what the customer was looking for.

i had read in 'amps: the other half of rock and roll' that ampeg vigorously tested their amps before shipment to ensure that they didn't distort. i think i now know why!

when the amp was finished i ran it over to my friends house and we fired it up (using an SWR LA-15 as a speaker cabinet). it was incredibly versatile! it had a clean, deep bass with beautiful high frequency content. the switches, though some seemed redundant, added some great options to the sound. the bass distortion didn't sound good to me but my friend indicated that his brother loved that sound.

i thought the amp sounded even better as a guitar amplifier, even though it was being tested with a bass speaker. as i stated before the amp is very versatile.


i supplied my friend with some marshall smallbox headcab plans. he designed a custom cabinet, using mdf. it was decorated with bed liner for a spray can and some oxblood grill cloth. i think the amp turned out looking great! (if only i could find a good source for a custom faceplate)


the overall cost of the amp was just over $500. i deem it a great success. it was also the first amplifier that i had made using 3d mechanical design software, something i will definitely be using more in the future.


ampeg B-15n (1968) preamp

after i had a co-worker's son build a ga5f1 kit, his uncle plugged his bass into it and loved the sound he was getting. he asked this co-worker if i could build him an amp. his current bass rig was a peavey 2 channel amp with one channel being vacuum tube driven. this was then run into a 500 watt solid state power amp.

knowing this, i set out to create a vacuum tube bass preamp he could use with his rig. since we all know the fender bassman is a guitar amp, i decided to build an ampeg b-15n (1968 circuit). now onto some more background.

a few years ago, i picked up a grommes 20w monblock hi-fi amplifier for $1. it was buried inside some old scientific equipment ( a sonometer, i think). it always reminded me of an ampeg b-15 with its huge transformers and the 6l6 tubes.

here is a pic of the grommes, followed by the ampeg b-15:

the grommes lacked a few things: a power-switch, three prong power chord and a standard 1/4" input jack. i took some measurements of the chassis and realized that this amp was only about 1/4" smaller than a standard hammond 13x2x5 chassis. upon this discovery, i was set on mating the ampeg preamp and the grommes power amp together. if my test preamp sounded good with the grommes then i would build one for my customer.

the grommes provided power for an external preamp, i tapped into this node to power my preamp, saving the cost of an additional transformer.

the preamp circuitry consisted of one channel from the b-15n, with the addition of a -10dB switch (customer request), output volume control, output mute switch and a preamp-out/power-amp-in jack.

the preamp board layout was quite simple, the tone circuitry however was a little complicated. it was easier to mount the components onto the pots with the aid of a terminal strip. here is the layout and a photo of the stuffed board

here is a photo of the installed preamp board with the tone circuitry:

the font panel layout and implementation:

the preamp-out/power-amp-in jack is a stereo jack which allows two options when used with a mono cable. insert it two-clicks and it functions as a preamp output without interrupting the signal to the power amp; insert it three clicks and it mutes the preamp signal but injects an external signal into the power amp. i had used the grommes before as a power amp for my matchless hotbox tube preamp pedal and it sounded fantastic.

here is the amp before and after the two chassis were mated:

after the preamp was completed, i invited my co-worker over for the test run. he brought a video camera to record the demonstration. his brother loved the video and asked that i build him a 100 watt version. (write-up on this one coming soon)


paradigm shift

i was able to accomplish two main things over the 2009-2010 holiday season. one was to build an ampeg b-15 preamp for use with a tube power amp that was just lying around. the other was to learn a bit of 3d mechanical design.

when i build the b-15 preamp, i had made a few mistakes. once again, i failed to plan enough clearance between the turret board and the input jacks (circled in green). i had to chisel away part of the circuit board board to get it to fit.

i had it all roughly sketched up on graph paper. this has been my quick and dirty method for laying out an amp chassis, except when i want to change anything i need to take a new piece of graph paper and start all over again.

3d mechanical design is changing all of that for me. i signed up for a 30 day evaluation of alibre design, a low-cost ($97 usd for the standard package) 3d design and 2d drafting solution. this 30-day evaluation is actually of alibre professional which includes sheet-metal capability.

i started by creating a sheet metal version of the chassis i used for my ga5f1 build and punched all of the mounting holes in it. then i started modeling each of the parts that went into the amp and assembled it. cool. now i have a 3d model of an amp that i have already built.

now when i did my layouts for the ga5f1, i used a graphic aers program as a 2d cad program. it was very time-consuming it took me hours to create the parts and incorporate them into a layout.

i took my sheet metal chassis model and created a 2 dimensional drawing from it of the front panel layout. i printed it off and it was precisely to scale. it took me only a few minutes to do this. i could easily use this template to drill a front panel on a new amp with accuracy.

nex,t i thought i see just how much time this software might save me. i recently bought another vox pathfinder amp to build a 5w version of the vox heritage amp line into. i looked for a chassis that would fit without modification, but i could not find any. so i took some dimensions from the original pathfinder chassis, translating all the mounting holes to a 3d sheet metal model. now i had a 3d model of the blank chassis i needed. i imported this model into the 2d design, added dimensions, hole centers and callouts as needed. now i had all the information a metal shop would need to fabricate the chassis for me. this entire process took me 10 minutes and 35 seconds!! (yes i timed it)

in the meantime, i've had an opportunity to take a 3d modeling training class at work and i've been applying what i have learned on the design of a 100 watt(ish) b-15 that i am helping a coworker build.

this has been an amazing experience. i am using a blank jtm45 chassis for the build and have to do all the chassis work myself. i modeled up the chassis and have been modeling up parts as needed. this has been a complete lifesaver! i have identified many mistakes that could have been disastrous to the amp build. now as i find these, i can make adjustments to eliminate them. here are two examples:

1. over the holidays, i had my coworker build and stuff the turret board for his amp. i had provided 4 mounting holes on the layout for the turret board. when i modeled these mounting holes and placed them in the blank chassis, it became clear that two of these mounting screws would interfere with the output transformer. now, i could have simply offset the transformer from the chassis using an extra nut. instead we identified 2 alternate locations for the mounting holes which, after adjusting the position of the output transformer gave us enough clearance.

2. the power transformer we are using is quite large for a jtm45 chassis. the transformer was modeled to be mounted near the power and standby switches. after inserting the models for the switches and power transformer into the assembly there was clearly going to be interference, and in a bad way. i moved the power transformer more near the power tubes until there was about 1/4" clearance between them, but the switches wouldn't fit. Looking closer at the switch datasheet, there was a switch available with solder tabs that were shorter than the quick disconnect switch contacts. i updated the model to reflect this and gained the clearance i needed. tragedy averted again!

i would highly recommended taking the time to install and check out the trial version of alibre design. if it saves you time as well, buy the standard version. i did and i feel it was well worth it!

here is the current (not completed yet) design of the b-15-100:

once the mechanical design is completed i'll just need to print out drill templates, order parts and start building!


on the horizon....

upcoming projects:

i've decided that i will be building some 5w single ended versions of classic amplifiers that could also be used as preamps running into my all-tube 20w hi-fi power amp (grommes 221.)

here are a few projects that i have made preliminary plans for:

- marshall jtm-5
- orange or-5
- vox ac5

mostly this stems from some of the above companies relatively week attempts at a 5w amp. the class marshall class 5 at least has two preamp tubes, the others are essentially champ variants.

i am also helping a friend build a high-power ampeg b-15n. i will be building myself a preamp to use with my grommes at the same time.

i will document each project as they complete.

i have included an option to email me questions. i would certainly prefer to answer questions directly to the persons who have asked them. i will also be creating a forum where questions about some of these projects can be discussed.


marshall 20w channel-switching amp

this amp came about after building the marshall 20w lead and bass (model 2061) and not being completely satisfied with the results. while i loved the clean sound from the lead channel on that amp, i was hoping for more of a marshall 50w sound out of the distortion channel. this is just not possible because the 20w lacks the extra gain stage of the 50w. i decided to create an amp that combined the sounds of the 20w's lead channel clean and the 50w's distortion channel.

i had already purchased an 18watt chassis with 20w style faceplate from mojotone. with this chassis i can use the 18watt transformer set, also. the 18w transformer has just the right voltages for the 20w. the 18W chassis is nice because it already has the extra preamp tube slot that I needed and will accommodate the 18W transformer set and 18W cabinets without modification.

the first circuit that i built was a stock 2061 circuit. for this implementation, i used the same circuit board that would be used for the channel switching amp, but only wired up the components that made it a stock 2061. playing it with proper transformers didn't produce much of a different sound that my initial 20w build that i documented here - i definitely wanted that extra gain stage in there.

my next attempt was to implement a stock 20 w amp with a switchable gain stage in series in the circuit. this didn't work well because of the cascaded volume controls. it was impossible to get a good clean sound and a get distortion sound that was at the same volume. also, any change in the first gain stage was sent downstream to the second gain stage. looks like i'll need some more modifications....

at this point i determined to redesign the amp entirely to get just what i wanted- a 20w lead channel for clean and switchable 50w channel for distortion, with the switching handled by a relay. since i always run my 50w controls at 10, i wouldn't need the tonestack and would use separate volume controls for both the normal and high treble channel to change to sound of the distortion channel. this way, the amp would essentially be switchable between a 50w circuit and 20w circuit, with the exception of the 20w bass channel and the 50w tone controls.

the power supply was based on the 20w amp to get the voltages right, while the phase inverter was derived from the 50w phase inverter. the amp would also be fixed biased instead of cathode biased because today's tubes can be pretty hard to tame in cathode biased amps to avoid red-plating. i used the bias circuit found at 18watt.com for the fixed bias. there would also be a separate power supply for the relay circuit fed from the 18w transformer tap for the tube rectifier.

the implementation of the channel switching circuit wasn't without its problems. there was certainly an evolution that occurred. attenuating the signal in the extra gain stages to get the distortion to sound right was one trick. i had also used the same gain stage for 50w normal and 20w lead channels, since the volume controls are after these channels in the circuit and their first gain stages are identical. this cause some signal bleeding between the 20w lead and 50w normal channels. i used the relay to ground out the signal from the 50w normal pot when the 20w clean channel was engaged.

here's a closer look at the relay power supply:

on the 50w channel, i implemented a pre-phase-inverter master volume. this helped to even out the volume with the clean channel. i noticed that when i had this volume low, i wasn't getting the distortion i wanted out of the amp. i soon realized that even the extra gain of the phase inverter had an astounding effect on the amp, so i used another tube stage to simulate the gain of the phase inverter before this volume control. the results were wonderful! i had learned the difference between a "master volume" plexi and a "non-master volume" plexi.

i also had a problem with the 50w normal and high treble channels acting like they were wired in parallel. it turned out that i had mistakenly installed 470 ohm mixer resistors instead on 470k mixing resistors. installing the proper resistors fixed this issue.

once all the channels were isolated the amp was sounding great - good enough to have my mother-in-law comment on how nice the clean channel sounded and have my friends and other guitar freaks to comment on the sweetness of the distortion. my buddy who had just bought a new amp liked it better than his own.

here is a peek at the inside. its a bit of a mess thanks to all the modifications. you can see how some modifications were done right on the tube sockets and were potted in hot glue to eliminate vibration or shorting of signals. the relay power supply is to the right of the chassis. i tried to do a nice job with the power transformer wiring on this amp. mojo seemed to have flipped the main power and standby switches on the 20w faceplate, so i had to rearrange my nice wiring into a mess.

i had also implemented a post-phase-inverter master volume (the rich mod) to keep the volume under control when practicing at home. It was quite a stretch to implement all these features into the minimal 20w front panel, but it was certainly done!

here is an explanation of how the controls are set up:

in the end, the amp was mounted in an 18w style 1x12 cabinet loaded with a celestion well broken-in g12h30 speaker (i pealed the sticker off of an avatar hellatone 30, really).

i came to realize that what i now had was very close to my first tube amp - the marshall jcm2000 dsl401. A fixed-biased el84 two channel switching amp, except this one was much better.

now i finally have an amp that gets the classic marshall sound - well, actually, two of them with the ability to switch between them easily - the stock 20w clean channel and a custom 50w 1987 normal and high treble (bridged) channel.

here are some schematics of the build:

marshall 20w model 2061

when marshall released their 20w handwired lead and bass amp (model 2061x) i really wanted to try one out. unfortunately, none of the guitar shops in my area had one in stock. i had heard them described as a mini-plexi and had found a few sound clips on the internet that sounded pretty good, but i had to build one myself to give it a fair try.

i was hoping to make a portable 1x12 combo amp that i could lug around for practices without bringing my head and 2x12 cabinet and still get a similar sound.

i began doing research to discover the details of the 20w circuit and layout, collecting pictures and schematics. i started this thread over at the plexi palace forum for anyone who was looking for the same info or had something to add. (If you're interested in building a 20w amp check it out, i posted some layouts for classic 20w circuits.)

this amp would start out much like other experimental amps: in a cakepan chassis. i had a spare power transformer i picked up from some old tube-driven hewlitt-packard scientific instrument for $1 at a surplus sale and a 15w hammond transformer left over from my 18w build. the remainder of the parts cost me about $60.

i followed the classic 2ow layout as closely as possible. i used grade xx garolite (paper phenolic) in this build. it was much easier to machine than the g-10/fr-4 boards i had used in the past.

i made one modification to the circuit, adding a matchless dc-30 style master volume.
when i powered on the amp, i found that i really had to turn up the master volume to get a good sound out of it. i later discovered that this type of master volume is best implemented on an amp with matching plate load resistors in the phase inverter, like the vox ac30 and ac15.

you'll notice some rather crude looking construction methods in the above picture. again, the experiment here was to build the classic circuit for as little as possible. the circuit is electrically fine, but it does leave something to be desired in the area of looks.

i really loved the clean tones that i was getting from the lead channel. the bass channel didn't seem capable of much more than clean tones and did not have as much volume as the clean channel. it was nice to be able to blend in the bass channel to change the overall tone with the channels linked.

when the channels were linked and the volumes were cranked the amp started to break up. the distortion seemed rather brash than smooth and i'd always thought there must have been something wrong with amp. i eventually found the schematic for the scientific instrument which seemed to indicate that the power transformer may have been underpowered. measuring the transformer showed that it was the exact size of a vintage fender champ transformer.

i decided the give the amp another shot with a proper set of transformers, this time with a nice 18w set from heyboer. i'd see if an improvement was made and if not i had other plans for the circuit- afterall, i did love the clean channel. i got a proper chassis and faceplate set from mojotone also.

unfortunately, the new transformer set didn't yield much improvement. i had hoped for a sound more similar to my marshall 50w. comparing the two circuits, i determined that i liked the sound of the extra gain stage in the 50w for distortion. the marshall 18w and 20 w circuits only have one gain stage before being sent to the phase inverter.

i decided to give the circuit another try, making the extra gain stage switchable. the amp this circuit would become is outlined in my marshall 20 watt channel-switching amp post


free el84 retainers

these next few post are some simple tips that i've shared with others on amp forums.

when looking around for an el84 retainer one day, a spring fell out of one of my j-slot preamp tube shields. i got to thinking that maybe this could function as a el84 retainer. i dinked around for a few moments trying to figure out how to mount it, finally coming up with the arrangement below:

i really didn't think this would hold up, but after trying to pull off the spring and lifting the entire amp off the table without the spring coming off, i determined this was actually going to work, after all!

an alternate and likely more stable arrangement would be to bend a small hook in the end of the spring and hook it into the slot next to the retainer pin that it is being supported by in the picture.

Parallel / Series Switch for extenstion cabinets

i had a few people ask about this on some guitar forums. They wanted to be able to use their 2x12" cabinet with one amp at 16ohms and another amp (like a fender bassman) at 4 ohms. this is possible with a cabinet having two 8 ohm speakers using a dpdt switch as shown below:

be sure you use a heavy duty switch that can handle the current. it may also be helpful to do this on a new jack plate so this modification can be reversed, if desired.

note that this switch can also be used with a 4x12 cab with 16 ohm speakers, since two 16ohm speakers in parallel are the equivalent impedance of one 8 ohm speakers. simply follow the diagram for the 8 ohm speakers and then wire a speaker in parallel with each speaker on the diagram.