E-Mu Morpheus Prototype?

I have a friend in Santa Cruz who is always finding strange toys to play with. She recently called me to have me service her E-Mu Proteus. She gave me an extra Proteus Custom she had sitting around (yay!) and I also got a weird ass rack with no faceplate decals on it. It looked to me like a piece of beat up hardware with nothing remarkable about it. Until she turned it over and I saw the business card for Dana Massey, the DSP engineer at E-mu systems during the ’90s. It suddenly dawned on me that I was looking at a factory E-Mu test bed. I was excited to get it home to see what kind of patches were saved on the unit from the lab. I figured I might be able to find some cool patches and stuff.

Alas, upon inspecting it while plugged into my soundsystem with a MIDI controller, I realized the patches were just the same shitty factory patches Dave Rossum had shipped with the original units. I was unable to find anything impressive or interesting about it while I played it. It seems like most of Rossum’s rack products: useful when it came out, and then within months or a year, each model became a thing of the past as newer and easier to use systems hit the market. The other thing about the Morpheus and Proteus modules are that they are ROM-plers, as in “samplers you can’t sample with,” as all of their sounds are just ROMs sampled at the factory. I personally like synthesizers and samplers much better, but ROM-plers like the E-Mu Proteus are still used all the time to make HipHop and Electronica.

Z-plane filters, which are Rossum’s unique contribution to the field of sound synthesis, are pretty cool. The Z-plane is a type of digital filter whose cutoff frequencies change over time. They’re usually 18 poles, as opposed to the 4 pole analog filter that Moog popularized. The Z-planes sound spacious and very pretty. The tones they produce can be bell like or very ambient and pretty. The Z-plane filter itself was always the domain of rack mounted E-Mu systems modules in 19 inch racks, until Rossum released the Eurorack Z-plane filter for modular synths. The Z-plane filter module for Eurorack is clearly his best iteration of this software, as it is the most versatile, useful with other hardware, and you can drive the module to do crazy shit the original rack synths couldn’t do. I also should note: I really don’t care for digital filters in general, so take my synopsis with a grain of salt.

From what I can tell, it looks like Dana Massey or whoever at E-Mu, was blowing up some ICs from time to time inside the unit. I can see where the quad flat pack ICs for interfacing with the slot inside have been resoldered. I can also see that the GAL/PAL ICs are socketed and most are unlabeled with the usual sticker. There are also several EPROM ICs with their windows uncovered on the PCB. I surmise that this unit was used as a test platform for Morpheus EPROMs or as a test bed for the slot on the PCB. The slot looks like it could be a slot for an E-Mu ROM expansion, but I looked into it and nowhere does it advertise the Morpheus as expandable, so I assume it could have been the programming or testing interface for mass production. I’m not sure.

Starving Musician in Santa Cruz apparently gave my friend Christina this unit for free, as it did not have a label on the face and also seemed to have Dana Massey’s business card on the bottom. The guys at the shop said they couldn’t sell it, so it went to Christina, and now to me. A fun little project, but altogether a little disappointing. The Morpheus sounds whack in my opinion, and who in God’s name wants to menu dive for 5 minutes for each parameter tweak. The Z-plane is interesting, and pretty cool. I was hoping to find some really crazy ass patches onboard, but no such luck, all patches were factory standard. If I were to get a Z-Plane device for my studio I’d be interested in the Eurorack module but not much else. This test bed goes back to Christina!

Below are some more photos:

You can see everything is Rev. A, 1.0, etc. Pretty cool. EPROMs still have the original labels and stickers. Everything looks great inside despite it’s beat up exterior.

Above are more Atmel programmable devices in their sockets, with their windows exposing the bare IC dies inside of the ICs.

Above is the Analog Devices ADSP-2115 DSP IC on it’s breakout board with it’s complimentary programmable devices on the right and what I assume is flash on the left.

Above is the output section, with it’s Analog Designs DACs, low noise N-FET output op-amps, big thick SMT resistors and caps, a Zettler relay (just like in an electric car!) and finally the E-mu ASICs on the bottom of the shot. Pretty cool.

A friend of mine recently went to work for Dave Rossum. He’s very very happy, as anyone who’s ever used an E-Mu product would be. Dave’s work with samplers gave the world hiphop. Basically, all on his own (aided by the team of professionals who enabled his vision.) Thank you, Dave Rossum. We are in your debt.

Dave is certainly an OG in Audio design, and love or hate the products, he’s worthy of praise as a designer, marketeer, and visionary in electronic audio. Just like Dave Smith, Tom Oberheim, Don Buchla, Bob Moog, and all the myriad of new designers who are now creating tomorrow’s classic hardware. But the when and where of Dave’s samplers, especially, had a very very deep impact on music production and changed the whole world. He was capable of seeing samplers as a new instrument, not a computer. An instrument that I believe still has not had it’s limits properly philosophically tested.

Fender VibroChamp Restoration

Behold! The Fender Vibro Champ Amp of legend. The OG, not the re issue. Why, then, does she look brand new? Good question! I restored the faceplate, baffle, internal electronics, and gave it a handle. This amplifier has had a long, strange, nebulous history. And it’s only stranger and stranger the more I know about it. I’ve had it kicking around for at least 5 years, maybe 10. I got it from someone in Humboldt who owed me money. It had no faceplate. No baffle. No power switch. No bulb for power indication. The two original octal tube sockets were tweakerized and removed by some nice meth head at some point too. It was also missing most of it’s resistors.

The tag inside says Petent 2817708, Des. 192859.

The transformer’s EIA 606 code says 828, which puts this date code to 1968, I believe.

Since I didn’t know what to do aesthetically about this little amp, I just did what I always do: I took a shot in the dark and ordered a baffle for the speaker, and a faceplate for the amp chassis itself. I wasn’t sure if this was a blackface or a silverface amp when it was conceived. Since it had been tweakerized and it’s old parts had been gishtolen, I figured it was time to just pick a faceplate to buy a remanufacture of. So I picked the blackface faceplate. I then went to Mojotone and got the matching baffle made by them for the speaker.

I then went about replacing the tube sockets, and resistors that some nice meth head had removed from it at some point before I ever got it. The RCA connectors for the speaker and vibrato footswitch were also shit housed so I replaced them with 1/4 inch jacks. I put in high precision metal film resistors and a nice silver mica cap for the 6v6GT. I also had to reinstall the power switch, and cut out the white electrolytic capacitors that had been in circuit for 40+ years. I put in Sprague Atom brand capacitors instead of cheap ones. I also took a couple of additional capacitors out and replaced them with Cornelli Dubilier brand Orange Drop capacitors that were closer to the original specs.

The multicore can capacitor and the transformers are still original. The cabinet’s tolex is still original. So are the potentiometers and the 3 Fender style knobs. The tremolo circuit is not modified from the stock with a vactrol, so it sounds like the OG Fender tremolo. I put Tung-Sol rectifier and power tubes in and Mesa SP12AX7’s in for the preamp and tremolo.

This is an amplifier that the original owner had told me was a 1990’s re-issue, and of no real value. He was an alcoholic and a fool. I believe he now sleeps in a tent in Arcata somewhere. Good times. Anyhow, he gave me the amp when he owed me some money, and I have had it since through multiple moves and years of sitting on it waiting to actually repair it. I just got around to it about a month ago and it looks and feels like the original Fender Vibro Champ. It sounds, however, just a little different because the metal film resistors have a different sound quality from the old carbon composition ones. I love it. This amp is a unique sounding repair-mod but still recognizable as a Fender VC, for sure.

I’m going to sell it. I’m not a guitarist, and this baby needs to be played. I’ll get rid of it on Craigslist, and if not there, then at Reverb or eBay or wherever. I can be patient. Hope yall like the pics!

Bryston 3B Service

A friend gave me this Bryston 3B rack for 250$ because it had a hum issue. I took it off his hands and looked into the chassis. Super sexy. Plitron transformers for power (one for each channel!) and the very simple circuit inside. Everything looked visually intact. I did notice though that the capacitors inside were clearly past their prime, and probably had gone bad. Thankfully, the circuit uses very few caps. The major ones are the 2x ones used for each power transformer that are 5700uF at 55v. I replaced them with Kemet brand caps that are the same size but 10000uF/63v instead of 5700uF/55v. The caps on the channel boards are not so substantial, but since there aren’t that many in the circuit, I wanted to put Sprague Atom caps in for those. So I went a little overboard with my components I chose to use. I’m replacing the 220uF/10v caps with 500uF/16v caps. I’m pretty sure it won’t change the character of the audio path at all.

Above is the shot of the left PCB having been denuded of it’s 1N973B zeners, 1.5uF/35v capacitor, 470uF/16v Siemens capacitor, and its 1N4007 diodes. The one on the right has it’s old components in the photo. I missed the 200uF caps in blue on the left before the photo. D’oh!

As you can see above, the new capacitors from Kemet are far more tightly packed and higher voltage tolerance than the old ones from the 80s were.

Above the two GIGANTIC Plitron power transformers are visible- one for each channel. The Bryston 3B is essentially 2x complete monoblocks inside of a brick shithouse of a chassis. It’s impressively heavy. Plitron is good Iron too- they’re reputable and do high quality work for the cost of their wares. The channel separation is complete and total: each channel has it’s own linear power supply and it’s own rectifier and fuse. It is truly two amps in one. The shot above also shows the new black Kemet brand caps I installed. The two resistor on the output board, which are 5.6kΩ, are Ashy Larry. No bueno, amigo.

Yeah. I know this symptom. It’s probably dissipating between 1/2 Watt and .8 W or so. And as opposed to seeing that dissipation only sometimes, the engineers probably said, “well, it’s going to see that power dissipation pretty much constantly, but we have to go to production, so fuck it they’ll probably be fine.” Over 20+ years of 40+ hours a week use, these resistors are toasted with the mosted- like Cypress Hill toasted. Ashy Larry toasted. Gishtolen like Amish Dave’s butterchurn. They’ll have to go.

Ideally a 5.6kΩ cement bathtub resistor pair at 5W or so could be used. I’d lean more toward using non-inductive wire wound resistors rated to 3-5W myself, but bathtubs would be cheaper. I do have some 1% tolerant 5.6kΩ KoaSpeer metal film resistors, rated to 1W, but I hesitated to use a 1W resistor in place of where the previous ones went slowly ashy. Maybe the metal film will hold up better than the carbon film type under so much heat dissipation? I’ll try em out. I want to fire this thing up, dammit!

I’m really excited because not only did I replace the caps with Sprague Atom Capacitors, but I also replaced the 1N973B zeners with JANTX1N973B’s from Mouser, which were crazy ass expensive, but they’re also mil spec so I’ll take it. I replaced the old and somewhat played out 1N4004’s in circuit too, but with UF4007 diodes instead of 1N4004’s. The difference should be even stiffer, quieter power supply rails than the stock units had: new reference diodes, new power diodes, new and better electrolytic capacitors, and new power supply ripple capacitors with better specs have all been added during the restoration process.

Above the Plitron transformers, Kemet caps, and the new 5.6k resistors are all visible. The amp would benefit from the replacement of the last 8 1N4004 diodes on the rear common PCB assembly with new UF4007’s like each channel PCB, but they don’t look decayed the same way the channel PCB diodes did so I left them. The rear board is more densely packed and a little harder to deal with because of it’s wiring. I may eventually do those diodes too before I sell it. Power supplies sit at +/- 55.6 at first power up but after a few minutes they settle at +/- 55.15 or so. Good God, they fucking put 55 Volt DC capacitors in a circuit that sits idle at +/- 55v DC. How did these function so long with the capacitors so close to the edge of what they were designed to handle? I suppose it’s due to the capacitors’ construction and the quality of the low transient AC coming from the Plitron transformers. I assume if there were more transients present the caps would have developed their 60Hz hum much sooner than 20 years after it was made.

For having been used for decades, this monster is in exquisitely rockin shape. Not the easiest to service, but certainly not the hardest either. On trip advisor, I’d give it 2 stars. The repair was good, but not that good.

Seriously- this amp was made the year I was born, and I was excited to compare it to my old Phase Linear amp, which was great but a little “soggy” sounding in the bass, and had lots of high frequency noise in the treble bands. The comparison made between the PL and the Bryston went terribly at first. The Bryston sounded great, but with all of this hashy high frequency noise and I hated it. That was two nights ago. I left it on to burn in the new components and give the old ones a chance to breathe some electrons for a while. Below are photos of the PCBs for each channel with the new Atom caps installed.

After a 48 hour burn in for the Sprague Atom capacitors, I have to say: this amp sounds fantastic. The Bryston sounds radically different from my personally designed tube amps made to the old school high fidelity specifications. It has no noise floor that I can hear. It’s sensitive and responsive to every subtlety of the input signal. It also lacks something difficult to verbalize, but I’ll try: it lacks detail here and there, and it sounds sterile as opposed to “warm.” Since both amps are analog, and the the Bryston’s power rails are totally isolated to each channel, both the Bryston and the push-pulls I build are both very punchy sounding. The Bryston lacks something hard to describe, but the trade off is clearly worth it if you need 100+ watts of output power to drive something, and can’t afford the tube alternative at the cost of ~3,000$ per monoblock for the cheaper 100+ watt tube amps. The Bryston can do about 100w into 8Ω and about 200w into 4Ω, per channel. So for the cost of $700-1,200 for a fully refurbished and properly modified stereo unit, the Bryston is a pretty good buy for entry level hifi stuff. Like every other amp, there are people who really love Bryston- and now I know why. I really love it’s design and sound. Its simple, elegant, and effective like anything well designed, and it’s made beautifully too. Mad props, Bryston. Well done

I made the comparison between my transformer coupled 6V6GT push pull amp and the Bryston 3B amplifier. I drove both amps using my Rev 2 preamplifier, which yes, will totally drive the Bryston no problem.

B&W Model 210FA Oscillator Restoration, Part II

Ahhh, yes. I posted some photos of this baby when I got it a few weeks back. It’s a 3 valve oscillator/ amplifier that’ll do a wide range of frequencies at high stability and has about a watt of output power. It does a pretty good AC sine wave at 60Hz for testing purposes requiring low current at a low AC voltage.

When I got it, the oscilloscope image of the wave was all folded up on itself and distorted. The light bulb used for the current limiter inside was clearly an open circuit (filament was gone.) I started there. I replaced the bulb with some easily found GE candelabra style bulbs of the exact same type. Sheer luck, thanks eBay. After replacing the Aerovox caps inside with metalized polypropylene, and replacing the signal caps with Sprague electrolytics, I was still seeing a ton of distortion on the output. My thought was immediately that the big dry multicore can capacitors were bad. So… time to drill em out and replace em before I do the panel connectors.

One multicore cap, before surgery.

The other, also before surgery.

Above is the terminal strip I secured with 4-40 screws and nuts. I used cheapo capacitors from Jameco, as that was all I had. I also increased all of the 10uF values to 22uF. They look pretty good for being a 1hr quick fix it job. Keeping the ground insulated from the chassis is important in this case, so care was taken not to connect ground to the chassis.

Below you can see where I removed the original cement 10W 2.5kΩ resistor in the power supply in favor of 2x 10W 1.3kΩ chassis mount resistors soldered in upright off of the terminal strip. I didn’t have axial leaded 47uF/450v caps, so I used radials.

The ripple observed at the output is basically gone. What remains is likely the result of the 1N1696 Si power rectifiers on the power transformer. I plan to replace them with Fast Recovery Epitaxial Diodes this weekend when my Mouser box gets here. But basically, that’s it. The old binding posts (banana jacks) up on the front panel were gross. I replaced them with the same colors as the originals, and only later noticed that the manual shows black in the center. I’ll have to trace the wires back and make sure someone else didn’t miswire it before I bought it. But the result looks great:

After removing the pile below, the unit is lighter weight (although not as sexy as with the big cans still intact.)

Below is a shot of the unit now that it’s had it’s caps replaced. There’s even room for a choke on the power supply on the top deck now… Hmmmmm….. Further 60Hz ripple reduction…?

That’s all! I’ll post pics of the scoped wave after I finish the restoration next week. Before you laugh at the “no tubes” note, I didn’t put it there. It’s on the enclosure for the variable capacitor at the heart of the tuned oscillator.

Tare-down: Sony TC-666 Tape Recorder (Reel-to-reel)

Unfortunately for the greatest named model of tape recorder ever, the TC-666, it is crap. Not worth much more than 30-50$ in near new condition, this one… is not in new condition. It’s “all fucked up” to use a classic engineering term. It must die. But we’ll gut it for it’s mic transformers, and some Sony branded 1970’s big package transistors, and look at some interesting OEMs inside before it goes to the E Waste pile. Below is a shot of the corrosion on the interface. Sad face.

Below 2x small PCB mounted transformers are visible through the jungle of wire tied inside.

Below is the underside rear of the preamp section- all electromechanical and clusterfucked. Plus it doesn’t work properly either, one channel’s blown. Probably a cap problem. I likely won’t get the meters or preamp PCB out, but such is life.

Below is a shot of the transistor PCB in circuit with the motor run caps behind. The green circle is a diode. Whacky stuff. Sony OEM transistors too.

Below: In circuit- the Japanese OEM transformer. Very sexy looking- almost looks like a Hashimoto stylke (which it couldn’t have been, but certainly similar style.) On the left is one of three big assed DC motors inside. The coils are in very good shape.

Below is the transformer removed from circuit. Very nice.It’s a little dirty, but in great shape. Japanese OEM, haven’t googled the OEM#. 0-75v winding is used with others, and the high voltage AC mains. Should be a good one for a DIY Solid State power driver.

Below is the three transistor driver PCB, probably for the solenoids or motors. Turquoise ceramic inductor visible with hand painted value marks in the center. Sony OEM tunable inductor, probably for +/- 60kHz AC tape bias.

Below, the lower right there are two yellow striped solid state diodes. OEM# 1R81

Below we can see one of the PSR .1uF 250v/120Ω has bowed out from shorting internally. Old caps do that.

Below are Sony OEM transistors, probably NEC or possibly JRC from Japan, with Sony’s name stamped on them. I actually don’t know if Sony ever had a fabrication plant of their own for silicon. These haven’t been tested, but if they work, might be a great set of drivers to build a solid state amplifier of some kind out of.

Below is a mystery board. 2SC633 and similar OEM transistors from Japan, with ultra thin brass leads, are used throughout all the PCBs in the unit. Never seen ’em like that before. Likewise, with the black arrow shaped diodes on end with the single grey dot on them and the black tarnished wire (probably silver?)

Below is “the money” as they say. I didn’t bother with a manual with schematics and the like for this one, but I know a couple of mic transformers when I see em. I should look up the schematics so I can see if their impedances are spelled out. I could just LCR test them out of circuit too. Yay! A 30 minute tare-down now worth it. They’ll have to be carefully extracted, cleaned up a little bit, and tested before a small preamp is designed for their re-use.

Below are the 4x tuned mica compression capacitors used to tune something in the circuit. And on the left are 2x tuned Sony inductors as well. 2x of everything? Looks like we’re seeing the “L” and “R” channels for the mic input, but can’t be sure. Weird ass shit to see compression caps in a solid state machine.

That’s all, folks! Before E-wasting stuff, I try to repurpose it. In this case, we reclaimed some transistors, some compression capacitors, mic transformers, power transformer, and some whacky inductors to play with. Worth 20 min for sure.

Friden Singer 1117 Calculator (Nixie)

Such a nice little stamped piece of metal. On my giant calculator. As every hipster accountant knows, the bigger your calculator is, the better a human being you are. This one is roughly the footprint of my 2014 MacBook Pro, so I’m pretty great.

This is my Singer Friden Electronic Calculator. Electroniuc Calculator, of course as opposed to the computator, the mechanical calculator. Mechanical calculators were actually very very common once upon a time. Look up “mechanical calculator” and you’ll be blown away. Suddenly learning math using an abacus, as I did as a kid, seems practical in comparison to the mechanical calculator.

This device was part of the “calculator wars” of the late ’60s and early ’70s on both sides of both of America’s oceans: Japan, the USA, and the UK all competing to create the first LSI (“Large Scale Integration”) electronic calculator. The Sumlock ANITA, from the UK was an electronic calculator, but not a LSI based one. Their first forays into the market were based on mini DecaTron valve counters to do it’s mathematical operations. The first real LSI calculators were Rockwell “calculator on chip” based designs. That’s why this big boy is so big: no “1-chip” silicon solution on the market at the time.

This Friden 1117 is actually a sort of neither discrete logic nor fully LSI. It is in fact a very very early LSI calculator, yes, but it’s not integrated into a single chip. Rather, it’s spread throughout the ceramic substrate IC’s inside. An old timer I work with pointed out that their die lids are spot soldered onto the IC’s inconsistently, which happened because people were inserting the dies and finishing these early logic IC’s by hand inside a vacuum chamber. Crazy. Below are a few images of the main PCB inside. I believe the OEM stamps from the fab are Hitachi from Japan (unit says “made in Japan” as well)

“Large Scale Integration” in the ’70s was really just saying “not discrete logic,” not “system-on chip” integration that’s been happening since. “Integrated?” Like, into 10 chips, instead of 96? Hilarious. Hand drawn PCB traces. Card Edge Connectors to connect the bottom and top PCBs. Phenolic board. For once, I’m taking photos before I put new capacitors inside.

Above is the keypad connector. Wild shit. Above the brass screw slightly to the right you can see there is a single key pressed down, with a ferrite bead containing a tiny magnet sitting next to the blueish glass package perpendicular to the black plastic key slot. The blue devices are glass encapsulated hall sensors, and there is one for each key. It makes the calculator prone to missing keystrokes, if you type more than one or too fast, but it has a major design advantage- they’ll simply never stop working short of major physical damage. Beautiful.

Here’s the whole assembly. When I recap this thing, I’ll pull the panel PCB and inspect it for caps. I don’t currently have schematics, but recapping by eye is fine with smaller devices like this Hand wired rotary switches on the right. Cables all hand tied. This thing is a real labor of will and love.

Above is the peculiar looking panel: 0-9 on a rotary switch, I assume as a memory recall function. These early calculators often advertised memory functions. As I currently don’t own a manual, I can’t tell y’all how it works yet. The 5/4 key is a rounding key: up, off, or down. It’s used for doing decimals. For example, with the switch down, if I do 63/2 it gives me 31. The same operation yields 32 with the switch up. Crazy.

Multiplying and dividing by decimals is a non-starter. It throws all 0 no matter what. It also takes a measurable second or so to do division and multiplication as well, flashing gibberish to the display for a split second as the IC’s run their calculation inside. If it overflows, an indicator comes on. I can’t wait to get a manual and schematics!

If it were ever to die, I’d make a PCB to re-use the nixies (actually called Numitron valves) as a display for my MIDI record time on Logic Pro X. Or something else cool. I put new capacitors in it, so it aught to last for a few more decades. It may need some new diodes at some point though. I can’t explain why I find these fascinating, but I do. There ya go- my überhipster accountant’s wet dream.

Preamp Repurposing

This preamp needed to be rebuilt in a big way. I plan to use the EM84 bargraph valves to indicate amplification in an amplifier soon, and it was just put in a badly conceived housing, so I dissected it.

As I am accustomed to Printed Circuit Boards (PCBs) being the norm for every build I’ve ever done, I am still adjusting to the process of layout and placement for point to point wiring vacuum devices. The spaghetti above is a perfect example. It looks like a Dog’s Breakfast, as Dave Jones would say.

The new design will also have a hexfred rectifier, and I will be doing two ECC803s tubes instead of one with the triodes split to different channels. I believe I’m going to be doing each dual triode as a gain stage followed by a cathode follower, but I’m not sure yet. I intend to use a choke in the power supply, and chokes in series with the plate resistors as well. After using an oscilloscope and doing some experimenting with my new EL84 amp, I’ve concluded that RC (Resistor-Capacitor) filters for ripple are not the way to go. Inductors are harder than resistors to understand, but their effect on current pulses and ripple noise is invaluable, and after all is said and done, the LC (Inductor-Capacitor) filter is far more cost and space effective. The guys at Silvercore Audio have a great blog about using LCLC filtering wherein the second inductor is coreless, and how this topology is (in their opinion) the best way to filter supply rails for a project (48 coils and xfrmrs in their particular design, however, so we stick with LC filtering.)

My company accidentally ordered some nice inductors when we were spec’ing parts for a new power supply, and since they were going to get thrown out, I commandeered them on their way to the dumpster. I also saw some bad ass really cheap, though unshielded, chokes on Mouser being made by Triad Magnetics, which are decent for the price as long as they are inside of the chassis. I need to finish some things before I buy them though. So for now, I’ll just use what I got.