My friend Jeremy has a 1995 Mustang that had the factory premium sound system in it when he bought the car used. The CD player was broken and he had the head unit replaced within a couple of weeks of owning the car. He later added a subwoofer.

I’ve always thought the stereo lacked clarity in the bass, and the head unit and EQ have had some quirks. Recently Jeremy pulled the head unit and found all sorts of interesting techniques used by the aftermarket installer that will be the subject of a later monologue … but one of the things we discovered is that the amplifier for the door woofers wasn’t working at all. Swapping it with the amp for the rear deck woofers caused them to go silent and the (shot) door woofers to work again (after resoldering their cut cables).

Turned out to be a delightfully easy fix.

The black connector on the bottom is input and the grey connector on top is power in and speaker out. Although both connectors are pinned for two channels, only one channel is actually connected — this is a mono amp. Interestingly, all of the audio interconnects in the factory sound system are balanced, which should help cut out alternator noise and clicks and pops from other electrical systems — a luxury neither of us has seen in aftermarket sound systems.

When first I opened the unit, I was disappointed not to find any smoked components. Since we got no output whatsoever, I really expected to find a melted power transistor or something else dramatic.

However, it didn’t take long to discover the dramatically cracked solder joint on this lead,

which turned out to be the inductor in the pi filter on the power input,

so named because of the schematic’s resemblance to the letter π (my favorite vitamin).

I wanted to be sure to get all oxidation and dirt out of the joint; so instead of merely reheating it and adding more (flux-core) solder, I scraped both the pad and the lead clean with a chisel tip. While doing so, I noticed that the whole inductor felt loose and discovered the less visually obvious second cracked solder joint.

My assessment is that the inductor was installed somewhat carelessly at the factory and left sitting slightly above the PCB. Lacking contact with the board to prevent downward motion, a decade (or perhaps less than a year) of vibration from the rear end of a sporty coupe with a tight suspension broke the solder joints, thereby breaking delivery of power and disabling the whole amp.

After cleaning both joints, I reinstalled the inductor, this time installing it flush to the board as well as bending the leads over for increased contact area before resoldering, in hopes of increasing the mechanical strength of the joints.

The capacitors to the left were already hot-glued in place to prevent damage from vibration, and before reassembling the amp, I glued down the whole pi filter.

I got in touch with Jeremy to let him know exactly how lucky he was, and last night we reinstalled the repaired amp and confirmed that it now works completely.

Credits

Thanks to flemworld.com for providing the name of the factory sound system, as well as wiring diagrams invaluable in the larger project but not directly relevant to this repair.

Both my Crumar T1 and T2 “portable” organs (the lower two cases in the stack) came to me without swell (volume) pedals. Each has a rotary potentiometer on its control panel for master volume, but I really want to be able to change the volume dynamically while playing. I’ve been using a Dunlop volume pedal (built into a rocker case identical to the CryBaby wah) on the organ’s output; but (at least when used with the organ) all of the pedal’s action is in about the lower quarter of its physical range, so it’s very finicky to use.

I recently bought this original T1 swell pedal on eBay, listed as untested / project. That usually means tested / didn’t work / can get more money if I don’t admit that I already know it doesn’t work; but I figured I could fix whatever was wrong with it. And I have.

Optical Volume Control

When I plugged the pedal into the organ, the pedal had no effect on the volume, regardless of its position. Opening the pedal, I found this optical volume assembly, with the lamp dark. I measured across the lamp’s leads and got about 14VDC, so power was definitely present and the lamp was definitely burned out. The volume of the organ changed as I exposed the photocell to light and shaded it with my hand; so the connector, cable, and photocell were all still working and the pedal should be easily reparable.

To back up a bit and explain what’s going on, some rocker pedals (Crybaby wahs) use a rack and pinion gear or a string and pulley to turn a rotary potentiometer as the pedal is rocked back and forth. These have the disadvantages of being relatively costly to construct, having the mechanical pot get scratchy as it ages, and being a pain to service.

The Crumar’s swell pedal — and older console organs’ swell pedals, from what I’ve been told — uses a photocell (light-dependent resistor) instead. The pedal’s rocker action moves an opaque tab with a V-shaped notch up and down between a light and the photocell. The width of the part of the notch currently positioned between the lamp and the photocell controls the amount of light falling on the photoresistor, changing its resistance and changing the volume.

It should have been easy to replace the burned-out bulb, but I couldn’t find a part number on it. I could search for a lamp with the same physical form factor, but I have the impression that the same size bulb is available in different power ratings. Without the part number, I wouldn’t know which brightness to buy; and a wrong brightness would change the behavior of the pedal.

LED Replacement

Rather than mess with trying to find the right bulb and risk installing a bulb too bright or dim, I decided to replace it with an LED. I could easily adjust the LED current (hence brightness); and without a filament to burn up, the LED could last even longer than a replacement incandescent bulb.

I started by sanity-checking the photocell so that when I got the LED installed I’d know how well I had done. With the pedal disconnected from power, I got easily 1MΩ resistance with the photocell as dark as I could shade it with my hand while keeping the ohmmeter probes on it, and easily 1kΩ resistance putting it directly under a light.

About this time, I noticed that the V-notch was oriented with its wide end between the light and the photocell when the pedal was rocked back to the minimum volume position. That is, quiet == lots of light == low resistance.

That makes sense. With the pedal not connected, the organ is at full volume. The pedal’s photocell is probably an optional parallel element in the lower half of a volume voltage divider, and lowering its resistance lowers the output voltage.

I grabbed a yellow high-brightness 10mm LED with a far narrower viewing angle than I want to use for anything else (purchased before I started paying attention to the difference between mcd and lumens). It came packed with 470Ω resistors to use in 12VDC applications, and I figured that was close enough to the 14VDC I had measured when it was plugged into the organ.

The original lamp holder wouldn’t really work for the LED, and I may design and print a new one on the CupCake for later. But for now, I used a piece of solid 14-gauge wire as a bracket to get from the mounting screw to the LED, then cheated and doubled that as the ground connection. I added a 1kΩ potentiometer (wired as a variable resistor) in series with the LED’s current-limiting resistor so I could dim it if 22mA turned out to be too bright for the photocell.

Testing

I connected my bench power supply to the connector’s lamp leads and dialed it up to 14V to sanity-test the LED, shown above. I then held the back cover on the case and measured the range of the photocell resistance while rocking the pedal back and forth. In the real dark inside the case, the photocell got up to 7MΩ resistance; with the LED shining through the widest part of the V-notch, it got down to 750Ω resistance. These seemed well in line with what I had tested initially.

I mounted the original lamp assembly to an existing 4mm stud inside the case so a dedicated restorer has the option to put it back to the way it used to be, then closed up the case and went to test with the organ.

I started by setting the pedal to full volume and connecting it to and disconnecting it from the organ. I couldn’t hear any difference in volume with it connected, so it appeared to be set okay on the high end. I then rocked the pedal and got a very nice response curve, giving me smooth control throughout the range of motion, going to not quite silent at the lowest end.

It would be nice to adjust it to go completely quiet — but that would mean lower photocell resistance, requiring more light, requiring more LED current — and it’s already running at 22mA, slightly over its nominal current rating. I’d need to swap to a different LED, or possibly move the LED closer to the notch so the notch intersects the cone of light closer to the apex and allows more light through to the photocell. Then I’d be lowering the photocell resistance below its current minimum of 750Ω, and I’d start wondering whether I needed to inspect the voltage divider and make sure I’m not driving more power through the photocell than I really should.

I’m already delighted with the improvement and I can live with the current behavior. And I can certainly remove the unnecessary potentiometer if I ever have reason to go inside the pedal again.

I recently ran across on Craigslist a

Bad horsie 2 that was plugged into the wrong power supply and messed up, and needs some minor electronic work.

I was intrigued by the challenge (I’m such a sucker for broken things, dang it) and bought it. When the seller and I exchanged the pedal for my cash, he remarked that he read on a forum that it probably just needed a resistor changed, and that if I were handy with a soldering gun I could probably do it myself.

Uh huh. Resistor.

Let’s dig in.

The circuit board has a hole in the top for a foam battery “cage” attached to the enclosure, something clever that I haven’t seen before. And it had no obviously damaged components.

Testing the Power Supply

Like most (all?) guitar pedals, the input jack is a switching jack — the middle leaf doesn’t contact the plug, but is a normally-open switch in the power supply circuit so the pedal is powered off until the guitar cable is plugged in.

Tracing connections on the solder side reveals an interesting tidbit — the external power supply jack is not switched; only the battery is. According to the one-page owner’s manual, the pedal only draws 16mA; so (A) it should last a long time on a 9V battery, and (B) the pedal’s current draw is probably insignificant compared to the inefficiency of the wall wart.

The power supply jack is also switched — the lower and right terminals are normally-closed, powering the pedal’s positive rail from the battery by default but disconnecting the battery when the external power supply is used.

Finally, both the battery and power jack run through a series 1N4003 diode to protect the pedal from improperly-connected power sources. If this pedal was damaged by being connected to a wrong source, it was not merely wired in reverse — that would have been blocked by the diode. It must have been connected to a higher voltage than intended.

Since I didn’t know quite what to expect from the pedal and didn’t feel like setting up an instrument and amp to test it, all of my troubleshooting was at a purely electrical level.

With a battery connected and a plug in the input jack, I measured .87V across the battery terminals, where I should have had nearly 8.4V from my NiMH 9V-form-factor battery. I also measured a fairly low resistance across the two power leads (with power disconnected). Something in the circuit had shorted and stayed shorted (instead of blowing open); and either way, it should be fairly easy to find once I knew what I was looking for.

Removing ICs

My experience with circuits damaged by power surges or other excess power is that it’s almost always the ICs — the most complex semiconductors — that blow first. So my first troubleshooting step is always to remove them from the circuit and recheck the resistance across power.

The two ICs (a TL074 low-noise quad op-amp and CD4069 hex inverter) weren’t socketed, so I needed to desolder them in order to recheck the circuit with them absent. Rather than use solder braid, I tried out my new hot air pencil that Cort bought for me (and one for himself) in exchange for building him a hotplate (that I need to assemble and deliver).

I turned the pencil up to about 350°C and waved it around the solder side of each IC while rocking an IC puller on the component side. After a little time to warm up the solder, the ICs pulled loose very cleanly.

The hot air pencil scorched the residual flux from the factory assembly,

but after wicking the remaining solder and cleaning the flux, the board looks pretty good. The vicinity of one of the two ICs shows slight damage to the epoxy (?) with the fiber substrate showing through.

Testing with ICs Removed

With the ICs out, the resistance across the power leads was very high, as I would expect it to be.

I soldered in IC sockets and measured again with the ICs reinstalled in the circuit one at a time. The CD4069 was obviously bad and the TL074 was not obviously bad. Joel had a CD4069 on hand; I installed it; and all of the pedal’s logic appeared to work — although I didn’t know yet whether the audio was working properly.

The bad 4069 does have a shiny spot on it. At first I assumed it was flux splash; but I’ve seen melt like that on ICs where the magic smoke had escaped. Maybe the spot really is a symptom of the damage.

Plugging the NiMH battery back in and powering up the pedal, I measured 3.5V across the battery leads, using either the original TL074 op-amp or a TL084 that I had on hand. This still seems like an awful lot of voltage drop for an allegedly 16mA pedal (even given the internal resistance of an aging NiMH battery); but any further diagnosis was going to require more information about symptoms.

Reassembling and Testing

Taking a chance on the TL074 (in part because I didn’t have any on hand), I reassembled the pedal with the original op-amp installed and took it to Jeremy’s house last night for him to try out. (He’s on kind of a wah pedal kick right now, and I’m feeding his habit.)

He grabbed the first power supply he saw, plugged it in, and started playing. It worked, but we had massive 60Hz hum, which didn’t really make sense — why would a fault in the pedal that I hadn’t found and fixed cause that much hum? The power supply section doesn’t have much by way of filtering, the audio section shouldn’t be specifically filtering out 60Hz, and the whole thing shouldn’t be more susceptible to hum than usual because of a broken component.

I suggested that we try a different 9VDC power supply, and went to change it out. Then I found that Jeremy had plugged my 9VAC x0xb0x power supply into my 9VDC Bad Horsie pedal. That’s right, he put 9VAC into the pedal and it still worked, albeit with hum. It was half-wave rectifying AC and that single capacitor filter was enough to get that baby going.

Which means whatever the previous owner did to it involved more than 9VAC — probably quite a bit more (or quite a bit of DC), since the CD4069 is rated for up to 15V supply. I’ll never know — although I really wish I would — exactly what he did to it. Maybe a 24V inkjet printer power supply???

After swapping in the right power supply, the pedal appeared to work correctly — although Jeremy didn’t especially like the sound running his guitar straight into the pedal. I found forum posts agreeing with his opinion that it really needs distortion before the Bad Horsie, which he played with more tonight.

While looking up the pedal’s manual today to find exactly what the controls do, I saw that Morley provides the schematic on their web site. If Jeremy decides that the audio portion of the pedal isn’t working correctly, or if I get too annoyed thinking about how much more current it’s drawing than it’s supposed to, I’ll definitely use the schematic to continue repairing the pedal. Still, this goes to show that one can do a reasonable amount of troubleshooting just by examining the construction and behavior of a circuit without having or using a schematic.

Planning the Paths

Besides the obvious goal of keeping the wires tidy, my main goal was to position slack in the cables so that the back panel could be lifted out of the case and set to the rear, allowing full access to both circuit boards, without disconnecting anything.

Before doing any real lacing, I mocked up the cable path to make sure everything would work. My first attempt, here, had enough length for the J3 bundle from the lower right of the main board to the right end of the I/O board; but J7′s wires coming from the left edge of the main board looped around too much before heading to the I/O board and didn’t reach their destination.

My second attempt worked much better. I brought J3′s large bundle up in a clockwise loop and all the other connectors’ bundles in an overlapping counterclockwise loop. All the cables reached where they needed to go, all the three-wire cables got tied together into a larger bundle that would hold its shape well, and the two loops flexed well to allow the back panel to be moved back and forth.

Attaching the Cables

I removed all the cables, then started with the longest run, from J7′s connector on the main board to its solder connections on the I/O board. I then used its path as a guide to size J5′s cable to length, soldered in J5′s wires, and tied the two together.

I repeated for J4 and J6, tying up after each to ensure that every successive bundle would follow its intended path and get sized correctly.

After the CCW section was finished, I soldered the J3 bundle to the right end of the I/O board and tied the two bundles together into the overlapping loop.

Since I now had everything attached, I figured I’d test the x0xb0x to make sure it worked before lacing everything into place. It was at this point that I found I couldn’t close the case because of J3′s connector being at the wrong end.

Besides having to desolder J3 at both ends and reattach it, this broke my plan to have the overlapping loops tied together. It wouldn’t do much good to have wires soldered to the I/O board tied to wires soldered to the main board. I could still leave the overlapping loops; I just couldn’t tie them together.

Lacing the Harnesses

After desoldering J3 and setting it aside, I laced the harness that holds everything else. I started at the I/O board, locking and wrapping every time another cable joined or left the bundle and removing my temporary ties along the way. Because of the number of branches, the usual advice of two and a half times as much lacing cord as the length of the main trunk left me short, and I had to run the main cord up to J4-J6 and start a new cord to finish out the run to J7-J5 instead of the other way ’round.

Note that I ended up routing J7 as a short branch off the run to J5 rather than splitting the two in a Y like I mocked up.

I understand that it would be more proper also to tie every bundle of three wires after it splits off the main trunk; but with only about an inch on each after the branch, I’m satisfied with the results.

The harness holds its shape well and it’s easy to see where every three-pin connector is supposed to plug in.

After reattaching the J3 buncle in the proper orientation, I laid out its path and made temporary ties to hold it in position, then started at the main board end and worked toward the I/O board. I finished with the wrap that holds the wires in shape as they fall to the left of the J3 connector on the I/O board.

Again, I’m pleased with how well the J3 harness holds its shape, especially as it doesn’t tie to any points along the way. The two harnesses’ loops overlap nicely in the middle — slightly differently than I had routed them here — and the back panel is very easy to put into and take out of the case.

Adafruit produced x0xb0x kits in batches of 100 as Limor was able to track down enough “rare parts,” order circuit boards, and assemble the common parts into kits. I’d been on her waiting list since 2008, so was terribly disappointed when she officially announced what we had all come to realize anyway — that tracking down the rare parts was becoming enough of a hassle, she wasn’t having any fun doing it and wasn’t going to produce any more kits.

Happily, Limor announced shortly thereafter that thanks to Adafruit’s open-source hardware license, James Wilsey of Willzyx Music in Taiwan has taken up the torch and would shortly be offering x0xb0x kits.

I browsed the Willzyx web site and saw an Express Kit with rare parts, circuit boards, and panels, but no case and common parts. I emailed James to thank him for making the kits available and ask whether he’d be producing full kits. He replied that he would shortly; and in spite of having a no-preorder policy, offered to hold one for me as soon as they were ready if I was interested.

I said yes, he soon told me that the kit was ready (even though the updated instructions weren’t yet), I sent payment, and he sent my kit, very nicely packed and with quick shipping and international tracking. James has been an absolute pleasure to deal with, pre-, mid-, and post-order.

What follows are a few notes from the build — not a time-lapse of everything all the way through, but things that struck me as unusual or interesting. All mistakes I made were my own fault for not waiting for the updated instructions.

Testing Each Stage

I’m just a bit too young to have assembled a Heathkit, so Adafruit’s instructions for the first several stages were novel to me — she identifies a section to assemble, then stops to have you test what you’ve just built. The power supply section, for example, includes notes for testing AC voltage and regulated DC voltage at two different points. Modular testing is a great way to make sure you don’t spend 40 hours putting your kit together and then have no idea why it doesn’t work.

Willzyx, by the way, provides sockets for all the ICs (which I hadn’t noticed when I soldered the op-amp above) and nice connectors for the inter-board jumpers (which I hadn’t noticed when I provided my own). Using his instructions will be much easier than desoldering these parts later.

Diffusing the LEDs

The kit includes high-brightness, non-diffused red LEDs. I don’t like the look of non-diffused LEDs, especially as panel indicators, because their low viewing angle makes them hard to see unless you’re looking at them dead-on (at which point they’re too bright). So I used my technique from four years ago (when I had just started the blog) and frosted them.

The diffused LEDs are noticeably translucent instead of water-clear like the unmodified originals.

And they provide a nice, even illumination across wide viewing angles. (Note that they’re really quite red when you look at them with anything other than this camera.)

Once I had all the LEDs prepared, to solder them with proper alignment to fit the front panel, I inserted them in their standoffs and did the old trick of tack-solder one leg, then adjust the LED while reheating the joint. Then I put the front panel over the LEDs (further on than it actually fits in final assembly) to hold them all securely in position while soldering the cathodes and resoldering the anodes.

J3 Plug Goes on the I/O Board

I had the x0xb0x nearly completed and ready to install all the cable bundles when I noticed that Willzyx provided wiring harnesses with a mating connector at one end to make servicing the x0xb0x easier. (The original kit just provided ribbon cables to solder at both ends.) Even after emailing James to confirm which board each plug was supposed to go on, I got this one wrong.

And when you get it wrong, the case doesn’t close. This is at the narrow front end of the case, and the connector is too tall. This was a nuisance to desolder and redo — but I ended up with tidier wiring the second time around, so it wasn’t all bad.

Shorten the Rotary Switch Shafts

The original Adafruit pre-assembly preparation instructions say to cut all the potentiometer shafts to the same length. Willzyx provides potentiometers that already have same-sized shafts (thank you!!!), so I disregarded this instruction — at my own peril.

Yeah, the potentiometers all have same-sized shafts, but the two rotary switches have shafts way too long for the knobs to reach the ground. These would have been much easier to cut before installing. Just another case where the updated instructions will solve the problem for everyone else, and I should have looked at the parts more carefully since I knew I was building without up-to-date information.

Note that the flats on the shafts (which I refiled for the knobs’ set screws) are opposite the selected function. I powered it up to double-check before filing.

Done!

After two weekends of intermittent work, it’s done and 100% functional. I’ve been having fun this afternoon tweaking knobs to the bass line of “Sweet Emotion,” the most interesting one-bar pattern that I had stuck in my head. Thanks, Limor Fried and James Wilsey!

• Capture a short sample of a MIDI performance, including key velocity data.
• Quantize to a tempo set by a “tap tempo” pedal continually and dynamically throughout the capture, rather than to an LED or click track.
• Loop and play back, by default to the last tempo seen but honoring continuous “tap tempo” data from the same pedal.

Using a MIDI sequencer with these “tap tempo” features should give greater flexibility for capture and playback during a live ensemble performance than using a traditional audio looper, which requires the whole ensemble to play to the tempo recorded in the loop.

But my real motivation is to be able to play a pattern and then make gradual, multi-bar changes to the analog character of the sound without having to continue playing with one hand and turn knobs with the other.

Record a one-bar pattern on a MIDI keyboard driving a x0xb0x (or a real TB-303, if you’re filthy rich enough to have one and a DIN-sync MIDI adapter to go with it), then play it back and slowly tweak the knobs while everyone else jams on for a bit.

Am I going to find that all of this functionality already exists within the x0xb0x? (It looks like it might be close — MIDI ports; internal sequencer; variable tempo, although perhaps not that sophisticated.) Alternatively, are there MIDI sequencers that do all of this? Is this de rigeuer for every sequencer under the sun?

I’ve always been curious about the construction quality of commercial patch cords — just how good are the connections buried under those lovely molded jackets and strain reliefs?

Naturally, the faulty end was the last one I disassembled. (Logic joke!)

A failure rate of one in twenty-fivish (I have three bundles of patch cords) is quite a bit higher than I expected, especially as the fault was a solder joint broken clear off of the center pin.

Notice how little of the braided shield wire remains. This doesn’t speak well to the overall quality and reliability of the patch cords — even though this part of the wire is cast inside the rubber housing and can’t flex.

Note also that the cable has two signal wires. The manufacturer apparently uses the same cable for TS (tip-shield, for unbalanced signals) patch cords as they do for TRS (tip-ring-shield, for balanced signals) patch cords. (The same bulk cable could be used for stereo unbalanced transmission — although I’d prefer to see each signal wire individually coaxially shielded.)

At this point, my curiosity shifted to whether I could fix up the cable and put it back together in a way that I found aesthetically acceptable.

When I mounted the plug to desolder the braid so I could strip back the jacket and start with fresh sections of copper, I noticed that the solder was very grey. I suspect it’s lead-free and I suspect that the greater challenges of lead-free solder contributed to the broken solder joint.

I reformed and tinned the shield wire and stripped and tinned the tip wires. I also tinned the shield and pin so the joints would go together well — which they did. And I tested for continuity and shorts after soldering, to be sure of my rework.

I slipped the broken pieces of insulating plastic back between the wires so the shield wires wouldn’t short out onto the pins. I then carved out the interior of the molded jackets, since I wasn’t going to be able to realign the insulation and wires exactly with the recesses from when the housings were originally formed.

After test-fitting and carving out more of the molded covers a few times, I got a fit I was happy with. I put the covers back on and sealed the deal with heat-shrink, which did an admirable job of conforming to the shape and texture of the original molding.

Making the repair was never an economic decision; but having fixed the patch cord, I may as well use it.

Who Made It?

I could never find a manufacturer’s logo on the patch cord or molded ends, nor does the cord look identical to the current catalog entry for the patch cord bundle I think it is. I’ll avoid buying anything more from the brand I suspect; but because I’m not certain of it, I’m not comfortable mentioning it here.

It’d make an interesting demonstration to buy a couple of packages each of several manufacturers’ patch cords and dissassemble and inspect them like this, to assess the build quality. Maybe a musicians magazine has already done that review?

My x0xb0x kit has arrived from Taiwan!!!

Has light scratches, small amount of rack rash, in perfect working condition- no issues whatsoever. Has been used in my guitar rig for the past several years with no problems.

It arrived oddly but adequately packed and … as you can see, not in perfect condition. I would go so far as to say it had issues. I suspect had I tried to use it, I would have had problems.

Well … I could complain to the seller, who would tell me it was damaged in shipping, and then I could try to deal with the USPS who I don’t think broke it, and I could spend a lot of time and frustration and maybe get some money back and probably end up with no compressor. Or I could just fix it myself and have a little fun in the process.

There’s not much inside modern rack equipment — it’s all about room on the front for controls and the back for connections; the depth is determined largely by the placement of the power supply. And maybe by needing to be large enough to make consumers feel like they got their money’s worth.

The compressor had a snap-in C-14 connector with spade terminals. I had a snap-in connector from a floor wart, but it had narrower terminals that the DBX’s spade connectors wouldn’t engage securely. I had scads of C-14s from PC power supplies, but they were all screw-mount instead of snap-in and I didn’t feel like modifying the DBX quite that much.

Okay, back to the source. The first two dead power supplies I pulled out of my boxes had snap-in C-14s.

The donor jack has spade terminals the same size as the broken original and even had a ground wire already soldered on. Sweet!

Reassembly is the reverse of disassembly. Piece of cake! It worked great this afternoon at Ron’s studio.

Bent Rack Ear

And by the way, is this the light scratches or the rack rash? I’d love to be generous and write it off to shipping damage, but I really think it was packed well enough to avoid damage in transit so I have to believe it started that way.

Again, it felt like a better ROI of my time to fix it myself than complain. I pulled all the knobs, unbolted the panel, wrapped it in card stock to protect the finish, and straightened it in the bench vise. It’s all good.

Crackly Volume Knobs and Stuck Master Tuning Potentiometer

Several of the volume knobs were pretty crackly.

Most Crumar keyboards are wonderful to service because of how easy it is to get inside. After removing a few screws, the top panel lifts back on its rear hinge, without even having to take the knobs off all the controls.

Crackly controls need wiper cleaner sprayed into them to loosen and remove the gunked-up graphite and grease built up on their resistive paths. As my can of wiper cleaner turned out to be empty, I used a spray can of silicone lubricant instead — it’s good at softening gunk and of course also at lubricating. Note that something in silicone lubricant (I’m guessing the solvent carrier) dissolves some kinds of plastic — this could be heartbreaking if used on something that mattered (like, say, the odometer from my ’67 Fairlane).

The panel-mounted controls (including the stuck master tuner) were easy to get to with the silicone spray, after which I worked them back and forth over their range to use the potentiometers’ wipers to work loose the now softened gunk. After a couple of applications, they now operate quietly.

Although I used a rag to protect the panel from overspray, I can see in the photo that silicone/solvent still oversprayed or later dripped out of the potentiometers and stained the interior of the cover.

Crackly Keys (Noticeable When Using Overdrive Effect)

Several keys on the upper manual cut in and out and crackle a bit, which is especially noticeable when running the keyboard through an overdrive processor.

In addition to top panels, Crumar manuals (keyboards) are typically hinged as well, lifting up to allow easy access to the underside of the keys and to circuitry underneath. (I love my Crumars!)

Hm, the undersides of the keys have downward-pointing tabs holding one end of springs that are probably being pushed down to make contact with a bus wire running the length of the keyboard. That looks like something I really don’t want to completely disassemble, clean and reassemble. Maybe I’ll fix some other things before worrying about the key crackle.

D♯ / E♭

Up and down the keyboard, the D♯ / E♭ key didn’t make any sound unless (A) I had the 1′ or 2′ drawbars pulled out and pressed one of the highest E♭s on the keyboard or (B) I had a non-octave drawbar pulled out. In other words, only the highest E♭ the organ could make was producing sound; lower octaves of E♭ were not.

Sidebar on Hammond organ drawbars and additive synthesis: Hammond organs and clones have drawbars labeled in feet (16′, 5 1/3′, 8′, 4′, 2 2/3′, 2′, 1 3/5′, 1 1/3′, 1′), with 8′ being the approximate length of pipe needed to make the pitch of the lowest note on the keyboard of a pipe organ and the other lengths representing lower and higher pitches. Each drawbar has nine positions (full off to full on) and mixes the amount of the corresponding pitch (fundamental, octave above, octave and a fifth above, etc.) into the sound created when you press a key, allowing you to create different timbres of sound (hollow, reedy, cathedral organ, etc.) for the same pitch.

Thus a failure of all the E♭s on the keyboard seemed like a problem with the tone generation, drawbar (selection), and/or mixing circuits. The manuals’ wiring harnesses break out to edge connectors:

Each edge connector appears to correspond to an octave’s worth of keys. The vertical backplane’s traces go down to a horizontal backplane and continue toward the front of the organ:

Where they disappear beneath a black cover, which I removed.

Yow, look at all those little circuit boards! Some have one IC on them and some have two; they’re divided into two banks (left and right) that match the wiring harnesses that coming from the upper and lower manuals; there are an awful lot of these little boards and there are an awful lot of keys on the manuals.

Hm, let’s have a look at what’s on them.

Looks like a TDA0470D with supporting resistors. So what’s a TDA0470D? Google gives:

• A Polish IC sales site listing it as a transistor array
• The Vintagechip site listing it as being used in Crumar T1s, among other things, and claiming that it’s a 10-transistor array and that at datasheet is available
• An edaboard.com post describing it in more detail as a multi-gate chip and claiming that it’s equivalent to the TBA470
• The MIDI Gadgets Boutique product selector page listing the TBA470 with datasheet

Transcribed:

Gate for Electronic Organs

Monolithic integrated circuit in bipolar technique, designed primarily for use in electronic organs. The device incorporates ten transistors, each replacing a mechanical key contact. Thus it is possible to reduce the numerous mechanical key-contacts on conventional organs (up to ten per key) to one single contact per key.

Each of the ten emitters may be driven by a tone-signal. The sum of all signals will be derived from the common collector (terminal 14) or if the signals are supplied into the base terminals, via an integrated diode from terminal 1. Any undesired peaks caused by blocked transistors are suppressed by this diode and an external capacitor.

So … the tones are presented on all of the inputs, a single key gates them, and this IC mixes them together? Sounds like these chips are the keys’ gate mechanisms for the drawbars, which is what the datasheet suggests. If each IC corresponds to one key and since the T2 had multiple keys with the same problem, I figured I needed to look further upstream toward the tone generation.

While looking at the gate PCBs, I noticed that just behind the long, front PCB, the backplane was labeled Do – Re – Mi – Fa – So – La – Si, with the positions of the notes approximately corresponding (with appropriate gaps for sharps) to the positions of the vertical traces on the front PCB. Perhaps the front PCB is involved in tone generation, then?

I pulled out one of the front PCB’s ICs (with the power off, then powered back up) and a whole swath of keys across the keyboard quit working. Ah ha! A little experimentation yielded that the leftmost IC makes all of the C – C♯ keys (not) work, the remaining upper ICs make all of the D – F♯ keys (not) work, and the lower ICs make all of the G – B keys (not) work. Now I was getting somewhere!

Further experimentation yielded that the upper of the highlighted ICs was the culprit; I could move it around to the other notes’ banks and the problem followed the IC.

So who is this mysterious stranger? A 4727BPC. Back to Google:

• At Datasheet Archive, a listing as a binary up counter claiming 15V supply
• A Synthforum post suggesting that it was CMOS
• And back to Vintage Chip, the HBF4727 and a claim that a datasheet was available.

Hoping to find more information about exactly what the chip did, I emailed Vintage Chip inquiring about a datasheet and was delighted to receive a reply within hours from Valter with a copy of the pinout attached. He doesn’t have a full datasheet but is expecting one within a month or two and will email me when he has it.

From the pinout, I could tell that the 4727BPC / HBF4727 is a bunch of independent toggle flip-flops used for octave division, which makes sense — the organ would have a circuit to generate all the pitches of its highest octave, then divide them into progressively lower octaves. The failure of one of the divider chips would cause lower octaves to stop producing tones, but the highest (original) octave would still be accessible from the highest drawbars of the highest keys.

Curiously, when I moved the faulty IC around within the range of ICs dividing D – F♯, I expected that I would get progressively more of the higher E♭s to work and only the lowest wouldn’t (because only the lowest weren’t being divided / generated), but my recollection is that all of the E♭s on the keyboard continued to not work. This seems odd.

In any case, because I want my dual-manual T2 working so I can practice on it, I borrowed a 4727BPC from my T1 and all the T2 keys now work fine.

Remaining Issues

Some of the keys are still crackly. I’m afraid I may have to disassemble all the keys and clean the springs and contacts.

When using the Line 6 POD guitar amp emulator to approximate John Lord’s overdriven Marshall stack, I hear a lot of static coming from the organ that I don’t hear with anything else plugged into the POD. I notice that all the op-amps in the T2 are 741s — maybe it’s time for an upgrade to TL081s (which as far as I can tell are pin-compatible, including the offset null).

I’ve left my T1 inoperable by stealing the 4727BPC from it. I could buy one to repair the T1, but I wish I could find an entire cosmetically trashed T1 or T2 to buy for salvage parts. Right around the time I was working on this, I saw one on eBay but let it slip away, and I’ve been kicking myself ever since.