In the days of the early to mid 1970s when rather harsh sounding cone and early dome tweeters dominated, a physicist by the name of Arnie Nudell sought to create a far more revealing, more extended, and smoother sounding tweeter that possessed all the advantages of electrostatic designs without their weaknesses and fragility. The result was the Electro-Magnetic Inductive Tweeter (EMIT), a hybrid planar-magnetic/quasi-ribbon design that incorporated the brand new technologies of medical grade samarium cobalt magnetics and Jim Winey's speaker advances. When these EMITs, remarkable drivers for their era and still excellent to this day, were incorporated into his speaker designs, the result was some of the first truly modern speakers that embodied flat, accurate sound with an extended frequency range most new speakers still cannot recreate. These thirty year old designs (starting with the Servo-Statik and advanced with the Quantum series) can still hold their own with today's high end and trounce many designs costing upwards of 10-20x the original prices of these old designs.

Unfortunately, while three decades is a remarkable lifespan for any excellent speaker, the years age the components just as they do the owner. The most common issues to be encountered over the years are the foam on the woofers rotting, the potentiometers corroding, and the midrange dome tinsels breaking due to a hardening of their coating. I've already written articles concerning how to fix these. However, another common problem is with the EMIT tweeters seemingly failing both in their ability to produce sound or with the magnets breaking loose (typically both). So, I will use this thread to show how to affect the two most commonly needed repairs.

- JP

Quite often, one will encounter EMITs that seemingly have "gone bad" in that they do not produce sound. The common explanation is that the diaphragm has somehow been "blown" possibly due to a clipping amplifier in the past. When the EMIT is removed and measured for DC resistance, it will read quite high (40-50Ω or even more) and the owner will assume that it is "open". A replacement will then be sought, either as one of the ever increasingly fewer and thus more expensive NOS replacement diaphragms or more typically as an EMIT removed from another parted out speaker (typically QLS-1s as they maximize the profit for greedy owners) and in trying to save one set of old speakers, a larger rarer set has been sacrificed. This needn't happen and for most "bad" EMITs, the repairs are quite simple and take a minimum of technical knowledge and almost zero cost.

The EMIT in the attached picture is a rear firing unit from a set of very early Infinity Quantum 2s. It is fairly square with a 4x4 slot arrangement that differs from the 2x8 of the much more common taller EMITs used to create the line array on the front baffle. The repairs are the same for either, but this squarer unit, due to the power of the rare earth magnets presents greater difficulties at dis and reassembly and tend to be more prone to magnet issues to be discussed.

The first step towards EMIT repair is the most obvious, the driver must be removed from the speaker. Most EMITs are held to the baffle with four 1/4" hex headed screws, which may be removed using a 1/4" hex tool (typically those screwdrivers with interchangeable bits are 1/4"), socket, or wrench. Carefully pull the driver out after the screw are removed and check to see if there is any cloth tape holding it to other units (typical of the Quantum series) and make sure you aren't straining the connection wires too much. For, the connections are lugs soldered to a circuit trace on the diaphragm that can torn from the driver making the repair a bit more difficult. Take a soldering iron and quickly disconnect the wires, making sure you observe which wire (plain or with a black stripe) goes to which connection.

- JP

Next comes disassembly. Remember, the Samarium Cobalt magnets used in the EMITs are very powerful and if you simply remove the six screws on the faceplate holding the front and back plates together, the two halves will twist, slide, and snap in ways that can damage the diaphragm, magnets, and your fingers. The best way to part the two halves is to first remove the two screws in the middle and replace them with longer (1.5"-2" is fine) screws of the same thread type as shown in the attached picture. These longer screws will serve as guides to hold the halves in alignment where the magnets repel one another rather than slightly off to either side where they will snap together very tightly. Once these screws are in place, remove the remaining four and creep the front plate upwards until you can feel a great weakening in the magnetic force. Then you may carefully remove the two screws and set the halves on your workbench. Watch for the two resin/fibreglass spacers and the diaphragm that will now be loose.

- JP

Make sure you mark which way is up on both halves for later reassembly as the magnets need to be arranged so their poles align to properly cause the diaphragm to behave as a quasi-ribbon. In the attached picture, you will see the diaphragm sitting upon the backplate. Also visible are the small punctures and scuffs caused by broken magnets that was part of the reason for the repair of this particular driver. The penciled number at the top is the DC resistance of the diaphragm as measured at the time of manufacture. In this picture, the voice coil trace is visible in its entirety so the principle of operation becomes evident.

- JP

The diaphragm is held in place by simple pressure between the halves, so it may be gently picked up to be removed, same with the grey spacer. The magnets pictured on the backplate tend to be firmly mounted and do not suffer the same issues as the front magnets. The felt is there to absorb the backwave thus making a naturally dipole design into a monopole. If all looks fine upon visual inspection, set this plate aside.

- JP

The trouble with these drivers is that the glue that holds the powerful magnets to their respective plates breaks down over three decades. Typically, the pressure of the screws holding the plates together is sufficient to keep the magnets in place. Sometimes, if the speaker is dropped or struck near an EMIT experiencing glue break down, one or more of the magnets can start slipping and the force of the field will cause the bar to fracture and the pieces be drawn to an adjacent intact bar. The result can be some damage of the diaphragm's Mylar. Sometimes these fragments can appear to be intact, but upon disassembly of the driver to solve the high resistance problem they can fly about when freed and shred the Mylar if one is not careful.

- JP

The spacer removed and the other fragments of the shattered magnet stuck to the plate. Unlike the back plate, the magnets on the faceplate are a bit narrower and are single continuous bars rather than two bars set in alignment. The yellow plastic coating on the front magnets serves to insulate the conductive magnet material from the voice coil trace (on the front of the diaphragm only) should they touch.

- JP

To repair the magnets, basic methods appear to be sufficient. It is important to determine how the magnet was originally oriented. To reglue the magnet in the opposite way can cause either the diaphragm in that section to fail to move properly to generate sound or do so out of phase with the rest. When you are certain how the magnet originally went and how the fragments fit together, make note of it.

Before you can glue the magnets back into place, appropriate spacers will be required else the fragments will attempt to twist and slide towards adjacent magnets in accordance with their poles. Use the spacing between the intact magnets to construct layered cardboard spacers (or wood if you have thin pieces) of the appropriate thickness. Ideally, the magnet bars will sit equidistant between the slots and will thus be the same for the voice coil traces.

For this particular unit, two magnets were broken, so three were required to hold all the fragments in place while the glue dried. To that end, I used a polyurethane based glue that seemed to perform its function well in a thin layer without expansion as glues that expand while drying can cause the plates to not fit together properly during reassembly. Move the spacers slightly after a little while and every so often to prevent their being glued to the plate as well as the magnet. Be sure to block the magnet ends to keep from misaligning them -- the ends must be aligned so that the spacer will fit about the array.

- JP

The result might not be good looking particularly due to the stretching of the plastic insulation layer when the magnets fractured, but it works and is better than parting out a good speaker for a driver that may experience the same issues in short order.

- JP

Now that the magnets are repaired, we can move on to the primary and most common issue with the EMIT driver, the high-resistance or open situation. The most typical cause of this is not due to the traces burning out as is often the cited problem, but with the solder joints on the diaphragm losing continuity due to age (fractures). In the attached picture, the solder joints mentioned can be seen at the top of the image ("bottom" of the diaphragm). Using a low wattage soldering iron (I recommend a 15watt to limit the risk of melting any portion of the diaphragm), melt the connection and add a little bit of new solder. Don't use too much solder as the joints must fit in the holes of the spacers. I recommend that you solder the diaphragm while it is sitting flat on your workbench (preferably on some aluminum foil or the like to keep from burning the bench surface) else there's a good possibility that the solder will flow out of the connection and cause the "open" situation to remain.

- JP

It is easier to re-solder the connections from the back of the diaphragm and the phenolic frame for the Mylar allows a little more room for error should the soldering iron's tip move where it shouldn't. Afterwards, take a multi-meter and measure the resistance of the diaphragm. Rather than reading no continuity or high resistance, the figure should now equal the penciled measure from before. You have accomplished the repair and can now reassemble the driver.

To reassemble the driver, just reverse the disassembly process. Make sure that the solder connections on the diaphragm align with the holes in the spacers and that the metal trace faces forward. Then with the tops of the plates aligned, use the long screw method to gently force the faceplate back down onto the diaphragm so that the magnets align and the traces are centered in the slots. Replace the screws in the corners, remove the long screws, and screw the originals back in. Reinstall the drivers into the speaker after soldering the wires back to their appropriate terminals observing the red pen mark indicating the positive terminal.

Occasionally, the voice coil trace on the diaphragm can be damaged by clipping amplifiers. Fortunately, this is repairable as well, though I do not have pictures of the process. For a burnt area of the trace, a rear window defroster kit can be used to bridge the gap or some manner of equally conductive paint will work. Should the trace have lifted from the Mylar, it can be reattached with a light coat of clear fingernail polish. If the trace stretched slightly when it lifted, the excess can simply be folded and secured with the fingernail polish again. Should the diaphragm be torn, a patch can be cut from Scotch transparent tape slightly larger than the affected area and be affixed over the tear using fingernail polish, but with the adhesive side out.

Finally, listen to make sure the repair works and kick back to enjoy the music.

- JP

Edit: I just noticed that I neglected to add a picture of the other side of the diaphragm.