Greetings, Earthlings! This month, I've got a voltage tweak for Alesis ADATs, a headphone amp boosting mod for HHB's PDR1000 PortaDAT and tips on replacing surface-mount chips using conventional tools.
Last year, one of Paul Lehrman's “Insider Audio” columns discussed the effect of mass production and miniaturization on the repair industry. While we've enjoyed more powerful and affordable toys, the miniaturization process has reduced the serviceability (or financial practicality of repairing) many high-tech products. This trend hardly encourages people to consider service as a profession. On the surface, this may seem an acceptable trade-off, but many audio designers and manufacturers began their careers in the service industry, ours in particular, because of vintage circuitry's popularity.
The current technological wave squeezes entire circuit boards of discrete components — circa 1990 — to less than a handful of ICs, most in the form of Surface-Mount Devices (SMDs). These components can be difficult to replace — especially when compared to a vacuum tube or an IC op amp — but the next wave (micro SMD) may render circuit board repairs a thing of the past. See Fig. 1 for the evolution.
Last year, one of my online geeks sent me a can of nuts followed by a Sony instructional video about using conventional tools to replace surface-mount components. Sure, 99.99% of you will never change an SMD, and it's not my goal in life, but the refined skill has saved money on specialized tools and minimized the need to replace very expensive circuit boards. I will share some SMD tips later in this column. The repair biz ain't over yet, but if you're over 40, go to the drug store and buy some reading glasses…
ADAT PS TIPS
Fortunately, most repairs don't require swapping SMDs. Here's a fairly simple solution to a common ADAT service issue.
Whenever multiple ADATs are in the shop at the same time, I check them as a System in a Master/Slave configuration. In a recent case, I immediately noticed that the machines didn't fast-wind at the same speed. Once mechanical variables are eliminated — rubber tire in good condition, no reel-table friction (see November 2001 Mix) and a happy reel motor — then the problem may be caused by the power supply voltage, which determines fast-wind speed. A variation of 1 volt in either direction can have a dramatic effect.
Figure 2 shows the front section of an ADAT-XT's main processor board, viewed from the left side. Connect a voltmeter's negative probe to either of the black wires on the power connector and measure each color pair; the voltages should be very close to those indicated here and on the figure (-31.16, +8.4 and +20.5 volts), the 20.5V orange pair being responsible for reel-motor speed. One pot on the power supply board adjusts all voltages at once, but don't make any adjustments yet.
Note: The power supply is in a separate compartment on the machine's underside. The voltage adjustment is typically glued. Dissolving the glue can damage the pot, so it is highly recommended that a new 1k-ohm pot be installed first. The supply will protect itself from an over-voltage condition, and the unit will not power up if the voltage is too low, but a scratchy pot will cause erratic, potentially damaging behavior. Users without technical expertise should forward this knowledge to their local service professional. Replace the pot rather than risk damage with a glue-dissolving solvent. This adjustment applies to all ADATs, except the M20 and Studer V-Eight. Additional details for this and other models in a future article.
HHB PDR1000 HEADPHONE AMP
My first experience with the HHB PDR1000 PortaDAT was a product review. Because NXT Generation (www.nxtgentech.com) is the authorized U.S. service center for HHB products, my contact with the machine has been minimal. Lenny Manzo of Film Services Inc., a Massachusetts-based rental company, contacted me because users were complaining that the headphone amplifier was not loud enough before distorting. This apparently has been a complaint for some time, especially when using the popular Sony MDR-7506 headphones. HHB once published a list explaining that lower impedance headphones would yield higher output levels but also higher distortion; high-impedance headphones would deliver less level and less distortion — but never a solution. The table shows mostly current headphone models and how their impedance varies.
Dennis Charney at NXT kindly forwarded the documentation. In Fig. 3, notice the “build-out” resistors (R-493/R-605 and R-494/R-6076) between the headphone amp and the jack. Each pair adds up to 32 ohms; the resistors serve a dual role — as protection against short circuits as well as being part of a muting circuit. Using 32-ohm headphones will reduce the available voltage by 50% or 6 dB.
In rental applications, Sony MDR-7506 headphones are the norm, so Lenny Manzo asked if a modification would make the PortaDAT drive them better. The DC resistance of the Sony “cans” is 70 ohms. That, plus the 32 ohms internal, is 102 ohms total, creating a voltage divider that delivers about 70% of the signal to the MDR-7506. That's not really the problem. The headphone amp is a surface-mount, dual op amp operating from a single 5-volt supply. (Note that resistors R-310 and R-311 “bias” the op amp to live halfway between 5 volts and ground. Original voltages are shown at the lower op amp in reverse video.)
Without resistors, the amplifier is capable of delivering 44.6 milliwatts (RMS) to 70-ohm headphones. With resistors, the power drops 3 dB to 21.8 mW, not enough of a difference to justify removing the parts in question; doing so would compromise the integrity of the unit and disable the Muting feature. What the unit really needed was a greater voltage swing, and within minutes of perusing the schematics, I had an alternative.
As you may know, most audio op amps run on bi-polar power, typically plus and minus 5 to 24 volts. The chip in question — NJM3414M — can run on 15 volts total or ±7.5 volts. Because other op amps in the PDR-1000 use bi-polar 5 volts, I borrowed “-5 volts” from pin 13 of PK-301 (the DC-to-DC converter) connecting it to the “-V” leg of the headphone amp (pin 4), removing R-310 and R-311 before powering up. Now the op amp swings nearly 10 volts peak to peak, delivering 87 mW to the Sony cans a full 6 dB louder than the unmodified version. This should satisfy most metal heads on location.
Note: This is not a user mod, because it requires major disassembly; in fact, the level of miniaturization could cause severe brain fatigue. In addition, power conservation is ultra-critical in a portable unit, and no study was made to determine how this mod affects battery life.
SERMON ON THE MOUNT
In an ideal world, all major components would be in sockets, but a cheap socket combined with poor soldering can decrease reliability. Compared to replacing a vacuum tube or a “socketed” IC op amp, a large surface-mount device has many more connections handling frequencies beyond dog and bat ears. Confidence in your troubleshooting skills is one matter — a willingness to purchase an 80-plus pin IC, plus all the gear necessary to replace, it is quite another.
Surface-mount removal equipment can range from $100 to $1,000, depending on your needs and skills. Peter Florance — of www.audio-services.com in Virginia Beach, Va. — forwarded a Sony technical video on this subject that shows several methods for SMD replacement. He also suggested the free sample kit available from www.chipquikinc.com, or a full-sized kit can be purchased from most suppliers (such as MCM Electronics). I chose the least-destructive technique — one that you would never have suspected would work, starting with Fig. 4a. This is but one method of removing and replacing surface-mount devices.
I start with Chem-Wik™ rosin-treated, copper-braided mesh that is one-tenth of an inch wide (part number 10-100L). Placing the braid between the soldering iron and the chip legs, the rosin helps solder flow away from the Printed Circuit Board (PCB). The degree of success depends on a number of factors, of which the quality of the PCB cannot be discounted. Time and pressure must be minimal with no movement, especially with cheaper boards, otherwise traces will be damaged. The task is to remove solder, but sometimes adding solder to the iron tip improves heat transfer and wicking action, while minimizing the urge to apply pressure.
The next step is to determine the efficacy of the solder-removing process. Using a magnifier, I place a jeweler's screwdriver between the IC's legs. If just a little twist pops them loose, then the chip is ready for removal. If not, then more wicking may be required. A heat gun can also be used to warm the area; a prying tool under one edge of the chip helps to let you know when the chip is ready to pop. Be gentle. Any radical force could pull traces off the board. (See Figs. 4b and 4c.)
Clean off the area if necessary, and apply flux to all traces; then “tin” the two diagonal traces with solder (Figs. 4d and 4e). Align the new chip, solder the two diagonal legs, and then apply additional flux to all legs. The iron tip should be clean and have fresh solder. The trick is to quickly “kiss” the legs with the iron tip, modulating toward and away from the legs (Fig. 4f). Again, this is not a “friction” thing — no pressure required. Solder will flow to the desired area, and if you're a fast “kisser,” then it will do so without shorting the pins. Again, you would never think this technique would work, but it does. Shorts can be undone with the help of gravity and improved technique, or with the copper braid.
A ROSE BY ANY OTHER NAME
Troubleshooting at the surface-mount level requires a bit more chutzpah because more is on the line. Large SMD chips are not cheap — $50 to $250 — but whole PCBs can be $400 to $800! There is always a chance that the new part won't solve the problem, but “progress” can be defined as not going backward but rather teaching an old dog a new trick.
My thanks to Dennis Charney and Peter Florance for their help.
Eddie can be found trying out new magnifying glasses atwww.tangible-technology.com.