Published in the October 2000 issue of ...
Analog Maintenance and Upgrade Considerations Part-1: Caps and OpAmps
I get plenty of e-mail regarding "modifications and upgrades" to recording consoles and other pieces of audio gear. I chose those words carefully to make a point. More often than not, modifications can be as expensive as a new piece of outboard gear, or worse, a major detour into oscillation-land. Why modify a console mic preamp when there’s already a product who’s designer auditioned each component for its sound, or lack of same? By contrast, the decision to upgrade should be borne from the maintenance history of the console or tape machine. If you want to know a little more about basic maintenance, enough to make educated decisions or possibly go a little further, ya gotta be in the ballpark to see the game and ya gotta practice to play the game. So let’s find our seats and check out the opening line-up. What better way to demonstrate the process of comparative analysis than by testing a recording console or tape machine, good channel next to bad channel? Capacitors: Square Biz Geeks in training need to start with maintenance fundamentals and this includes familiarity with test equipment and electronic components. (A list of test gear was detailed in the May 2000 issue.) Capacitors — the weak link in almost any aged product — are failure prone over time making travel for low frequencies difficult. The end result is "thin" sound. You can test for bad capacitors without pulling a module; the fastest way requires a square-wave oscillator and an oscilloscope. Figure One shows a family of square-waves starting with the 500Hz reference wave (a), bass roll-off simulates the tilt caused by bad capacitors (b), while bass boost tilts the wave in the opposite direction (c). (Any midrange frequency will yield similar results.) This test is particularly useful when checking old Pultec and Neve modules or any equalizer that has switched EQ settings. Not only do you want to know that all frequency options are functional, but also that the center détente is really "flat" and that the output transformer (if applicable) is properly terminated. Also included are snapshots of treble boost (d) and cut (e). All "visual effects" were produced with a shelving equalizer where the slope is a gentle 6dB per octave. A low-pass (high-cut) filter has a much steeper slope, at least 12 dB / octave leaving more bandwidth untouched and leaving a recognizable footprint as seen in Figure –1 (f). Notice how the rise time of the square wave is affected. The images will vary depending on the slope of the EQ curve. Boosting the midrange EQ at the same frequency as the oscillator will turn it into a near-sinewave.
Once the bad modules reveal themselves, it’s time to find the bad caps. Be sure to mute the speakers and power down the console when removing and replacing modules. An extender card makes it easy to probe individual modules. If you don’t have that luxury, pull a few modules to clear some real estate. A schematic will help locate input and output caps, most of the time the circuit board is silk-screened with part designations. Using a scope probe, locate a cap at the input or output of an amplifier and check each leg for the "before and after." You might find that many of the same caps in each module have failed. The decision is yours to change only those that typically fail or go the "wholesale replacement" route. Caps can fail in batches, some are more stressed than others — and then there are caps that just plain suck. Once you begin to recognize them, order a quantity to have on hand when some "downtime" opens up. Digi-Key is a good place to buy mass quantities. Note: Both HFS and HFQ series caps by Panasonic were being discontinued at the time this article was published in October 2000. The "replacement" is the "FC" series. One reader also recommended the Sanyo Os-Con and Elna Starget. Anyone with recommendations is encouraged to drop an e-mail and share their knowledge.
I like to replace electrolytic inter-stage (coupling) caps with Panasonic’s HFS series, designed to withstand the abuse of switching power supplies and rated for 105° C. The difficulty with selecting replacement caps is that many were originally chosen for their small footprint, which is mostly determined by selecting (some would say compromising) voltage and capacitance values. Construction materials also affect size. The ambient temperature inside the product must also be considered. Heat shortens life, so the 105° C rating is a safer bet than the more typical 85° C. As a rule of thumb, the power supply rail(s) should be within 70% of cap’s DC voltage rating. Quite often caps ride the rails much closer than that (not good). For example, a 24-volt rail wants a 35-volt cap. Using one rail of a bipolar supply as reference, a 16-volt rail wants a 22.8-volt cap, which you’re not going to find. Capacitor voltage ratings are stepped — 10v, 25v, 35v — so the next highest value is a good place to start. For additional comfort margin, add the bipolar rails together, but remember that larger capacitance values and voltage ratings make bigger caps. Good compact caps cost even more. Yes, you can argue that different capacitor materials have a sound,
Dave Hill of Crane Song once told me a story about replacing the Tantalum
caps in an old Neve module. Apparently the non-tantalum replacements changed
the sound of the module. Dave’s research determined that Tantalum caps
contributed more even-order harmonic distortion, not as clean, but pleasing.
|
Op amps |
|
|
|
|
1.4 ~ 2.5 per amp |
at unity gain |
|
|
|
|
|
|
|
|
|
|
|
|
|
Op amps |
|
|
|
|
1.25~2.8 per amp |
|
|
|
1.4 ~ 2.5 per amp |
at unity gain |
|
|
|
50V/m S decomp |
Motorola www.onsemi.com |
Table Two: Quad opamp comparisons
GROUND ZERO |
|
|
|
|
|