Music & Gear

Battery replacement on Yamaha FS1R

Montag, August 17th, 2009

Yesterday my FS1R came up with a „Battery Low!“ warning after I had already done the replacement on my M1R and TX802 one year ago. Fortunately again, the CR2032 button cell sits in a holder. Note the Sanyo LC78834 DACs.

RNC 1773 repair

Sonntag, Juli 5th, 2009

Five years ago I ordered an FMR Audio RNP 8380 preamp and RNC 1773 compressor bundled with a Funk Logic FRT-8373 rack tray from Mercenary Audio. Mercenary is located overseas from me, as I live in Germany, so the weak Dollar in relation to the Euro was the main reason to order directly from the U.S.; aided by the fact that local prices for FMR Audio gear were comparatively unattractive (virtually not nice) at that time. This has changed meanwhile.

Well, although I’ve not the least doubt that Mercenary Audio would do everything to help me in case of a damage or malfunction, there is still the shipping time and cost to and from overseas if a repair becomes necessary. And last tuesday evening, when I hung out in my vocal booth, the worst case became true: all gain reduction LEDs of my RNC started to light up without an audio signal applied. After I switched it off and on again the LEDs still kept lit. Then I observed that turning the RATIO knob straight to the left made them turn off, but with just a few degrees turned to the right they all lit up again.

Would the german distributor be delighted to fix it for me? I did not ask but switched to plan B: become a geek and fix it by myself.

Easier said than done. After one hour of component checking I sat there without having a clue. Before, when I opened the housing and first saw the three voltage regulators, I would have bet dollars euros to donuts that either one of the two 15 volts regulators is broken. But they were not.

After having put a pot of coffee on, downloading datasheets and digging in somewhat deeper into the circuit, I stumbled upon a 5.37 volts input signal on port 16 (AD0) of the CPU. Up to four ports of the 68HC705 can be used for A/D conversion, and all of them are utilized that way in the RNC; exactly matching it’s block diagram, the three pots RATIO, ATTACK and RELEASE are connected to AD1 to AD3, and AD0 seems to be feeded with the equivalent audio input sum signal brought over from output 3 of U4 (TL074, not in the main audio path, mentioned for the enthusiasts).

Back to the 5.37 volts on AD0: do they make sense if the reference voltage on pin 15 of the 68HC705 is exactly 5.0 volts? Probably not. Tracing back the path to U4 made me say „thank you“ to diode D3 for its job to protect the CPU from being blown: output 3 of U4 showed about +10 volts, with the non-inverting input 3 connected to ground and the inverting input 3 hanging around at +5 volts.

So I removed U4, which was btw more delicate to desolder than expected due to the surrounding resistor networks on the component side, and replaced it with a new TL074 in a precision ic-socket. Luckily that did the trick and the RNC is now fully operational again.

Freshman’s 01/W refurb

Mittwoch, Dezember 10th, 2008

The following is a set of photos from Freshman’s 01/W refurbishment. In fact, his 01/W suffered from the „no sound“ syndrome, the „sticky keys“ syndrome, lots of sluggish panel switches and a dim display. Freshman fixed it all in one repair which took several days, but in return his 01/W now operates like brand new.

Read the rest of this entry »


Mittwoch, Oktober 1st, 2008

Als Bastler ist man oftmals geneigt, sich zu verbasteln. Kleine Dinge werden dann groß, und einfache Dinge werden kompliziert. So kam es, dass ich kürzlich für einen schlichten Lautsprecherschalter nicht weniger als einen Microcontroller und vier Relais verbaut habe. Das Ganze, verpackt in ein mediokres Plastikgehäuse, führte zu einem Tischgerät im bewährten Design eines FeTAp 61; einer optischen Erscheinung also, die sich perfekt in jede Derrick-Kulisse eingefügt hätte.

Andererseits erfüllt das Gerät seinen Zweck sehr gut und konnte endlich meinen mechanischen Lautsprecherschalter mit Wippschaltern ersetzen. Der litt zunehmend an Kontaktschwäche, sodass jede Schaltung zu einer nervigen Herumdrückerei geriet. Auf eine Kontaktreinigung hatte ich schon deswegen keine Lust, weil dieser von Anfang an nur als eine preisgünstige Verlegenheitslösung gedacht war. Eine Schaltung per Relais musste her.

Einfache Relais benötigen eine permanente Stromversorgung, und sofern dafür nicht das x-te Steckernetzteil auf die volle Steckdosenleiste gesteckt werden soll, bleibt nur ein Akku- beziehungsweise Batteriebetrieb. Kräftigere Relais, wie sie hier verwendet werden sollen, saugen gängige Batterien aber innerhalb von Stunden leer. Daher habe ich mich für bistabile Relais entschieden, die nur jeweils kurzzeitig für die Umschaltung Strom benötigen. Damit scheiden wiederum Schalter an der Bedienseite aus, denn es soll ja nur kurz getastet werden. Nun muss beim direkten Ansteuern von Relais mit Tastern ein Kontaktprellen unterbunden werden, und außerdem ist es für Batterien und Relais besser, wenn die Erregungsdauer (Dauer des Umschaltimpulses) unabhängig vom Tastendruck fest vorgegeben ist.

Dafür hätte es zu Zeiten des FeTAp 61 sicherlich tolle analoge Lösungen gegeben, aber heutzutage ist ein kleiner Microcontroller das Mittel meiner Wahl. Da kann das Gehirn beim Basteln ausgeschaltet bleiben, und nach der Fertigstellung gibt es mit Bugfixes und Firmware-Updates gute Gelegenheiten für selbst bescherte Glücksmomente. Damit nicht der Microcontroller seinerseits die Batterien entleert, wird dieser nur in der Sekunde eines Tastendrucks aktiviert und verbleibt ansonsten in einem Schlafmodus. So ist die Lebensdauer der Batterien fast nur durch deren übliche Alterung begrenzt.

Für alle Komponenten, bis auf die Relais, wollte ich Teile aus der Bastelkiste verwenden. Die vielen Schraubklemmen auf der Platine spiegeln die Unentschlossenheit des Konstrukteurs wider, wie das Gerät am Ende genau aussehen soll. Die Klemmanschlüsse auf der Gehäuserückseite sind in erster Linie billig, aber leicht zu montieren und für Leistungen im Nahfeld-Monitoring ausreichend. Vier hochwertige EAO-Taster machen Freude beim Betätigen und entstammen alten Bedienkassetten aus der Rundfunktechnik. Deren 24V-Lämpchen habe ich durch neue 12V-Lämpchen ausgetauscht, damit die optische Rückmeldung bei der Schaltung richtig zur Geltung kommt. Lämpchen und Relais werden über je ein Treiber-IC ULN2803 angesteuert, welches auch Freilaufdioden enthält. Damit jeder Tastendruck einen externen Interrupt am Microcontroller auslösen kann, sind die Tasterkontakte mit Dioden auf einen Eingangs-Pin „verODERt“. Die Versorgungsspannung von 4,5V für den Microcontroller wird von drei der insgesamt acht Batterien abgezweigt.

Korg 01/W „no sound“ repair

Sonntag, August 3rd, 2008

A few weeks ago my 01/W started sounding „hissy“ and distorted and the output level dropped audibly. After a short power off it recovered without further action. But then, on the next day, all analog outputs were dead.

When I opened the housing I found two leaked electrolytic capacitors on the mainboard’s sub-PSU for the DAC and analog stages. This ±5 volts sub-PSU with two 100 mA voltage regulators is disposed behind the ±12 volts stabilisation circuitry of the main PSU and supplies stable and noise-free local voltages.

The caps C95 and C96 had ruptured and the sub-PSU and even more distant parts were spattered with electrolyt. And the really bad thing about leaked electrolyt on a printed circuit board is that it causes corrosion and damage of metallic parts and circuit traces. See the „overgrown“ capacitor C100 or the diode D3 on the photos for an impression of how this affects and proceeds.

To clean this type of mess I prefer cotton swabs heavily soaked with isopropanol (isopropyl alcohol). There’s no need to spare with the isopropanol, it does not do any harm to electronic components and evaporates without residue. And the cleaning should be repeated a few times with fresh swabs.

What has been left over from the „acid attack“ on my 01/W did not look dramatic at first. I replaced the two evil caps, connected the mainboard and powered up the 01/W, but there was no sound. While measuring I found that the negative leg around the IC52 (79L05) was operational but IC53 (78L05) floated around at 0 volts on it’s output and input. Then I noticed that one of the C100 capacitor’s leads had been „eaten“ by the acid and — more important — the +12 volt trace below the lead had also been etched away.

So I replaced C100 and patched the broken trace by scraping off a bit of solder resist and bridging it with solder. As you can see on the photos I also replaced IC53, which stabilised at 4.5 volts and not 5.0 volts. Of course, it is not state of the art to replace an SOT89 SMD component with a TO92 through-hole component, but firstly it would have taken some amount of effort to get the SOT89 variant and secondly it is yet a robust and well-functioning replacement in this case (fitting pinout and leads, no thermal or mechanical aspects to be considered).

The other capacitors on my 01/W were apparently in a good condition, so I would not estimate this to be a general case of „bad-caps-alert“ for all the other 01/Ws around. However, in the recent past I read about 01/W owners on the net who have asked for advice on their whatever natured „no sound“ problem, and so this article may give you a hint towards the possible cause.

Update (October 07, 2008):

J-man sent me photos of his „no sound“ repair on a 01R/W, the rackmount variant of the 01/W. He writes: „I’m sending you the pictures of my fixed 01R/W, before and after the repair. […] I’m sorry about the quality of the pictures before the fixing: they are quite blurred, expecially one of them.“