Luna II - dead?

[Nebukadneser]

After months of inactivity, I switched on my DAW only to witness the possible death of my Luna II. I opened a project and streamed some audio from Spotify, and with frequent intervals I could hear crackles. After a while the audio stopped and the diodes on my Luna box started blikning red. I restarted my system, and now I cant open any project, I keep getting the message of "dsp overload" no matter which project I try to open. To search for failure, I took out my cards (Luna II, Pulsar 2, Pulsar 2) and inserted one by one. The fail appears to be the Luna. On startup, the DSP meters show currect numbers of DSPs active, but a message reads that the system times out waiting for response from DSP 0. Is this the end of the Luna? Has anyone solved similar problems?

Neb

[garyb]

yes, it probably is.
it may be a bad capacitor...

[Peter Drake]

Yes, I've had similar symptoms with a Pulsar 2 and fixed it by replacing the bad capacitor. It's not too challenging if you have basic soldering skills. Any competent repair tech should be able to do it, but pm me if you have too hard a time convincing someone to take it on. Good luck.

[astroman]

in fact most of these errors are related to caps, either failing or (more often) just falling off their solder pad.
Look at the bottom part of the card and at the bottom edge of each DSP for missing caps, it's quite easy to spot.

[at0m]

Got a similar problem with my luna2496: All the time when the Luna2496 is connected, Scope's Samplerate Settings panel shows the Z-Link LED blinking intermittently and irregularly. Sometimes the LED remains on somewhat longer (seconds), other times it blinks every ~100ms. On the Luna2496 itself, I see all LED's turn red briefly (for 100-300ms?) right at the moment the firewire cable is connected or disconnected, then these LED's go out again immediately.
Visual inspection was ok, guess i'll go measure out some things one of these, and replace if i find something obviously borked.

[Nebukadneser]

Hmmm, got the Luna II and the Luna box running again. I have used the system for a while by deactivating the Luna II card, using only the 2x Pulsar II cards and the analog in/out of the cards itself. Now I have forgotten what tests I did last time that made me conclude the card and box are dead, so felt that I should check again. I took out the card and fitted it again in the same PCI slot, then I loosened the STDM cable and fitted it again across the 3 cards.

I can say for sure that all DSPs are now active, and the time out issue waiting for DSP 0 is no longer an issue. Default project is loaded without hickups.

The green led on the Luna box is now on.

Then I tried to build up a new project from scratch. I use SCOPE 7, by the way, and inserted the Z-link modules. This caused a cryptic error message to occur, including a message saying that no corresponding board can be found. In the dialog box none of my 3 cards were listed. This error shut off the led on the box. Also, some effects were no longer available, possibly due to the missing link between SW and HW.

Then I attempted to load old projects from 5.0 / 5.1 that included the Z-link box. To my great surprise this works fine, and I can record and playback using the Luna Box!

I have not yet stress tested the system, but I believe that some of the trouble I have had is related to installation of v. 7 and the registration of software and corresponding hardware. Has anyone experienced similar problems?

I should add that my Luna card dates back to 2002(!) and that I am extremely happy that I can still use a system based on components 14-18 years old. For comparison, the Scope system gives a much smoother experience and lower latency than when I switch to my new Steinberg UR242 external USB sound card. I did this comparison when working on fairly large projects in Ableton 10.0.

Neb

[Nebukadneser]

Aaaahhhh, much too optimistic!

I had the pleasure of enjoying a fully re-installed Scope 7 on my W7 64 machine with a Luna II, and 2x Pulsar II. Even the presumed dead Luna box was fully operating using the 8 ins for my old hardware synths. All separately purchased plugins were installed. Then I left the system on for a while, and as I returned the system was unstable. Upon reloading the default project, error messages appeared saying "big modules don't fit" etc, which indicates that not all dsps are available since there were no longer capacity to load a previously working, big project.

I opened up the machine and re-plugged the cards. My working hypothesis was that the error is caused by malfunctioning STDM cables. I realise that the cables are bent into place to connect the cards. After 15-18 years, it is not unlikely that cables become worn. So I replaced one cable entirely, and shifted connections on the one cable where I only use 2 out of 3 plugs (Luna + Pulsar). Now it does work. It did load light projects and the big one that did not load before I swapped cables.

Crossing fingers now. The alternative is to replace one of the cards, if I find out that there are capacitators that are broken.

We'll see.

Neb

[garyb]

sounds like you fixed it.

[Nebukadneser]

Nah, it works for a while, but then the connection to the cards gets lost again with certain intervals. I realised that if I shut down my computer and also turn off the switch on the powersupply, then the system will work for a short time upon restart before it begins loosing the connection again. This may indicate that there are capacitators on the Luna that are gone.

I will now buy a Behringer ADA8200 to replace the Luna box for an 8x8 I/O using the ADAT port of my Pulsar II.

Neb

[valis]

Solder breaks or aging caps.

[Nebukadneser]

Solder breaks or aging caps.

I guess you are right. Perhaps most of the old CW PCI cards are nearing the end of their lifetime? I am a little worried paying big money for hardware made in 2001.

ANyway, it appeared that a friend of mine had a Pulsar II laying around, that I bought for an acceptable price. Now I have a setup with 18 DSPs across 3x Pulsar II. An ADAT in/out solution will make the system complete.

Neb

[valis]

It's not hard to check solder joints and caps and such, the biggest issue is likely the cost of shipping & possibly downtime for the roundtrip. I had my Virus kB checked last summer and was happy to find it only had 1 component out of tolerance so it's aging well. That one component turned out to be rather important, but thankfully keeping my room a constant temp over the years has paid off otherwise.

Ie, solder cracks due to heating/cooling cycles such as power on/off, and for failed caps this can cause issues as well. These are state-change boundaries or more correctly transient response issues, if you follow my signal (sic) drift. Heating and cooling state changes occur over longer periods than power on & off obviously, so the micro-fractures that occur take longer to reveal themselves. But when another component fails during a power cycle or etc, you can be sure that was just the thing that sent it 'over the edge'.

And so intermittent behavior like you're describing is eventually followed by complete failure. With anything I have that is merely for passion, this is tragic but can occur. For items that contribute to my bottom line and stability it's probably wise to handle it before that occurs, or even invest in future-proofing the stability by having the item serviced properly and tested good.