PlanetZ forums - scopeusers.com

In my previous messages in "Help! Lost Installation CD Pulsar II" thread I mention the following errors, but they belong in another thread:

Unexpected DSP overload

sample rate settings window:
all red leds on the left side are on, and sometimes briefly off.
I shake the pulsar board a little up and down in the slot, and the situation normalizes.

third crash, almost 2 hours later.
this time: one click in the speakers, 30 seconds nothing, then again crackle noise galore!
This time when I lever the pulsar board nothing happens.
I unplugged the luna box, plugged it back in, the noise has stopped but the master meters of the STM mixer keep going crazy.
In samplerate window the usual situation.
I set 'slave', then back to 'master' and everything works again.

">you could try to clean the slots itselves."
It's a new computer with practically unused slots

">or try to make the card work in another computer."
this is the 4th computer in which I have the same symptoms
...

Now yet another one : "A communication error with the Pulsar board has occured" (or something alike, don't remember exactly...) showed up right after windows startup. The Scope platform didn't load.
Restarted PC again, now SFP starts up, but with a 'unexpected dsp overload'. Again the same actions as above solve the problem.

Btw, I wonder if I'm not mistaking in assuming the problem is the Pulsar board rather than the Luna Box attached to it.

anything I could try appart from getting a new soundcard ?

thanks. Tim

Contact CW or your dealer before your warranty expires...it might be also a defective board.

You should uninstall one of them in turn and see how the other one alone works.

If it only happens with one precise board it might be a hw issue.

yes, first of all isolate the devices to verify if it's the card or the box.
second, if warranty applies only do a couple of tests (see below) to outrule you've overlooked something trivial (not likely with such a persisitent trouble, but there may be exceptions...)

I assume you have just one card, a Pulsar 2 (?)
remove the IO-plate and put it back in place (there have been contact problems reported with this item)
disconnect the z-link box AND the breakout cable
verify the card is properly seated and has good contact to the metal of the case

does it work now ?
if not:
have you had all 4 computers used in the same location, with the same powersource ?
Are you certain that the electrical installation is appropriate (grounded) and there are no heavy electrical disturbances on your lines, as are refridgerators, dimmers, elevators, passing trains or trams nearby ?

in a badly grounded environment these sources may build up significant induction pulses, that may irritate the clock circuitry.

dip a finger into water, so it's a little wet and then touch a metal part of the computer case very, very gently - is there a needle-like feel at your fingertip ?
it's harmless if you don't hold a waterline or a powersource in your other hand simultaneously

Is the computercase grounded ? does the blackplate have solid contact (sometimes the coating of the case prevents this)

To clarify these points will not interfere with your warranty (which might be applicable)
If you can outrule the afforementioned error sources (doing the test WITHOUT peripherals !) and the card still behaves strange and warranty applies, send the card back - there's no more you can safely do on your own.

If the card is beyond warranty (and you outruled etc...) you may inspect the electrolytic capacitors (the small colored tubes mounted vertically) - any of the loose, showing traces of leakage, looking bulky ?
Don't bet on it, but there have been bad solder connections reported on those.
If you have few to no experiences with electronics your options are limited of course.
If you have a digicam with macro you may post a couple of hires pics as attachment and I could tell if anything looks suspective.

cheers, Tom

Hi guys,

thanks for the replies.

I found at least one reproducable problem:
the z-link cable that connects the luna box to the pulsar io plate is very sensitive to movement. I suspect bad contact.
I now connected the grounding of the two (pulsar and lunabox) with an extra cable and as a result the z-link cable can be jerked around without causing dropouts.

I assume you have just one card, a Pulsar 2 (?)

yes, one Pulsar2

remove the IO-plate and put it back in place

I didnt do that yet. But it is true that the io plate is placed by myself at the time, and maybe if i go back in my memory that was about the time the problems started...coincidence ?

disconnect the z-link box AND the breakout cable
verify the card is properly seated and has good contact to the metal of the case

I did not yet put the screw in the card...got me there too...
I sometimes have this bad habbit of not screwing cards...you don't want the explanation...
anyways, grounding is as of now assured.

have you had all 4 computers used in the same location, with the same powersource ?

no, different situations... many...

dip a finger into water, so it's a little wet and then touch a metal part of the computer case very, very gently - is there a needle-like feel at your fingertip ?

I feel absolutely nothing when I do that. Is that good or bad ?

Is the computercase grounded ? does the blackplate have solid contact (sometimes the coating of the case prevents this)

what exactly is the backplate ?
The electricity cable that powers the computer has 3 pins, one of which is grounding, (and it goes to a real grounding in the basement)

If the card is beyond warranty (and you outruled etc...) you may inspect the electrolytic capacitors (the small colored tubes mounted vertically) - any of the loose, showing traces of leakage, looking bulky ?
If you have a digicam with macro you may post a couple of hires pics as attachment and I could tell if anything looks suspective

I'll do that as soon as I am sure the problem is not solved by improving the grounding like I did now. At the moment things seem more stable, but it can take hours, days, weeks for the problem to show up again,
which of course I hope it won't....

tim

the backplate means just the backside of the computercase, where you fix the cards by screws.
Afaik the cable whips of all Pulsar/Scope boards are identical, except for the 'plus' variant. Imho (I've had a look inside a Pulsar One cablewhip) it's absolutely essential that it's properly fixed to the computer case and the latter has a proper ground connection, too.
All ground wires of the various signal are collected in the 'frame ground' of the connector, which is supposed to be fixed to the case.

If it is not, then a capacitance may be induced, (effective) ground level on the card might be drifting, the typical hum and buzzing noise etc.
Those effects will be emphasized from a not properly grounded item on which sometimes you can sense the flow of current by that fingerprint propbe
It doesn't mean too much, though - but a 'negative' result is more appreciated in this case.

If you suspect the firewire cable, then get rid of it immediately - there are power supply lines inside that may distribute their potential anywhere if it breaks.

cheers, Tom

The card has been stable for a few hours now, and I can reproduce the crash by jerking the z-link firewire cable. I think the contacts are worn out on the board as well as the luna box.
Tomorrow I'll also get a new firewire cable.
Now I hope that at least I have pinpointed the cause of the problems, but still there is a chance that bad z-link cable connection is not the only one.
I don't really feel like taking off the io plate just yet though...

A few days went by...
The problem is clearly NOT solved.
Still random crashes.
It can work for hours, then suddenly crash.
when a crash occurred i tried the following (always followed by clicking 'Slave' ant then 'Master' in Samplerate Settings):

-pull out the z-link cable to the luna box, plug it back in after 20 seconds or so.
result: sometimes fixed it, but last 2 times I tried it: nothing

-lever the pulsar board gently up and down in the slot
result: sometimes seems to restore the situation

-squeeze the pulsar board and the i/o plate a little together.
result: restored the situation 2 times in a row (this is the last method I came up with, since someone replied there could be a problem with the i/o plate contacts...)

Have also tried without the luna box, using the pulsar analog dest output. Doesn't seem to change much.
Connected the luna box case with the computer case with a cable for extra grounding.
Still have to replace the firewire cable...

Tim

tim167 wrote:...pull out the z-link cable to the luna box, plug it back in after 20 seconds or so.
result: sometimes fixed it, but last 2 times I tried it: nothing...

dunno if the z-link driving routines can deal with such a disconnect-reconnect situation at all - I rather doubt as the unit is also powered this way.

regarding the connectors of the IOplate it may of course develop corrosion, but it's not very likely - unless you live close to the sea or the air is chemically aggressive in another way, which I don't hope for you...

this could also point to a bad solder connection (most likely) in the inner layers of the board (there may be 4 to 6).
you should not stress your luck by overstressing these delicate items... you've noticed it, and that's ok...

You may inspect the solder points, in particular around the big parts (mostly caps), but it could as well apply to any connection between both sides of the board.
It sometimes helps to suck out the old solder with a special type of wire and resolder, but only (!) if there is evidence for a bad connection - and it's not guaranteed to help.
It IS GUARANTEED that it will ruin the board if someone unexperienced tries it, it's mentioned here just in case you know a technician in your area (like a TV serviceman or radio amateur).

You may also look if any of the surface mounted parts have come lose, which sometimes happens when the board is accidently hit by another part when (for example) installing a new drive or something like that.
It's not easy to see, good light and a magnifier required.

the hardware procs are just mentioned for completeness, in case it's difficult (or too expensive) to send the board to the appropriate service.
I'm in no way responsible if anything goes wrong - I personally enjoy the technical challenge spiced with the expectations that a few hundred Euros may be down the drain if I fail, but I wouldn't generalize that...

cheers, Tom

I decided to switch to an rme hammerfall dsp multiface.
My pulsar is not sold or dumped yet, so maybe when I have time I'll look at it again...
Thanks all for the help anyway.

Tim