micro-editor won't start

[jimchism]

I havn't run micro-editor on this machine for a couple of months. Today when I clicked on the icon, micro-editor flashed on the screen and then disappeared. Tried it again, same result.

I tried un-installing and re-installing software and unplugging the machine for a few minutes. No luck. I also tried to re-install the Krystal drivers.

I am running 5.3 and Krystal1200.

Any suggestions?

[Rich LePage]

A few things you might try:

1. Did you have the Krystal out of the computer, since
you mentioned you hadn't used it in a while? If so, check
that it's seated OK. I've seen your symptoms in my systems
several times when Krystal was almost- but not quite-
seated. Many case/mboard combos easily permit that.

2. If 1 above is OK, what about the interrupt for the slot Krystal is running in? Under 5.3 I think a dedicated IRQ was still required.
You can often hard-map an IRQ in most BIOSes to the slot-- but some BIOS won't allow doing that. There's stuff in the Medit manual about that as well as IRQ sharing within Win98-- but you need the manual that has the Win98 tweaks in it, not the Win95 ones. Drove me nuts once which is why I mention it. If you don't have the Win98 version of the manual, contact MTU about getting that.

3. With 5.3 I think you can still run the Krystal and motherboard tests that should be in your 5.3 directory. Note that if you installed the IRQ-less drivers for say 5.4 I don't think that runs correctly in that case. If you can run it, do so and see if you are getting good numbers and results from the motherboard and Krystal.

Hope some of that helps.

[geggyboy]

Had the same a few times recently. This anom. generally isnt anything to worry about as a reboot (off and on) sorts it.
It's as Rich says, when the card (dsp download) is not detected on boot. Dust on the PCI seating, PCI slot itself, duff card or.... MTU told me this once before when I had a run of it. The MTU screen flashes up and nada, nichts, zilch after that. As far as I can tell they are slightly different faults.. Once windows is up, if you get "UNABLE TO DOWNLOAD DSP" (if that is constant.. oooops! new card) and just MTU screen flashes with no other warning, then restarts (not reboots) fixes it. In your case, that doesnt happen it seems, so look deeper.

A. On several of my systems, it appeared to be a powersupply oddity rather than fault; whether the power supply itself or a motherboard quirk. I eventually upped the powersupply to a 300 Watt. (Since they are dirt cheap) I'm running a 350 watt now.. and I have not had that problem. Who knows?

B. If your card is naffed, then you have probs. obvioiusly. If the PCI bus doesnt see the card initially, you cant run the tests to prove it either. Dilema
Briefly put the card in another system and see if windows detects something.. No need to go full installation.

g

[Rich LePage]

Geg's suggestion about the power supply is also a very good
one. Most power supplies are rated in PEAK watts, not sustainable ones over long haul all day long.

So if yours is running near its continuous max, that could also
be the problem. And it's pretty hard to know if it is, since they don't rate 'em that way. So it's better to over-spec the supply to be on safe side. Like Geg said, over-300 watters are pretty cheap right now, especially 350 or so watts. I bought a $35 case the other day that had a 350watt P4 power supply included.

Especially with newer AMD chips, you really need a beefy power supply. By the time you get all those coolers and fans into the case, you've soaked up a lot of the power right there -- not the mention a pile of drives and of course the CPU itself.

[jimchism]

Thanks for the help. The machine is up and running again. Don't know what the problem was yet, but I had to completely uninstall the software and hardware, and then re-install. I had forgotten about an added adaptec USB2/FIREWIRE board which I also took out and have not reinstalled. I will wait until I finish this round of projects before attempting that re-install. I'm thinking I had a conflict with that board. If that turns out to be the case, I may be looking for some more help.

Thanks again, Jim ---12 years with MicroSound and it's still the best.

[Rich LePage]

The USB2/Firewire is a good candidate for having been your culprit. I ran into this with a maxed out Athlon recently when trying to add a Firewire card. It just wouldn't work, as it was unhappy sharing any IRQ and there were no more dedicated ones available (and Krystal was hard-mapped to an IRQ in the BIOS).

Result was that Medit wasn't working right, and a whole lot else wasn't either. So I ripped it out, though did not need to reinstall anything.

I'm going to try it again when I get more time and defeat one of the machine's serial ports in BIOS and then hard assign that IRQ to the slot for the Firewire board. Or maybe defeat the LPT port and map that IRQ to the Firewire card, as my audio systems mostly don't have printers on them.

By the way, PC Mag proved that though the spec for Firewire is slightly slower than for USB 2, Firewire transfers large files faster.
I was able to verify that using an outboard box that has both USB2 and Firewire interfaces and an Asus P4PE Pent 4 machine which also has both on the m/board. I took a typical mix of MTU files (sev large and a few smaller SFs, a medium sized SF2 of about 600Mb, etc) and tryied both ways. I had an outboard Maxtor 40 gig drive in the box. I transferred same set of files to 2 diff logical drives in the Pent 4, one USB2 and one Firewire. Firewire seemed to get it done about 20% faster in my test.

I also deliberately fragmented 2 diff logical drives created on the outboard and then attempted to defrag them using just the Win98SE defrag (FAT 32), doing one drive each way after re-boot of course. I used same file sets, and deleted some files so the drives would be fragmented. Result was again that Firewire was about 15% faster than USB2 in defragging under these conditions.

If looking at these outboard boxes, the ones to get seem to have the Oxford chipset - all the reviews seem to say that one offers the best USB 2 and Firewire implementation. The boxes sell under several brands, but mostly all look the same so prob all come from same place. They run about $65 to $75 at several Internet sources. Also for USB 2, be sure to use a USB2 rated cable. Many are not.

One more tip- before I install anything new in machines I need to rely on, I run (and verify) a C drive backup. This is true for ANY patch or program and ANY new hardware! I use BackupMyPC (formerly Backup Exec) to do that, though I guess you could use newer Roxio stuff since that now will verify. I keep the boot drives small on Win98SE machines-- and I almost never allow them to go on the Internet -- I d/l any patches to other machines, copy to CD, and install that way. Only exception to that is a very few Win patches that you can't-- but since most are browser and Internet patches, I don't use those in the audio machines. Keeps Win from being a bigger resource hog in my view. Most of my audio systems have been online maybe an hour tops in their entire lives, and a couple don't even have modems. (I'll hook up an external one to a serial port if absolutely necessary).

With Backup My PC, you can create 2 emergency floppies, and if the drive is small, you can fit the C drive's contents on a CDRW. So when problems arise, I simply wipe and re-format the C drive using the emerg. floppies, then restore the whole known-good configuration from the CDRW.

This has saved my tail numerous times.

[MTUSUPPORT]

It sounds like it was an IRQ conflict with this Adaptec board. Make sure that its not on the same IRQ as the Krystal board, under 98 and ME Krystal couldn't share any IRQ's, as now in 2000 and XP it can share correctly and work.