My emu esi 4000 is not working it turns on but the screen has little bars on it

[djkrev520]

I bought a emu esi 4000 and it's not working it turns on but the screen just has little bars on it and all the buttons have the lights on I saw a post about this problem but no definite answer for it. if anyone knows if it a power supply issue or a main board issue. Please help me Thanks

[crochambeau]

Did you list it on reverb as broken? If so I may have just purchased it.

If not, I'll have one that exhibits the same or similar issues underfoot in the time it takes for a parcel to cross a continent. Making me interested in the subject matter as well..

[madtheory]

This is an educated guess, but I’d suspect a PSU issue. Is it outputting the correct voltages? Check the main caps. The PSU itself is a Meanwell PT-45B, you can purchase a replacement.

But honestly you’re better off with an EIV.

[crochambeau]

Awesome, thanks for the tip! It is marvelous that they used a general purpose, off the shelf supply.

ETA for mine is Monday, then I'll be in a better condition to assess its health. An EIV would be cool, but I also like the ugly ducklings, at least that's what I keep telling myself.

[madtheory]

[crochambeau]

The positive voltages are a touch on the high side.

+13.75
+5.4
-12

Does the boot cycle take into account out of tolerance voltages, and freeze when flagged?

I also had it powered while trying to set up a camera to grab a pic of the bars, and I noticed a momentary backlight burst followed by the screen looking like this:



The goofy lighting is me holding a flashlight on it..

I power cycled again, and after pressing various buttons I achieved a screen that was mostly musical notes, affixed to a machine still dead to the world.

The Master/Global LED goes out when the screen changes.

If this rings a bell with anyone I'd love to hear about it, meanwhile I've got plenty to do.

[crochambeau]

Reseated everything internal.

Powered up.

EXHIBITED PATIENCE. (arguably the most important part of this process, some stuff takes minutes to wake up)

..and it appears as if I have a working sampler. Next step, proving it.

Now to spend a few weeks wallowing in the mud for familiarity sake. My only other EMU is a Morpheus, so hopefully some of the logic structure there will help (hahaha).

[madtheory]

Great! The user manual is excellent. Very easy to learn. After you've done that, do this (I wrote it):
https://www.soundonsound.com/techniques ... asterclass

[crochambeau]

Great! The user manual is excellent. Very easy to learn. After you've done that, do this (I wrote it):
https://www.soundonsound.com/techniques ... asterclass

Excellent! Bookmarked, thank you!

.....I'm not out of the woods yet.

So after being able to observe life in this, I ordered up a ZuluSCSI and pretty much waited. At this point, I had observed the LED hang and black bars on the screen for a span of time, 30 seconds to a minute, and then it would come around.

In my inexperience with this device, I accepted that as typical behaviour.

Yesterday the SCSI emulator arrived, and I set about installing it.





Only to have the unit hang on me again, and remain bricked.

I pulled the PSU and watched it (unloaded) on my scope. All the rails are nice and prompt at power up. I reinstalled the PSU and watched it on my scope (one rail at a time) and again, each rail solidly finds its position and stays there on power up.

I adjusted the voltages with the single VR control to bring the +5 volt rail down to +5.2 (as opposed to +5.4) which pulled the positive rail from +13.75 to +12.6. What I would give for per rail voltage adjustment.

Only once have I been able to get this to boot, and that was after unplugging the bus ribbon the feeds SCSI output board and effects board, and unplugging the SCSI that feeds the emulator.

The boot was prompt, but while I was trying to figure out how to reset the eprom (following instruction on this dubious sweetwater link: https://www.sweetwater.com/sweetcare/ar ... t-for-emu/

The unit froze up and became unresponsive.

Since then it's only been black bars and stuck LEDs.

Anyone have a line on an ESI 4000 service manual?

[crochambeau]

Surface mount tantalum (?) capacitors.

+ side of both C35 and C47 read around +1.8-2 volts
+ side of all others are in the +5 volt region, with exception of C64 which reads 0 volts on + side, and -5 volts on - side.

C75 has a fair bit (100 mV) of signal in the ~ 100 mHz region, the rest of the caps seem to be doing their job.

Has anyone played in these fields? I'm curious about the +2 state, maybe a virtual ground for splitting analog signal at the ADC? I think I'll pull C75 and test it, 100 mV is a lot of chatter on a power rail.

[madtheory]

I would think the problem is the firmware. 3.0.1 was buggy for SCSI IIRC. 3.0.2 was the latest. According to this site, it adds support for more SCSI drives, but I recall there being more issues than that at the time. But that was 25 years ago (!) so I could be misremembering.

https://monotanz.de/product/emu-e-mu-es ... ic-puller/

[crochambeau]

Update.

3.02 firmware acquired from monotanz, installed, no change.

I went through several bouts of frustration with this thing, never reaching a solid conclusion until today.

1st hint: the boot hang has all LED except "transpose" lit. This indicates a failure in diagnostic cycle at the display (per diagnostic manual, which is the nearest thing to a proper service manual I could find).

This morning I was able to get the unit to wake up long enough to run some diagnostics on it, until it failed again - display going dark, unit freezing.

The thing is, the unit froze with characters indicating a mid point in one of the RAM tests, complete with percentile indicator. It was stuck at 18% far longer than previous increments..

..and when I physically knocked it, the backlight relit and the test resumed from where it had left off.

So I infer that a constant communication with the display module is imperative for the units operation. It's like the display is a link in the chain, without which, nothing works.

After it failed again, I poked and prodded more. Finding it more difficult to rouse with a simple knock.

After some mental gymnastics, I tried running a test lead jumper from the chassis to a mounting tab/exposed common pad on the display module PCB and lo and behold the display fired right up and the test resumed from where it had left off.

So, my theory is: with the display module mounted on a plastic faceplate, this presumably crucial link in an operative chain is at the mercy of potentially dodgy ground connection, and that confirming ground between the display module and rest of circuit might help anyone else in the same boat, since this seems a relatively common failure mode.

Is mine fixed? Only time will tell.

[madtheory]

Glad you've found a fix. When I read the bit about the RAM test, I was gonna suggest removing the SIMMS, cleaning the contacts and reseating them. I think that still might be worth doing.