Reply 80 of 104, by gdjacobs
- Rank
- l33t++
When thou dost find things not working in proper measure, first see if the supply is of juice most sweet and wholesome.
All hail the Great Capacitor Brand Finder
When thou dost find things not working in proper measure, first see if the supply is of juice most sweet and wholesome.
All hail the Great Capacitor Brand Finder
wrote:When thou dost find things not working in proper measure, first see if the supply is of juice most sweet and wholesome.
I don't quite get it.
Here you can get fantastic wallpapers created by a friend of mine: patreon.com/Unpocodrillo
Check your voltage supply rails.
All hail the Great Capacitor Brand Finder
I have set up another test bench but the memory error remains the same. Seems it's either soundcard or memory itself.
Here you can get fantastic wallpapers created by a friend of mine: patreon.com/Unpocodrillo
Is the memory 70ns or faster?
wrote:Is the memory 70ns or faster?
I don't know. How do I determine that?
Update: seems it's 80ns.
Here you can get fantastic wallpapers created by a friend of mine: patreon.com/Unpocodrillo
I doubt that your modules are defective. Maybe you have the opportunity to test the modules in a SB32/AWE32 card (The DOS based DIAG tool tests the cards DRAM).
I just ran MAUIDIAG on one of my Mauis:
After loading the DIAGMAUI.MOT firmware the program performed the tests and displayed no errors. The test machine was an AT&T Globalyst Socket 4 computer with Pentium 66. PCI / VLB / ISA type, non PnP. MAUIs addresses are set to 330h / i2/9 (standard). The tested memory were two different sets, 2x 4MB modules each, one 3 chip configuration (with parity), 60ns and one 9 chip configuration (with parity), 70ns. One pair had gold plated contacts, one pair had tin plated contacts. With both pairs the reported memory size was: 8650752 bytes.
Literally "hours" later, all tests were "passed OK".
//edit:
I got an error with another pair of RAM that has been tested with memtest before:
www.AmoRetro.de Visit my huge hardware gallery with many historic items from 16MHz 286 to 1000MHz Slot A. Includes more than 80 soundcards and a growing Wavetable Recording section with more than 300 recordings.
This is not a memory error, but UART external loopback error, as seen in error description. I get this UART error everytime, which is normal.
Ok, seems either my Maui is faulty, or 80ns memory is too slow for it.
Here you can get fantastic wallpapers created by a friend of mine: patreon.com/Unpocodrillo
I didn't get the timeout error before but maybe because I had a Korg AudioGallery hooked up and tested the MPU before.
I could test with some 1MB 80ns chips. It may help to evaluate if the speed is really causing you the error.
www.AmoRetro.de Visit my huge hardware gallery with many historic items from 16MHz 286 to 1000MHz Slot A. Includes more than 80 soundcards and a growing Wavetable Recording section with more than 300 recordings.
wrote:I could test with some 1MB 80ns chips. It may help to evaluate if the speed is really causing you the error.
Please do it then. I have to decide whether I should buy faster memory.
I can buy a 16mb 60ns module, but will it work as 8 Mb, or only as 4 Mb? Anybody knows that?
According to FAQ there may be problems with memory slower than 70 ns:
h) What kind of SIMMs can the Maui use?
The Maui uses 30 pin, 9 or 3 chip SIMMs (Standard PC 30 pin SIMMs),
70 ns or faster. Slower SIMMS (ie, >70ns) may or may not work.
The Maui requires parity, according to Turtle Beach (ie, no 2 chip
or 8 chip SIMMs, the kind Macintoshes use)
Although ICS2115 specification says that 80ns is acceptable:
It seems that a maximum of 4Mb per memory stick is supported. So, I can't just grab a 16Mb stick and have all the 8 Mb of sample RAM.
Here you can get fantastic wallpapers created by a friend of mine: patreon.com/Unpocodrillo
How many chips do your 4MB SIMMs have (eight or nine, two or three?) Nine or three chips means you have parity, which is required for the Maui. Luckily I have parity SIMMs. I also have 60ns 16MB SIMMs with parity that I can test. That way you don't order them if you don't need them.
wrote:How many chips do your 4MB SIMMs have (eight or nine, two or three?) Nine or three chips means you have parity, which is required for the Maui. Luckily I have parity SIMMs. I also have 60ns 16MB SIMMs with parity that I can test. That way you don't order them if you don't need them.
These 4 Mb 80ns sticks have 9 chips on them. Please test your 16 Mb modules so that I don't throw $30 on something that won't help 😀
Here you can get fantastic wallpapers created by a friend of mine: patreon.com/Unpocodrillo
Ok. I just tested 2x 16MB 60 parity and it did not work. Here is the result:
With a single 16MB 60 parity stick.
It shows up as 4MB and shows a similar error.
edit:
Don't buy 16MB sticks; they won't work.
Maybe your SIMMs are 8MB or different capacities? Please take a detailed picture of them and post them. Maybe someone who knows more about RAM can help.
wrote:Maybe your SIMMs are 8MB or different capacities? Please take a detailed picture of them and post them. Maybe someone who knows more about RAM can help.
http://geryon.ru/mem1.jpg
http://geryon.ru/mem2.jpg
Some people already said that it's a 4 Mb FPM 80ns stick.
Thanks for your efforts, FuzzyLogic. This inspires some hope - we see that I have the same errors, so perhaps it's memory after all, not the soundcard itself.
Meanwhile, I installed 5 Mb of RAM in my Maui and fixed three WFB banks of lesser size, all about 2 Mb.
http://geryon.ru/chaos-b.wfb
http://geryon.ru/fine_b.wfb
http://geryon.ru/ksgm_b.wfb
The corresponding SF2s, along with Fanta GM 8: http://geryon.ru/sf2s.zip
What is clear is that Awave Studio does something really weird with WFB banks. It adds some regions inexistant in original SF2s with first bank sample.
Example: a rock organ program in Fine GM bank
When I save the bank as WFB, it's still the same:
But when I close that WFB and reopen it, it's another story:
Some black magic is going on.
We see that weird regions with first sample is added to some bank programs:
But not every bank program:
This is a mess. I'll look into it later.
Here you can get fantastic wallpapers created by a friend of mine: patreon.com/Unpocodrillo
yeah, that's right. sad I didn't participate on this project earlier 🙁 been_there_done_this around 1998 when sound fonts were easy to obtain and and I could buy CR-ROMs with WF banks at a store one block from my apartment.
First of all, the WaveFront synthesizer was made 1st as a Sampler inspired by this, 2nd as a Editable Synthesizer, 3rd General MIDI compatible Synth with the possibility of downloading sound banks.
This 3rd priority is what you are trying to achieve.
So this task is not easy due the design in mind of these cards.
A few tips/things:
Yes, one midi instrument (except drums channel) can be made up to 4 layers of .wavs, layers can sound simultaneously or be spread that will take 4 sampling slots.
In the drums channel you can place any .wav in any key.
The "normal" process to create a new instrument in a WaveFront ICS would be:
raw wav --> Wave SE editor --> upload to synt --> WavePath to adjust different settings --> save WFP or WFB --> use WF control panel to load next time you want to use the bank or the DOS utility.
WaveSE looks like a .wav editor but it also can send that edited wav into the Synths ram directly. SO it's an important tool (you can only set loop points using this and it can also add effects)
First just try a Drum Kit it's easier to handle:
Get a decent Roland 808 or 909 sound font, convert it with Awave to .WFD upload it using WF control panel or WavePatch.
Try it with some game midis
one more thing:
if you really really really really really want a GM compatible sound bank to replace the Voice Crystal on the Turtles, the best way is to do it manually with Wavepatch, one instrument at a time, it's really not that hard at least for me. You can cut corners by converting a ready made GM bank and then again (can't stress this enough) using WavePatch the debug the missing/wrong sounding instruments.
You've got to learn how to use WavePatch 1.3
Trailing edge computing.
If anyone interested, just tried these modules: http://www.ebay.com/itm/2x-4MB-8MB-30pin-SIMM … HIAAOxy2YtRrRJe
Card recognizes half of actual capacity, i.e. 6mb. Modules shows full capacity on Gus pnp though.
Will try to solder parity chips and see what happens
Does anyone knows how to configure ics on irq other than 9 in pure dos? Looks like my mobo consume irq 9 even if I turn acpi off in bios
wrote:If anyone interested, just tried these modules: http://www.ebay.com/itm/2x-4MB-8MB-30pin-SIMM … HIAAOxy2YtRrRJe Card recognizes […]
If anyone interested, just tried these modules: http://www.ebay.com/itm/2x-4MB-8MB-30pin-SIMM … HIAAOxy2YtRrRJe
Card recognizes half of actual capacity, i.e. 6mb. Modules shows full capacity on Gus pnp though.Will try to solder parity chips and see what happens
Does anyone knows how to configure ics on irq other than 9 in pure dos? Looks like my mobo consume irq 9 even if I turn acpi off in bios
Check out the last few posts of this thread I started on the same topic: Intel SE440BX-2 ("Seattle") & IRQ 9
I finally got my hands on a 30-pin SIMM with 4MB of parity RAM (60ns) and was able to test one of the WFB banks. I used fine_b.wfb, and it loaded without errors.
The music is a bit off in places (some tracks in the Tyrian jukebox don't play at all), but it sounds okay with some tests and pretty good in some cases. I own a lot of ISA cards, and it's pretty neat to be able to load soundfonts from DOS, even if we are in the testing stage.
I'm using the card on an SiS 85C496 Socket 3 board with a Cyrix 5x86-120, btw.
wrote:Check out the last few posts of this thread I started on the same topic: Intel SE440BX-2 ("Seattle") & IRQ 9
Thanks. Tried /i parameter, but setupsnd refused to accept it. It said don't know such parameter. My card is tropes plus
I will check my manual later and let you know what are the possible settings under DOS.
Trailing edge computing.