Cypress AN2131SC ?

[Jidis]

Sorry for the minor question flood. 8).

I was thinking about throwing Thorsten's USB interface in with my next etch. I was wondering if that main processor was something you might find hanging around on scrap USB devices or boards, or if it's one of those "specialty items".

George

PS- If TK sees this-

1) Is that "still under development" note over with (is it still wise to make the original)?

2) I thought I remembered you having some pretty good details on the reliability and latency of a few different commercial interfaces, of various connection types. In your own experience, how would you say your's stacked up, and has it been pretty reliable about remaining online,etc.?

( I'm seriously looking for something I can count on for studio duty, with just a few ports. Mainly for controller data, but funky timing might also be a problem if I needed it for modules or clock.)

-Thanks Again! 

[raphael]

Hey Jidis,

I was thinking about throwing Thorsten's USB interface in with my next etch. I was wondering if that main processor was something you might find hanging around on scrap USB devices or boards, or if it's one of those "specialty items".

Sorry, I've never seen it in a device so far. Maybe someone else?

PS- If TK sees this-

I`m ot TK but maybe I can answer your questions...

1) Is that "still under development" note over with (is it still wise to make the original)?

The EUSART bug still exists (shame on Microchip!):

http://www.midibox.org/forum/index.php?topic=6151.0

2) I thought I remembered you having some pretty good details on the reliability and latency of a few different commercial interfaces, of various connection types. In your own experience, how would you say your's stacked up, and has it been pretty reliable about remaining online,etc.?

Yes, Thorsten wrote a benchmark app (http://www.ucapps.de/mios/midi_benchmark.zip). Details can be found here: http://www.midibox.org/forum/index.php?topic=2342.0

RME Hammerfall wins, followed by the MB_USB ;D (I`ve build already two of them and they are absolutely reliable).

Raphael

[Jidis]

Thanks again Raphael,

That's great to hear about TK's circuit. Unless something comes in from him to change my mind, I'll probably go ahead and get one or two of those chips. I've been dragging my feet on putting together a DigiKey order anyway.

It would be cool to see a version with >2x2 i/o. My main concerns are not only the latency and multi-port bandwidth, but also being confident that nothing will interrupt any heavy incoming audio channels during multi-track recording. I'd still love to find a good parallel interface circuit, but it's not looking too promising. Also still wondering who the guy was in here that did something like that.

Or was that all in my mind? :o

I may even try tracking down the guy that did Maximum MIDI (Paul Messick). He had a smaller serial one or something that looked cool, but it died off long ago. :'(

Maybe he could just rearrange some code.

It's surprising there isn't a multi-port DIY parallel out there. There's even a couple obsolete, vintage patchers and MIDI interfaces out there which would be a perfect host chassis for one. I've got an 8x8 Yamaha MIDI patchbay and an Opcode 2x6 Studio3 interface here. I may consider putting a Thorsten USB circuit in the Opcode box and mult'ing the outputs to 6. The body is in mint condition and the things are just about worthless now.

BTW- I mailed Uwe Beis (the S/PDIF guy - http://www.beis.de/Elektronik/Electronics.html), after seeing a link in that Micromixer thread recently, to thank him for the info he's compiled and created. He turns out to be cool as s*** (thanks for the tip Jack)! I was bugging him about S/PDIF VU meters and DIY ADAT. He apparently has been working on a full-blown multi-channel USB2 audio interface with mic pre's,etc., but hasn't had enough time to finish it up lately (the good ones never do ;)). Oddly enough, the analog parts of his stuff sound like the areas he's actually dissatisfied with. I wish he had taken on a good MIDI interface at some point.

-George

PS- Raphael, I couldn't help but notice just now that you and Jack both are in a related thread from the Deutsch forum back in 2004 ("8x8 USB MIDI interface + MIDImon + Serial?"). How did that end up turning out, and more importantly, what exactly was it about? (for us primitive "unilingualists" ;D) 

[raphael]

Hi George,

PS- Raphael, I couldn't help but notice just now that you and Jack both are in a related thread from the Deutsch forum back in 2004 ("8x8 USB MIDI interface + MIDImon + Serial?"). How did that end up turning out, and more importantly, what exactly was it about? (for us primitive "unilingualists" ) 

Jack wanted to integrate 4 MBUSB interfaces and a USB hub in one case. In addition he wanted to integrate a core with a serial port and a midi monitor. The hole discussion was about how to get the data for the midi monitor from the different channels so you can see all messages from all channels on the 32 nokia LCDs.

Jidis wanted to get them directely from each AN2131 via the I2C bus (J3 on the MUSB).

Thorsten pointed to the problem, that the free KEIL compiler only supports up to 4k binarys which is already used by the MBUSB firmware (the AN2131 itsself supports 8k). Jack wanted to try it anyway. Perhaps you can sen him a PM if he was successful?

(In his first post Jack mentioned a serial interface but it was nor part of the discussion)

Raphael

[Jidis]

Raphael,

Thanks for the translation! That's way heavier than anything I'd need. I was wondering if a multiple MBHP USB system like that would involve multiple individual modules, or if it's all integrated in one, with one USB line, and just additional AN2131 chips. Sounds like something which would go over quite well, not only with the MB community, but all DIY audio people. I guess Thorsten or anybody probably wouldn't want to invest much time in it though if the chip is now E.O.L., or has been replaced by something stronger.

It did seem like a lack of that sort of stuff on the web. There's usually a decent price jump from any current multi-port MIDI rack unit, and the small 2x2 and 4 port boxes. There's also usually additional time code and sync stuff, which many of us don't need these days. I'm getting more into the idea of replacing an obsolete or broken unit's guts with something like that. I was looking at mine last night thinking about it. The Opcode thing would be perfect for it, especially if the panel lights were replaced with i/o status lights from the new one, and there's already a whole row of all the appropriate back panel MIDI holes and PCB jacks. Just looked at a few on completed eBays. Anything with Mac 8 pin serial is pretty much a trash can item, and all the driver support for the others dropped off when Opcode went under, so there's probably no 2K/XP. The Studio3 looks like it can go for 5 bucks at times. There was one for two of them, which went for 13 dollars, a Studio 4 with 14(!) rear jacks and 4 front panel for 10 dollars, a Studio64x with 10 jacks or something for 9.99. You'd almost spend that much just for some decent PCB DIN sockets. Some of them have PC serial jacks on the back, plus they've all got SMPTE and footswitch stuff, so you could throw something else in there, like some audio i/o or an internal sound module. ::)

I hate to keep coming back to this, but parallel DIY would rock. There'd probably be enough feedback and outside development with an open source driver that it would live forever through multiple OS's. As sought after as those are these days in the studio world, it seems most of the available ones aren't very well supported. The wonderful looking feature-filled MOTU parallels are a total crapshoot under 2K/XP, and even if you can make them work, as with many other pieces of MOTU's sh.... um, I mean hardware, they run a chance of disappearing on you mid-session. >:(

I wish I knew enough to try to work on one, but on my level, I'm a long ways off. Please keep me informed if you should happen to run across one.

Take Care,

George 

[raphael]

I was wondering if a multiple MBHP USB system like that would involve multiple individual modules, or if it's all integrated in one, with one USB line, and just additional AN2131 chips.

USB is a point-to point design, not a bus. So you need a USB hub integrated in your box to run multiple MBUSB interfaces with only one USB connection to your computer.

Concerning the parallel interface: I don't know any DIY design. Parallel ports are already disappearing (most laptops don't have one and even on some new mainboards they are left out :() So I don`t know if someone will do a new disign for it. In addition you would have to write your own driver for W2K and XP which is not so easy. So I think you're out of luck with this.

Raphael

[Jidis]

Thanks again Raphael,

If worse comes to worst, 2-2x2 USB's may be enough. I may look into their effect on heavy multi-track streams before keeping them up while tracking. I'm considering putting a few simple (standard pot) MB controllers in the control room for some of the basic processor plugs, and would probably be running more than one core. I'm guessing I wouldn't want it all squashed into one or two serial MIDI lines. ;D

Also, if you or anyone else knows- Assuming the USB bandwidth is high enough, the timing and latency of the driver you guys are using is adequate, and the stream doesn't have any drastic affect on the audio activity: Are a bunch of 31.x kHz MIDI signals squashed into a USB line (like with a hub) much different than one or two in that regard? Also guessing this is all well within USB 1.0 spec., and there would be no point in running a fast hub for these chips(?). My boards do both formats, and have multiple jacks as well as a couple separate USB busses, so I could also split a multi-MBHP USB arrangement to more than one USB bus or port on the machine if it made any difference. The backs of the machines and racks are pretty well buried, so extra cable mess wouldn't make much difference.

I still want that parallel thread, and I'm still pretty sure it happened in here, but I've done just about every search I could think of. Maybe the guy that was working on that one will sense the presence of this discussion and come flying in from wherever it is he went to, like a superhero or something. :D

The driver issue is obviously way over my head, so if I ever find a circuit, I would assume that whoever had designed it would also have written one, or be using an existing one.

FWIW- I don't run laptops or Macs now, and most of my studio boards seem to be the type where they just keep adding to the available port types as the technology advances, so I don't see parallel or serial disappearing on the ones here for a while (knock on wood). I'm also sort of hoping that Synsopos, the great god of USB dongles, will be merciful and grant me the ability to run a couple USB MIDI circuits if I need to.

Take Care,

George

PS- Has anyone, by chance, had that MB USB interface (or the driver) go undetected or temporarily vanish on them while using it? It had happened with my MIDISport, and might be a bit embarrassing if all my controls went out in front of anybody.       

[moebius]

Also, if you or anyone else knows- Assuming the USB bandwidth is high enough, the timing and latency of the driver you guys are using is adequate, and the stream doesn't have any drastic affect on the audio activity: Are a bunch of 31.x kHz MIDI signals squashed into a USB line (like with a hub) much different than one or two in that regard?

Absolutely not. USB has "build in" jitter and latency for any kind of data. ;D Mixing stream will not affect this or make it much worse. Buahahahaa!

Moebius - soon to be happy owner of Parallel port Miditimepiece AV *grin*

[Jidis]

Moebius,

I guess you mean it's already screwed up, so more serial data muxed in won't make it much worse >:(?

Man, I really hate to burst your bubble on this, but if you haven't got any prior experience with them, or haven't been reading a lot of web info, try not to get overly "high hopes" on that MTPAV. The Micro Express parallel was one of my biggest gear disappointments in memory, and MOTU unfortunately seems to have decided on a "unified" driver version for their USB and COM port MIDI devices, neither of which appear to be 100% trustworthy, and the driver is already a couple years old, despite all the problems and complaints. FWIW, I think I've seen less complaints on the MTP's than the smaller ones, but it may just be the quantities in service. We ran the MTPAV at a place I worked at a few years back and it did OK, but that was back in the 98 days, and the main machine it was on was actually a beige G3 (8 pin serial). I think I had some of the notorious MOTU trouble with it on occasion, maybe more on the PC side of the chain, but I doubted my own experience with such an elaborate setup too much to realize it was actually their fault. ;D

If you want an idea of some of the issues, drink a bunch of cough medicine or something to relax, and then go read this-

http://caps.pcrecruiter.com/MOTU/

It's old info, and some believe it has been resolved, but many have realized that it hasn't. I don't think the guy that wrote it had even run his box for long after whatever driver came out back then, so he probably hadn't yet hit anything. Although the info is good, and I appreciate the time and knowledge he invested, I still don't think some of it is a 100% cure. I've seen users post of how they got their parallel MOTU box working by a specific, yet convoluted, startup sequence, but I fear they were overly optimistic after seeing it work a few times. Like others, even when I got mine out of "MIDI Flyer" mode, it had the ability to screw up or disappear while it was on the system. I'd almost rather have it go undetected. The SysEx thing mentioned on that site was also a serious problem here, and his suggestion to change channels or something, wasn't possible with the output of my JLCooper.

It's a real shame that their software priorities and support are so horrible, especially for those MIDI units, as they have so many nice non-USB ones. The Micro was 4x6 or something, with a pair of front ports, had individual status lights, and was a half rack desktop unit, which probably would have been perfect for what I needed. It worked OK in 98, probably on the Mac as well, so it's obviously capable "hardware-wise". I'm keeping it around in hopes that someday they'll get off their asses and fix some of it, but it doesn't look promising. They invest far too much in rushing out new hardware and advertising it, so I'm guessing that keeping any previous customers happy isn't at the top of their list.

If (or when) you start hitting any bumps with it, you may want to write them and make sure they know what's happening, and are aware of your interest in parallel MIDI and intention of remaining on a PC. That companies priorities and the treatment of it's customers has really begun to disgust me. When I first gave up on the Micro, I wrote them a nice polite letter explaining most of that, and sent it to customer relations. So far, there's been nothing, and the support email correspondence ended after the first "try a different cable" type crap response they gave me didn't work. I've taken every opportunity to let people know what the common experiences are with their hardware's reliability and their poor Windows support. Some are under the impression that it's changed, but fishing around would prove otherwise.

Here's another bunch of related comments (most of their MIDI stuff has the same quirks):

http://www.sonicstate.com/digital/comments.cfm?modelid=152&srow=1

and another depressing, but interesting article about some of their company policies:

http://www.soniccontrol.com/blog/index.php?itemid=6

and a somewhat "young" forum (mainly about XP & MOTU's audio boxes)

http://s15.invisionfree.com/MOTUXP

Then of course there's UnicorNation, which would probably be more helpful if it's info were open for unregistered public viewing and would show on search engines.

I'll probably be seeing you in there. ;)

Wish you luck with it!

George