Interface for FS and hardware

[Guest LeoL]

@Manuel>The cheapest PIC over here in Germany costs just 1.50EUR (and thats one-off, end user price).WOW! I should buy my parts from Germany. A PIC18F costs me $15-20 CAD depending on quantity this side of the Atlantic!>True that logic chips CAN be faster clocked, but in our>applications they rarely are. In my project, the fastest chip>(microcontroller) runs 40MHz. For some logic families this is>already pretty fast.>>But there's not much externally running faster than say 1-4>MHz.True especially since there's EMI/EMC considerations when running an external clock line higher than 4-6Mhz anyway so its best to keep things slow. 40Mhz is fine in your case for the micro since that is likely an internal freq.>If I'd start over designing cockpit interface stuff, I'd>seriously look into CPLDs and FPGAs. All you'd need there>would be one chip and possibly driver chips for the stuff that>draws more current than a few mA.I've considered FPGA's myself for my designs and although they represent an elegant solution, the cost of one FPGA can exceed the entire cost of a module using discrete ICs. They may be cheaper in Germany, but definitely not here. It's a great solution if you're building something custom for yourself where price is not a factor, but for commercially sold boards the price will likely be out of reach for most builders.-Leo

[Guest LeoL]

>I know USB can do 12MBit/s but this is just the raw value and is only good for marketing purposes.Marketing purposes indeed!I can't speak for Ferdie's design, but I've used the same USB chip in my modules and can tell you that transfer rates are 8 bytes / 10 ms frame (800 Bytes/s) using a single pipe since the PIC16C745/765 is a Low Speed USB chip.The control pipe can transfer 64Bytes / ms (64kBytes/s), but it is not recommended practice.A Full Speed chip can transfer up to 64Bytes / ms (64kBytes/s) using a non-isochronous pipe...that is if your PC application can keep up with 1ms frames.These values are guaranteed for non-isochronous mode and defined by the USB specs. I don't have the numbers on other transfer modes such as "bulk" and isochronous but I think you can push it up to 128Bytes/ms (128kBytes/s) and is dependant on availability of USB bus bandwidth. More devices plugged in, less bandwidth available for isochronous modes.-Leo

[tigert 0]

>Hello my fellow simmers.>>I'd like to know if anyone, i say again, anyone has completed>a cockpit using any of the systems.>>As far as I know, only EPIC people have completed their>simulators. There are maaaaany project hardware and software>available, but seeing any of these WORKING on a full system>setup is a thing yet to be seen.>>I say again, Anyone has COMPLETED his cockpit using anything>but EPIC? And I mean Full instrument panel with CDUs,>throttles, speedbrakes, landing gear, autoflight (MCP,>autothrottle, FMS), radios, instructor station with weather,>time and system control, and overhead panel with complete>systems (fire protection, electrical, hydraulic, lights, IRS,>pneumatic, fault detection, and so on...)>>Cheers.We have a completed sim - but it is not a big iron, however, but a C172. Still it has a IFR-capable radiostack with 7-segment displays (FSBUS) that works very well, I dont quite understand the "speed" issue here, since most of the time this is not some overclocking fest of PIC vs logic chips :) The radio displays update just fine, no issues with delay.Our instruments are SimKits gauges, since we thought that sure, we could do it all ourselves, but it'd take over a year or more, and FSBUS currently has the limit of 31 or so expansion cards, one servo controller can take 8 servos, and one needs a lot for a full analog instrument panel -> no go. Thus SimKits - we saved time, and it seems to perform pretty satisfactorily. Some gauges are not super smooth, but they have improved their software over time, so stuff works pretty nicely.What else.. hmm. All switches for lights, magnetos, fuel valve, parking brake etc, go via FSBUS. Pedals and toe brakes go via FSBUS AD-module, throttle, mixture, etc go via the same.For the yoke we built the "Baby Plasma" or "MiniMe" or what it is called now. This is incredibly good choice - the control is precise and smooth and no spikes whatsoever.So, here you have a complete sim. Even for a C172 there is a LOT of stuff. Took us about 2 years to finish it, and we still have stuff to do - we need to design and implement the elevator trim with yoke centering and also we have 2 TFT screens for side windows, that is something we'll wonder about in the fall when the weather gets bad and there is no real flying to be done for a while.Like you see, there is no "magic wand" to do everything, but we used several technologies to create the simulator. The goal has been achieved though - no mouse / keyboard once you step inside the cockpit.Was it insane? Yes. Was it easy? No. Lots of sweat, though the building was fun. But REMEMBER, this is not something you do over the summer holiday to impress your friends. It takes an *INSANE* amount of work, so if you can, try to plan in steps so you can take a break and FLY the sim when (yes, when) you get tired of building. You can see that when you look at the number of home cockpit sites that were last updated in 2000 or 2001 - or the ones that were alive when we started in 2002, and havent been updated since. Build a MCP first. Dont take a too big bite at once, it does not necessarily need to be a *replica* of a real airplane, a MCP for those "big jet" guys is probably the best thing you can get for starters, then a throttle and a yoke. Build in steps :)Okay, time to shut up, but you get the idea. :) This is an awesome hobby, but do not take too big of a step for starters. Remember, it took 5 guys almost 2 years to build a C172. How long do you think it takes to design, build, wire, solder, program and debug a modern airliner jet? It is not just a nice purchase from FlightDeckSolutions, a website with a cool domain name :) Plan realistically, and keep your feet on the ground while your head and heart is soaring :)Sure, we could have done it faster, but we had a budget of roughly 2000 euros from the club. Simkits was 1200, we got a bit of stuff donated but I think the total is way under

[Guest High Octane]

>EPIC is out of this age. The choice nowadays is FSBUS or >OpenCockpits IOCards. "Never judge book by its cover!" Meaning?? nothing about card'sage, its how interface card can statisfy cockpit builder's dream. I have tried FSBUS, IOcards, KE-72. And I choose EPIC as number one.

[Guest High Octane]

>EPIC nowadays does not have any functionality that cannot be made by >a IOCard circuit.Are you sure? If I were you, I would check out the info about EPIC and IOcards again.I m not trying to causes any "battle-debates" here about EPIC compared to other interface cards, I m here trying to point out the reason why not so many builders out there is buying EPIC card. Its not about the ages or anything, its based on QUALITY of the card. Like I metioned early to Kevin.. "never judge book by it's cover" All interface card has almost same features as EPIC, but only difference is the quality. So its not based what it have, or how it can work. Its based on cockpit builder's perference. Some like to buy cheap sone like to buy expensive they can still work same.Its like comparing between Mercedes and Kia , nothing different between them except with quality. since they have engine, tires, light and etc. But of course mercedes is much more elite than Kia which makes it expensiver than kia. So is EPIC compared to other Interface cards.

[Guest mbessler]

Hi Leo,>WOW! I should buy my parts from Germany. A PIC18F costs me $15-20 CAD depending on quantity this side of the Atlantic! The cheapest (PIC16F630) costs 1.45 EUR.An 18F252 costs 6.55EURThe 16F627 costs 2.45EURThose are all single quantity. Since you use PICs commerically, have you considered buying them directly from Microchip ?When I checked with buy.microchip.com a few weeks back, their prices seemed pretty good (similar to my prices here in Germany) >True especially since there's EMI/EMC considerations when running an external clock line higher than 4-6Mhz anyway so its best to keep things slow. 40Mhz is fine in your case for the micro since that is likely an internal freq.Yes, the 40MHz are interal. 10MHz multiplied by a 4xPLL. >I can't speak for Ferdie's design, but I've used the same USB chip in my modules and can tell you that transfer rates are 8 bytes / 10 ms frame (800 Bytes/s) using a single pipe since the PIC16C745/765 is a Low Speed USB chip.Oh, the C745/765 is low speed ... I have looked at the datasheet before but forgot its only Low speed. The control pipe can transfer 64Bytes / ms (64kBytes/s), but it is not recommended practice.>A Full Speed chip can transfer up to 64Bytes / ms (64kBytes/s) using a non-isochronous pipe...that is if your PC application can keep up with 1ms frames.>These values are guaranteed for non-isochronous mode and defined by the USB specs. I don't have the numbers on other transfer modes such as "bulk" and isochronous but I think you can push it up to 128Bytes/ms (128kBytes/s) and is dependant on availability of USB bus bandwidth. More devices plugged in, less bandwidth available for isochronous modes.Thanks for that info. :-)Have you ever experimented with the FTDI232 chips ? The Serial<->USB converter chip. The data sheet seems to suggest it can do Full Speed. Something I found last week, have you seen this: http://www.cesko.host.sk/IgorPlugUSB/IgorP...20(AVR)_eng.htm That guy used an AVR without USB capability and made it USB capable with two resistors and a USB implementation in firmware. Manuel

[Guest mbessler]

Hi Ferdie,>IOCard not only a free project, is a OPEN project and nowadays>have many people developing, or write manuals or tutorials. InI don't think IOCards can be considered a free project (free as in freedom) since you disallow commercial use.If someone wanted to use it in a cockpit that he charges money for to fly (for fun of for "training pilots") he would not be allowed to do this. So I wouldn't consider it "free".Manuel

[Guest LeoL]

>Since you use PICs commerically, have you considered buying>them directly from Microchip ?"commercially" is a relative term in my case. Lets just say it