Post by thread:[PIC]: Modem suggestions

PCI is _much_ more complicated than ISA. I'd avoid this like a plague! ----- Original Message ----- From: Alan B. Pearce <spam_OUTA.B.PearceTakeThisOuTRL.AC.UK> To: <.....PICLISTKILLspam@spam@MITVMA.MIT.EDU> Sent: Monday, October 30, 2000 3:52 AM Subject: Re: Modem suggestions >Anyone tried interfacing a PIC to the PCI bus? I think I would only try this after reverse engineering the software for your winmodem of choice. That seems to be all that is available these days for PCI bus. -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

On Mon, 30 Oct 2000 06:58:27 -0500 Bob Ammerman <RAMMERMANKILLspamPRODIGY.NET> writes: > PCI is _much_ more complicated than ISA. > > I'd avoid this like a plague! > And since ISA stuff is going away, we're stuck with expensive chipsets or modules when we want to integrate a modem into our product... Harold FCC Rules Online at http://hallikainen.com/FccRules Lighting control for theatre and television at http://www.dovesystems.com ________________________________________________________________ YOU'RE PAYING TOO MUCH FOR THE INTERNET! Juno now offers FREE Internet Access! Try it today - there's no risk! For your FREE software, visit: dl.http://www.juno.com/get/tagj. -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

Harold Hallikainen wrote: >On Mon, 30 Oct 2000 06:58:27 -0500 Bob Ammerman <.....RAMMERMANKILLspam.....PRODIGY.NET> >writes: >> PCI is _much_ more complicated than ISA. >> >> I'd avoid this like a plague! >> > > And since ISA stuff is going away, we're stuck with expensive chipsets >or modules when we want to integrate a modem into our product... > Several people have bad-mouthed the idea, but what the heck is the big problem with interfacing the PIC with an "external" modem? They are more expensive than internal, and probably are "not" going to go away in the near future, but this sure sounds 100X easier to deal with than trying to tap an ISA or PCI bus edge connector. RS-232 channel plus 2 or 3 handshakes. ??????????????? - danM -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

On Mon, 30 Oct 2000, Dan Michaels wrote: {Quote hidden}> Harold Hallikainen wrote: > >On Mon, 30 Oct 2000 06:58:27 -0500 Bob Ammerman <EraseMERAMMERMANspam_OUTTakeThisOuTPRODIGY.NET> > >writes: > >> PCI is _much_ more complicated than ISA. > >> > >> I'd avoid this like a plague! > >> > > > > And since ISA stuff is going away, we're stuck with expensive chipsets > >or modules when we want to integrate a modem into our product... > > > > > Several people have bad-mouthed the idea, but what the heck is the > big problem with interfacing the PIC with an "external" modem? > > They are more expensive than internal, and probably are "not" going > to go away in the near future, but this sure sounds 100X easier to > deal with than trying to tap an ISA or PCI bus edge connector. > RS-232 channel plus 2 or 3 handshakes. ??????????????? External modems are getting pretty pricey if bought new, last I checked. If you're using used modems, there are issues with getting enough the same make/model/rev level to respond predictably. In my case, I find the app will most likely result in a bunch of different modem types, which means coming up with an init string for each one. Plus you have an external device under its own power, doing its own thing... the code required to do error detection and handling, try to reset the modem, etc is very expensive in terms of program memory. I've also had many, many instances of standalone modems needing to be power cycled to get them out of a non-responsive hung condition. Without a lot of board space and components devoted to controlling the power to the external modem (which may be AC or DC, either polarity, any of several different connector styles which manufacturers change at whim, etc) you stand a good chance of ending up with a unit in a useless state needing manual intervention, in my case at a very remote location requiring a lot of very expensive labor to reach. And the worst part -- no way to detect that the remote is down without polling frequently, which is simply not an option at 250 long distance calls a couple of times a day. So those are my reasons for not liking to use standalone modems, at least the main ones. If this were a onesie-twosie hobby project or had an unlimited budget, yeah, buy a new external modem, write the code for that modem, all's well. This is for a larger project, though, with cost constraints and deployment in remote areas. Dale --- The most exciting phrase to hear in science, the one that heralds new discoveries, is not "Eureka!" (I found it!) but "That's funny ..." -- Isaac Asimov -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

Dale Botkin wrote: >External modems are getting pretty pricey if bought new, last I checked. >If you're using used modems, there are issues with getting enough the same >make/model/rev level to respond predictably. In my case, I find the app >will most likely result in a bunch of different modem types, which means >coming up with an init string for each one. Plus you have an external >device under its own power, doing its own thing... the code required to >do error detection and handling, try to reset the modem, etc is very >expensive in terms of program memory. > I would think that pretty much everything you mentioned here pertains to internal ISA/PCI modems, as well as external - except h.w. reset, which could be done with a relay, and one power supply for both PIC and modem. So I guess in your project you're going towards full integration rather than tapping a PC-internal job. Problem here of course, is modem chips/DAA modules cost so much more than any kind of commodity PC jobs - esp to get anything approaching 56 kpbs. Would be interesting to hear what you finally decide on. - danM -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

Bob Ammerman wrote: >Negatives for external modems: > >-Size >-Power supply >-Two boxes >-"clunkiness" >-Room for end users to goof things up >-Cost > Negatives for internal modems: -Size -Power supply -Two pcbs wired together -"awkwardness" -Pain-ina-butt interfacing to PCI or ISA bus -Writing modem-specific s.w. -Cost [cheap - only advantage I can see] Negatives for integrating chip modems: -Finding a chip <= 64-pin smt -Writing "all" the s.w. -Designing complex multi-layer pcb -Designing complex analog-digital-filtering ckt -Integrating DAA -Getting FCC approval -Cost [chips ain't cheap in < 100,000 qty] -Cost [prohibitive when using available DAA modules] Seems every route has its own personal set of problems. - danM -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

Uh oh, Dan, we agree again. As to Bob's "two boxes" and "power supply" I suggest that you get one bigger box, bolt the external modem into it, and bolt the application board somwhere else in the box. Or maybe inside the modem, if there's room. The power can be handled smartly by having the input jack from the wall wart plug into the application board, which then draws its power from it and passes the rest on to the modem. If the original wall wart is incapable of supplying adequate power, just get a bigger one instead. Or you could charge a (internal) battery and let the modem draw power from there when it needs to run the modem. The modem, after all, only needs power when it's in use. Andy Dan Michaels <oricomspam_OUTUSWEST.NET> on 10/30/2000 02:39:43 PM Please respond to pic microcontroller discussion list <@spam@PICLISTKILLspamMITVMA.MIT.EDU> To: KILLspamPICLISTKILLspamMITVMA.MIT.EDU cc: (bcc: Andrew Kunz/TDI_NOTES) Subject: Re: [PIC]: Modem suggestions Bob Ammerman wrote: {Quote hidden}>Negatives for external modems: > >-Size >-Power supply >-Two boxes >-"clunkiness" >-Room for end users to goof things up >-Cost > Negatives for internal modems: -Size -Power supply -Two pcbs wired together -"awkwardness" -Pain-ina-butt interfacing to PCI or ISA bus -Writing modem-specific s.w. -Cost [cheap - only advantage I can see] Negatives for integrating chip modems: -Finding a chip <= 64-pin smt -Writing "all" the s.w. -Designing complex multi-layer pcb -Designing complex analog-digital-filtering ckt -Integrating DAA -Getting FCC approval -Cost [chips ain't cheap in < 100,000 qty] -Cost [prohibitive when using available DAA modules] Seems every route has its own personal set of problems. - danM -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

30 Oct 00, Dan Michaels writes to All: P> Negatives for integrating chip modems: P> -Finding a chip <= 64-pin smt P> -Writing "all" the s.w. P> -Designing complex multi-layer pcb P> -Designing complex analog-digital-filtering ckt P> -Integrating DAA P> -Getting FCC approval P> -Cost [chips ain't cheap in < 100,000 qty] P> -Cost [prohibitive when using available DAA modules] Did you check Si2400 from Silicon Labs ? http://www.silabs.com/products/isomodem.asp They have nice 2-chip solution (SOIC16) with integrated DAA, 3.3/5V 60 mW, up to 2400 bit, check also Microchip Appnote AN731 for Si2400 application. They quote $8.50 for chipset in 25000 quantity, it means about $15...$25 in small qty. If V.23 or Bell202 (1200 bit) will be enough for your project, you can find another solution - software modem on the Scenix chip. www.scenix.com/virtual/download/v23Kit.html V.23 demo kit available for $89 from Monterrey Tools Probably one can manufacture similar modem on SMD components for the same $15...$25 in qty 100. And you will have some processor resources left for another task... P> Seems every route has its own personal set of problems. For low-speed mid-volume and high-volume applications integrated modem seems the most adequate solution. >> Negatives for external modems: >> -Size >> -Power supply >> -Two boxes >> -"clunkiness" >> -Room for end users to goof things up >> -Cost External modem will be the best solution, when you need several experimental devices. P> Negatives for internal modems: P> -Size P> -Power supply P> -Two pcbs wired together P> -"awkwardness" P> -Pain-ina-butt interfacing to PCI or ISA bus P> -Writing modem-specific s.w. P> -Cost [cheap - only advantage I can see] Possible application for internal modems - cheap multiport router under Linux - put 4 modems into old desktop with 486 processor, 8/16 meg and small hard drive and run Linux on it - it can be the cheapest solution for 4-port dial-up server. However, it is not so reliable, as Cisco router, but works OK for beginners. I know many internet providers, who start from this solution... You can also found this solutions working today in some not so advanced countries... I am not sure, that internal ISA or PCI modems can be used successfully in embedded applications. Best regards Alexey -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

On Mon, 30 Oct 2000, Dan Michaels wrote: > Dale Botkin wrote: > > >External modems are getting pretty pricey if bought new, last I checked. > >If you're using used modems, there are issues with getting enough the same > >make/model/rev level to respond predictably. In my case, I find the app > >will most likely result in a bunch of different modem types, which means > >coming up with an init string for each one. Plus you have an external > >device under its own power, doing its own thing... the code required to > >do error detection and handling, try to reset the modem, etc is very > >expensive in terms of program memory. > > > > I would think that pretty much everything you mentioned here pertains > to internal ISA/PCI modems, as well as external - except h.w. reset, > which could be done with a relay, and one power supply for both PIC and > modem. To some degree, yes. The difference is that I *could* easily control power and therefore the known state of an ISA modem. Either ISA or PCI is pretty much out for this project, though, as I don't have six months to figure out how to interface with PCI and don'thave enough spare I/O pins for either. > So I guess in your project you're going towards full integration rather > than tapping a PC-internal job. Problem here of course, is modem chips/DAA > modules cost so much more than any kind of commodity PC jobs - esp to get > anything approaching 56 kpbs. Would be interesting to hear what you > finally decide on. Well, I would if I could get any manufacturers to sell integrated modems in small quantities, or a distributor who had anything for a reasonable price. I've had several people poitn out really nice looking modem chipsets that would be great if I were planning to build a few thousand units, but if it doesn't have a low cost, low volume US distributor I can't use it right now. I don't need anything even remotely approaching 56K, in fact I would prefer a 212A modem. I've not seen a 1200 connection fail for a long time, but V.90 sucks for reliability on anything but clean lines. Remember this is a PIC driving it, not me sitting at the keyboard knowing that the supposed 40K connection is useless when it doesn't respond. These things will dial in once in a great while and send a couple of lines of text, so even 1200 is overkill. 300 would do. I'm trying one more time to get a response out of Cermetek, since the others seem to be two to three times more expensive and way out of budget for this. Mean time, I'm rewriting the modem handler to be much, much more generic, and thinking seriously about an external power control relay for the PIC to be able to power-cycle an external modem. Dale --- The most exciting phrase to hear in science, the one that heralds new discoveries, is not "Eureka!" (I found it!) but "That's funny ..." -- Isaac Asimov -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

On Mon, 30 Oct 2000, Dan Michaels wrote: > Seems every route has its own personal set of problems. Ain't THAT the truth!!! I'm beginning to question the wisdon of tkaing on this project. Seemed so simple at the time! I think I'm about a forty hours further into this than I planned... but I'm learning a WHOLE lot. I suspect I'll end up with a very fine general purpose modem driver .h file written for CCS C, made as generic as humanly possible, that will do what's needed with just about any Rockwell or USR based modem. When I arrive at that happy point I will, of course, share the code. Dale --- The most exciting phrase to hear in science, the one that heralds new discoveries, is not "Eureka!" (I found it!) but "That's funny ..." -- Isaac Asimov -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

> >I don't need anything even remotely approaching 56K, in fact I would >prefer a 212A modem. I've not seen a 1200 connection fail for a long >time, but V.90 sucks for reliability on anything but clean lines. >Remember this is a PIC driving it, not me sitting at the keyboard knowing >that the supposed 40K connection is useless when it doesn't respond. >These things will dial in once in a great while and send a couple of lines >of text, so even 1200 is overkill. 300 would do. If you're doing credit card transactions, anything other than a 212a modem will just slow them down. The hosts are still 212A because the call establishment of the higher speed modes takes too long. -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

On Mon, 30 Oct 2000, David VanHorn wrote: > So there is this external modem, which is specifically designed to do > credit card transactions, with a true bell 212 modem, which is what you > need for this. > > It has a serial port, which you can easily connect to your system. > > It has a display, keyboard, and card reader, but these can be enabled or > disabled as you like. > > It's called a Tranz 330. Yes, but I could buy Cermetek or Xecom integrated modems cheaper from an inflated distributor. Even used Verifone units go for $50 and up, usually over $100, from what I have seen -- unless you know where a bunch could be had in the $30-40 range? That's where we need to be. Dale --- The most exciting phrase to hear in science, the one that heralds new discoveries, is not "Eureka!" (I found it!) but "That's funny ..." -- Isaac Asimov -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

On Mon, 30 Oct 2000, Alexey Vladimirov wrote: > Did you check Si2400 from Silicon Labs ? > http://www.silabs.com/products/isomodem.asp > They have nice 2-chip solution (SOIC16) with integrated DAA, 3.3/5V 60 mW, > up to 2400 bit, check also Microchip Appnote AN731 for Si2400 application. > They quote $8.50 for chipset in 25000 quantity, it means about $15...$25 in > small qty. > If V.23 or Bell202 (1200 bit) will be enough for your project, you can find > another solution - software modem on the Scenix chip. > www.scenix.com/virtual/download/v23Kit.html > V.23 demo kit available for $89 from Monterrey Tools > Probably one can manufacture similar modem on SMD components > for the same $15...$25 in qty 100. And you will have some processor resources > left for another task... And either way I get to design and FCC certify a new modem. That's why I was looking to Cermetek or similar products, or a standalone unit. I don't want to try to design and build a line interface for this project, it's not my area of expertise. Dale --- The most exciting phrase to hear in science, the one that heralds new discoveries, is not "Eureka!" (I found it!) but "That's funny ..." -- Isaac Asimov -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

unless you know where a bunch could be had in the $30-40 range? That's where we need to be. Customers for embedded products (and the developers too) need to be aware that you just don't GET the same economies of scale in an embedded system as you would in a "commodity" PC. Just because you can get a PC modem, ethernet, USB, or other IO card for $30 doesn't mean that you can put the same sort of functionality into a non-commodity platform for the same price. Even if your platform is a PC, you're faced with any particular technology going in&out of "commodity" status very quickly. (for example, "thinnnet" ethernet boards are getting hard to find...) You might want to look into some PCMCIA modems (which will hopefully outlive the ISA bus by a bit.) Don't some PCMCIA modems essentially bypass most of the PCMCIA bus and just sit off a serial port? Also, perhaps internal modems aimed at mac or older PC laptops might be useful. BillW -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

>Several people have bad-mouthed the idea, but what the heck is the >big problem with interfacing the PIC with an "external" modem? >They are more expensive than internal, and probably are "not" going >to go away in the near future, but this sure sounds 100X easier to >deal with than trying to tap an ISA or PCI bus edge connector. >RS-232 channel plus 2 or 3 handshakes. ??????????????? Hey, why use an ISA modem? Gather a RC2496 from Rockwell and some glue logic. http://www.ustr.com has some nice info on it --------------8<-------Corte aqui-------8<-------------- All the best!!! Alexandre Souza RemoveMExandinhoTakeThisOuTinterlink.com.br --------------8<-------Corte aqui-------8<-------------- -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

>for this. Mean time, I'm rewriting the modem handler to be much, much >more generic, and thinking seriously about an external power control relay >for the PIC to be able to power-cycle an external modem. If you have an external modem and level translators then you can implement a modem power switch using the BREAK condition on TxD. It'a a simple RC combination driving a bipolar transistor. I did this a long time ago, using two transistors and a relay. It steals power from the modem wall wart and the relay cuts the power to the modem after the circuit. The two transistors implement a standard Schmitt Trigger taken from a book. You can use a monoflop instead with a 555 or such. This works with any computer not just PICs. Set BREAK for 1 second - hop the modem is off. You can leave the line in the BREAK state forever to avoid the modem taking power (and answering the phone or bugging the lines if it is electrically 'crazy' if you use the second pair of relay contacts to separate the phone line). This solution was used in a modem pool (many modems) where a hanged modem usually meant service on Sunday. Peter -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! use TakeThisOuTlistservEraseMEspam_OUTmitvma.mit.edu?body=SET%20PICList%20DIGEST

On Wed, 1 Nov 2000, Peter L. Peres wrote: > >for this. Mean time, I'm rewriting the modem handler to be much, much > >more generic, and thinking seriously about an external power control relay > >for the PIC to be able to power-cycle an external modem. > > If you have an external modem and level translators then you can implement > a modem power switch using the BREAK condition on TxD. It'a a simple RC > combination driving a bipolar transistor. I did this a long time ago, > using two transistors and a relay. It steals power from the modem wall > wart and the relay cuts the power to the modem after the circuit. The two > transistors implement a standard Schmitt Trigger taken from a book. You > can use a monoflop instead with a 555 or such. This works with any > computer not just PICs. Set BREAK for 1 second - hop the modem is off. You > can leave the line in the BREAK state forever to avoid the modem taking > power (and answering the phone or bugging the lines if it is electrically > 'crazy' if you use the second pair of relay contacts to separate the phone > line). This solution was used in a modem pool (many modems) where a hanged > modem usually meant service on Sunday. > Now, there's a good idea. I could use a 4PST relay to shut off the modem *and* disconnect it from the phone line when not in use, thus preventing 99.999% of the modem-destroying phone line transients as well! What a great idea. Dale --- The most exciting phrase to hear in science, the one that heralds new discoveries, is not "Eureka!" (I found it!) but "That's funny ..." -- Isaac Asimov -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! use RemoveMElistservTakeThisOuTmitvma.mit.edu?body=SET%20PICList%20DIGEST

----- Original Message ----- From: "William Chops Westfield" <billwEraseME.....CISCO.COM> To: <EraseMEPICLISTMITVMA.MIT.EDU> Sent: Monday, October 30, 2000 10:52 PM Subject: Re: [PIC]: Modem suggestions > You might want to look into some PCMCIA modems (which will hopefully outlive > the ISA bus by a bit.) Don't some PCMCIA modems essentially bypass most of > the PCMCIA bus and just sit off a serial port? Also, perhaps internal > modems aimed at mac or older PC laptops might be useful. Does anyone have any experience of using a PCMCIA modem with a microcontroller? Can it be done without implementing the whole card/socket services shebang? If so it could save a lot of messing around with external modems which, because of the painful telecom approval procedures over here, is often the "best" solution for low volume products. . -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

On Mon, 6 Nov 2000 01:20:51 -0000, you wrote: >----- Original Message ----- >From: "William Chops Westfield" <RemoveMEbillwEraseMEEraseMECISCO.COM> >To: <RemoveMEPICLISTspam_OUTKILLspamMITVMA.MIT.EDU> >Sent: Monday, October 30, 2000 10:52 PM >Subject: Re: [PIC]: Modem suggestions > > > >> You might want to look into some PCMCIA modems (which will hopefully >outlive >> the ISA bus by a bit.) Don't some PCMCIA modems essentially bypass most >of >> the PCMCIA bus and just sit off a serial port? Also, perhaps internal >> modems aimed at mac or older PC laptops might be useful. > Yes - with most PCMCIA modems you are effectively talking to a 16550-style uart with a hayes-type modem connected to it. About the only part of the PCMCIA spec you need to know is the pinout! Almost all of the card/socket services stuff is to do with mapping the card into the host machine and so is not required for a MCU application. Unfortunately you won't be able to get at the serial part directly (in modern modems this probably doesn't physically exist), but talking to the uart should not be at all difficult. One minor caveat, though - don't assume that a simple interface will necessarily work with ALL PCMCIA modems - there may be some that are implemented differently. This should not be a major problem - just be sure to test any modems you plan to use, and don't let users plug just any old modem in and expect them to work. -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! use RemoveMElistservTakeThisOuTspammitvma.mit.edu?bodyT%20PICList%20DIGEST

{Quote hidden}> -----Original Message----- > From: Mike Harrison [SMTP:EraseMEmikespamspamBeGoneWHITEWING.CO.UK] > Sent: Monday, November 06, 2000 12:59 PM > To: RemoveMEPICLISTKILLspamMITVMA.MIT.EDU > Subject: Re: [PIC]: Modem suggestions > > On Mon, 6 Nov 2000 01:20:51 -0000, you wrote: > > >----- Original Message ----- > >From: "William Chops Westfield" <billwSTOPspamspam_OUTCISCO.COM> > >To: <spamBeGonePICLISTSTOPspamEraseMEMITVMA.MIT.EDU> > >Sent: Monday, October 30, 2000 10:52 PM > >Subject: Re: [PIC]: Modem suggestions > > > > > > > >> You might want to look into some PCMCIA modems (which will hopefully > >outlive > >> the ISA bus by a bit.) Don't some PCMCIA modems essentially bypass > most > >of > >> the PCMCIA bus and just sit off a serial port? Also, perhaps internal > >> modems aimed at mac or older PC laptops might be useful. > > > Yes - with most PCMCIA modems you are effectively talking to a > 16550-style uart with a hayes-type modem connected to it. About the > only part of the PCMCIA spec you need to know is the pinout! Almost > all of the card/socket services stuff is to do with mapping the card > into the host machine and so is not required for a MCU application. > Unfortunately you won't be able to get at the serial part directly (in > modern modems this probably doesn't physically exist), but talking to > the uart should not be at all difficult. > One minor caveat, though - don't assume that a simple interface will > necessarily work with ALL PCMCIA modems - there may be some that are > implemented differently. This should not be a major problem - just be > sure to test any modems you plan to use, and don't let users plug just > any old modem in and expect them to work. > There are also Winmodems available in PCMCIA form. I have an IBM 10BaseT/V90 Modem which I had no end of trouble getting working. I only found out after extensive digging that it was a winmodem. Needless to say, not suitable for use with a PIC! Mike -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! use KILLspamlistservspamBeGonemitvma.mit.edu?body=SET%20PICList%20DIGEST