Lars, an all interested parties I guess not doing a good job of explaining the issues that revolve around current limiting. What I'm trying to say is:
1. The performance of the modem should not be compromised just to satisfy the requirements of the PTT. 2. The physical limitations of a PCMCIA card does not allow the card to dissipate 2 watts. Using components of the size required to efficiently dissipate the heat generated and protect heat sensitive modem components is not possible. A PCMCIA PCB is 3 X 1.85 inches and 0.015" thick and a poor heatsink. The DAA generally takes up 1/3 of the card area on both sides of the board. The card is housed between 2 metal covers. The card is inserted into a connector housing that is mounted inside of a laptop. The connector housing is plastic. The laptop is designed to dissispate the heat it is generating and some of the heat the inserted cards generate. I doubt the laptop designers had in mind for the card to generate 2 Watts. 3. A V.34 modem under the best of conditions will have a connect rate at 28.8 bps of 80 t0 82% <28.8 bps line rate>. To insure the modem only has to deal with impairments present by the PSTN the DAA is designed not to add to impairments. 4. The PCMCIA cards I tested that were limiting the current and dissipating the heat generated by more than 0.8 W exhibited reduced performance and throughput. This was caused by two things: One the elevated temperature of the transformer. Two, the current density a hot Darlington transistor. 5. The cards I tested under limiting current conditions had the the 28.8 connect rate reduced to as low as 42%. As Joe Randolph pointed out in his email, 60mA current limiting under prTBR 21 requires the a PCMCIA card to dissipate 2 watts. Joe also pointed out solid-state SLIC circuits by there nature limit the current for self preservation and conserving battery life. 6. Compliance for protection of the PSTN equipment at the expense of modem performance is a real problem and needs to be addressed. There are certain individuals in the world that design DAAs for compliance only and never consider the performance aspect. I have had to fix few. This was done at the expense of my time and a second round of compliance testing. Not to mention all the time and sales the customer lost, as they watched the market window close. During my 27 years as an engineer in the modem business I've was always tried to design a modem with the best performance as possible with the given specs. I'm still under the impression modem performance is paramount. I hope this ranting was not too long, windy and/or pedantic. Regards, Duane Marcroft Telecom Consultant On Tue, 27 Aug 1996, Jon D Curtis wrote: > From: [email protected] (Lars Poulsen) > Newsgroups: list.treg > Subject: Re: 60 mA current limiting in prTBR 21 > Date: 26 Aug 1996 18:51:09 -0700 > Organization: RNS / Meret Communications > Lines: 22 > Message-Id: <[email protected]> > References: <[email protected]> > > In article <[email protected]> > [email protected] (Duane J Marcroft) writes: > >I'm glad to see someone is ranting about the blatant disregard for the > >physical limitations of a PCMCIA Card DAA design. > > Hello, > > I'm a lowly software engineering manager lurking in this environment, > and while I try to understand what the issues are, I do have some holes > in my technical knowledge. > > If a DAA were to limit the current to 20 mA, would that violate the > French requirement ? (Since if you are at 20 mA, you are still limiting > the current to 60 mA or less.) If someone with the documents can quote > the wording of the French current limit specification, I promise not to > follow up and produce more noise. > > -- > / Lars Poulsen Internet E-mail: [email protected] > RNS / Meret Communications Phone: +1-805-562-3158 > 7402 Hollister Avenue Telefax: +1-805-968-8256 > Santa Barbara, CA 93117 Internets: designed and built while you wait > >
