Depending on the communication program you are using, there are a few things you can do to check your system. Remember, though, that data throughput is the real test, not the connect speed. The last section of this article explains the relationship between connection speed, transmission errors, and throughput.
Apple Remote Access (ARA)
If you are using Apple Remote Access, you can try connecting to the ARA test server that Apple has set up. Compare the connection speed you get when connecting to Apple's ARA test server with that of other ARA servers you connect to.
To perform this test, select the PowerBook 3400 Internal modem. Use a new connection document with these settings:
- Connect As Guest
- For connection to Apple's ARA 3.0 test server, the phone number is 512-908-8039. You can call into this server with either 2.0 and 3.0 ARA client software.
If you can successfully connect to Apple's ARA test server, and consistently get faster connection speeds than with other ARA servers you connect to, then check these things on the ARA servers you connect to:
- what version of Apple Remote Access is running on the server(s).
- what CCL is selected on the server(s).
- what type of modem. (If possible, try another modem on the server.)
Applications With Modifiable Initialization Strings
If you are using an application that allows you to modify the modem initialization string, try using the string AT&F, which resets the modem parameters to their factory default.
Negotiation Error Fallback Algorithm
Another reason why a PowerBook 3400 internal modem might connect at slower speeds than external modems has to do with transmission errors and error correction. When one modem calls another modem, they usually negotiate a connection speed that produces a minimal number of transmission errors. When a transmission error occurs, the modems must re-send the data. The PowerBook 3400 internal modem, however, uses a negotiation error fallback algorithm which negotiates a connection speed that produces no transmission errors. A connection with no transmission errors typically must be made at a slightly slower speed than a connection with some transmission errors.
As mentioned at the beginning of this article, though, actual data throughput is what is most important to you. A connection at a slightly slower speed but with no transmission errors can have the same data throughput as a connection made at a slightly higher speed but with occasional transmission errors.