Data Communications: Market Competition 1969-1972
Modems and Multiplexers
Encounters Unexpected Problems: 1969
In early 1969, Codex used
money from its public offering to fund a promotions campaign for its AE-96
modem. Mindful that if they could sell 50 modems in the coming year that
they would equal all of last year’s sales, they were both stunned and ecstatic
when they received 8,000 inquiries. It seemed as though all they had to
do was scale up manufacturing and begin filling orders. That was until
they had experience with customers using their modem. The AE-96 would not
stay working. Carr remembers:
used to joke that we had made something less than 100 AE-96's and shipped
several hundred of them, because they kept coming back and going out
and coming back and going out again."
With each passing month, Carr increasingly bore the brunt
of Jim Cryer’s frustrations. Cryer, Carr’s boss and President, read the
overwhelming interest in the AE-96 as customers ready to buy and didn’t
want to hear Carr’s excuse of 'but it doesn't work' for not meeting sales
objectives. Since it had been Cryer’s decision, or mistake as was becoming
more apparent, to bet the future of Codex on the AE-96, he leaned on his
management to get results, at least as much as his emotions would let him.
Being an engineer, this was difficult. And when results were not forthcoming,
he felt betrayed and ever more isolated as he watched the ramped up operating
expenses depleting the dwindling cash remaining from the Initial Public
Offering. Crisis invaded every conversation and decision, suffocating any
joy from their work.
Solving the cash crisis meant solving the AE-96 problem.
No one understood the problem, much less the solution. Holsinger was stumped.
Kohlenberg, bereft of ideas, turned in frustration to G. Dave Forney, who
worked exclusively on R&D development contracts, and asked him to drop
what he was doing to help. Forney, who had joined Codex in 1965 at the
suggestion of Gallager, earned his Ph.D. from MIT in information theory
and had virtually no modem experience.
A frustrated Holsinger identified the problem of the
AE-96 as “phase jitter” -- a problem he thought he had solved at DRC. Phase
jitter, while endemic to telephone lines, remained largely unrecognized
for it had no affect on voice communications or the slower speed modems
of the day. Since
telephone lines were not controlled for phase, every circuit could differ
as to being either in phase or not. Phase fluctuations -- hence jitter
-- caused modems that seemed equalized to the circuit’s characteristics
to lose equalization. Since the AE-96s could neither detect nor correct
phase problems, if phase jitter occurred, they ceased working.
Forney quickly, and cleverly, conceived of how to detect
and correct phase jitter errors enough of the time to stabilize the performance
of the AE-96. With Forney’s innovation, AE-96’s once equalized to a circuit
could adapt to subsequent phase jitter. The required electronics forced
the addition of another printed circuit board that had to be hung under
the lid of the modem for lack of space. Named the Threshold Decision Computer
(TDC) for marketing reasons, it pulled out as if from a drawer and had
a red light that flashed every time it corrected an error. Carr recalls:
"Well, it worked. In fact, I remember
going to Air France in Paris, which was the first transatlantic installation
of 9600 bit per second traffic, and they used to run it with the drawer
out all the time because it somehow gave them comfort to see this red light
blinking, and I'd stand there watching the red light blinking and I said
to myself -- this was a machine that I had now observed going out and coming
back because the equalizer wasn't converging, and they were telling me
how happy they were with it, and how wonderful it was, and how much money
it was saving them, and I was standing there, shaking my head and looking
at this light blinking on and off, and saying: "I really don't believe
I'm here." You talk about pioneering days, it was really pretty bizarre."
John Pugh, Director of Product
Marketing since January 1969 when recruited from 3C by his former and now
current boss, Carr, wanted to include the TDC with every AE-96 without
charge, arguing that the modem did not work without it. But Cryer, facing
a survival-threatening cash crisis, insisted it be sold as an extra. So
the TDC, which made for a stable and working AE-96, sold for $2,000 and
shipped with every modem.
Next, working AE-96s made
obvious the problem that commercial customers did not have peripherals
that communicated at 9600 bps. Recalling their days at 3C when they sold
minicomputers to multiplex many incoming communication lines to host computers,
Carr and Pugh turned to the idea of using a multiplexer. If Codex had a
small, inexpensive multiplexer to multiplex and demultiplex eight 1200
bps lines, or four 2400 bps lines, onto one 9600 bps line, then the AE-96
would provide the transmission speeds customers wanted. As if by magic,
a multiplexer of eight or four lines solved another problem thought to
stand in the way of success. The added expense of the multiplexers, as
well as the premium paid for the AE-96s, were recovered easily through
reduced telephone line costs, for in the case of eight-line multiplexers
and AE-96s, only one telephone line would be needed, not eight.
Carr and Pugh contacted
ADS seeking to buy multiplexers they could resell, thus becoming an OEM
to ADS – an original equipment manufacturer (OEM) sells product that others
resell, either with or without attribution. ADS had absolutely no interest
in a OEM arrangement, strained as they were to engineer all the products
in their development queue. Carr and Pugh then began arguing their case
to Cryer and the engineers. Relieved to have at least reason for optimism,
Cryer authorized an internal development effort to develop a multiplexer.
(See Appendix 3. Codex Marketing literature.)
Having persuaded Cryer of
the logic of multiplexers, Carr next pressed his case for a 4800 bps modem.
Only this time his reasoning fell on deaf ears, for building a product
already being sold did not conform to the Codex credo of creating technically
challenging products. But Carr refused to give up; every day his salesmen
complained of having sold customers on upgrading to higher-speed modems,
only to have them buy perceived less risky 4800 bps modems, often from
Milgo, a competitor of growing concern to Carr. Certain he was right, Carr
pressed Cryer at every opportunity: "A,
it will work. B, I can sell it, and it would be nice if we had some money
coming into the place!" Yet as hard as he lobbied and fought
for a 4800, Cryer remained unyielding.
By November, Holsinger no
longer could repress his inner calling to start his own company, and he
resigned from Codex. The man who first proved a 9600 bps modem possible
had seen his knowledge and expertise institutionalized as Codex’s growing
engineering department. Lacking challenges and sensing a boring future
engineering slower speed modems, as being argued for by Carr and Pugh,
Holsinger concluded he could do the same in a company of his own. In 1970,
using the profits from the sale of his Codex stock, Holsinger, the entrepreneur,
joined a long list of start-ups founding Intertel.
transmit data at 9600 bps, data had to be encoded not only using
the amplitude and frequency of the analog signal but also its phase.