It’s certainly been a few decades, but plenty of us remember a time before widespread access to broadband internet, when connections were generally made over phone lines using acoustic modems. In the 90s these could connect you to AOL and Napster well enough, but in the early 80s the speeds were barely enough to read text as it loaded. A company called Hayes set out to change this with some of the first useful, widely-available modems for the PCs at the time. While they couldn’t keep up with the changing times there’s still a retro community that has these antiques, and to modernize it a bit this drop-in replacement for the PCBs replicates these old modems almost exactly.
The new PCB is equipped with everything needed to get a retro computer online again, including all the ports to connect a computer without any further modifications. It houses a few modern upgrades beyond its on-board processors, though. Rather than needing an actual acoustic coupled phone, this one has an ESP32 which gives it wireless capability. But the replacement PCB maintains the look and feel of the original hardware by replicating the red status LEDs at the front, fitting into the original Hayes cases with no modifications needed at all, and even includes a small speaker through which it can replicate the various tones, handshakes, and other audio cues that those of us nostalgic for this new online era remember quite well.
For those looking for a retro feel without the hassle of getting antique networking equipment functional again, this type of upgrade that preserves the essence of the original hardware is an excellent way of keeping retro computers functional on modern networking equipment. But if you absolutely must get the networking equipment exactly right down to the last patch cable, you might end up having to build your own ISP from scratch.
Looks like an ESP8266 to me.
Indeed so. After checking the high resolution images on the Tattler website it shows as “ESP-12F” etched on the module.
So this is the legendary Hayes Smartmodem.
To be honest, I’m a bit irritated about how simple and lightweight the chassis looks.
It’s as if the maker simply slid a modem PCB in a random *industrial chassis.
By contrast, I’ve been used to use a Dr.Neuhaus modem for my BBS once and that one seemed more heavy/sturdy built.
t looked a bit like an US Robotics Courier series model. All black and heavy metal plate chassis.
*Makes me wonder why to bother to dismantle the original here, even.
The upgrade board could be shipping with a frontpanel for a standard chassis, just as well.
Yes, if you plan to hit it with a hammer, otherwise their age and the fact they are still in one piece proves they are sturdy enough…
I didn’t mean to be disrespectful.
It’s just reminding me of hobby projects.
Such basic industrial shells made of aluminium were often being used by small companies and hobbyists.
And for prototyping.
That’s why it’s irritating.
Some TNCs are made same way, also.
This one comes to mind: http://www.symek.com/g/t2hfaq.html
That small but once popular company used a standard chassis, not something custom.
Using a standard chassis isn’t bad, by any means.
I just find it to be an indication for lack of creativity and lack of dedication.
If you’re believing in your product, your creation, you would make it to want stay out of the gray mass.
I would *love* it if more commercial products use a generic chassis! We would never have to buy/make enclosures for our projects again! Imagine if you could repurpose e-waste by just slipping in an updated PCB or slapping a new panel on the front.
I believe they still in use in sole olde security systems. I sold many of them on eBay for under $100. Hayes 2400
*in some older security systems for remote access
WOPR
Greetings Professor Falken
Fun fact: These modems were sized to fit under a Western Electric 500-series desk phone (the standard at the time) without taking up additional desktop real estate.
Another fun fact: There were other devices built in the same style case designed to be stacked along with the modem (and telephone). One that I’m familiar with was the Hayes Chronograph. It was a quartz clock with a VF display in the front, and connected to the computer via RS-232C. Besides being a clock on your desk, it provided time functions to computers before most had their own time-of-day clocks.
I had forgotten about that. I think the Chronograph was the only one that made it to market? And now we have NTP, which would have been an excellent addition to that product…
So this is the same Hayes who invented the AT modem protocol. Which still lives on to this day.
Think I still have one somewhere up in the attic. The last modem I used was a US Robotics 56k. How far we have come…currently on 300mbit/s fiber.
Hi, the true problem or limiting factor wasn’t the modem technology of back then.
Rather contrary, these serial modems were highly complex DSP devices.
The real problem was the poor infrastructure.
The analog landlines were as horrible as shortwave communications.
They were noisy, unreliable and narrow in bandwidth. A few KHz, at best (less than 10).
Today, on shortwave (~3 KHz SSB channels), among best technology possible is PACTOR 4, which uses proprietary compression (thus not allowed on ham bands).
Using most modern technology (that modem/PTC costs about 1800 USD),
it can reach speeds not much better than the average modem of the 90s.
DSL goes another route and uses much higher bandwith on the same old twisted-pair cables.
It’s no longer audio spectrum, as with the
modems of the day.
If 56k modem technology had been allowed to use a similar bandwidth to DSL, they would have done wonders back in the day.
Back in the 1980s I used a Hayes compatible 1200 bps modem to access work. The was an apparently a span somewhere between us joined by a D4 channel bank set for loop timing instead of being slaved to the BSRF timing signal. That caused a frame slip which would cause a series of ~r every 17 seconds to march across the screen when the line was idle in the mark hold condition. The scrambler polynomial to disperse energy is 16 bits long, thus it always causes a glitch in two adjacent characters. Those were the days before auto error correction. At 2400 bps, you’d get a string of _w marching across the screen.
Trying to explain what a frame slip is to customer service representatives who might provoke action to be taken was basically impossible. My solution was to just keep redialing until I would finally randomly hit a route which was clean. I eventually got a 9600 bps modem with error correction and that made life better. You could still see the EC light flash and there would be a short pause like clockwork due to the bad span.
They couldn’t “use a similar bandwidth to DSL” back in the day. The telephone lines would have probably supported it, but the signal processing needed wasn’t cheap enough.
DSL is only possible because of extremely cheap and extremely capable digital signal processors. That didn’texist back in the days of analog modems – and surely not back in the days of the 1200 and 2400 baud modems in this article.
The 1200 and 2400 baud modems did some pretty hefty signal processing for their time, but there was nothing comparable to the processors in a modern DSL modem available to use in those analog modems.
The problem was pure and simple the technology – it just wasn’t there.
Analog telephone lines didn’t just magically get better and allow DSL. The phone lines are the same crummy non-twisted pair wiring that it was way back when. It is just that today we have DSPs that can handle very wide band signals in real time. That allows a DSL modem to handle all the crap on the analog lines.
“They couldn’t “use a similar bandwidth to DSL” back in the day. The telephone lines would have probably supported it, but the signal processing needed wasn’t cheap enough.”
Hi, I’ve meant the 90s, the 56k modem days.
Back in the early 80s, the 300 to 1200 or 2400 Baud days, telephone modes still had used Bell norm or FSK/AFSK like on Packet Radio.
Two audio tones, essentially, with a static analog filters and no error correction.
These was modified RTTY equipment, essentially.
By mid-90s, most modems were being based on DSP technology.
Some used sound card technology, even. Those WinModems come mind..
These modems could be flashed to the newest V-something standard (V.90, V.92 etc).
They were a far cry from original Hayes modem.
Also, by the late 80s, there was ISDN already. A digital telephone line.
It had certain features that DSL+VOIPdidn’t have, for example.
Low latency and fine audio quality, for example.
The control channels could be repurposed for X.25 communication, even.
Sorry for not being more clear about it. 😅
Weren’t WinModems predominantly software based? They relied on a certain amount of main CPU grunt to do the signal processing, hence they were only properly supported in Windows, at least as far as I can remember.
Horrid things.
The sounds of dialup… ahh.. the memories.. I had an acoustic coupler modem at one point. Made me feel like David Lightman.
“Weren’t WinModems predominantly software based? ”
Yes. I talked about two different things.
USB or serial modems had their own on-board computer.
The WinModem thing was about PCI modems (not all) and AMR modems (proprietary slot).
https://en.wikipedia.org/wiki/Softmodem
https://en.wikipedia.org/wiki/Audio/modem_riser
“The problem was pure and simple the technology – it just wasn’t there.”
Oh yeah, that old story again! 😃
Not in consumer segment, maybe.
There had been flight simulators in the 70s or HD CAD/CAM workstations in the 1980s, already.
https://www.youtube.com/watch?v=jLOIoS5sJ4E
https://www.youtube.com/watch?v=9EEY87HAHzk
https://www.youtube.com/watch?v=u7GuwCDq620
Very expensive and bulky, yes, but technically available.
And sometimes superior to todays technology.
Especially those SGI graphics workstations. Very good programming (excellent!) of the software.
Anyway, just saying. I find it funny how people always think that technology “wasn’t there” because they haven’t seen more than a poor C64 at home. 😂
No seriously, history is not always as it seems.
There are so many early technologies that were mature, but hadn’t been recognized by their society.
Example:
https://en.wikipedia.org/wiki/PLATO_(computer_system)
https://www.youtube.com/watch?v=5k_QQV9sj4I
https://en.wikipedia.org/wiki/Tektronix_4050
https://www.youtube.com/watch?v=a94uJjTOjTY
https://en.wikipedia.org/wiki/Xerox_Alto
https://www.youtube.com/watch?v=9H79_kKzmFs
^This is impressive 1970s technology, with high resolution graphics.
Nothing for kids and home computer fans, though.
Another phenomenon is “false belief in progress”.
It’s a concept about progress being a one way street. Ie, everything gets better and better.
People believe in it, in order to give their meaningless life a purpose.
The possession of the smartphone is an example for this.
It gives them a false feeling of superiority.
I don’t share this belief, though. There are moments in time, were earlier inventions were more sophisticated that what comes after.
Let’s just take the pyramids, for example.
Or old cathedrals made centuries ago with “primitive” tools.
Or the concrete mixture the Romans had used.
These buildings or things are being unrivaled yet.
Best regards.
Another example pertinent to the 1990’s would be IBM’s MWave. Somewhere around the mid-90’s I had a card in my system (and was available on Thinkpads, IIR) that was a sound card with wavetable and a high speed modem all in one thanks to the DSP. Prototyped in 1981 on the 5150, on the market in ’91 or ’92 for Sun, OS/2 and Windows.
I had a couple of those back in the day before I moved to the USR’s for speed for the BBS…
DSL relies on the ability to pull the signals out of the analog domain somewhere between the subscriber and the Central Office, typically less than 10k cable feet from the subscriber. Past the DSLAM, it’s all packet switched data.
Analog modems rode a circuit switched audio channel from one endpoint to the other (56k service required a modem connected to an ISDN channel at the far end).
Seems like a missed opportunity for flexible baud rates. The ESP should be able to do other standard bitrates. I like the idea of making old equipment useful again though this only uses the original chassis.
Retromodem 1200 is just the name. It supports baud rates from 300 to at least 115200. The simulated connect sound played is the one from a 1200 baud modem regardless of the serial baud rate. The DTMF digits played are the IP address of the dialed website.
> The DTMF digits played are the IP address of the dialed website.
It’s that sort of attention to detail that makes a product stand out. Very cool!
We still use a Hayes 2400 smartmodem at work for remotely programming old equipment. Can’t get rid of it till those systems go extinct. I did however upgrade the user experience a bit by making an auto installer with a custom configured DosBox environment with two DOS programs so you can program the systems from Win 10/11. COVID downtime at the office was kinda productive a few years ago🤔 plus I’m the only capable person in the company that could have even attempted it.