What does ManT1S-Bridge add?

The problem

In the previous guide I explain the basic concept of a T1S network or "mixing segment": the really convenient concept of just being able to daisy-chain a bunch of boards together with a pair of wires to create a network. I did wonder for a while why they introduced a new term for this, but after I received my first ManT1S prototypes in 2024, and I started to play with them, I think I figured it out.

The hard truth is that such a T1S mixing segment, even though it's a TCP/IP "network", is really an island unto itself. In this day and age, when we think of a "network", we think of something that has worldwide reach, something that we can easily talk to from our computers or cell phones, and can make use of convenient network infrastructure we usually take for granted, such as DHCP and DNS.

With that backdrop, working with the T1S network seemed really inconvenient. No DHCP server to hand out IP addresses automatically, unless I decided to run one on a ManT1S. No lookup of other devices by name, unless I decided to run my own DNS server on a ManT1S as well. Probably I'd have to write this software myself and waste resources on a node that should be doing something application related instead. Also, there's no way to reach any Internet resources for simple things like getting the current time or to implement convenient system monitoring. The "network" was just a few boards with fixed IP addresses that could only talk to each other. Without the interconnectivity we have come to expect, it was really no more convenient than something like CAN.

Now I realize that there are perfectly valid use cases for having a networked island once a system is deployed, don't get me wrong. But there are plenty of use cases where connectivity to a "normal network" is either very convenient, or absolutely required. One such important use case is software development. Even though these were networked boards, I still needed physical access with a serial programmer to load code on them. This just seemed very wrong. There was also no easy way to test an API on the ManT1S from my PC with something like curl, or to monitor what was going on in the system.

That's when I started to look for any T1S-to-standard-Ethernet bridges on the market. I couldn't find any. So I decided that this problem could be an opportunity if I could provide this functionality to my customers at a reasonable price, and be able to deliver a complete package for easy T1S development.

The ManT1S-Bridge solution

It was a winding road to get there (that also exposed some gaps in my understanding of networking, which will be the subject of a future update), but eventually I managed to make it work. And once I had a working ManT1S-Bridge prototype, everything about working with the ManT1S changed!

Suddenly, these boards could just use DHCP and get an IP address from my router. Also, by setting a hostname on the ManT1S, the router would receive this hostname during DHCP, while other devices could now use the router's DNS service to look up the device by name. No more hardcoded IPs! Also, all ManT1S boards were now part of my normal network and could be reached from my computer. This opened up possibilities like testing APIs with curl and even remotely programming software to the boards over the network with MicroPython's WebREPL or ESPHome's OTA update capability. Where previously these boards were hiding out on their own little island, suddenly they became part of a much bigger world where getting full visibility into the system was just a few simple API calls away!

Using the bridge

So how do you use the ManT1S-Bridge? You just daisy-chain it to the T1S mixing segment, just like any other T1S board and, on the other side you connect an RJ45 cable that goes to your network. Simple!

In the diagram, the ManT1S-Bridge is connected as an end node, but it doesn't have to be. You can connect it anywhere in the chain, it really does not matter.

And have you noticed that compared to the diagram in the previous guide, I'm not showing extra wires to inject power into the network? That's because when you're using the ManT1S-Bridge, if it's connected to a PoE switch, you can have it inject the 48 V PoE voltage into the T1S network! Up to 30 W is available when connected to a Class 4 capable switch, which immediately solves the power for your system as well. If you don't have PoE, or the 48 V PoE voltage is not suitable for your application, you can disable this and still inject your own power at any node just as shown in the previous update.

Another thing enabled by the ManT1S-Bridge is building heterogeneous distributed systems. If the distances between parts of your ManT1S-based system nodes are farther than the 25 m allowed, or there are more nodes than allowed on one mixing segment, or if certain nodes are better served by being based on a different board like the wESP32 or a Raspberry Pi, you can build the system out of a variety of T1S mixing segments (bridged to Ethernet) and other Ethernet/PoE capable boards, all networked together and powered by a PoE switch. When all nodes speak TCP/IP, your system can be made from very different parts that can all talk to each other and literally span the entire world!

I do not exaggerate when I say that without the ManT1S-Bridge, the ManT1S might not have become a reality. I do not like to release any product that feels half-baked or is irritating to use. And to be honest, before the ManT1S-Bridge existed, that's what developing code for the ManT1S felt like. But because of the ManT1S-Bridge, the ManT1S can really shine!