Understanding passive TTL termination

Question

I've been working with 1970s devices and an unterminated system bus. I am adding termination to that but want to first understand a circuit that I have seen to make sure I understand what's going on. This is part of a schematic (the context isn't super important except that it's all 74xx series TTL). The DRFAL and TR0* signals coming in from the left are coming in over a ribbon cable from a connected disk drive mechanism, into a controller board (which is what this schematic is for):

Seeing the 220/330 ohm resistors in that configuration on those lines indicates a termination to me. I note that the signals continue beyond the resistor pair and into other logic.

Elsewhere, in this 1979 Kilobaud article "How Important Is Proper Termination" (regarding system busses), the following quotation and diagram are given:

"The standard TTL termination is a passive 2.6 V reference, composed of a 360 Ohm and 390 Ohm resistor in series across the power supply"

This shows a signal line "ending" at the resistor pair.

My questions:

1. In the first example, is it correct that the intention of those resistors is to provide termination of some kind on the signal coming in from the other device?

2. This seems like a dumb question, but is it the case that despite how they are drawn in this particular schematic (as if the signal continues "beyond" the termination), you could redraw this such that the resistors were at the "end" of the path, too-- thus that doesn't really matter, right?

3. The Kilobaud article says that a 360/390 pair is "standard". The drive controller circuit uses 220/330, for a 3V reference. Was that likely calculated from actual impedance of the line, or is this just another standard?

If my questions imply that I'm really misunderstanding anything or making bad assumptions as well I'm happy to know that.

Thanks!

Answer 1

1. Yes, that seems like a "standard" termination for TTL buses. The 220/330 are very typical values and very often seen. But they are not the only values.

2. Yes, schematics describe the logical connections between components, so they could be drawn anywhere and it would not change the circuit logically. It just may change how much it makes sense to humans, so same circuit can be drawn in a way that is easy to understand or difficult. The physical instace of a ciruit of course has to make sense, so it might be that the signam goes to two chips that are right next to each other, and the termination network is very near those two chips.

3. The word "standard" does not really mean there are actual standards for these - there's just a lot of typical values you see a lot everywhere, and the values depend on the circuitry requirements, environment, or the designer just making a good guess. The actual selection may be based on something we can't see from the schematics. For example, the 220/330 pair has impedance of 132 ohms and does terminate to 3V. The 360/390 pair termintes to 2.6V with 187 ohms. Now, there is less margin, but the driver needs to sink less current, so weaker driver can be used. The signal rises slower back to high level from bus capacitance.

So the selection may be a trade-off. Sometimes you have a fast bus, sometimes a slow bus is enough. Fast buses need more current, slow buses can live with weaker current. Long buses may require better termination than short buses which can just live with good enough termination. It also depends on the cable type, as different cabling will have different impedance and work best when impedance is matched.

Sometimes the resistor values are set by just by having those values in the circuit anyway, as it makes no sense to have a dozen different resistance values on your board, if you only have one or two of them per board. If you can round off and only use half a dozen different resistances the board will be cheaper. At least in the 1970s that might have made sense.