We just noticed an issue with a particular analysis where many of the intersections and movements had low volumes. As such, we would expect that an optimization would revert to the minimum splits programmed, and it did.

The issue is how the minimum splits are entered, and adjusted within the program. When we calculate a minimum split to program into a signal controller, we add the minimum initial (or green in some controllers) the yellow and all-red clearance times OR the ped. walk/FDW plus the yellow and all-red clearance times, and use the higher of the two.

We recently ran across a case in Synchro where we first had no ped. phases, and changed the minimum splits to 14, which is the min. initial, yellow and all-red (7,5,2) That's a common setting for us. We then clicked on the ped. phase box located on the phasing window. With a default minimum walk/FDW time of 16 (11,5) it changed the minimum initial to 23. All of this makes sense. But then we determined that no ped. phase was needed, and clicked it off. We would have expected the min. split to revert back to 14, however it stayed at 23. Not noticing that, we then optimized, and those intersections basically "lost" 9 seconds of effective time, since the program was forced to use the 23 instead of 14.

In addition, we had programmed in our usual minimum greens for main street throughs at 12 (it varies for us by speed limit) but then when we globally changed the minimum splits back to 14 (to "correct" the ped. phase change) it did change the min. splits, but also changed the min. initial back to 7, rather than staying with our entered values.

My question/request is this. It seems to me that the minimum split is always a calculated value (see above.) Why is it the item that can be globally changed on the network settings window. Shouldn't it always be calculated by the methods identified above, and if we need to increase it, we should program in a different minimum initial? That way the minimum split is always truly a minimum, and hasn't been "artifically" changed by other things. I would change the network settings page to globally change the minimum initials, rather than the minimum splits. (Yes, we do have different min. initials for thrus and lefts, so having two settings, as there are now, would still be helpful.)

Hi James,

This is a good summary of the Minimum Split value. You've hit it right on. Basically, as noted, the Minimum Split could update (increase) if you increase the Y, AR, W, FDW, or Min Initial. But, as you have noted, it does not reduce if you reduce any of these values.

I understand the desire to have this reduce when you change the values. But, what Synchro does not know is if the Minimum Split has been manually set by the user. For instance, say you have an actuated movement with a min green of 6, Y+AR = 5. The lowest Minimum Split would need to be 11. But, lets say you know that on occasion, there is a spike on some cycles and you want to ensure the split is long enough to accomodate. Then, you'd set the Minimum Split higher than calculated to ensure it never drops below this value. Since it is actuated, it will gap-out if it doesn't need it.

When you are setting your values and you want to calculate the Minimum Split, click in the Minimum Split row and press [F12]. This will set the value to the lowest value.

Thanks,

John.

For a non-coordinated signal, we would change the maximum green in the controller. For coordinated signals, we would do cycle/offset/split which would have the same effect, but that wouldn't change the minimum splits. What you are describing in your "spike" is the "total" split in Synchro, but it should never change the minimum split. (As you said, if the phase didn't need the time that cycle, it would clear and go to the next phase, unless coordination settings override that.)

My point was that the minimum split is always a calculated value, there so I don't see any reason that this value can't be "protected" and not allowed to be changed.

Of course, the engineer entering the data should always check these values, but the reality is that in the rush to finish projects, this doesn't always happen. I see the minimum split to be something that could be automatically calculated, and would be an item that the engineer could assume is always correct (assuming that he/she has checked the minimum initial, clearances and ped timing, which he/she would need to do anyway.)