# How to get rid of or reduce the gap of an equation break?

I am writing a long equation and I have to use the align environment to break it.

Here is the code:

\begin{align*}
A_C\Delta\phi_{2\,C}^\star - \sum_{i\in\mathcal{N}}A_i\Delta\phi_{2\,i}^\star = b_{C} +& A_C^x(\phi_{1\,C}^\star - \phi_{2\,C}^\star) - \\
& A_W(\phi_{1\,W}^\star - \phi_{2\,W}^\star) - \\
& A_E(\phi_{1\,E}^\star - \phi_{2\,E}^\star),
\end{align*}


It looks like this:

The issue comes that there is a "big" jump between the first line and the second line of the previous equation due to the sum index. How to eliminate (or reduce) this gap between lines of a broken equation?

Thanks

• The minus signs at the right are also too close to the preceding material. It's better (more conventional) to place such signs of operation at the beginning of continuation lines. Commented Dec 17, 2019 at 15:57

Simply put the \sum in a \smash[b]{...} command. I took the opportunity to propose a slightly different layout for this equation:

\documentclass{article}

\usepackage{amsmath, amssymb}

\begin{document}

\begin{align*}
A_C\Delta\phi_{2\,C}^\star -\smash[b]{\sum_{i\in\mathcal{N}}}A_i\Delta\phi_{2\,i}^\star = b_{C} +{}& A_C^x(\phi_{1\,C}^\star - \phi_{2\,C}^\star) - \\
& A_W(\phi_{1\,W}^\star - \phi_{2\,W}^\star) - \\
& A_E(\phi_{1\,E}^\star - \phi_{2\,E}^\star),
\end{align*}

\begin{align*}
A_C\Delta\phi_{2\,C}^\star -\smash[b]{\sum_{i\in\mathcal{N}}}A_i\Delta\phi_{2\,i}^\star = b_{C} & +A_C^x(\phi_{1\,C}^\star - \phi_{2\,C}^\star) \\
& -A_W(\phi_{1\,W}^\star - \phi_{2\,W}^\star) \\
&- A_E(\phi_{1\,E}^\star - \phi_{2\,E}^\star),
\end{align*}

\end{document}


• \smash[b]{\sum_{i\in\mathcal{N}}A_i\Delta\phi_{2\,i}^\star} would use the correct spacing after the summation. Commented Dec 17, 2019 at 13:21
• @egreg: Yes, I wrote [b] purposely, as there might be a full line of text above the sum. Commented Dec 17, 2019 at 13:28
• @campa: it seems I misunderstood. Anyway the larger spacing was due to the subscript being wider than the sum symbol, and my code has the same effect as the \smashoperator command from mathtools. Personally, I prefer it like that. Commented Dec 17, 2019 at 14:02
• I does have the same effect, at least in this case, for a reason I don't understand – I used both to test, and the measured difference is less than 0.1pt. Commented Dec 17, 2019 at 15:18
\begin{align*}
A_C\Delta\phi_{2\,C}^\star - \sum_{i\in\mathcal{N}}A_i\Delta\phi_{2\,i}^\star = b_{C} &+ A_C^x(\phi_{1\,C}^\star - \phi_{2\,C}^\star)  \\[-.75pc]
&- A_W(\phi_{1\,W}^\star - \phi_{2\,W}^\star)  \\
&- A_E(\phi_{1\,E}^\star - \phi_{2\,E}^\star),
\end{align*}


Try out this!