4 Link vs 3 Link: How Many Links, and Where?

Part of the Crawler Tools & Guides series. Pairs with DIY Custom Links and Bent Links.

The three-link vs four-link question has been asked on every crawler forum ever built, and it never dies because the answers people get are conclusions with no reasoning attached. "Four-link is better." "Panhard is more scale." "Some comp guys run chassis servos." All partly true, none of it explains anything. The reasoning is one idea, and once you have it, every layout choice in crawling makes sense: a solid axle needs its links to do exactly two jobs, and the layouts are just different ways of splitting up the work.

TL;DR

• Links do two jobs: stop the axle from rotating (axle wrap) and stop it from sliding sideways (lateral location), while leaving it free to travel and articulate.
• A triangulated four-link does both jobs with the same four links. The axle stays centered through travel, articulation is symmetric, nothing extra to buy or tune. That's why it's the default on nearly every modern crawler, front and rear.
• A three-link + Panhard hands the sideways job to a dedicated bar. It works, but the axle swings on the bar's arc, so it shifts sideways a little through travel. And at 1/10 scale the bar is short, which makes that shift proportionally bigger than on a full-size rig.
• The real reason to run a Panhard isn't the links. It's the servo. Chassis-mount the servo and a Panhard matched to the drag link is how you keep bump steer under control.
• Axle-mounted servo + four-link means the whole steering system rides on the axle, so suspension movement can't steer the truck. Zero bump steer, no bar to tune. And a behind-the-axle mount keeps it looking scale.

That's the whole post. The rest is the why, for those of us who want the reasoning behind it.

Two jobs, one axle

A solid axle hanging under a crawler wants to move six different ways. Two of them you want: up-and-down travel, and articulation (one wheel up, one wheel down). Four of them you don't: sliding forward and back, sliding sideways, rotating around its own centerline (axle wrap, the pinion climbing under throttle), and steering itself under the truck.

Your suspension links exist to allow the two good motions and shut down the four bad ones. Fore-aft and wrap get handled by any sane layout; that's what the lower links plus something up top are doing. The job that separates the layouts is the sideways one: lateral location. Every solid-axle suspension ever designed is an answer to "what keeps the axle centered under the truck?"

How a four-link answers it

Look at a crawler four-link from above and the links aren't parallel. They angle inward from the axle toward the skid. That plan-view angle is triangulation, and it's the whole trick. Angled links can't all get longer or shorter at once, so the axle physically cannot slide sideways. The same four links that control wrap also lock the lateral position, and they do it symmetrically: the axle stays centered whether the suspension is compressed, drooped, or crossed up on a ledge.

This is exactly what our Geometry Tool is checking when it looks at your plan angles. Run your links dead parallel top and bottom and the tool throws its bluntest warning: "axle is laterally unlocated (needs triangulation or panhard)." That's not a style complaint. Parallel links in plan view genuinely cannot resist a side load; the axle will walk sideways until something binds. The tool's green band wants at least around 8° of plan angle on the lowers, enough triangulation to take lateral hits without loading the rod ends sideways.

So the four-link's case is short and boring, which is exactly what you want from suspension: both jobs done, no extra parts, no arc geometry to think about, axle centered through every millimeter of travel. Boring won.

(One terminology note before the forum pedants arrive: some rigs run a Y-link, two lowers plus a single wishbone-shaped upper whose apex mounts on the axle. That apex also locates the axle laterally, so a Y-link is its own answer, not a Panhard setup. Axial's own suspension guide notes the Y-link's trade: the axle rotates around that single moving pivot during articulation, which can bring some axle sway with it. Different animal, different post.)

How a three-link + Panhard answers it

Strip the triangulation out, run the links more or less straight front-to-back, and the sideways job is unemployed. The Panhard bar is who you hire: a single link running across the truck, one end on the chassis, one end on the axle. Now the links handle travel and wrap, and the bar handles lateral. Clean division of labor, and it locates the axle firmly.

The catch is geometric. The axle end of the bar swings on an arc, and the radius of that arc is the bar's length. So through suspension travel the axle doesn't move straight up and down relative to the chassis. It also shifts sideways along the arc. A longer bar flattens the arc; a shorter bar curls it tighter and shifts the axle more. Full-size designers fight this by making the bar as long as the chassis allows. And here's the scale problem: a 1/10 crawler's Panhard is maybe 100-150mm long, so for the same proportional suspension travel, the sideways shift eats a bigger slice of the track width than it does on a full-size rig running a meter-long bar. The arc is the price of the bar, and at our scale the price runs higher.

Is that price disqualifying? No. Plenty of good trucks run it, and we'll get to why. But stack the two layouts side by side on pure axle control and the four-link wins every line: centered vs. arcing, symmetric vs. not, four links vs. four links plus a bar. If the links were the whole story, the Panhard would be a footnote.

The links were never the whole story: it's the servo

Here's the part the "which is better" threads usually skip. Nobody chooses a Panhard because they love the bar. They choose it because of where the steering servo lives, and the servo decision drags the suspension layout along behind it.

Mount the servo on the axle and the entire steering system (servo, horn, drag link, tie rod) lives on the axle and moves with it. The suspension can cycle all day and the steering geometry never changes, because nothing in the steering path crosses the gap between chassis and axle. Bump steer isn't small on this layout; it's structurally zero. Bonus: the servo's weight rides on the front axle, and on a crawler chasing 55-60% front bias for climbing, that's weight in a useful spot. Low on the axle beats high on the chassis.

Mount the servo on the chassis and you get real benefits too: the mass sits low and central, the front axle sheds some weight, and the truck looks the part, because a 1:1 truck doesn't have a servo bolted to the diff. But now the drag link has to reach from the chassis down to the steering knuckle. It crosses the gap. As the suspension cycles, the axle moves along the path the links dictate while the drag link pushes the knuckle along its own arc. Any mismatch between those two paths steers the truck without asking you first. That's bump steer, and on a chassis-servo rig with no countermeasures it's baked in.

The countermeasure is the Panhard. Make the bar the same length as the drag link, mount it parallel to it, and the axle's sideways arc matches the drag link's arc. The two swing together and the steering stops fighting the suspension. This is why chassis-mounted servo kits ship with a Panhard bar in the box. The bar isn't the point. The bar is the fix that makes the servo relocation work.

Where does comp land on this? Depends on the class, honestly. Some classes require a scale look, some score scale points, some dont allow,  andsome don't care at all. So  some "comp" guys run chassis servos and Panhards, some SOA/BTA, and what they run is usually dictated by the rulebook as much as the physics.

Now for the part we lived through, so forgive us for having opinions. For years the RTR and kit market only had chassis-mounted servos and Panhard. Short of the bomber/wraith and Gen7 there just wasn't many SOA options and BTA wasn't a thing. 

I didn't like doing panhards. So I started building servo-on-axle and 4-link converters for the Gen7 portal and TRX-4 portal axles, then brought out the behind-the-axle (BTA) mounts giving a bit better scale look and better performance. You no longer had to pick between looking right and steering right.

Five years and about ten thousand orders later, it turns out a lot of other people didn't like doing Panhards either. These days Vanquish, Axial, Redcat and others all make BTA mounts that look, well, familiar. Like we tell our customers: y'all helped change the market.

So the honest comparison is never "three-link vs four-link" by itself. It's axle servo + four-link vs. chassis servo + three-link + Panhard. Two complete packages, each one internally consistent. Pick where the servo lives and the rest follows.

Full-size runs the same play, mirrored

Full-size solid-axle rigs hit the same fork in the road, just from the other side. A conventional steering box bolts to the frame, which makes it the full-size equivalent of a chassis-mounted servo. So the front end nearly always carries a track bar (that's a Panhard by another name) matched to the drag link, for exactly the bump-steer reasons above. The rear has no steering to package, so it gets the triangulated four-link and no bar. And when full-size crawlers switch to full hydraulic steering, with the ram mounted right on the axle, the builders' advice flips to running a four-link up front, same as the rear.

Read that again from an RC seat: our little trucks get to run the layout full-size builders reach for whenever steering packaging allows it, on both ends, because a servo is small enough to bolt to an axle and an engine-driven steering box never was. Being 1/10 scale has its perks.

What the Geometry Tool can and can't tell you here

The tool models a symmetric four-link at each axle, so a four-link rig maps straight in: your plan angles, link separation, roll centers, and anti-squat all come out of your actual mount coordinates, and the lateral-location warning tells you if your triangulation has gone missing. A Panhard rig doesn't map directly. The bar is a single asymmetric link, and the tool's symmetric-pair math doesn't represent it (its roll center comes from the bar height where it crosses the truck's centerline, which is one of the tuning levers the follow-up post digs into). If you're running a Panhard, use the tool for the jobs it does model: CG, weight bias, tip angles, and the fore-aft link geometry. Treat the lateral picture separately.

The decision rule

Running an axle-mounted servo, or don't care where the servo lives? Four-link. Done. Both jobs handled, zero bump steer, nothing to match or tune. There's no crawling performance waiting for you in a Panhard conversion by itself.

Chassis-mounting the servo mainly for looks? A BTA mount gets you the clean front end without the bar, and that's not a hypothetical anymore; every major axle maker offers one now. But if your build or your class genuinely wants the servo on the chassis, then three-link + Panhard is the proven pattern, not a compromise to apologize for. Just go in knowing the two rules: the bar and the drag link must be matched (same length, mounted parallel) or you're installing bump steer on purpose, and the lateral arc is real, so run the longest bar the chassis allows.

What's next

If you're running a Panhard, or this post talked you into one anyway, the bar itself is a tuning instrument: its height sets your roll center, and its angle and length set how much the axle shifts and which way bump steer pushes when the matching isn't perfect. Panhard geometry and how to set it up right is coming up next in this series.

More setup math and calculators live on the Crawler Tools & Guides page, including the Crawler Geometry Tool for link geometry and the Gearing Calculator for drivetrain numbers.