story by Don Alexander • Illustration by Ruben Cueto
Picture this: At the race track, you’re unloaded and ready to go for the first practice session. After the session is over, you feel lost because your car’s handling is way off. You make changes that should help, but in the next session, the car feels even worse. The same scenario …
I call bullshit on the big changes.
I love making big changes to see if that's where the problem is, and then bracketing in to somewhere in between.
I think having a baseline is one of the best tuning tools there is. When things start to get stupid, and what should have been the correct adjustment doesn't have the effect you thought it would, re-setting to the known basically correct-but-untuned settings can point to the problem. This logic works for a great many things other than suspension.
"Don't try to tune something that binds"
AKA - Don't try to tune an RX-7. The front suspension binds by design and the rear suspension binds by design. That's why the racers back in the day basically made it so stiff that it couldn't move (a suspension that can't move, can't move badly) and adjusted handling with alignment. Or just drove around the problems.
This is why I am infuriated by Techno Toy Tuning or whatever their name is. They may know Toyotas but they certainly don't understand Mazdas. They make a lot of parts for RX-7s that either don't address any real problem, or they actively make the chassis' problems worse. They actually sell Heim-jointed rear suspension links as an upgrade, when non-compliant bushings is exactly the last thing you want in the rear suspension. Absolutely unforgivable!
In reply to Knurled:
I've tried to explain this to people when they attempt to "tune" whatever chassis they have.
Nothing like adding stiffer bushings to a sway bar that is essentially a cross-connected torsion bar. Or worse making this bar larger.
Its also interesting to note how many people relate body roll and responsiveness to "good handling" especially after watching professional race cars bouncing around on super stiff suspension, not realizing that much of that is due to the amount of aerodynamically induced load that is placed on the suspension and that maintaining the car's ride height is very important for under car aerodynamics.
Knurled wrote: "Don't try to tune something that binds" AKA - Don't try to tune an RX-7. The front suspension binds by design and the rear suspension binds by design. That's why the racers back in the day basically made it so stiff that it couldn't move (a suspension that can't move, can't move badly) and adjusted handling with alignment. Or just drove around the problems. This is why I am infuriated by Techno Toy Tuning or whatever their name is. They may know Toyotas but they certainly don't understand Mazdas. They make a lot of parts for RX-7s that either don't address any real problem, or they actively make the chassis' problems worse. They actually sell Heim-jointed rear suspension links as an *upgrade*, when non-compliant bushings is exactly the last thing you want in the rear suspension. Absolutely unforgivable!
Why wouldn't you want accurate, controlled movement in your rear suspension? Or is that sarcasm?
sesto elemento wrote:Knurled wrote: "Don't try to tune something that binds" AKA - Don't try to tune an RX-7. The front suspension binds by design and the rear suspension binds by design. That's why the racers back in the day basically made it so stiff that it couldn't move (a suspension that can't move, can't move badly) and adjusted handling with alignment. Or just drove around the problems. This is why I am infuriated by Techno Toy Tuning or whatever their name is. They may know Toyotas but they certainly don't understand Mazdas. They make a lot of parts for RX-7s that either don't address any real problem, or they actively make the chassis' problems worse. They actually sell Heim-jointed rear suspension links as an *upgrade*, when non-compliant bushings is exactly the last thing you want in the rear suspension. Absolutely unforgivable!Why wouldn't you want accurate, controlled movement in your rear suspension? Or is that sarcasm?
Assuming that this question is serious and not sarcastic:
Yes, accurate and controlled movement is important, when a system is designed to move properly. The problem occurs when a manufacturer makes compromises, usually for packaging or for cost. Many times, the link lengths or the design of the linkage keeps it from moving freely throughout the range of motion that it would move through under normal operation. The solution that is normally used is to install soft bushings, which makes the links "variable length". That's okay for getting groceries, but not good when you start to really put a load on things. People install stiff bushings to help that problem, but now the design issues show up, and you end up with two suspension rates. The first little bit of roll is controlled by the springs and dampers, but then the linkage binds, and the roll stiffness can increase by orders of magnitude. At this point, you are basically driving a different car. If you aren't expecting it (and sometimes if you are), it can cause bad things to happen from a car control standpoint.
When replacing any bushing, etc. with a different material or a heim joint, it's always important to look at how it moves as the suspension cycles. You want to pick the type of joint / bushing that will allow it to move as freely as possible in the directions it needs to move and as little as possible in other directions.
Binding is definitely a common issue with sway bars. The common stick-type sway bar end links with sandwich bushings at each end (shown below) bind horribly on any suspension that isn't very short travel. In the case of my Jeep, the suspension behaved drastically better after removing the binding-prone rear sway bar setup and swapping in stiffer rear springs to re-gain the lost roll stiffness. Overall handling balance was about the same, but it was more predictable and rode better too.
The tough thing for this article it has to deal with generalities;
Newbie to mid level drivers tend to crank the rebound on dampers and use higher spring rates as it gives instant feed back rather than leaving them softer where they build more grip but have less feel. Now if the car has a big splitter and rear wing throw that out the window because maintianing optimal ride heights net more lap time than that what you would gain with suspension compliance.
@mister fister yeah sometimes making big changes will at least get you out of the woods / pointed in the right direction BUT I've raced several cars and bikes that have a narrow range.
The aspect of the article I whole heartedly agree with is write everything down and take as many notes as you can.
sesto elemento wrote:Knurled wrote: "Don't try to tune something that binds" AKA - Don't try to tune an RX-7. The front suspension binds by design and the rear suspension binds by design. That's why the racers back in the day basically made it so stiff that it couldn't move (a suspension that can't move, can't move badly) and adjusted handling with alignment. Or just drove around the problems. This is why I am infuriated by Techno Toy Tuning or whatever their name is. They may know Toyotas but they certainly don't understand Mazdas. They make a lot of parts for RX-7s that either don't address any real problem, or they actively make the chassis' problems worse. They actually sell Heim-jointed rear suspension links as an *upgrade*, when non-compliant bushings is exactly the last thing you want in the rear suspension. Absolutely unforgivable!Why wouldn't you want accurate, controlled movement in your rear suspension? Or is that sarcasm?
If you replaced all of the suspension links with Heims, the suspension will only want to move straight up and down. There are serious geometrical issues in the 4-link and compliant rubber is required if you want the suspension to articulate. This is a large reason why the cars had a habit of ripping the upper mounts off of the chassis and/or twisting the rearend housings so the pinion angle points up, if they were driven hard.
Bored one day with a stock chassis and with the springs out, I unbolted one upper link and articulated the axle. The link was something like 2" too short when drooped and 1.5" too long when compressed, supporting the axle in its center. If you don't have soft rubber bushings in the rear suspension, there will be nothing to allow that misalignment but chassis flex and rear axle twist, and the handling gets... scary. Now that I have a stock chassis again, I may make a YouTube video demonstrating this.
This is exactly my point - the parts that this company sells are not designed with the car's quirks at all, but they prey on the uninformed consumer who sees something shiny and makes sense on the surface (compliance is bad, right?) so they go for it. And that, IMO, is scamming people.
That is only one example, and the most dangerous. Most of the product they sell for the Mazdas is inherently flawed in one way or another. Sadly, all of the "traditional" aftermarket for these cars is long gone since they're too old/rare/outclassed to be viably competitive cars anymore, so all you have left is this hard-parker junk.
Thanks for clarifying, I wasn't familiar with the car in question's suspension design.
In reply to sesto elemento:
You're welcome. I may be a little overenthusiastic about it, but people selling snake oil is something that REALLY gets my goat.
And the chassis does work wonderfully once you go with a 3 link arrangement, the old E/Production method of eliminating bind. Call the 3rd link a "traction bar" (legal mod) and replace the upper links' bushings with roll bar padding (bushing material is free) and suddenly there's no more binding in the suspension and you can TUNE it, instead of stiffening everything up so much it doesn't move...
Knurled wrote:sesto elemento wrote:If you replaced all of the suspension links with Heims, the suspension will only want to move straight up and down. There are serious geometrical issues in the 4-link and compliant rubber is required if you want the suspension to articulate. This is a large reason why the cars had a habit of ripping the upper mounts off of the chassis and/or twisting the rearend housings so the pinion angle points up, if they were driven hard. Bored one day with a stock chassis and with the springs out, I unbolted one upper link and articulated the axle. The link was something like 2" too short when drooped and 1.5" too long when compressed, supporting the axle in its center. If you don't have soft rubber bushings in the rear suspension, there will be nothing to allow that misalignment but chassis flex and rear axle twist, and the handling gets... scary. Now that I have a stock chassis again, I may make a YouTube video demonstrating this. This is exactly my point - the parts that this company sells are not designed with the car's quirks at all, but they prey on the uninformed consumer who sees something shiny and makes sense on the surface (compliance is bad, right?) so they go for it. And that, IMO, is scamming people. That is only one example, and the most dangerous. Most of the product they sell for the Mazdas is inherently flawed in one way or another. Sadly, all of the "traditional" aftermarket for these cars is long gone since they're too old/rare/outclassed to be viably competitive cars anymore, so all you have left is this hard-parker junk.Knurled wrote: "Don't try to tune something that binds" AKA - Don't try to tune an RX-7. The front suspension binds by design and the rear suspension binds by design. That's why the racers back in the day basically made it so stiff that it couldn't move (a suspension that can't move, can't move badly) and adjusted handling with alignment. Or just drove around the problems. This is why I am infuriated by Techno Toy Tuning or whatever their name is. They may know Toyotas but they certainly don't understand Mazdas. They make a lot of parts for RX-7s that either don't address any real problem, or they actively make the chassis' problems worse. They actually sell Heim-jointed rear suspension links as an *upgrade*, when non-compliant bushings is exactly the last thing you want in the rear suspension. Absolutely unforgivable!Why wouldn't you want accurate, controlled movement in your rear suspension? Or is that sarcasm?
The things you say about the RX7 suspension are interesting to me. From what I gather in your posts they are a basically a 4-link on a live axle, is that correct? I've always understood that normal triangulated 4-links don't have any bind at all when using rod-ends and will move freely in roll and bump to the limits of suspension or rod end travel. I have even modeled this on solidworks and the axle moves freely. It is also my understanding that parallel four links like drag cars are guaranteed to bind in roll. It is my understanding that introducing normal bushings like an OEM would are what actually introduce binding, ala the mustang "quadrabind". Although these are my beliefs about four links, I still have a piece of reservation in me that accepts there could be more special cases from link length/angles/etc that turns a four link into a useless bind monster. This is a special interest to me since I would someday like to replace the leaf springs on my MGB with a four link (Satchell) someday in the future.
Do you have a good picture or diagram of how these RX7s are laid out?
All of my experience is in the Ford Fox chassis, with its extremely crappy triangulated 4-link rear suspension. And, while I never actually tried it myself, I'm given to understand that putting rod ends on both ends of all the arms will cause a bind in roll because the plan view rod lengths are not the same. I switched to a 3-link/Watts setup (with all rod ends, except for the rear lower control arm bushings-they're Delrin) to get around that problem. The rear seems to not have roll bind issues.
snailmont5oh wrote: All of my experience is in the Ford Fox chassis, with its extremely crappy triangulated 4-link rear suspension. And, while I never actually tried it myself, I'm given to understand that putting rod ends on both ends of all the arms will cause a bind in roll because the plan view rod lengths are not the same. I switched to a 3-link/Watts setup (with all rod ends, except for the rear lower control arm bushings-they're Delrin) to get around that problem. The rear seems to not have roll bind issues.
See I've always been led to believe that the bind with the rubber bushed 4-link was caused by the fact that rubber bushings really only like to flex in one direction, twisting. While a single rubber bushing on its own can generally accommodate multidirectional motion, an entire collection of rubber bushings as used in a 4-link are all kind of fighting each other as the suspension moves.
Are you saying that it binds because the upper and lower pairs of links are not the same length, or all links are different lengths. Either way it should still work in theory.
Also, what kind of bushings does it use OEM? Are they bonded "metalastic" or free spinning? I can definitely see how someone might confuse suspension bind for bonded bushings winding up.
freetors wrote: Do you have a good picture or diagram of how these RX7s are laid out?
Not offhand, and I don't do the GIS/hotlink game at work. It's a semi-parallel 4 link, the lower links are parallel to each other, the uppers are not parallel to each other in plan view, maybe 2" further out at the rear. Uppers and lowers are parallel at ride height in side view. The uppers are also almost half the length of the lowers.
There is also a Watts link for lateral location, and its mounting is somewhat too high for good handling. (This is why properly engineered lowering springs don't go very low in the rear, and are generally 10-20% softer than stock. Roll stiffness goes up greatly as suspension compresses, the opposite of the front)
Knurled wrote:freetors wrote: Do you have a good picture or diagram of how these RX7s are laid out?Not offhand, and I don't do the GIS/hotlink game at work. It's a semi-parallel 4 link, the lower links are parallel to each other, the uppers are not parallel to each other in plan view, maybe 2" further out at the rear. Uppers and lowers are parallel at ride height in side view. The uppers are also almost half the length of the lowers. There is also a Watts link for lateral location, and its mounting is somewhat too high for good handling. (This is why properly engineered lowering springs don't go very low in the rear, and are generally 10-20% *softer* than stock. Roll stiffness goes up greatly as suspension compresses, the opposite of the front)
It sounds like they intended to have have a triangulated 4-link, ran into packaging constraints, and then gave it some half assed geometry to make it fit. Then they must have figured out the axle was poorly constrained side to side and the forces imposed on the angled links were huge because of not enough angle between them. To fix that they put on a half assed watts link which overconstrained the whole darn thing...
24 seems like a lot of deadly sins. Even fictional biblical types knew to keep that E36 M3 to 10 or less. And 7 of those were total bullshyt.
Imagine I wrote an article for Vogue called "The 2247 Ways You Are berkeleying Wrong". Technically, it could be correct but there is probably a common theme the author missed that could maybe have a root cause, no?
You guys used to be a step above click-bait or, if you weren't - it wasn't retarded. WTF?
Uh, stow your anger there, Huck. The article is from 1999--before clickbait was even invented.
Margie
lol...by '99, we had hardly figured out how to make interweb pr0n work. What we've learned since then will shock you!
freetors wrote: See I've always been led to believe that the bind with the rubber bushed 4-link was caused by the fact that rubber bushings really only like to flex in one direction, twisting. While a single rubber bushing on its own can generally accommodate multidirectional motion, an entire collection of rubber bushings as used in a 4-link are all kind of fighting each other as the suspension moves. Are you saying that it binds because the upper and lower pairs of links are not the same length, or all links are different lengths. Either way it should still work in theory. Also, what kind of bushings does it use OEM? Are they bonded "metalastic" or free spinning? I can definitely see how someone might confuse suspension bind for bonded bushings winding up.
The OEM bushings were bonded.
Here's the sequence of events that took place in my car, a Fairmont:
Original suspension: squatty, floppy. Crap.
Step 1: Install lowering springs. Lower, floppy. Crap.
Step 2: Install "polygraphite" bushings on rear in stock arms (because they don't need lubed all the time like regular polyurethane. Result: Initial turn-in understeer, then snap oversteer.
Step 3: Replace rear lower control arms with arms that have a rod end at the front, and a solid Delrin bushing at the back (supposed to keep the axle side-to-side under control). Result: not as much wheel hop, still snaps loose on cornering.
Step 4: Install adjustable length rear upper control arms with rod ends on the body side and poly bushings on the axle side. Result: oil-canning of the rear seat floor depending on acceleration load or braking! Wicked loose.
Step 5: Remove upper control arms, replace with forward-facing single arm with rod ends. Install Watts link with rod ends. Result: hoses poorly driven C-4 Vettes, keeps up with well driven C-4s, average C-5s and Spec Miatas. On 225-50 tires, exhibits throttle-off oversteer that settles, then plants, when throttle is increased, until the tires spin.
So, the short version of the story is that roll bind isn't bad with soft factory bushings, is horrible the stiffer you make the stock setup, and goes away if you fix the design. I think that the factory bushings deflect in length and twisting.
Pfft, just increase the spring and damping rates in the front, that'll fix everything.
Excess use of the word 'thou' me thinks. But good advice.
snailmont5oh said:Step 3: Replace rear lower control arms with arms that have a rod end at the front, and a solid Delrin bushing at the back (supposed to keep the axle side-to-side under control). Result: not as much wheel hop, still snaps loose on cornering.
Step 4: Install adjustable length rear upper control arms with rod ends on the body side and poly bushings on the axle side. Result: oil-canning of the rear seat floor depending on acceleration load or braking! Wicked loose.
...
So, the short version of the story is that roll bind isn't bad with soft factory bushings, is horrible the stiffer you make the stock setup, and goes away if you fix the design. I think that the factory bushings deflect in length and twisting.
I know it's a thread brought back from the dead for no apparent reason, but this jumped out at me and might be worth noting. Using rod ends on only one end of the control arm, and solid or nearly solid bushings on the other, still inherently binds. To allow full free 3-dimensional articulation, both ends of the arms on a triangulated 4-link (with no additional lateral locating links) need to accommodate multi-axis angular misalignment.
Of course, the uppers being the triangulated pair also generally makes for a rather high roll center, which doesn't help matters any either. This is why the "Satchell link" setup triangulates the lower links instead.
So while you could have gotten bind-free motion with the 4-link by using rod ends on both ends of all four links, the 3-link also allowed you to more readily lower the rear roll center as well...The combination of which apparently worked well together to tame the rear end.
@Driven5:
I feel happy, since you understood exactly the point I was trying to make.
The use of rod ends at both ends of all the original links would have removed the bind, but would have also necessitated the use of a coil-over setup on the rear, since the rear spring perches are on the lower control arms.
You'll need to log in to post.
