Electronic Design

What's All This Double-Clutching Stuff, Anyhow?

Download this article in .PDF format
This file type includes high resolution graphics and schematics when applicable.

Recently \\[in 1992\\], a pair of California Condors were released into the wilds of the San Gorgonio wilderness area of Southern California. This was a significant event, because there had been Condors flying around those mountains from the year 60,000,000 B.C. to 1986 A.D. But in 1986, the naturalists convinced the California Fish and Game Dept. that to prevent the Condors from going extinct, they had to capture all of them and put them into protective custody until the breeding stock had reached a viable level.

For six years, these naturalists tried various approaches to bring the number of Condors to a healthy status, and they did finally succeed. One of the important tricks they used was whenever Mama Condor would lay two eggs, one of these naturalists would sneak in through a trapdoor, remove an egg, and put it in an incubator. The Mama would look down and decide to lay another egg, since the number of eggs didn't look like a very large number. (Apparently Condors aren't as smart as crows, which can count up to 5 or 6 fairly consistently.) Anyhow, by this procedure, the Condor families were tricked into raising two clutches of eggs—one that the family would raise, and another that would be hatched and raised by the naturalists, behind the scenes. This trick was called, "Double-clutching."

Now, if you have driven old cars, or trucks, or sports or racing cars, you will be amused at this play on words. That's because double-clutching is also a procedure to save wear and tear on your gearbox, and on your clutch, at the expense of a little extra work for your left leg. What exactly is double-clutching (or, as the British call it, "double de-clutching"), and why is it significant?

Let's say you're accelerating in second gear. When you're ready to shift into third, you decide not to use the standard approach, which is just to tromp on the clutch, take your foot off the gas, shove the shift lever into third, and let the clutch back out. Instead, you opt to double-clutch properly:

First, you take your foot off the gas and kick in the clutch. You shift into neutral, AND let the clutch out. You wait perhaps 0.3 to 1.0 seconds for the engine to slow down from its high revs, depending on how fast you were revving when you started to shift and how much inertia the engine has to slow down. THEN you kick in the clutch and shift into third, and let the clutch out quickly, feeding the gas appropriately. If you have judged it right, when you let the clutch out, there isn't any JERK. And when you shove the lever into third, the gears and engine are at a synchronized speed, so there's minimum wear on the synchronizers, which are the tiny clutches that bring the clutch plate and the gears into smooth synchrony. There's also usually less wear on the clutch plates.

What's the big deal? The main point is that when you try to shift into third gear, the engine has slowed itself and the clutch plate down to the right speed—just about the same speed as the gearbox—so it saves wear on the synchronizers and the clutch. It can also save shock and strain on the whole drive-train, because the speeds are just about synchronized when you let out the clutch.

Well, big deal, you say. Cars haven't needed double-clutching since the synchromesh transmission was popularized 50 years ago. Why bother? Why fool around with anachronistic motions? Isn't it just buying trouble? Even Tom and Ray Magliozzi* claim that double-clutching is silly and stupid and wasteful of energy. Ah, but I can give you reasons why it is beneficial.

First, in most cars, the actual gears are always in constant mesh, and the synchronizers only decide which pair of gears to connect to their shaft. But many trucks and some racing cars are still set up with a non-synchromesh gearbox. With trucks, because they have so many gears, it's noticeably more efficient not to have all of the gears in mesh all of the time. So with the "crash-box," you HAVE TO double-clutch, or you will not be able to shift. The same holds true for racing cars—to gain the last couple percent of efficiency, only one set of gears is in mesh at any time, and you have to actually synchronize their speeds or you can't get it in gear. Despite the obvious drawbacks of having to double-clutch, the gearbox is stronger and more efficient than a comparable synchromesh one, and has less tendency to overheat.

Other reasons for double-clutching: Because it is the right way to operate the clutch. Because it saves wear and tear on your synchronizers in the long run, if you're planning to run your car over 200,000 miles, as I do. Because it is fun to do. Because in very cold weather, (-10° F, for example) you may have to double-clutch to shift gears at all, at least for the first few miles.

One very important reason is that, if your clutch linkage ever fails, you can still shift and get home by double-clutching, getting the engine and gears' speeds synchronized and then just EASING the shifter into the right gear. In the last 1,050,000 miles of driving VWs, I have lost my clutch about 3 times, and each time, with careful planning, I've been able to drive home safely. One time I pulled into the Customs House at Calais, Maine, and discovered my clutch was out. I eased along carefully and managed to get all the way home, 350 miles, to Boston, where it was convenient to put the car in the shop to have the clutch repaired—much more convenient than in the middle of a vacation, or the middle of Maine.

Another reason is that on some old cars, first gear isn't synchromesh, so if you need to shift into first without coming to a full stop, you have to double-clutch. Also, a lot of cars these days are made with weak, chintzy synchros, so they soon wear out, and to drive them gracefully, you need to double-clutch.

Note, when down-shifting, you have to shift into neutral and then blip the throttle momentarily before you shift into the lower gear. It requires practice and a good feel, a good touch, to do it right, especially considering the embarrassing noises you make if you miss your shift into a low gear on a crash-box. For example, you should aim to have the revs just a little high, so if you miss, the engine will soon slow down, and then gears will be at the right speed to mesh and the cogs will slip in....

OKAY, Pease, I'll try this double-clutching some day; but why do you bring up all this stuff in an electronics magazine? Ah, there's an excellent analogy: In most conventional switching regulators, the power transistor turns on while there's lots of voltage across it, and after it turns off, the voltage usually increases to a large voltage. When the transistors turn on, the diodes are already carrying significant current, and the transistors have to turn the diodes off. This is all somewhat stressful, and causes the transistors and diodes to have large turn-on and turn-off surges—pulses of power on every cycle. Of course, diodes and transistors have been designed to withstand these stresses and surges with excellent reliability; we see them all the time.

Still, people have specially designed "resonant mode" switchers to have zero-voltage and zero-current switching. In these regulators, most turn-on and turn-off stresses are eliminated, because the transistor is at a very low voltage when you turn it on, and at a very low current when you're ready to turn it off. Consequently, most voltage and current transients are greatly decreased. Less filtering and shielding is therefore required, enabling the complete regulator to have low Radio Frequency Interference (RFI). Now, to design such a supply takes a more complicated controller IC, more expensive parts, a very careful layout, and a lot of expertise in the electrical design. So while you get some advantages, you have to pay for them.

Now, when you want to build a compact, high-performance, switch-mode regulator at switching frequencies up to about 1 MHz, conventional switchers can do at least as well as resonant ones in terms of cost, size, and performance. But if you need a switcher even smaller and faster than that (most users do not), when the switching frequency rises above 2 MHz, the resonant-mode switchers begin to show real advantages.

At this time, National doesn't make these resonant-mode switchers, so I can't offer you any detailed info about all of their advantages and disadvantages. But I have explained most of their key features. And now you can see why the smooth, stressless turn-on and turn-off of the transistors and diodes in these resonant-mode switchers are analogous to double-clutching your shifts.

Comments invited! / RAP
Robert A. Pease / Engineer

P.S. Even if you double-clutch your shifts most of the time, as I do, do you know when it's a good idea NOT to double-clutch? My primary answer would be either when you're in complicated traffic and you don't want to fool around, or when you're really struggling on an upgrade and a speed-shift prevents you from losing speed. So, as it is with every rule, you should be aware that there are times when the rule doesn't apply. Some day I'm going to write a column about that topic....

*CAR-TALK, on many National Public Radio stations, is a wild mixture of automotive wisdom and entertaining banter about cars that's hosted by Tom and Ray Magliozzi. I tune in nearly every Saturday morning. Ask your local NPR station for their broadcast time—if you like my stuff, you'll probably like theirs.

Originally published in ELECTRONIC DESIGN, August 5, 1993

RAP's 2000 comments: After I published this, I was wondering if I really do double-clutch most of my shifts. One morning I got my shoelaces wet before I left for work. I noticed that every time I shifted, my wet shoelaces slapped my left calf TWO times. Since then I am more observant. About 1% of the time I just speed-shift when I'm in a big hurry. About 10% of the time, I do a "normal" shift. About 40% of the time, I double-clutch. About 48% of the time, I "single-clutch"—I shift OUT of the gear without using the clutch, but then complete the double-clutch. And about 1% of the time, I shift from gear to gear without depressing the clutch at all. In fact, one time, I realized that I had downshifted from 4th to 3rd without using the clutch, without planning that, and without even thinking about what I was going to do. So, just as shifting can become an "automatic" habit, so can double-clutching.

One guy asked me if double-clutching wouldn't wear out the clutch cable twice as fast?! Good question! I think not. Because when the cable is new and well-lubricated, a large number of clutch-depressions does not wear out the cable. Only when it gets old and dry and rusty does it start to wear out rapidly. Besides, if I "single-clutch" as described above, I only depress the clutch once per shift, but I get all the advantages of double-clutching. Right now my '68 Beetle is up at 365k miles. The clutch is showing no problems after 132k miles.—rap

P.S. This column is featured as Appendix M in my new book, How to Drive Into ACCIDENTS...And How NOT To. See www.transtronix.com or e-mail rap at [email protected]

Download this article in .PDF format
This file type includes high resolution graphics and schematics when applicable.
Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.