DS Anti-roll prototypes
  • Help
Results 1 to 12 of 12
  1. #1
    Fellow Frogger! Andy N's Avatar
    Join Date
    Nov 2004
    Location
    NSW
    Posts
    816

    Default DS Anti-roll prototypes

    The link is to Citroenet which seems to have lots of hidden pages like this one:
    http://www.citroen.mb.ca/cItROeNeT/m...raulics-2.html
    If there were 6 of these DS, are there any still in use or were they destroyed?
    Does anyone know how this was incorporated into the D's suspension without electronics? Amid so many restrictions the DS was still ahead of its time even in 1975, perhaps a car that didn't roll and lurch back then would've been too revolutionary?? I would love to get my hands on one of the technical drawings!!
    Andy

    Advertisement
    Attached Thumbnails Attached Thumbnails DS Anti-roll prototypes-activa-ds.jpg  

  2. #2
    Fellow Frogger! Paul Smith's Avatar
    Join Date
    May 2000
    Location
    Camperdown, NSW, Australia
    Posts
    624

    Icon12

    Andy,

    I believe they used something like a pendulum device mounted on the inside of the firewall - I would assume the system tried to keep it vertical - clever, but even for Citroen engineers probably very fiddly.

    Paul
    Paul Smith

    1974 DS23 EFI BW Auto
    1974 G Special 1220


    http://www.simplicitas.com.au

  3. #3
    Fellow Frogger!
    Join Date
    Sep 2005
    Location
    Scotland
    Posts
    169

    Default

    I don't see any reason why the basic anti-roll system couldn't have been the same as the Xantia Activa, minus the computer. (Naturally) I doubt it would have used a pendulum as this would have been useless on any kind of sloping ground...

    Although the computer in the Activa decides when to stiffen the suspension for cornering in the Activa, (based on inputs such as steering wheel angle and road speed etc) the actual roll is measured and counteracted purely mechanically and hydraulically by a "roll corrector" which is basically a standard height corrector with the delay chamber removed.

    This is connected by means of a differential spring and rod linkage to the front lower control arms, allowing it to measure differential height of the front suspension arms (eg body roll) without responding to vertical height changes.

    The roll corrector is then piped to the rollbar rams - all perfectly doable in the days of the DS.

    My guess is the problem that prevented them from putting it into production (apart from additional expense and complication) is exactly the problem the computer solves - deciding *when* to be trying to compensate for roll.

    Without the computer deciding that the car is cornering, the roll compensator would be trying to operate when the car is driving straight ahead on uneven ground as well - causing the car to rock left and right as the front suspension travels over uneven ground, and the roll corrector instantly "compensates" for any differential movement of the front wheels.

    This causes a big loss in ride comfort and stability in the roll axis on uneven surfaces. In the Activa whenever the car is driving straight ahead (or near enough) the Activa sphere is switched in, which provides a compliance in the hydraulic circuit that the rams are connected to, allowing the wheels to articulate freely even when the roll corrector thinks there is roll that needs correcting. (But which is really just the front travelling over uneven ground - since the activa only measures the roll at the front, for simplicity)

    Without a computer they would have had to at minimum measured the steering wheel angle and road speed hydraulically and calculated whether to enable the anti-roll system based on the hydraulic equivalent of an analog computer - not impossible, (as a similar thing is done in earlier auto gearboxes) but certainly not simple or reliable.

    Another way would have been to use an actual solid state analog computer (which were possible in the 60's and later, as used in the early fuel injection systems) to measure the speed and steering wheel angle and do the neccessary calculations, but now you need electrically operated solenoids - much like the ones in the Hydractive 2/Activa.
    Just imagine if they had managed to get it into production way back then though

    Regards,
    Simon
    Last edited by Mandrake; 28th January 2006 at 08:49 PM.
    1998 Xantia Mk2 V6 Auto Exclusive

  4. #4
    Fellow Frogger! Andy N's Avatar
    Join Date
    Nov 2004
    Location
    NSW
    Posts
    816

    Default

    Thanks Simon and Paul for your responses.

    It must still be top secret engineering for you guys to be speculating about how it worked, I guess there'd be very few people who really know.
    It is interesting that there was also an anti-roll prototype 2cv in 1946, and we all know the 2cv wasn't hydraulic. Could this suggest that the Anti-roll was mechanical wizardry?

    It took me a while to work out how a pendulum mounted to the body would work. This has made me wonder if the lateral shift in weight could be used to maintain a level ride under all forces.

    I have only thought of this now but I can imagine some kind of weights (pendulums) located between four height correctors located at each wheel that move in accordance to lateral forces allowing valves to open in the direction of force and letting fluid into that sphere to compensate for loss of height due to compression from pitch, roll, etc. e.g. The car brakes which causes both weights between front and rear wheels to move forward thus opening the valves in the front height correctors and allowing fluid into the front spheres therefore compensating for spring compression and maintaining correct height. At the same time the fluid is drawn from the rear suspension. When the car accelerates the opposite happens. At the same time and in the same way the weights between left and right wheels keep the car level in both roll situations and slope. e.g The car steers right and the weight moves left forcing fluid into the left spheres to maintain level height. The car goes off the road and the left wheels fall into a shallow ditch which also causes the weight to move to the left and the wheels are forced downwards to level the car. The speed and amount of correction would need to be within specified limits different for each height corrector so I guess this is where it gets difficult. There would have to be a balance between keeping the car level without aggressive height changes i.e accelerating up a steep hill would cause a lot of fluid to be forced into the rear suspension so in this case as well as a downward slope the height correctors should allow as much fluid as needed to maintain only a slightly higher ride height. To alleviate rocking motions and in the previous offroad ditch example there would have to be an allowance for massive height differences between left and right wheels. Could this be how the Anti-roll DS worked or am I thinking of an entirely different system. I have a feeling that there may be something obvious that I've missed which makes this system impossible to be useful.

    Andy

  5. #5
    Fellow Frogger!
    Join Date
    Sep 2005
    Location
    Scotland
    Posts
    169

    Default

    Like I say, I don't think pendulums would have been used, because they would measure the body movement relative to "down" rather than relative to the nominal angle of the road surface.

    So you drive the car on a sideways slope and it tries to compensate for the slope of the road - not good

    Furthermore, measuring the sideways acceleration doesn't give you enough info to accurately correct the roll - because you don't know exactly how much roll you're going to get from a certain amount of G-force as this will vary depending on the condition of the spheres (how well gassed they are) the amount of weight in the car, the centre of gravity of the extra weight, etc.

    So you could be under or over compensating on a system like that which doesn't have any feedback. With the roll corrector on the Activa since it is measuring *actual* roll (at least on relatively flat roads) it is a closed loop feedback system that will will compensate by whatever amount is necessary to cancel the roll.

    The photos of the anti-roll DS show it cornering perfectly flat, and IMHO this could have only been acheived if it actually measured the roll in a feedback system.

    The Activa system is not without its limitations though and there is still room for a lot of improvements.... one of them is the fact that it only measures the roll at the front - so it can't tell the difference between actual body roll, and suspension warp movements where you're simply traveling over uneven ground.

    To do this you'd need to measure the "roll" at the rear as well and compare them - if both ends of the car roll the same way you know it really is body roll and you should be compensating it, but if both ends "roll" in opposite directions you know it is really warp movements and you shouldn't be compensating it.

    I can't think of a simple way to do this using a purely mechanical roll corrector and linkage system as used on the Activa, however it should be dead easy to do on the Hydractive 3+ platform as used on the C5 etc, as it has an electronic height sensor on EACH wheel, so a computer algorithm could easy seperate body roll from warp movements and make the appropriate corrections.

    On the simpler system on the Activa, once you're cornering and the anti-roll system is active, the ride becomes very firm and the problem I mentioned in the previous message of the car rocking in the roll axis on uneven ground would become a problem. There are other ways around this that would allow the ride to be maintained reasonably soft while cornerning but they involve even more complication than the existing Activa setup...

    Regards,
    Simon
    Last edited by Mandrake; 29th January 2006 at 07:53 AM.
    1998 Xantia Mk2 V6 Auto Exclusive

  6. #6
    1000+ Posts cruiserman's Avatar
    Join Date
    Aug 2002
    Location
    Canberra
    Posts
    1,290

    Default

    Quote Originally Posted by Andy N
    It is interesting that there was also an anti-roll prototype 2cv in 1946, and we all know the 2cv wasn't hydraulic. Could this suggest that the Anti-roll was mechanical wizardry?



    Andy
    I think that the important word here is prototype whilst 2CV's arent hydraulic there is nothing to say that the prototype wasnt.
    Neil
    '94 405 SRBT

    '82 CX 2400 Pallas 5spd

    '77 CX 2400 Pallas 4spd manual Import

    '76 CX Super

    '05 WL Statesman

    '92 HZJ80 Landcruiser

  7. #7
    1000+ Posts pottsy's Avatar
    Join Date
    Sep 2002
    Location
    Melbourne, Northern Suburbs
    Posts
    1,136

    Default

    As I understand the section in "L'Album de la DS" (A must-have book for any DS freak, albeit in French) the sensing was accomplished by a pendulum assembly that was centred by springs, presumably to avoid the angle anomaly mentioned above.

    The pendulum was connected to a set of variable resistors which were connected in a type of Wien Bridge formation. This resulted in a a three step signal being produced from the sensor.

    Zero output corresponded with the pendulum being in the central position as in normal driving. A hard cornering force would produce a positive or a negative "error" signal depending on the direction of turn. This error signal was then used to operate solenoid valves to stiffen the suspension on the inside of the turn.

    A simple piece of very intelligent design. Fairly typical of the Citroen brand of inventive thinking too.

    I stress that this is simply my interpretation of the French text accompanying the section, and may differ from the reality, but it makes perfect sense to me. Bear in mind that I last studied the French language in year 11 some 40 odd years ago!

    As to why it was never adopted, I suspect that there may have been issues with the response times of the solenoids, and given that they were switching hydraulics at suspension pressures, there may have also been production cost limitations as well.

    The same sort of timing issues stopped Bendix from further development of their excellent electronic fuel injection system in the 50's. They were using state of the art Mil Spec vacuum tubes, but the average American Joe would not wait for a minute for the electronics in the engine to warm up before starting his car! It took Bosch to adapt the new-fangled transistor before the D-Jetronic system was perfected and practical.

    Cheers, Pottsy.
    Buvito Ergo Sum!

    The Fleet:

    2018 C3 Shine ("Oscar" Mrs P's) Grandma's Taxi
    1974 D Special Manual Sedan ("Moby Dick") Weekend Warrior
    2015 C4 Grand Picasso ("Pablo") Workhorse & Grandpa's Taxi
    1982 2CV6 ("Gaston") Seasoned Traveller
    1975 GS Pallas ("GiSelle") Bonsai CX
    And for Fun, 1968 Mini Deluxe ("The R & D'luxe" 1078cc, Grin Factor 100!)

  8. #8
    Member Sturla's Avatar
    Join Date
    Nov 2005
    Location
    Bergen, Norway
    Posts
    73

    Default

    It appears to be feasible as a retrofit, on http://www.citroensm.net/under interviews is found an, yes, Interview with Lasse Frykholm who has applied parts of the activa system to his SM. Should be equally applicable to do so on a D or a CX. As a starter he fitted the ball valve to prevent the hydraulic fluid from shifting from one side of the car to the other in curves.
    Last edited by Sturla; 30th January 2006 at 08:30 PM.

  9. #9
    Fellow Frogger! Andy N's Avatar
    Join Date
    Nov 2004
    Location
    NSW
    Posts
    816

    Default

    Thanks guys, especially for Pottsy's translation of a French text that sheds more light on the pendulum theory. So I take it that this system was partly electronic (resistors, solenoids and sensors)? I was hoping that it was purely mechanical and hydraulic genius!
    A retrofit to my CX would be great, especially for roundabouts which there are a lot of in Lismore and it takes some nerve to take on a roundabout with gusto in a CX! The body roll is massive to a point where it just holds on against the anti-roll bars with all 4 tyres glued to the road. The tyres can squeel a bit but that may be because I've got the pressure at a low 30 psi running large 195 tyres which look a bit flat under load.....beautiful ride though.
    Andy

  10. #10
    Fellow Frogger! tlampre's Avatar
    Join Date
    Sep 2005
    Location
    Melbourne
    Posts
    167

    Default

    The rest of this page is interesting too. 28 years and 1,130,000 km of testing on a hydrostatic transmission! To quote Thomas Edison: "Genius is 1% inspiration, 99% perspiration." No lack of sweat with these guys.


  11. #11
    Fellow Frogger! chris's Avatar
    Join Date
    Jan 2005
    Location
    Sydney/Marsfield
    Posts
    376

    Default

    Quote Originally Posted by Sturla
    It appears to be feasible as a retrofit, on http://www.citroensm.net/under interviews is found an, yes, Interview with Lasse Frykholm who has applied parts of the activa system to his SM. Should be equally applicable to do so on a D or a CX. As a starter he fitted the ball valve to prevent the hydraulic fluid from shifting from one side of the car to the other in curves.
    That sounds almost too easy... replace the 3-way union in the middle of the suspension circuit with the valve block from a Hydractive car, and you're away laughing. Now, where can I get me hands on a Hydractive valve block

    Chris
    GS 1220 break. Beige cars go faster

  12. #12
    Fellow Frogger!
    Join Date
    Sep 2005
    Location
    Scotland
    Posts
    169

    Default

    Quote Originally Posted by chris
    That sounds almost too easy... replace the 3-way union in the middle of the suspension circuit with the valve block from a Hydractive car, and you're away laughing. Now, where can I get me hands on a Hydractive valve block

    Chris
    It's not quite that easy....

    You'll need to make blanking plugs for the unused ports - you'll need a threaded plug to block the sphere port, another one for the HP input that normally feeds the electrovalve, and you'll probably need to pipe the electrovalve overflow back to the tank as well. (Or block it)

    As well as this you'll have to make some kind of adaptor from the large diameter 8x10 pipes that connect using flared joints to the small 2.6mm pipes going the suspension cylinders.

    Finally, while it should reduce body roll, it will also make the ride firmer in the process, and you'll probably find the car has more tendancy to rock sideways on uneven ground, although I would be interested in hearing back from someone that has tried it, as I also thought of a mod like this myself a few years ago.

    Regards,
    Simon
    1998 Xantia Mk2 V6 Auto Exclusive

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •