Clutch hydraulic damper / delay valve, considering deletion...

[johnl]

These things are generally known to cause problems with the clutch action, because they create a degree of disconnect between what the pedal does and what the clutch itself actually does. For a long time now I've been 90% convinced that the delay valve has been 'interfering' with with gear changes because the clutch is a bit vague and 'numb' in the way it works. It's harder than it should be to consistently perform clean gear changes, i.e. getting the timing just so.

So, today I pulled the battery tray to have a gander at the suspect problematic valve, to see how it might be deleted, bypassed, or in some way 'sabotaged'.

Firstly, I now want to shoot the person who designed the battery tray attachment, what a stupid way to do it. It's not so much that it's necessary to remove the fuse block from the fuse box, and to also disconnect wires from it to do so, and it's not so much the 10mm nut inside the box carefully hidden beneath a little plastic cover, it's the utterly hidden and not easily accessible 10mm nut beneath the the rear of fuse box itself. What a pain, but I digress...

I was initially hoping it might be possible to disconnect hoses and in some way be lucky enough that they could be reconnected in a manner that bypassed the valve, but of course it's not going to be that easy. One thing is clear, it will be necessary to retain the hard line on the outlet side of the valve (that is connected to the rubber hose that goes to the slave cylinder), because the hose is permanently attached to the hard line, and can't be unscrewed from the rubber hose (there is no threaded fitting, just a clamped fitting).

The hard line is also permanently attached to the delay valve, and can't be unscrewed from it. If one were to ever need to replace the rubber hose, then it looks like the hose, hard line, and valve would all need to be replaced as a unit. The valve can be unscrewed from the hard line that leads to the master cylinder (though it seems to attach to flexible braided hose near the MC rather than directly to the MC itself).

I'm now thinking that the path of least resistance might involve taking off the valve, rubber hose and hard line as a unit. It might then be possible to cut the end of the valve body off, delete its' guts, then weld a plate onto the end of the valve body, leaving an empty cylinder. The empty cylinder should then just fill with fluid, and cause no problems.

I'm thinking it would be best to purchase a used valve (and associated hard line and rubber hose), and modify this instead of risking encountering a problem while modifying the only valve etc that I currently have...

Any thoughts or red flags on this idea?

I have done something similar to this with the clutch delay valve in my old Accord, which did improve the clutch action and gear change precision (though the Accord delay valve was externally quite different to the Alfa one).

Regards,
John.

[Citroënbender]

Why not bypass it with new bubble flares in the existing hardlines and a short bridging length of flared tubing with female flares?

[johnl]

Thanks CB,
Possibly.

This had crossed my mind. The hard line in and the hard line out of the valve are different OD, and I don't know if this would limit the availability of adaptors(?). I'll measure more exactly and make some enquires, when I can find time.

The 'in' hard line will already have a flared fitting, so I'd like to use that if possible. So I'd need to disconnect it to establish exactly what kind of flare and thread is used to ensure it fits an adaptor. The hard line that is attached to the rubber hose would need to be cut as close to the valve as possible, and then be flared to suit a fitting and adaptor.

I doubt it would be neccesary to add any extra hard line to bridge a gap, as I think it should be possible to bend the existing lines to meet each other end on, even if this might mean needing to make a supporting bracket, though I suspect a support isn't required since the valve is currently unsupported other than by the 'in' hard line (though I wouldn't consider this 'best practice'...).

Regards,
John.

[Citroënbender]

It's a clutch, much less a life-or-death instrument than brakes. You could just cut, sleeve and hard solder too.

FWIW anyone with a Sele knows the battery tray routine, it's not that hard and is a good chance to clean up stuff around there from general accumulated grime.

[johnl]

It's a clutch, much less a life-or-death instrument than brakes. You could just cut, sleeve and hard solder too.

A few possibilities.

I had another look at the connections, it's a bit hard to see in poor light and shadow but the hard line going to the valve isn't (as I first thought) attached to a braided hose near the firewall, that's one of the brake lines. The clutch hard line screws into a metal fitting that itself attaches to what appears to be a plastic fitting protruding from the firewall. This then has a fitting that appears to be the same as (or very similar to) the fitting where the rubber hose attaches to the slave cylinder, i.e. a 'push - pull' fitting with a securing wire clip. So, to remove the hard line from the firewall; pull the clip out and then just pull out the hard line (I assume...).

This would be a good thing because there isn't much room to swing even a small spanner between the firewall and the inlet manifold (to undo the threaded fitting on the line itself), let alone get ones' hands in there. So, it seems like removing the hard line between the firewall and valve should probably be relatively quick and easy (where have I heard that before...), as should be taking out the valve and it's undetachable hard and rubber lines. Once all the lines and the valve are out it should be easy to unthread / cut the valve from the lines and then solder the lines together.

I'm not convinced that this would be less work than taking the valve off, cutting, gutting, and resealing it with a welded plate. It is however a viable Plan B if Plan A (eviscerating the valve) doesn't work out for whatever presently unkown / unexpected reason. I kind of like Plan A if for no other reason that it involves less dimantling (i.e. avoids removing the hard line from the firewall), thus less risk of something breaking, or refusing to re-assemble, or re-assembling but then leaking...

FWIW anyone with a Sele knows the battery tray routine, it's not that hard and is a good chance to clean up stuff around there from general accumulated grime.

The battery tray will be a lot easier next time, now that I know exactly where that damned 10mm nut is and how to get at it...

Regards,
John.

[bazzbazz]

The clutch hard line screws into a metal fitting that itself attaches to what appears to be a plastic fitting protruding from the firewall. This then has a fitting that appears to be the same as (or very similar to) the fitting where the rubber hose attaches to the slave cylinder, i.e. a 'push - pull' fitting with a securing wire clip. So, to remove the hard line from the firewall; pull the clip out and then just pull out the hard line (I assume...).

The plastic fitting protruding from the firewall is the back end of the Master Cylinder. And yes, pull clip out and pop off the line fitting, and yes, as usual, there is bugger all room. I suggest to save a lot of swearing when working here to undo upper, lower & front engine mounts and swing the engine forward to give some room.

Saves having to apologize to the neighbors for the language. 

[johnl]

The plastic fitting protruding from the firewall is the back end of the Master Cylinder. And yes, pull clip out and pop off the line fitting, and yes, as usual, there is bugger all room. I suggest to save a lot of swearing when working here to undo upper, lower & front engine mounts and swing the engine forward to give some room.

Saves having to apologize to the neighbors for the language. 

I have no neighbours near enough to disturb with the language and general loss of ones' 'cool' that can occur while battling with the stupidities inflicted upon us by car designers, but the family (including the dog) are a different matter...

It sounds as if my preference to avoid, if possible, disconnecting the hard line at the firewall is well enough founded.

Thinking about these 'delay valves', it has occurred to me that there is a possiblity that they could be implicated with cases of the clutch hydraulics 'sucking' air into the system. My theory is that the valve might in certain circumstances create a momentary low pressure inside the master cylinder that might allow air to enter the MC past a worn / marginal piston seal.

How does such a loss of 'normal' pressure in the MC occur? These delay valves are designed to be free flowing when the fluid is passing from the MC to the slave cylinder, but restrictive when flow is reversed. If the pedal is released quickly enough then the restriction at the valve could be great enough to cause a substantial drop in pressure between the valve and the MC, even if pressure is still fairly high between the valve and the slave cylinder. This could concievably result in an aged MC seal that starts to leak air into the MC well before it might have done if the valve didn't exist...

The MC will have a 'directional' piston seal, i.e. one that is far better at containing a high pressure inside the cylinder than it is at preventing air from entering the cylinder, if the internal pressure were ever to be low enough. This is because, due to the geometry of the seal shape, the seal will expand tightly against the cylinder wall when internal cylinder pressure is significantly higher than atmospheric, but doesn't seal nearly as strongly if the internal pressure is quite low (especially near to zero).

When there is only a small internal pressure, the seal has to largely rely upon the inherant tightness of its' fit inside the bore, and if the seal is old it will have hardened and compressed to some degree, and no longer statically press against the bore wall as strongly as it did when it was more youthful. Nor, due to age hardening, will it so readily expand with internal pressure, so internal pressure now has to be somewhat higher to push the seal lip against the bore as effectively.

If the valve didn't exist, then the MC piston seal would be exposed to the full system pressure for the entirety of its' return stroke, so there will be significant internal pressure acting on the seal nearly all the way to the pedal being completely 'up', assisting to prevent seal leakage by expanding the seal against the bore until the reservoir port is exposed. Of course the pressure would reduce the closer the pedal is to being fully released so that the affect (seal expansion) is relatively small near the end of piston motion, but it is still there until the piston seal passes the port into the reservoir.

On the other hand, if a delay valve is present and if the pedal is released quickly enough, then the internal MC pressure could potentially drop to near zero well before the pedal is fully released (due to the restiction at the valve). This would mean that the piston seal is moving in the bore for some distance without significant internal pressure acting to expand it against the bore wall, so the likelihood of air bypassing the seal would be increased. With a good seal this wouldn't be an issue, but if the seal is compressed and hardened by age, it might be significantly more likely...

Regards,
John.

[bazzbazz]

This may be of some interest -

https://www.competitionsupplies.com/clutches/tilton-clutch-flow-control-valve

[johnl]

This may be of some interest -

https://www.competitionsupplies.com/clutches/tilton-clutch-flow-control-valve

I can't imagine why anyone might actually want to fit one of these things if they are lucky enough to have a car not fitted with one. My experience with them is that they cause problems with gear shifting, albeit relatively subtle ones.

The story, possibly apochryphal, goes that some car makers started to use them as a result of drive train breakages during test drives by mechanically unsympatheitc journalists brutally abusing the clutch, and then writing that the drive train was a weak point on the car...

All they are supposed to do is retard / slow the re-engagement of the clutch, which can cause additional clutch slippage, and presumably wear over time. If the clutch is used sensibly then they fix a problem that doesn't really exist. Before I disabled the delay valve on my Accord, I found that the clutch re-engagement was somewhat erratic, sometimes fine, sometimes there would be unexpected slippage, sometimes it just felt a bit weird. I'd lose confidence that the clutch would do what I expected it to, and get a bit 'lost' in the pedal travel.

Oddly, considering that these valves supposedly don't affect flow from the MC to the slave cylinder (supposedly only the reverse flow), the disengagement point seemed to also change erratically. Sometimes the clutch would disengage with only a short pedal push, sometimes the pedal needed to go all the way to the floor before the gear would cleanly release or cleanly engage.

After disabling the valve this all got a lot better. The clutch behaviour became less erratic and significantly more predictable.

To me the clutch in the 147 feels quite similar to how the clutch in the Accord felt before I gutted the valve, which is why I have some hope that deleting or disabling the Alfas' delay valve might improve it... 

Regards,
John.

[Citroënbender]

Coming soon... Conversion to twin Stromberg CD175s and mechanical advance only dual point distributor?

More seriously, what if the clutch system is just deficient through age, wear or trapped air? The fact you can buy replacement hydraulic dampers for various other marques, suggest it may be a known failure point.

[poohbah]

Coming soon... Conversion to twin Stromberg CD175s and mechanical advance only dual point distributor?

Going slightly off topic (and because I have no technical input to this discussion), I watched a video the other day of Edd China helping some Scandinavian guy restore his Mk 1 Golf GTi, which included replacing the "i" with twin carbs. (Can't recall if Strombergs or Webers). Some people do the strangest things.

[johnl]

Coming soon... Conversion to twin Stromberg CD175s and mechanical advance only dual point distributor?

CB, are you suggesting that these valves are a major advance in automotive technology that we now can't do without? Respectfully I strongly disagree (the word 'bollocks' comes to mind, respectfully of course!).

Strombergs? What are you smoking?

Maybe a brace of DCOE Webers, or even DHLA Dell'Ortos (more or less a DCOE clone). Properly set up you'd get the same power (or at least near to it, though the graph might not be quite so smooth), and, they would sound fantastic! (compared to the muted and uninspiring induction sound of the EFI single TB). Keep the ignition timing as controlled by an ECU though.

With good carbs, the butterflys would do exactly what the driver wanted them to do, not what the ECU decides it will allow them to do. In particular the throttles would close sharply when the driver lifts off, not the lazy / slow closure (causing 'rev hang') that emmissons control requirements force upon us in order to reduce nitrogen oxides...

Call me a Luddite, but throttle response and 'drivability' would be much improved (Webers and Dell'Ortos specialise in throttle responsiveness), at the expense of a possible slight reduction in absolute power, and idle speed control (AC and electrical loads would affect idle, a small price to pay, though you could get around the worst of it with a throttle solenoid activated when the AC is on).

More seriously, what if the clutch system is just deficient through age, wear or trapped air? The fact you can buy replacement hydraulic dampers for various other marques, suggest it may be a known failure point.

The clutch itself works fine, no slippage or other issues other than what seem attributable to the delay valve. It was bled recently.

BMW drivers whinge about these things a lot, deletion is common, you can buy kits to do it.

Regards,
John.

[Citroënbender]

BMW drivers also think the cure for a flat SMG accumulator is replacement of the pump relay. 

I wasn't suggesting wear in the clutch friction but possibly seal wear in the master cylinder, assuming they also have a good shaft seal this might go unnoticed for a while.

[johnl]

BMW drivers also think the cure for a flat SMG accumulator is replacement of the pump relay. 

I'll just pretend that I know what an SMG accumulator actually is (since my car has a manual gearbox and I'm pretty sure doesn't have one, I'm guessing something to do with semi automatic gear selection and the clutch?).

I wasn't suggesting wear in the clutch friction but possibly seal wear in the master cylinder, assuming they also have a good shaft seal this might go unnoticed for a while.

A reasonable thought. Which prompts a speculation re a possible cause of the clutch engaging and disengaging at variable pedal heights:

If the main MC seal were to be marginally effective and as a result 'sucking' air on the pedal upstroke (but not neccesarily actually leaking fluid out, though it might), then considering that the MC is very high up in the system, some or all of this air might self bleed into the reservoir whilst the pedal is fully 'up'(?). Sometimes it might all self bleed out of the cylinder, other times it may not have time to completely self bleed...? As a result there could possibly be a variable amount of air present in the system on each pedal stroke, so clutch engagement and disengagement points in the pedal travel would become somewhat erratic. If such a leak became bad enough then not all air would self bleed regardless of how long the pedal is 'up', and so air might get pumped farther 'down' into the system, and become more permanently entrapped.

Re my earlier speculation about delay valves creating a momentary lesser pressure in the hydraulics between the MC and the valve (than would be the case if no valve were present); this may well make such a problem more likely, and / or more severe. It might be more likely because the  internal MC pressure would be lower for a greater distance in the piston back-stroke, and the longer a lower pressure exists in the MC the more likely it is that air will leak past a dodgy seal into the cylinder, especially while the seal is still moving in the cylinder bore...

Regards,
John.

[johnl]

So I did it today, that is, I disabled the clutch delay valve (CDV). On the face of it, this has worked very well. The clutch is now much nicer to use, seeming to disengage more fully and without subtle 'delays' in engagement and disengagement. Gear shifting is significantly improved with a more predictable clutch action. The pedal now feels noticeably 'springier', which is quite evident especially on the pedal upstroke.

There are of course two things that have occurred, the valve has been disabled so no longer does what it was designed to do (hooray), and the system has been bled as part of re-fitting it. Bleeding may account for what I'm feeling, but it was bled a couple of months ago and this did not have the same affect (a slight improvement in clutch action, but only slight). I suppose time will tell if the clutch action deteriorates from here due to ingress of air...

To disable the valve; firstly I did the tedious preliminary removal of the battery and tray. Then clamped the rubber hose leading from the reservoir to the MC. Then disconnected the single threaded hard line attachment at the valve. Then disconnected the fitting at the slave cylinder (removed wire clip and pulled it off). Now I had the valve and its' permanently attached hard line and rubber hose on the bench.

Next I ground off the swaged flange at the base of the valve body, until the steel disc came out (this disc is what you see when looking at the bottom of the valve, held in place and sealed by the flange, it has a hole in the center). There is an internal plastic cylinder that can now be extracted (this cylinder was held and sealed by the steel disc). Next I welded a steel plate onto the base of the valve body to reseal it. The valves' external cylinder is now empty, and the valve is disabled. Lastly, the valve was re-installed and the system bled (using my home made pressure bleeding thingamy). It was all very straightforward, nothing went wrong it and took about two hours from start to finish.

The CDV seems a very simple device, yet it's principal of operation isn't obvious (at least not to me...). It's just a cylinder with an inlet (from the MC line) at the top / side of the cylinder. The outlet tube (leading to the slave cylinder) projects about 1cm into the top of the cylinder void, with an open end that bears against the end of an internal hard plastic tube. You can see where the open end of the inlet tube presses against the end of the plastic tube, there is a tell tale 'ring' shaped mark in the plastic made by the open end of the inlet tube. The junction between the inlet tube and the plastic tube must be partially sealed because they are pressed against each other (hence the tell tale mark), but I can't imagine that it's a positive seal. The inlet tube must have some holes in it near where it starts to project into the cylinder void (above its' open end), because I could hold my finger over the fully open end of the inlet tube and was still able to blow air into the void (a hole or holes must exist in the inlet tube wall, but are not actually visible in there).

On the disc which can be seen at the end of the valve body, there is a hole which opens onto a void in the internal plastic cylinder, this void consequently being exposed to atmospheric pressure. The valve action must be associated with this void being exposed to atmosphere, and possibly by deformation of the plastic tube due to hydraulic pressure, but I'm not at all sure exactly how...

Regards,
John.