Saggy Expert brake pedal

[Wookey]

I have a now 23-yr old 1.9TD Expert and I often have to do some brake-fettling to pass an MOT.

The car hasn't been anywhere since March last year, except moved from drive to garden a few times to make room for building works. I checked it over in the summer thinking it was MOT time but decided that the brake pedal really did go a bit too far down to be convincing and I should try and get to the bottom of it. After reading up here (e.g 'Citroen Dispatch Brakes' viewtopic.php?t=38048, and 'Synergie - sinking brake pedal' viewtopic.php?t=30484), I suspected that my master cylinder may be at fault and I should test it. Then it was winter and far too miserable, and I still had nowhere to go, so only now, with some temps in double figures, did I get round to some spannering.

However the results are confusing, so I come to report and see what people think.

The main issue is that the brake pedal is nice and hard with engine off. If I start it, then the pedal sags about 3-4cm as it should, but if you keep pushing it goes further down. Not all the way to the floor, but further than feels 'good'. From this I conclude that my servo is working as it should. I have found no leaks anywhere, and it all got a good fettle last MOT (july 2019ish). Fresh fluid, a new rear cyclinder, a new (well s/h) compensator valve, some new pipe, full caliper service with new sliders and seals on nearside, and the year before that offside caliper service and new front flexpipes at front both sides.

To test the master cylinder I wanted to block off the ports and see how the pedal worked with the rest of the system disconnected. I tried putting rubber bungs under the brake nuts but that didn't work so I made up a couple of blocked-off nuts -i.e a brake nut with 3cm of pipe, flattened and folded over to make a stub. Putting those in the front two ports (and leaving the back two as-is) the pedal sagged much less - an obvious improvement (presumably despite a little bit of air in that stub?). OK, so one of those pipes is the nearside front and the other goes to the compensator. So this seems to say the issue lies in one of those. (The other two ports are much more of a pain to get to and thus block off and test, and there doesn't seem much need now). So I left the stub in the rear compensator port and reconnected the front nearside. Now the pedal seems to have gone back to sinking too far. Even though this is a very short pipe (1m?). I bled it as some air probably got in during these tests, but nothing seemed to change. It was getting late so I may have been a bit slapdash, but what might be going on?

It seems to me that my master cyclinder is in fact OK, given that blocking off two ports removes most of the sag. But if there really is an issue with the front nearside, where might the 'sag' be coming from? Am I missing something? Might my fairly new flex pipe be 'stretching' because it's too cheap and nasty? Something else?

What would you test next?

I guess I should put up some pics and a schematic of the brake layout, as this may not be totally clear. I can do more checks this w/e but I'm a bit confused at this point.

[Wookey]

Hmm. No bright ideas then? I'll dosome more testing next weekend and see if I can make any sense of it.

[CitroJim]

It's normal for the pedal to sag a little as the engine starts and the servo charges... Not by quite such an amount...

I'd have said flexi hoses but as they're new then maybe not... I think stubborn and air in the system and maybe old contaminated fluid...

Give it a really good bleed and flush all the old fluid out. If you've not done a full fluid change it'll certainly be due and could be the cause...

I'm not sure what effect old and contaminated fluid will have but I know from my own experience it can make a difference...

[Gibbo2286]

Has it got ABS? if so secondary bleeding with a diagnostics tool might be necessary.

[Wookey]

I did a full fluid refresh last (2019) MOT and it's done under 500 miles since then. So it shouldn't really need doing again already should it? Seems like a waste of practically-new fluid. I agree this is well worth doing if it's been there a few years and often got a bit brown and dirty. But the stuff I'm bleeding is nice and clean. I know they say '2 years or 25000 miles', but is it really going to collect a lot of water in 2 years of mostly sitting?
Any idea what the volume of fluid in the whole system is? My Russek book doesn't say. My impression is that's something like half a litre, maybe 700ml?

I could run another bottle through just to be sure I guess.

No ABS (too old and basic). The only complication is the rear compensator so there is extra pipe going all the way to the back and then back to the front, increasing system volume somewhat. (edit: this is wrong - the pipes just go to the back, through the compensator and to the wheels)

[CitroJim]

The full fluid change on my AX used around a litre so 700mL is a good estimate...

If it was only changed relatively recently and still looks clear and clean then it should be OK...

As to water absorption, I guess that depends upon where and in what sort of environment it's been stored.

[Gibbo2286]

Drum rear brakes? If so probably just the self adjusters not doing what they should do. There's a bit of info in this post that may help.

If you can pop the drums off and adjust them manually until the drum is a bit tight to get back on you might have a chance.

https://dispatchexpertscudo.org.uk/for ... ?pid=3812

[Wookey]

Yeah, drums on the back, and those adjusters are a pain. That is indeed a good suggestion, except that just reconnecting one front wheel to the system seemed to increase the brake travel. I'll give them a check/fettle anyway.

[Wookey]

OK. I've done some more fiddling in an attempt to make sense of this, and blocked off all the ports with stubs. If I block off the two master cylinder ports at the front (Front L and rear R) then the pedal doesn't sink too far. If I connect either one then it sinks much further. This this tells me something about what's happening inside the cyclinder. If I block off all four ports then I get a rock-hard brake pedal (with the engine on) like a well-fettled older-Citroen. The pedal sags only about 1cm when the servo comes on. I think this means that the master cylinder is _not_ letting fluid past the seals - if it was the pedal would still sag with all the ports blocked as it let fluid into the reservoir, right?

Attaching any of the brake pipes to the system gives me lots of pedal sag back - the more pipes the more sag, even after bleeding. Here are some measurements of pedal-height from floor when pushing hard.
Status Engine off (no vacuum), Engine on (with servo)
4 ports blocked 11cm, 10cm
Forward ports blocked (Front L, and Rear R) - 7cm
Front R wheel only connected 10cm, 7cm
Front R +Front L wheels connected 9cm, 6cm
Both front wheels+Rear L 8cm, 4cm
All pipes connected 7cm 3.5cm

So in every case the pedal went down some as the servo come on, but the more pipe/caliper/cyclinder connected, the soggier it is. I bled about 200ml through each brake, which should be enough to get the bubbles out (using an eezibleed at 12-20psi). The much longer rear pipes don't seem to add more sogginess than the quite short front pipes, which seems a little surprising.

The net result is that the brakes are significantly worse than when I started this. Soggy pedal and I can't lock up the wheels.
Do we agree that if it goes nice and hard with no pipework connected that it's not the master cylinder, and presumably I have air in the rest of the system (or stretchy flexible pipes or something else wrong like cylinders/calipers taking up a lot of movement), right? I bled most of a litre through the system so I'm a bit surprised if there are still bubbles in there, but clearly something isn't right. I'll try again tomorrow

I'll put another litre through, but it that doesn't fix it, and i can't find any leaks, then I'm kind of out of ideas (short of expensively replacing every damn component). Do we have reason to believe that higher-pressure bleeding using the brake pedal (as opposed to the steady push of the eezibleed) is sometimes effective in dislodging bubbles?

[Gibbo2286]

Try using a hose clamp on each of the individual brake hoses to see if you can identify which of the circuits are duff. I think your master cylinder is almost certainly ok.

[Wookey]

Good idea, but I don't have a hose clamp. I guess I can improvise one, but just using my mole grips on the pipe directly could be bad for it? (and it's got a wire spring round it that looks like it might get in the way).
I've just put 300ml of fluid through each wheel and got a couple of small bubbles out of one wheel but that's all I saw.
I've still got a very soft pedal that just keeps sagging if you keep pushing (with the engine on/servo energised) (not quite to the floor, but 'not good'). Does that mean that actually there might be something wrong with the servo? Where exactly is that travel coming from? I still can't lock up the wheels with heavy braking. The front discs get equally warm from trying and it feels balanced L/R. Not much warmth in the rear brakes, but then the van is nearly empty so that might just be the compensator doing its job.
This is all rather frustrating.

One odd thing I did notice. If, whilst pressure-bleeding, you push the brake pedal a couple of times, then (on the rear wheels) the fluid flow reduces to drips rather than a steady stream. Remove the eezibleed pressure and put it back on and the flow resumes. I've never noticed this before. It's as if the negative pressure from the brake pedal rising sucks something back in, largely blocking the flow. This doesn't seem to happen on the front wheels so I guess it might be something to do with the way the compensator valve works? Probably entirely normal, just not something one normally does.

[Wookey]

An update: I bought some pipe clamp tools, but they turned out to be too feeble to clamp brake pipes effectively so that didn't help much. With some heavy braking tests I found I could lock up the front wheels and it seemed adequately balanced so I sent it in to see what the MOT man said. It failed, on 50% imbalance at the front and zero action at the rear. (But everything other than brakes was fine Not a huge surprise.

Nothing at all at the back was disappointing. I took the hubs off, hoping to find seized cylinders, but they were both nice and slidy, so it's not that. Jacking the rear of the van up, with someone foot-braking, I can turn both rear wheels by hand, so that is essentially 'no brakes', (handbrake works fine) but then the compensator valve should be detecting 'negative weight' so maybe that's expected. So propped it up on the axle instead so that the weight of the van is on the axle. Still no braking worth mentioning. But there is some if I push the compensator valve lever hard. So I think the zero-point is set wrong. I don't understand how to set the weird spring arrangement - more on that later.

As a whole new set of discs and pads was £47 with a current sale code from ECP I decided to just replace the front. Did that today and found out why I was getting 50% imbalance on the front. One pad was completely bent! The 'ear' must have caught on the edge of the caliper so it had to bend when the piston operated! Guess that's my fault from last time I fiddled and didn't put it back together carefully enough. oops. Not good. It was working surprisingly well considering. Nice to have a reason at least. Pedal is still soggy though, so the bent pad wasn't the reason for that.


So this is what the compensator looks like:


Not a great image. But I hope you can see that the rod mechanism is loose between the panhard rod bracket and valve lever. No brake pressure will happen until the spring starts to push on the lever. There is a flange on the rod above the bracket which sticks through the bracket (you can see the fat bit) so it is not possible to tighten the bottom nuts up to stop it rattling about in there - that seem to be how it is designed for some reason (hysteresis?) The top nut is plastic and is also tightened fully against the top flange. Maybe the middle flange at the bottom of the spring can be adjusted (but it has no hex)? The Russek book just says 'take it to a specialist' to set the pressures.

So anyone know how to adjust this thing? or is it really too hard to do at home?

And I'm still none the wiser why my pedal keeps sinks too far. I'm suspecting the servo unit as there isn't much left to test. The pedal has been getting 'sinkier' for several years now.

[Gibbo2286]

Some stuff here from another forum.
https://www.diynot.com/diy/media/untitled.93658/

[Wookey]

Aha. that is really useful Gibbo. So the tool is 2 16mm shims. So there should be 16mm between the top nut and the top of the lever and 16mm below the bracket, then tighten the base flange down to there.I think I can do that. Will try on Sunday.

I wonder what the 'reference height' should be? height with an empty van? Height with a full van? Height with a half-full van? Ah I found on another (berlingo) thread that said it should be "wheels on ground but nothing extra in van".

[Wookey]

Just to complete this thread. With a new set of front disc+pads and an adjusted rear compensator, I got an MOT!

Setting the compensator was a bit tricky. I set it originally with a bit of stuff still in the van, and that resulted in removing all the rattle from the spring mechanism and when the pedal was pushed the wheels stopped turning. The garage must have set it wrong for last year's MOT so my rear brakes probably hadn't been doing anything all year unless the van was very loaded. About 18mm of thread remained visible at the bottom.

But then I took everything out and re-did the setting. That leaves the nut very close to the end of the thread - with only about 6mm showing. This seemed uncomfortably little so I wound it in to between those two positions with about 10mm showing. This seemed to do the trick, but I'm still not quite sure what the 'reference height' should be.

The brake pedal still sinks more than I'd like so the original issue remains, but clearly it was deemed 'good enough'.
I've just bought some insurance so can now finally find out if the wheel bearings I fitted 9 months ago cured the mysterious rumble in another thread or not...It's been a long wait.