ZF BW-7 gearbox. Clutch not disengaging?

[LittleSister]

It is difficult to pull my boat out of gear, to the extent that you have to pull so hard on the control lever that when it finally releases you tend to overshoot neutral and end up in the opposite direction gear, then have the same problem getting out of that one (all of which makes close quarters manoeuvring rather exciting!).

I think it also tends to graunch somewhat going into gear, even from neutral, but can't remember for sure. (My boat's been out of the water for several years due to an unfortunate series of life 'events', but hopefully back in service soon.)

I spent some time tweaking and replacing control cables, and greasing up the somewhat Heath-Robinson dual-station controls (wheelhouse and cockpit) divider mechanisms, to no avail. But when the engine is turned off moving the control levers, or the lever on the gearbox itself (with cables removed), is smooth and doesn't require much force at all.

It has woefully belatedly occurred to me that the problem may be in the gearbox, rather than the control cables, and that likely the clutch is not disengaging properly. I have never understood quite how marine gearbox clutches are operated (but I am only too familiar with various motorbike clutches). I have read numerous mentions of slipping boat gearbox clutches, but don't recall ever hearing of them not disengaging.

Is failure to disengage likely, and is there any fix, adjustment or mitigation, short of removing the ZF BW-7 gearbox and having it dismantled?

I have changed the gearbox oil, but only after the boat was taken out of the water (and hadn't had it that long before that), so don't know if that has made any difference. (Perhaps it had the wrong lubricant in before?).

 

[tillergirl]

Can you disconnect one station each in turn and see if the box engages sweetly with only one station connected? You might identify a connection issue rather than the clutch. Way back I had a BMC 1500 with a separate gearbox and clutch. The lever had a very Heath Robinson arrangement - as one moved the lever forward a cam disconnected the clutch, further forward engaged the gear and further more the cam was released and drive occurred. When the cam came loose it didn't disengage the clutch and the box didn't like to engage..... not at all.

 

[saxonpirate]

I'm not familiar with your box, but I've stripped and rebuilt quite a few Hurth boxes, and the actual engagement of the clutches for forward and reverse is actually quite a simple mechanism, and normally.. if the sintered bronze discs, wear plates, and clutch return springs are within tolerance.. along with correct oil, they'll engage and drive ok.

If the ZF-BW7 box works on a similar principle to the Hurth.. there's one little monkey that can play tricks and that's the cam on the back of the control lever. If its worn, it can make for very intermittent shifting. It doesn't take a lot, sometimes barely noticeable. That cam has to strike the perfect balance through its travel to engage with the clutch shoe. If its worn, as I said above it can allow you to select a gear, sometimes partially, and then not have the throw to get back to neutral. I would most definitely remove the plate and its control lever before delving into the box. As always.. simple stuff first..

 

[LittleSister]

thanks for the suggestions. Keep 'em coming, please.

Can you disconnect one station each in turn and see if the box engages sweetly with only one station connected? You might identify a connection issue rather than the clutch.

As I suggested above, I am fairly sure that I have eliminated the control cables as being the culprit. The problem occurs with both control stations, and the box 'engages sweetly' when the engine is not rotating, even just operating the lever on the gearbox by hand with the cable disconnected..

I'm not familiar with your box, but I've stripped and rebuilt quite a few Hurth boxes, and the actual engagement of the clutches for forward and reverse is actually quite a simple mechanism, and normally.. if the sintered bronze discs, wear plates, and clutch return springs are within tolerance.. along with correct oil, they'll engage and drive ok.

If the ZF-BW7 box works on a similar principle to the Hurth.. there's one little monkey that can play tricks and that's the cam on the back of the control lever. If its worn, it can make for very intermittent shifting. It doesn't take a lot, sometimes barely noticeable. That cam has to strike the perfect balance through its travel to engage with the clutch shoe. If its worn, as I said above it can allow you to select a gear, sometimes partially, and then not have the throw to get back to neutral. I would most definitely remove the plate and its control lever before delving into the box. As always.. simple stuff first..

I don't have a problem with intermittent shifting, nor with the clutch once in gear.. It always goes into and out of the relevant gear, just with significant difficulty getting it out of gear when the engine is running.


Update/further info:
The problem is the transition into and out of gear when the engine is on (i.e. rotating). It take a lot of pressure (far too much) to pull it out out gear (both forward and reverse), and though it doesn't require much pressure to put it into gear, IIRC it also graunches going into gear. There is no problem at all once in gear, nor when in neutral, nor even going into and out of gear when the engines is not running.

Last night I made a detailed study of the parts list/diagram and workshop manual, trying to deduce exactly how the clutch and gearbox works.

There are actually two clutches - one for forward, one for reverse. (It seems unlikely that both would be equally afflicted by deterioration of e.g. the plates..)

As far as I can make out (that's not very far!), putting the boat into gear is merely pushing the sliding collar a short distance in the relevant direction, which in turn pushes the relevant clutch pressure plate, which engages the relevant clutch, which is then connecting the relevant gear (pre-engaged and previously just spinning on its needle bearing) to the output shaft. For some reason that movement is impeded when the engine is turning, but unimpeded when it is stationary.

I cannot fathom out a clear mental picture of the how the 'roller support' moves and rotates, or doesn't, in relation to the sliding collar or output shaft - (i.e. how the rotation of the driven end of the clutch (the pressure plate) is transferred to the output shaft, and whether the 'roller support' slides back and forth with the sliding collar.

I fear that unless there is a miraculous effect from a change of oil and a bit more use, I am likely looking at a gearbox removal/overhaul/replacement, and all the financial pain and the practical aggravation that would involve.


Homework for those so inclined! -

Parts List (with exploded diagrams)
Pages 36 to 39 cover the BW-7 gearbox.
https://bukh.co.uk/wp-content/uploads/sites/8/2016/09/Spare-part-list-BUKH-DV-36-ME-BW7.pdf

Workshop Manual with photos
Section R (pages R1 to R33), towards the end of this Manual, covers the BW-7 gearbox. It has some clear photos, but no description of how it operates, and brief, enigmatic text on disassembly/assembly presumably translated from another language (and calls the parts different names from those used in the Parts List!).
https://bukh.dk/upload_dir/docs/FAQ...p manual/Work Shop Manual - BUKH DV 20 ME.pdf

 

[penfold]

I get the impression these are quite an old design, a google throws up similar forum postings suggesting parts are not available. They're a mechanical gearbox rather than hydraulic, so the actuating mechanism is likely to be a cam of some sort; if this has worn or been damaged it may cause the symptoms described.

 

[LittleSister]

It is quite an old design (and, indeed, an old gearbox!), but quite robust.

The gear change is cam operated, but there's nothing wrong with it being so, and I don't think the cam is the problem because it moves into and out of gear fine when the engine is not running, and performs fine in either gear.

It seems to me likely that the problem is that it is reluctant to disengage because the gears (or some other parts?) are still loaded against one another (because e.g. the clutches are not disengaging). Alternatively, there is something else impeding the return movement of the selector and clutch pressure plates when trying to disengage if the motor is running.

 

[ean_p]

thanks for the suggestions. Keep 'em coming, please.


As I suggested above, I am fairly sure that I have eliminated the control cables as being the culprit. The problem occurs with both control stations, and the box 'engages sweetly' when the engine is not rotating, even just operating the lever on the gearbox by hand with the cable disconnected..



I don't have a problem with intermittent shifting, nor with the clutch once in gear.. It always goes into and out of the relevant gear, just with significant difficulty getting it out of gear when the engine is running.


Update/further info:
The problem is the transition into and out of gear when the engine is on (i.e. rotating). It take a lot of pressure (far too much) to pull it out out gear (both forward and reverse), and though it doesn't require much pressure to put it into gear, IIRC it also graunches going into gear. There is no problem at all once in gear, nor when in neutral, nor even going into and out of gear when the engines is not running.

Last night I made a detailed study of the parts list/diagram and workshop manual, trying to deduce exactly how the clutch and gearbox works.

There are actually two clutches - one for forward, one for reverse. (It seems unlikely that both would be equally afflicted by deterioration of e.g. the plates..)

As far as I can make out (that's not very far!), putting the boat into gear is merely pushing the sliding collar a short distance in the relevant direction, which in turn pushes the relevant clutch pressure plate, which engages the relevant clutch, which is then connecting the relevant gear (pre-engaged and previously just spinning on its needle bearing) to the output shaft. For some reason that movement is impeded when the engine is turning, but unimpeded when it is stationary.

I cannot fathom out a clear mental picture of the how the 'roller support' moves and rotates, or doesn't, in relation to the sliding collar or output shaft - (i.e. how the rotation of the driven end of the clutch (the pressure plate) is transferred to the output shaft, and whether the 'roller support' slides back and forth with the sliding collar.

I fear that unless there is a miraculous effect from a change of oil and a bit more use, I am likely looking at a gearbox removal/overhaul/replacement, and all the financial pain and the practical aggravation that would involve.


Homework for those so inclined! -

Parts List (with exploded diagrams)
Pages 36 to 39 cover the BW-7 gearbox.
https://bukh.co.uk/wp-content/uploads/sites/8/2016/09/Spare-part-list-BUKH-DV-36-ME-BW7.pdf

Workshop Manual with photos
Section R (pages R1 to R33), towards the end of this Manual, covers the BW-7 gearbox. It has some clear photos, but no description of how it operates, and brief, enigmatic text on disassembly/assembly presumably translated from another language (and calls the parts different names from those used in the Parts List!).
https://bukh.dk/upload_dir/docs/FAQ/Bukh ME - Data/Bukh ME - 10-20 ME/Bukh 20 ME - Manuals/Works shop manual/Work Shop Manual - BUKH DV 20 ME.pdf

I think ( though could well be very wrong!) that the trick you are missing is that the 'roller support' is 'fixed' to the output shaft and that the drive to the output shaft is via the the three shafts /pegs that the tension rollers rotate around. These pegs are driven by the sliding collar that rotates then the same as the output shaft and 'within' the shift fork. There seems to be two models ( chain or intermediary gear models) and the rotation of the sliding collar is 'restrained' by the shift fork in different ways as indeed is the movement of the shifting fork! But going on with the drive, the pegs can slide axially in slots within the sliding collar so that when the shift fork is moved then the rotating sliding collar slides across these pegs either to the left or right of the roller support. In doing so it brings one or other of the pressure rings into contact with the tension rollers. Because these rollers restrict the axial movement of the pressure rings they then exert pressure on to the clutch plate pack which then stop 'sliding' and 'lock' up moving the drive from the gear spline to the plate tabs.
The direction of this 'drive is of course reversed and starts at the gear spline, through the clutch plates to the plate tabs then into the sliding collar and then into the pegs of the roller support which is the output shaft. Well I think thats how it is......
Now as to what might make disengagement harder....... sounds like one of the several sliding operations in and around the sliding collar and as it goes both ways then I would look at a common which would start with the 'wear' limits between the shifting fork and the sliding collar. After that I might look at the movement of the shaft that carries the shifting fork and after that it's possible that the issue is within the sliding collar, maybe the roller support pegs or the détente pins sticking when sideloaded. Beyond all this as a first step, for the few quid it would cost and ease of action, I would be inclined to drain the existing oil and refill with a known quantity of the OEM's recommended oil twice, once as a flush and second as the fill. Good luck

 

[LittleSister]

I think ( though could well be very wrong!) that the trick you are missing is that the 'roller support' is 'fixed' to the output shaft and that the drive to the output shaft is via the the three shafts /pegs that the tension rollers rotate around. These pegs are driven by the sliding collar that rotates then the same as the output shaft and 'within' the shift fork. There seems to be two models ( chain or intermediary gear models) and the rotation of the sliding collar is 'restrained' by the shift fork in different ways as indeed is the movement of the shifting fork! But going on with the drive, the pegs can slide axially in slots within the sliding collar so that when the shift fork is moved then the rotating sliding collar slides across these pegs either to the left or right of the roller support. In doing so it brings one or other of the pressure rings into contact with the tension rollers. Because these rollers restrict the axial movement of the pressure rings they then exert pressure on to the clutch plate pack which then stop 'sliding' and 'lock' up moving the drive from the gear spline to the plate tabs.
The direction of this 'drive is of course reversed and starts at the gear spline, through the clutch plates to the plate tabs then into the sliding collar and then into the pegs of the roller support which is the output shaft. Well I think thats how it is......
Now as to what might make disengagement harder....... sounds like one of the several sliding operations in and around the sliding collar and as it goes both ways then I would look at a common which would start with the 'wear' limits between the shifting fork and the sliding collar. After that I might look at the movement of the shaft that carries the shifting fork and after that it's possible that the issue is within the sliding collar, maybe the roller support pegs or the détente pins sticking when sideloaded. Beyond all this as a first step, for the few quid it would cost and ease of action, I would be inclined to drain the existing oil and refill with a known quantity of the OEM's recommended oil twice, once as a flush and second as the fill. Good luck

Thank you very much ean_p, that is very helpful indeed. As I said, I couldn't work out from the drawings quite how the sliding collar/roller support/clutch pressure plates/shaft worked together.

I'm still not completely clear, so could you please just further clarify what you say about that?

You say 'the rotating sliding collar slides across these pegs either to the left or right of the roller support. In doing so it brings one or other of the pressure rings into contact with the tension rollers'. I can't envisage how the rollers, which are fixed on the roller support, in turn fixed to the output shaft, can come into direct contact with one or other of the clutch pressure rings which are, in neutral or in the other gear, a small distance away from the rollers. (I am assuming the gears themselves do not move axially.)

I can imagine a situation in which a sliding ring itself could, when slid from the neutral position, engage with the clutch pressure ring by means of the curious projections inside of the collar, and transfer the drive (as the plates engage) from pressure ring, to sliding collar, to the pegs of the roller support, and hence to the shaft, but that would leave the rollers redundant, and the roller support pegs and slots they sit in in the sliding collar don't look suitable to take those sorts of loads.

Quite where the 'locking pin' on the pressure plate/thrust collar, and the sprung check pins in the roller support come into the equation, also remains a mystery to me.

Working out what is amiss.
I somehow had overlooked that the sliding collar would be rotating in the shifting sleeve/fork. That would seem very vulnerable to wear without some sort of intervening bearing. Workshop manual specifically says check the wear is within the specified limits, but as it goes in and out of gear fine when the engine is off, and engages properly in both gears when running, I am doubtful whether that is likely to be the problem.

I am more inclined to think there might be some wear or damage creating a notch in some part such as e.g. the slots in the sliding collar, which doesn't affect it when there is no load, but when there is a circumferential (?) load because the engine is on, it catches and makes axial movement difficult. Alternatively, perhaps the tension rollers are jammed (and perhaps worn eccentric as a result?), or there is something amiss with the clutch pressure plate locking pin or roller support check pins (whatever it is these parts do).

What to do about it.
I have already changed the gearbox oil. Changing oil again to flush is a good idea, and I'll do that. The old oil I took out looked clean, but I don't know if it was the correct type (which is just engine oil). Not sure whether if previous owner put in, say, ATF in error would make clutch plates plates less likely to slip.

Unfortunately boat is ashore at the moment, so I can't check the operation of the gearbox under load until it goes back in the water. I'll be able to run the engine soon, but will have no load on the propellor.

I increasingly suspect I'm needing a gearbox replacement or rebuild.

 

[ean_p]

Thank you very much ean_p, that is very helpful indeed. As I said, I couldn't work out from the drawings quite how the sliding collar/roller support/clutch pressure plates/shaft worked together.

I'm still not completely clear, so could you please just further clarify what you say about that?

You say 'the rotating sliding collar slides across these pegs either to the left or right of the roller support. In doing so it brings one or other of the pressure rings into contact with the tension rollers'. I can't envisage how the rollers, which are fixed on the roller support, in turn fixed to the output shaft, can come into direct contact with one or other of the clutch pressure rings which are, in neutral or in the other gear, a small distance away from the rollers. (I am assuming the gears themselves do not move axially.)

I can imagine a situation in which a sliding ring itself could, when slid from the neutral position, engage with the clutch pressure ring by means of the curious projections inside of the collar, and transfer the drive (as the plates engage) from pressure ring, to sliding collar, to the pegs of the roller support, and hence to the shaft, but that would leave the rollers redundant, and the roller support pegs and slots they sit in in the sliding collar don't look suitable to take those sorts of loads.

Quite where the 'locking pin' on the pressure plate/thrust collar, and the sprung check pins in the roller support come into the equation, also remains a mystery to me.

Working out what is amiss.
I somehow had overlooked that the sliding collar would be rotating in the shifting sleeve/fork. That would seem very vulnerable to wear without some sort of intervening bearing. Workshop manual specifically says check the wear is within the specified limits, but as it goes in and out of gear fine when the engine is off, and engages properly in both gears when running, I am doubtful whether that is likely to be the problem.

I am more inclined to think there might be some wear or damage creating a notch in some part such as e.g. the slots in the sliding collar, which doesn't affect it when there is no load, but when there is a circumferential (?) load because the engine is on, it catches and makes axial movement difficult. Alternatively, perhaps the tension rollers are jammed (and perhaps worn eccentric as a result?), or there is something amiss with the clutch pressure plate locking pin or roller support check pins (whatever it is these parts do).

What to do about it.
I have already changed the gearbox oil. Changing oil again to flush is a good idea, and I'll do that. The old oil I took out looked clean, but I don't know if it was the correct type (which is just engine oil). Not sure whether if previous owner put in, say, ATF in error would make clutch plates plates less likely to slip.

Unfortunately boat is ashore at the moment, so I can't check the operation of the gearbox under load until it goes back in the water. I'll be able to run the engine soon, but will have no load on the propellor.

I increasingly suspect I'm needing a gearbox replacement or rebuild.

Hi LS
I'm less familiar with this box than you yourself are and am 'assuming' the function from looking at the drawings etc that you provided. As above 'I think ( though could well be very wrong!)'. So take the above and the following as guestimation and maybe use it to inform your own thoughts.....
But to answer some of your questions....... the check pins ( spring loaded) are détente pins that position the sliding collar relative to the roller support by dropping into 'location' slots within the collar having been moved by the shift fork. Presumably there are three of them representing F, N & A.
The 'locking pins' appear to be conical (why??) studs that engage with slots also within the sliding collar and thereby center the pressure rings to the sliding collar radially but still allow limited movement axially.
So it then appears that when the sliding collar is moved say into astern ( to that position at which the detente's move from the neutral position to the 'astern' position then the pressure plate is moved closer to the roller support and the rollers which then put increased pressure on the pressure plate which in turn compresses the clutch plates together. The rollers are 'located / engaging ' in three corresponding concave depressions on the inner face of the pressure plate. I assume that because the pressure plate must still have limited radial movement then as the depressions try and move against the rollers ( effectively the rollers try to climb out of the depressions) the effect is to increase the axial thrust on the plate away from the roller support and towards the clutch pack and hence on the clutch pack, locking up that selected gear in drive. having compressed the clutch pack then I'd tend to think that they would also assist in the transmission.
My thoughts on the changing of the oil was in case that which was in was an incorrect spec that might have impacts on any and all sliding surfaces.
As for mechanical causes....I was tending to think that wear could cause something to stick when it was 'displaced' / cockled over slightly by having the drive through it or perhaps one of the détente pins is sticky....I would tend to think against a bruise (a notch) in one of the slots as that would tend to be specific to one direction or the other !

 

[LittleSister]

Hi LS
I'm less familiar with this box than you yourself are and am 'assuming' the function from looking at the drawings etc that you provided. . .

So it then appears that when the sliding collar is moved say into astern ( to that position at which the detente's move from the neutral position to the 'astern' position then the pressure plate is moved closer to the roller supportand the rollers which then put increased pressure on the pressure plate which in turn compresses the clutch plates together. . . .

I'm not familiar with any more than the outside of the box! I've previously dismantled clutches and engines, but never been forced to deal with the insides of a gearbox, so only have a theoretical idea of how they work.

I don't think, though, that the pressure plate can be moved closer to the roller support, (unless the gearwheel is also being moved, which I don't think is the case) as that would reduce, rather than increase the pressure on the clutch plates.

Anyway, I have become convinced through this thread that there is something amiss in my gearbox shifting mechanism that is beyond my abilities to rectify, and have bought a second-hand gearbox to replace it with. (Fingers crossed that is a good 'un.)

Thanks to everyone who has contributed for their suggestions and for engaging in the puzzle.

 

[pandos]

I must confess I have not read all of the posts so apologies if this is already considered.

Have you checked that the engine is dropping back to a proper tickover as you pass through neutral a friend had a similar issue in the past and after a rebuild and a new box his problem remained...it was eventually identified that the throttle cable was creating an issue..

 

[LittleSister]

I must confess I have not read all of the posts so apologies if this is already considered.

Have you checked that the engine is dropping back to a proper tickover as you pass through neutral a friend had a similar issue in the past and after a rebuild and a new box his problem remained...it was eventually identified that the throttle cable was creating an issue..

Thanks for the suggestion. I'm pretty sure the ticker is OK, but I will check.