I am trying to find a 'wave plate' for a 904 style LSD so we can look at re-manufacturing a small quantity .
I understand that this may be hard to find bit maybe, just maybe someone has one kicking around in a box :rolleyes:
Printable View
I am trying to find a 'wave plate' for a 904 style LSD so we can look at re-manufacturing a small quantity .
I understand that this may be hard to find bit maybe, just maybe someone has one kicking around in a box :rolleyes:
Let's say someone happens to have a box full and provides a wave plate, will you be able to supply the reproduced part or will they be like the Nadella parts you have?
I've disassembled dozens of the 904 diffs for gearbox rebuilds and have never seen a pressure wavy plate in those diffs, unlike the next generation which used the wavy plate for static preload. The outer and inner lamina in the 904's were flat and diffs had no preload. Gordon
We already have the 88mm plates from the later diffs but we have some 904 diffs without theses plates and I would like to fit them as they are shown in the Workshop manual.Quote:
Originally Posted by gled49
http://imagizer.imageshack.us/v2/xq90/922/KMy8pu.jpg
I think that the same basic tooling that is used to make the 88mm plate may work but I would like to have a spare early plate to be sure that we do the job properly.
I only plan to make around 6 or 8 plates so it is unlikely we will offer to sell them and I guess demand would be very limited.
Yes, there were 904 diffs with wave plates, but they were not the standard version offered or used in Porsches at the time.Quote:
Originally Posted by chris_seven
The ZF number was 4061 004 024, compared to the standard 004 004 version without wave discs.
In most cases it's no longer practical to rebuild these, since the diffs produced today are more efficient and you don't have to deal with worn bearing flanges, cracked housings etc..
Attachment 435623
Attachment 435624
Jon B.
Vista, CA
Jon,
We have never broken a 904 diff and the replacement plates we have are Moly Coated as were the later plates supplied by ZF in the '88mm' diff.
We do make an alternative diff but we still have a number of people racing with the original ZF and they just don't want to change.
As the Wave Plate is shown in the workshop manual I assumed that it had been fitted by Porsche in some cars although I am sure that it is not too common.
I guess we will have to make them to the same wave height as the later diff.
Chris, it's common for ZF diffs to crack where the machined housings were welded together.
An example of a complete failure, from a 904 transmission, is pictured below. This one should have been noticed well before it reached this stage :-(
It's also common for ZF bearing flanges to wear down, as the bearings begin to spin on the housings. While this usually doesn't result in failure, it will cause the adjustment shim to wear down as well.
I would be concerned too about the internal gears in well-used ZFs, regardless of passing crack tests.
If your customers are somewhat serious about their racing, the diffs produced today have variable ramps to alter the degree of lock in both power and deceleration phases.
They don't require any mods or alterations to the transmission, and will accept both Nadella and Löbro axle flanges.
A 904 ZF housing failure...
Attachment 435626
Jon B.
Vista, CA
The ZF 904 diff. manual show the welded end with bolts securing it. I wonder if adding bolts might prevent the weld from cracking. A well known parts dealer named Chuck told me the 904 lsd was never welded when he considered buying one from me. Oh well. The internal tracks where the plate tabs ride can get worn also. I wonder if the cracked end plate is even noticeable with the bearing preload as the pieces seem to key into each other pretty well.
I routinely replace the wave plates in the gen 2 ZF diffs with the 928 bevel (cone pressure plate) washer. Why don't you make a suitable pressure cone for the 904? Gordon
The ZF manual does show a diff that opens on the S1 (ring gear) side, but all period illustrations and photos of the 904 diff show housings that open on the S2 side.Quote:
Originally Posted by 66S
I don't think the 904 housings can be modified to bolt together where they've been welded together, the pieces involved would need to be designed for that.
The cracked diff in the photo was no longer a precise or stable platform for the ring gear, which was damaged as a consequence.
Chris, the chart I posted shows the release torque values for these diffs with or without wave discs and pre-tension: 1-1,5 mkp without wave discs, and 5-12 mkp with wave discs.
If you do make new wave discs or conical (Belleville) discs as Gordon suggests, you might use these values as a guide.
Jon B.
Vista, CA
http://imagizer.imageshack.us/v2/xq90/923/fZA8t9.jpg
Jon,
I really do appreciate the advice but we already make a diff with 2 x different angles machined into the ramp drivers and can custom manufacture any angles that are wanted.
The diff has 6 plates per side in total and does use a Belville for the preload.
As standard we set them to around 35 lbsft but we can increase this if needed.
The torque we always measure is the sliding value rather then the release value as we find that this is more consistent during bench testing and is more appropriate to the diff's function.
We are trying to design a system that will allow external preload adjustment without copying the system used by Drexler but it is quite tricky. Drexler 901 units currently cost around $3750
Our experience with the 88mm diff that we used in Ford Rally Cars has been that wave plates generally seem to offer more consistent characteristics and hold preload for longer than Belville's and the majority of the 'Atlas' diffs used in Escorts use these plates.
The first diffs used in the Factory Fords all had wave plates and this didn't change until the introduction of the 2.8 litre V6 Capri which seems to have used a diff that had been more aggressively cost engineered.
We are going to try a 904 diff with Belville's but I have a client that wants to use Wave Plates and is quite insistent.
I would much prefer to sell him one of our new units that has a Chromoly Body, no welds and a complete back up service in terms of spares but.... he is the customer :)
Very nice, Chris :-)
Which axle flanges does this unit accept?
I agree with you on the "sliding value", measured with a torque meter and not the type of torque wrench shown in the Porsche manual.
Jon B.
Vista, CA
We normally use a dial indicating torque wrench but I would really like to build an inverter driven test stand and measure just how the diffs bias torque with varying wheel speeds but to cost is just too great compared to the number of diffs we are able to sell.
We supply them to suit Lobro's generally but we can make adapters to suit Nadellas if needed.
We recently refurbished some competition style Nadellas
http://imagizer.imageshack.us/v2/xq90/922/AcUshj.jpg
We also made a small batch of Grade 5 Titanium Hub Nuts with washers made from 17-4PH Stainless Steel hardened to 42HRC and A4 cotter pins.
http://imagizer.imageshack.us/v2/xq90/924/sOk45V.jpg
Can I assume the preload is measured from flange to flange and not body to flange.
Yes you can :)
Gordon, the test method shown in the manuals was to hold either the differential housing or one axle flange in a vise, and then apply torque to a free flange.
Once enough torque is applied to slip the clutches, you'll achieve rotation. The torque needed to achieve or maintain rotation should be within the factory recommended range.
The early manuals illustrate the test using pre-set click type wrenches, but it's more accurate to use a torque meter, or a beam or indicating wrench.
Factory manual test from 1970
Attachment 435751
Factory manual test from 1976
Attachment 435752
A suitable torque meter
Attachment 435753
Jon B.
Vista, CA
.... or one of these
http://imagizer.imageshack.us/v2/xq90/922/mCC1Yz.jpg
Been using torque meter, flange to flange for 30 yrs, just curious. Never used housing to flange method as damage to ring mounting surface seemed inevitable. G