Engine upgrades

All items are for the 400/450, they may also be for the other models.

To confirm differences check out your bike's sector at the bottom.

Light blue for 470/501, light yellow for 550/650.

Also see Year recognition and changes and Common faults to incorperate any chnges into the rebuild.

Husaberg 400/450 Models

Main bearings

There have been problems with the mains on mostly the 550 and 650 machines, it matters not whether the oil is changed regularly, whether it's for SM - when they go, they go!

Firstly, in Lineaweaver's opinion - and it's one we trust here at UHE - you must have the cases checked using an Arbor (see one below) and bored if need be for alignment, the housings themselves specifically on the right - hand bearing (Lineaweaver) are often undersized by several microns making either type of bearing a force fit causing problems in themselves. The best approach for meticulous preparation would appear to be to use the arbor method (Lineaweaver) and use the extra material in order to machine a straight line.

Secondly, get the axial tolerance right. Around 0.3 - 0.4mm for 2 x ball, 0.5 - 0.6mm for a ball and roller mix, 0.6 - 0.8mm for a double roller crank (Enginehardwear of Sweden).

During '04 -'5 it would appear that engines were built with a lack of end float at the crank. Most of the problems will have worked their way through the system by now.

It would further appear that the crankshaft is slightly inbalanced in the way it is held by the two mains. The one on the right is against the flywheel cheek while the one on the left sits some 12mm away - the counter-balancer bearing sits between the casing and the crank. The crankshaft is 'adequete' but slight oscilation in an unsymmetrical way occurs. The roller bearings, when fitted can be an improvement but it has been found that sidefloat tolerance is critical. The rollers can catch a ringlip and start chaffing the edge, the chippings cause spawling and the rest - as they say - is history!

And thirdly, the bearing should only be a slight interference fit into the housing. The bearing should be 62mm dead to -10 microns, while the housing should be 62mm +2 to +11 microns for a good fit. While the crank journal should be 30mm +2 to 11 microns and the ID of the bearing being 30mm and nil to -10 microns.

Try to heat the cases in the oven whilst chilling the bearings before having them pressed in evenly. Without the balancer the crank should be rebalanced to a factor of 70% plus or minus 10% (Lineaweaver).

Ball v roller

Husaberg's come with SKF ball-race bearings as standard up until '05 when they changed to a SKF roller mains. Some of the roller mains were re-verted to ball race in an effort to cure the failures and just to confuse the issue: some earlier bikes have been converted to roller mains!

Left roller SKF, right OEM ball race. (Bergoroadracer)	Note the widths 20mm and 16mm (Bergoroadracer)	16mm ball vs 20mm roller. Note there are 13 rollers in cage not 12. (Weed)

Ball race

Standard 6206 mains (with 8 ball) can be replaced with the SKF 6206 ETN9 ball main which is the 9 ball bearing - it definetly has a higher load rating than NTN 6206 or SKF 6206. It has 600n higher static load rating (12900n), and 3100n (23400n) higher dynamic loading (JBS racing/Ben Ballard/Ausberg). Visit Ben's site here: http://www.blog.jbsracingpage.co.uk/ He really pushes the boat out.... The suffix ETN9 is not a reference to 9 balls. The E suffix stands for reinforced ball set and the TN9 stands for a type of reinforced polyamide cage and the idea is that if the bearing collapses the polymide won't destroy the rest of the engine but break up. The ball-race bearings can be evil to place and replace (Orangeberg).

Swedish SM team Enginehardwear have been using another mainbearing on the left side an NTN bearing: SKF TMB206/C3 JR2 - a ballbearing and still use the NJ206ECP/C3 (roller) on the right side. [b]After further tests their latest cure is: KTM Id no: left side, ballbearing from 610cc crate motor Part no. 06250.60206 (with KTM Id no: Right side, Roller bearing OEM part no. 800.30.019.000 )[/b]

The breakdown of the codes are thus: TMB is transmission steel which is case carborised, 206 means maxifill (as many balls as can be fitted in race, allowing more load but lower speed rating), C3 is the internal clearance (gap between the rolling elements and the inner and outer race (which allows for thermal expansion, C4= more expansion), JR2 denotes a steel cage (Wildman).

An FAG chart: Ball mains 19.3kn 6206 (Weed)

Lineaweaver arbor tool

Bearing notes:

The NJ206 series bearings have a rpm maximum of 11,000 rpm but only if they are in an oil bath mounted in the horizontal position. When in the vertical position the maximum for this bearing is 20% less, or 8,800 rpm to be exact, but only if the bearing is at least 30% submerged in oil. (NSman) Ball-race bearings have a max load factor of 19KN while the NJ206 roller bearings have a load rating of 42KN and should easily be able to handle the crank loads.

Roller type

This is Lineaweaver's prefered way of doing it. This became the OEM set-up from '05 onwards and is retro fitable to previous Husabergs of this engine design. KTM part no: Right side, Roller bearing OEM # 800.30.019.000 (with KTM Id no: left side, ballbearing from 610cc crate motor. part no: 06250.60206) will work fine, and keep same axial clearence as you had when running 2pc roller bearing. We have had varied quality on the SKF NJ206ECP/C3, so we now use these KTM OEM ones, they also have a harder surface on the inner race. We have now 18hrs on +73 RWHP engines (53 RWHP std), with this configuration and they're still running.(Enginehardwear)

Patent by Husaberg R&D Racing for safety with crankshaft bearings. (Bergoroadracer)

FAG chart of 16mm roller 45kn 206 (Weed)

20mm instead of 16mm roller

The twin-row counter-balancer should be removed and a wider (20mm roller instead of 16mm) roller bearing fitted with the appropriate spacer made to suit. This is a modification being tried by several race teams throughout the world. A spacer is generally made to suit the resulting gap OR a single row counter-balancer has been used with the later-day gear teeth design and again a small washer/spacer.

For Ben Ballard's JBS racing engine, Ben removed the balancer and instead took the chance to beef up the drive side bearing from 16mm width 44kN OEM Husaberg to 20mm width 55kN rating.

Crank seal

Crankshaft oil seal, metal vs. OEM rubber. (Bergoroadracer)

Conrod bearing

Left 05, right 04. (Bergoroadracer)

Top end issue

Cams
First an overall picture must be produced and the story is told as simply as can be. We'll start with the cams fitted from '00 up to '03 inclusive and these were called '01; '53; '55; or '08. The numbers can be seen right on the tip of the shaft at the chaindrive end. Now these were incredibly torquey but relied on late opening and early closing with viscious lift ramps. The cams create massive torque so much so that 4th gear can be used when anyone else would use 2nd. Alas the above is the cost. It causes heavy acceleration peaks in the valvegear.

Valve springs
To stop the valves launching into the piston but to make them come back, a very, very strong single conical spring was used. The springs were so strong that the violence of the cam whipping back and forth at especially low (and any other) speeds gave the camchain a torrid time jolting it about. Unrelated; the springs have a finite life and break so they're worth replacing with dual spring sets which are far softer and do the same job.

Broken valve spring on the left and guide seal wrecked

Camchain tensioner blade
The early yellow blades prove inadequate. So the later thick but straight black blade should be used. The chains needed but didn't get very good support.

Latest example (left), a crooked early yellow and the std yellow which was very weak (Weed)

Looks ordinary but adjuster has full adjustment. Pressure point has moved down after being 'modified' by Weed

Camchain
The camchain couldn't cope with the violent pressure of the cam whipping back and forth and the lack of support from the tensioner blade.

Camchain tensioner - and it's tensioner spring
With the black blades being so thick and strong a shorter tensioner is required and the simple way is to snap the alloy knob off the end. See below...and the tensioner spring may be shortened to put less pressure through to the blade. 50mm down to 30mm.

2 x tensioners

Cam follower bearings
The accelerated lift speeds of the viscious cam gave the cam follower bearings - CFB - that follow the lobes a torrid time as well and these would wear the tiny needle rollers inside and induce large tappet clearances.

Valves
The valves, specifically the inlets, also found themselves, launched off the top of the cam - 'bouncing' is the term - and despite the extra strong springs would, smack back into the seats. This causes the valves to try to sink into the ports by beating the valve's seating faces into a new shape known as 'tuliping'. If the valves 'tulip' back into the head then the tappet clearances disappear/close up and despite regular adjustments starting becomes poor.

Not so flat intake valve seats (Weed)

Cam follower bearings - 2
So if the valves sink and the clearances disappear we have a new problem! A CFB gains heat as it lifts and lowers the rocker arm under extreme pressure but when it hits the base circle of the cam we would normally have looseness in the tappets, oil would wash over and reduce the working temperature. So hot, cool down and repeat. Alas with no clearance the bearing breaks down due to heat and lack of oil - it never got a rest and never cools down! The Inlet CFB would go due to tuliping valves and the exhaust due to poor maintenance and the most heat to deal with.

CONCLUSION
If you change the cam to a post '03, the valves get an easier time, so the follower bearings get an easier time and with the aid of a later CTB so does the chain! There are now softer dual valve spring sets available that match that softer cam etc and make life even easier still for the valvetrain. All this means that if you do all this, you needn't upgrade those early rocker arms (8mm pins) so you save money!

The comments for each part below are taken in isolation and for use with the original cam and springs etc etc.

Camchain

This is for bikes up to '06. Wears very quickly. Aluminium shavings in oil: Probably the chain wearing against the water outlet in the head.

In the photo, the shiny chain is the standard camchain ('02) and the dull one is an early husey H/D. Two seasons of racing and the shiny one is absolutely shagged! Note how the pins are like 3mm out of synchro! The bikes came (from '06) with D.I.D as standard. Then other chains started to be found.

New chain v an old. Measured from the beginning of the boss on a link to the boss at the beginning of the 31st link. 64 link chain for a 400. This should however help anyone with a 501/550 chain as well as the 650.

New chain = 569mm

Old chain = 573mm.

Virtually the full width of the boss out in wear!

Camchain tensioner

In '01 and '02 the limited travel, alloy knob wears easily. Here's a Lineaweaver answer: Change to stainless steel and lengthen by +4mm to 15.5mm total length. Within weeks the original camchain tensioner can become fully extended, the alloy knob then frays badly causing alloy deposits in oil. Just a tensioned steel rod from '05 onwards with no knob on end, this shorter rod only has to push a thicker (see diagram below) stirrup balancing the dimensions out. So upgrade Camchain tensioner by removing the knob from the end. That's right! simply remove it and you have an '05 tensioner! to go with the better profiled stirrup from '04 onwards. Same design as KTM now.

Cam slipper (tensioner)

An upgrade for 2000-02. Timing Chain slipper (tension arm) 800.36.002.000. Stepped Bushing (requires modification) 800.36.002.050. In the diagrams below are the early yellow and then the banana black blades, both of which are now superceded.

Required modification. See this UHE Thread. (LeFrog)

So yes, I had to use a vinyl washer, drill it to fit the piece of bushing that ran through, and it fits very nicely. (LeFrog)	There is about 1 mm left when the new assembly is inserted. (LeFrog)	Yes, the new end bushing has to be machined badly. Very crude Dremel job, but guess what... this is not a part that can be seen, can it? (LeFrog)

The '03 is more arched up toward the upper timing sprocket. It uses all the same hardware as the old stirrup, but is keeping the chain far from the waterjacket. The steel pushbar mates perfectly.

It was upgraded in '05 for a better supporting pivot screw and nut.

When fitting the later blade, the shoulder can stay on but a deeper bush is required. The alternative is to file the shoulder off and put the shoulder of the bush against the head. There is then only enough thread on the other end for the nut (but no washer). However, given the work/task that the nut has to do it will be perfectly ok! The blade is now positioned prefectly for the chain run.

If fitting to a 400 or 470 the later blades are long enough to clip the top of a gear so remove 3mm from the tip (Husabutt).

Camwheel

The '01s were made of alloy, and if made in aluminium it’s trash... Replace with the hard steel. '02's were fitted with either steel or aluminium, '03+ = steel. If changing to steel shorten the bolts or they go through the thinner camwheel and impact the head (44104)

Lighter valves, springs, collets and caps

Dual straight valve springs are used from a KTM450SX. Kits are available from Kibblewhite's in the USA. code of the kit is 96 - 96031.

Reducing valve rates via the camshaft will allow 37 mm SS Intakes to see 10,500 RPM with a low pressure dual spring kit. Unless you plan on 10,000 RPM plus and using an OE camshaft Ti is not necessary. (Lineaweaver)

The seat pressure is the same and the peak pressure is 1/3 of OEM. This will increase the rocker arm cam follower bearing and cam lobe life (Lineaweaver). The springs should only be compressed to a certain height, this is known as the 'Installed/nominal spring height'. The installed height for the kit is 29.65 - 29.9mm. To measure this: fit a valve in the guide, a cap with collets and measure the gap between the spring's seat and the underside of the spring cap. With the use of the shims and a lathe you will get the correct installed height.

But....

The inside of the camcover on the '01-03s need the rotary file putting over a small area because the spring caps touch slightly (Belgiquebastard).

RD racing head: titanium valves, aluminium retainers, double springs. (Bergoroadracer)	Rocker cover before rotary (Belgiquebastard)	Rocker cover after rotary (Belgiquebastard)

Rocker-arm roller bearings and pins

This is for all '01 and '02 machines with rocker part#: Intake: 200 281-01 and exhaust: 200 283-01. An upgrade is VITAL or continual replacement every two years of average use (as a precation) and still using your ogiginal parts. The factory modification was to reduce the pin size from 8.0mm to 7.6mm for '03+ and this used a smaller needle rollers. This means that all '99 - '02 rockers can't be updated as the rocker arm holes are of course too big! They must either have (more of the same old, same old) new follower bearings fitted or a complete replacement rocker arm assembly to the later rocker arm with it's 7.6mm bearings. You can buy the complete arm or save your old tappet screw and locknut to save a little.

Rocker arms complete from the '03 fit '99 - '02 models as a direct upgrade. Part#: Intake 200 330-01 and Exhaust 200 331-02. Bearings fitted by the factory look like a pin inside a hollow dowel from the side (explode the below left photo to see it more clearly), for repeated procedure by mechanics etc the view is of a simple pin with punch marks etc to keep the pins from drifting during use (explode the photo below to see it more clearly).

For 2004, all Husaberg heads came with the smaller valves - 35/30 (as opposed to 37/30 in the 470+)and now share rocker arms with the later (see ++ below for old larger arms) model KTMs. Part#: Intake 590.36.060.144 and Exhaust 590.36.061.144. As mentioned in 'tuning', the inlet valves on all 470cc+ engines were 37mm and set wider. The 400 had the follower problems but the rocker arm size of today's bikes is still OK for it as it ran the 35/27 head.

Follower bearings need an end float of between 0.25 - 0.30mm. This can disappear when punching the edges of the pin so be careful!

For What it may be worth: I have on rare occasion surfaced the ID flats of the rocker arm to eliminate the needle roller wearing a ring. Such at best will get you through another bearing change prior to replacing the rocker arms (Lineaweaver).

++If you find it difficult to get the earlier rocker assemblies look up the 2000 exc 525KTM parts. The ktm has the same part # as the Berg. (Flatchat)

Manual de-comp

When operating the manual decomp on the electric starter, the point of contact is off centre line of the decomp shaft. The rocker arm therefore forces the decomp back closed and your hand starts a fluttering motion with each turn of the engine! The answer is to file a 45 degree edge to the shaft and this makes the contact point on the rocker arm directly under the middle line of the decomp shaft. It will now decomp the engine as you wish it too!

The dark area represents the 45 degree edge filed. It means that the "half-moon" is more on edge and isn't forced back against the cable. The 'edge' is about 2.5mm wide.

Decomp lobes

The difference between them to look at?

The old one has the end profile of 'a boat from above' and is symmetrical, and a 'D' profile where the return spring will slip onto.

The latest one has the end profile of 'a dutch clog from the side' and there is a single hole in the end profile. The spring sits in a gulley in the arm and the base of which is still a letter 'D'. The counterweight designs do differ but is too hard to explain!

Clog decomp	Berg 04 cam with latest decomp lobe. Note how close it is to the front compared to old.

This was particularly bad for machines through from '01 to as late as '03 but continued through to '06. Essentially, the decomp lobe would wear down and not lift the exhaust valve the .040" (1mm) to release some compression while the engine turned under cranking. There was also the problem that the lobe return spring would be either weak or unwound which equalled "ineffective". This meant the engine was trying to fight full compression and partial cure at the time was to add a 'blob' of weld onto the decomp and then re-profile. Husadawg has some great photos.

The auto de-comp. bleeds only a fraction of the trapped charge engaging near 85 BTDC and closing near 15 BTDC with a peak valve lift of less than .040". "E" Cranking compression should be roughly 130 PSI. (Lineaweaver)

In '03 Husaberg produced a new camshaft for the 650 and de-comp lobe (200 306-01), they produced another cam for the rest of the range but ommited the fact from the parts book. The new cam profiles appear to make life easier for the valve train. It's easier to follow by reading through an '04 parts manual than an '03!!!

In '04, everything got new part numbers and a new cam again for the 650. The decomp lobe code is now: 590.36.016.000 and is a complete cure for an ailing engine (Ady).

Valve length change

The later inlet valves were shortened by 1.25mm/0.050". The collet groove moved equadistant down the stem thus the valve spring is further compressed by 1.25mm on later models.

Camshaft profile

All cams up to '03 had a viscious ramp that made the valves bounce as well as tulip. The cam was changed for '03. Always rebuild with the later Cam coded '000' in the parts manual.

Weak and failing electric starter '00 - '02

For three years owners struggled with weak starters that were classified as 'restarters' and only to be used after the engine was warm. Typically the starter would spin through and not quite 'catch' the engine and batteries went flat quickly.

There was an error in the '03 parts manual whereby they simply failed to realise that the whole set up had been upgraded for that year. The parts book was corrected in '04 when Husaberg went to the present Parts # system.

By matching the old and new, it should be very easy to get the right parts. But remember, the sprague was only 40% of the problem for owners wishing to have good electric starts! The battery needs beefing, the starter motor wants changing. Then we have the modified decomp which matches a later ('03) camshaft.

WAS IN '02

9 250 247-01....Free wheel outer part 01- (outter)

10 250 245-01...Clutch mechanism free wheel 01- (sprague)

11 360 035-01...Needle bearing free wheel 01-

12 250 248-01...Free wheel inner part 01- (inner). Changed to part No.250273 for '03.

BECAME IN '04

20 800.40.020.000 .......03 FREE WHEEL GEAR 03 (outter)

24 0405 101310............02 NEEDLE BEARING K 10X13X10 02

25 800.40.025.000 ........03 FREE WHEEL HUB 03 (inner)

26 800.40.026.000........ 03 FREE WHEEL 03 (sprague)

Don't mix the 'sets' of parts! (read Engine whining ....below).

CBR600/Yam R6 Starter motor

Discovered by Faktor, the CBR600 starter motor is stronger and fits EXCEPT that the mounting bracket has to be fitted from the later model ('05 plus). It can be noted by the fact that the original mounting bolts are used but the starter is longer! Loonyberg fitted one and it's really easy!

Ned37 then found that infact the starter can fit the original bracket but you just need to drill fresh holes and a little filing. However, no big deal to do ok! Next Ned discovers that due to the windings in the Yam R6('99), the starter from the Yam spins far faster and enables his big 650 to start everytime.

Honda starter	Honyam starter	Honyam starter	Honyam starter mounted

The auto de-compression issue

The auto de-comp sits in the exhaust cam lobe on the underside and suffer from rapid wear on leading edge. the answer is to buy the latest decomp which has an end profile like a clog.

This was for machines through from '01 to as late as '03. Essentially, the decomp lobe would wear down and not lift the exhaust valve the .040"(1mm) to release some compression while the engine turned under starting. There was also the problem that the lobe return spring would be either weak or unwound which equalled "ineffective". This meant the engine was trying to fight full compression. The partial cure at the time was to add a 'blob' of weld onto the decomp and then re-profile. Husadawg has some great photos.

The auto de-comp. bleeds only a fraction of the trapped charge engaging near 85 BTDC and closing near 15 BTDC with a peak valve lift of less than .040". "E" Cranking compression should be roughly 130 PSI. (Lineaweaver)

Sprague bearing

How it works! (Neil E).

Gear wheel torque limiter

At the same time as the above was being 'beefed up' ('03), Husaberg changed the small idler wheel between the starter and the sprague assembly and created the 'gear wheel torque limiter'. It is designed to slip when faced with a massive surge such as a backfire etc and protect the starter motor. The assembly is built akin to an old drum brake with the friction surface on the inner wheel.

The slippage value is between 7 - 15 lb/foot2 (Boss). The gear was modified in '04 to part number: 21 800.40.021.100 .......04 GEAR WHEEL TORQUE LIMITER 04. The part can be disassembled and rebuilt with a larger shim: "The one I took off mine slipped at 5Nm, so when you undo the bolt that holds the two halves together there is a dished washer. All I did was to put a 0.015" shim in first then the dished washer, re-tighten the bolt and voila! it now slips at 25Nm! (Sparks). See the UHE Forum thread here.

Kickstart problems

Taff's crescent mod

For early kickstart engagement see the modification below. Take 2mm out of the crescent so that as the 'keyhole' part moves clockwise it can drop open and lock into the outer gear. You'll need to take a tiny, tiny bit off the case as well (the small area above the screw hole-in pic 3-only) using a flat faced rotary file. First pic shows the 'squarer' updated crescent.

Kickstart ratchet flies off!

It is often difficult to get the kickstart assembly to stay wound up and ready. The shoulder of the key only juts out 2-3mm and can easily slip inside the gear and then the gear also has say, 0.5mm slack out to the sidecover! Before tightening the cover the assembly has slipped off and flown undone. The answer is to add a little weld to the end of the keyhole end so that as much as possible juts out the back of the gear and can locate on the crescent's shoulder.

Kickstart key lifted

Kickstart welded key vertical

Whining noise/Racket

Noise from left side cover. Must get the correct sprague bearing for your model or they make a real rackit! Pre '03 spragues; #HB250248, post '02; #HB250273. See this UHE Forum Thread.

Balancer gear change - mid '02

Another gear that was changed was the counterbalance gear and that of the intermediate shaft gear with which it meshes. The factory made this change 'mid-year' 2002! = not good (for us!). The gear has a number stamped on it, get the new one wrong and a whining noise will result, so get the right parts ordered:

If it's stamped #130 you have a '01 engine and need balancer #210-255-01.

If it's #145 you have a '02 engine and need balancer#210-253-01.

Gearbox slack

Some owners may find that their gearbox is loose and may even whine! Dale Lineaweaver yet again knew what to do when he made these gearbox shims to take up the slack. "He made it 0.050" thick and slightly larger in O.D. than the O.E.M. 0.040" thick disc (JoeUSA).

"It was perfect, and in addition to correcting the axial location of the counter shaft gear axial location it corrected the excessive axial clearance of the counter sprocket".

Parts including custom stop disc (JoeUSA)	OEM & custom stop disc, shaft (JoeUSA)	OEM & custom stop disc (JoeUSA)	Stack assembled with custom stop disc. Perfect clearance for retaining ring. (JoeUSA)

Double-row counter balance bearings

The counterbalancer has gained weight in recent years and to remain reliable, the factory went to two ball-races side-by-side: a kind of 'duplex bearing' ('06+). Check out the '07 parts book for this upgrade. However there is concern about the whole crankshaft assembly and it's best to read page 4 and Enginehardwear's comments. The bearings can be very expensive to replace so see Ausbergs answer below!

2006 counter balancer, one piece bearing with two ball races	also close-up. (Dr C)

Counter balance bearing

The bearing can fail on the 501 engines whereas the 400 gets away with it's single row. But the factory solution is for the counterbalancer to contain two rows of ball bearings each at 6mm wide: a kind of 'duplex bearing'. Check out the '07 parts book for this upgrade.

The double row are infact two 9mm bearings ground down to 6mm. The cost of replacement for the originals is prohibitive so Ausberg simply ground a pair of 9mm bearings down by taking 1.5mm from each side. The oil seal lip has been machined away.

Sourcing replacements if nil stock and the new version balancer price scares you... (Ausberg)	8.9mm 6906 halfway to becoming a 5.98mm counterbalancer bearing. Seal retention lip is totally gone. (Ausberg)

Stuck in gear

The selector "claw" metal is too soft. Separate the "claw" from the plate, rub away edges and burrs. '04+ is made from hardened steel and has a darker colour (Dsducati). See old and new together below:

Note the two tiny screws (above). Because the newer spring makes the shifter taller, I initially tried a longer screw and a brass collar (on the right in the photo), but it was too tall and hit the underside/outer edge of the clutch gear teeth. So I had to keep the original screw, narrow the head down to just the width of the dowel.

Changing gear

Clutch centre gets grooved. A new one means easy changes again. 100% cure

Ignition timing marks

Find TDC using "dead-stop" method. Warning: 8d out on my bike. See this UHE Forum Thread.