Conversion from a Nissan 2 stage variable camshaft timing system to a fully variable system from a Mazda BP6D engine (page 2).
Solenoid housing design & manufacture
As I was installing the circular rear oil feeds I noticed a problem. The circular housing would jiggle slightly within the head, despite being a snug fit. As a rigid oil pressure line would have to be fitted to it as well, that would increase the tension even more, I decided to use a different way of providing oil to this part of the camshaft.
For the VG30DET(T) Nissan decided to use an additional camshaft bearing, that sits behind the last cam lobe. For the newer VQ V6 engine series Nissan removed that last bearing section as it was deemed unnecessary. This form of over-engineering can be used to my advantage. With the load on the camshaft on this final section being dispersed over 2 bearings one could miss a small amount of surface area where the oil channel is cut.
After meticulous measuring and aligning the galleries I came up with a single housing unit to house both the solenoid and magnetic sensor. In order to properly align the housing so that the magnetic sensor aligns exactly centered on top of the camshaft I designed the housing to sit on top of the 2 rear cam bearing caps. However, it’s not possible to bolt it right on top and use longer bolts. The bolts are regular M6 but they have a 6,40mm shank that centers the bearing cap. These bolts are hardened and impossible to reproduce in a lengthened version. To overcome this problem I cut 2,00mm off of the bolt surface area on top of the caps, and use a 2,50mm flange for the bolts to hold the housing down.
The housing that holds the solenoid, magnetic sensor and oil gallery was intentionally designed as a one-piece. On most factory engines, these housings sit on top of the valve cover. The Mazda BP6D for example has a ‘tower’ with internal oil galleries over which the solenoid housing slides as it’s mounted. This tower is closed off by several O-rings. This system allows for added horizontal play from valve cover gasket compression. However, I was unable to produce such a ‘tower’. The Mazda VVT housing is cast, and my housing needs to be machined as a billet. It’s impossible to cut the odd shapes of the oil galleries in a billet machined design. Additionally, valve covers for engines with bolt on solenoid housings usually have alignment dowels, so that the valve cover will mount back on the same location within a very small tolerance. The Nissan valve covers however, are quite loose which can easily cause misalignment issues during cam cover removal/installation. Worse, the magnetic sensor location would shift with every valve cover removal/installation causing valve timing changes. Installing alignment dowels was not possible with the Nissan cam covers.
However, the one-piece housing on top of the cam bearing caps has a drawback. Because the solenoid is quite long so is the chamber in which it sits. As a result, the housing must be fitted over the valve cover instead of vice versa. I decided to merge the cover and housing into a one-piece unit with JB weld, the 4 extra bolts from the cam caps allowing for perfect alignment. I’ve noticed that the factory valve cover gasket compresses to 1,50mm so I aligned everything to that height, while O-rings for the inlets allow for some horizontal play. But as the camshaft has to be mounted prior to the cover being fitted, the last 2 bearing caps can’t be torqued down. However, since this is a V6, the rear caps can be torqued down later if the cam lobes are installed upwards at these brackets. This way, no pressure is applied to the caps or camshaft meaning no chances of deformation.
The valve cover ‘welded’ with JB weld, and largely finished (will be sanded and painted later).
After both covers were finished, they were test fitted on a dummy engine to check for interference with the intake manifold. Unfortunately, or fortunately considering it didn’t happen with the engine running, the driver side JB welded cover cracked over it’s entire length. I ended up TIG welding both covers and used JB weld just to fill in non-structural spots that I can’t reach welding.
The VG30DET heads are adjusted as designed in Solidworks. The 1mm oil feeds for the front could not be drilled as the aluminium is 20mm thick up to the oil pressure gallery. Instead, I inserted spring pins, 2mm thick with a 1mm hole, as 2mm can be drilled in this thick material. The factory oil feed for the second to last cam bearing was blocked and two 2mm spring pins were also inserted to fully lubricate that cam bearing. The camshafts and heads were cut and adjusted on numerous places. The front bearings caps and heads were machined with a 1mm deep channel in them to achieve additional surface area. The factory VTC oil feed in the head was blocked, the front bearing now receives it’s oil pressure from the main gallery.
Synchronizing the trigger plates.
The left and right cylinder bank run each their own magnetic sensor, but share the crank sensor. As a result, the 2 banks must be angle matched exactly. If they’re not, the active adjustment limit will be substantially reduced if the banks have different limits. This will also cause both banks to require different settings, unnecessarily complicating their operation. In order to match them, I first measure the angle of the trigger plate on the BP6D head at the position Mazda indicates as TDC. To convert this to the VG30DET, by measuring the relation between valve lift and duration, the camshaft is set to start it’s 260° opening period at 5° ATDC for the 1st cylinder, the passenger side head. The trigger wheel is pressed on and locked with spring pins.
In order to synchronize the 2 camshafts, the above procedure is repeated for the driver side head, for the 4th cylinder, which mirrors cylinder 1. The trigger is carefully set in place. It’s a completely manual process that takes time, though it makes synchronizing the cams with the adjustment screws on the gears easy when the timing belt is installed. Normally, in order to synchronize the cams between 2 cylinder banks a similar procedure to the one above is required. However, with the cam covers open, all I need to do is fit 2 geo triangles to the trigger wheels and the difference can be seen instantly, The synchronizing turned out to be perfect, the VVTuner’s showed no discrepancy between the banks once the engine was running.
Modified oil galleries inside camshaft
The Mazda BP6D camshaft has double oil feeds to the advance and retard chambers in the phaser. At the smallest section they’re 3mm each. From the front of the camshaft I cut two 3,2mm holes next to each other (to avoid contact with the alignment dowel) and a single 4,8mm hole on the other side. I cut a special sleeve inserted to block the contact between the 2 galleries at the front oil feed. The thin front section allows the sleeve to bottom out without being a meaningful restriction.
In order to feed oil to the phaser in the same pattern as the Mazda camshaft delivers it I designed a unique spacer with internal oil galleries and mounting points for both the dowel in the camshaft and the phaser. This spacer also acts as the aligment spacer for the cam gears to the timing belt. By test fitting I was able to determine that the spacer should be 10mm thick for the exact same offset as the factory gears. The spacer was cut on the milling machine with a rotary table. For testing, this 7075 grade aluminium will do but a steel version will replace it.
At the front of the Mazda phaser unit is an overflow chamber, that is covered by the copper coloured plate as can be seen on page 1 of this project. The overflow chamber catches oil that escapes along the chambers. In order to evacuate this oil, a 3mm hole is drilled into the mounting bolt that holds the phaser down to the camshaft, and the threaded hole the bolt sits in has a passage drilled from just behind the oil seal back into the head. I designed it so that some of the escaping oil will seep in between the phaser and the side of the cam for added lubrication.
Eventually, after I assembled the complete system and applied air pressure to the retard or advance chamber the phasers operated properly. Only a small amount of air pressure seeped along the phaser housing into the overflow chamber as I expected. However, I noticed that once the maximum retarded setting is achieved and the lock pin locks, the pressure from the retard feed easily flows out into the overflow chamber. This is intentional; Mazda placed a plastic disk at the outer side of the phaser covering another overflow gallery from the retard feed to the overflow chamber that unlocks once the maximum retard setting is reached. Why is this? I think that in order to ensure the lock pin activates once the engine is shut off, all oil must drain as quickly as possible, and this helps.