1972 Triumph TR6 looking for a purpose

Yesterday and today were finishing the carpet install.  It's been a while since I installed a full carpet set (about 25 years), and I had forgotten how tedious this can be.  The worst part was fitting the carpet on the tunnel cover so that forcing the center console into place didn't open up a gap.  There were some tight fits, but it really wasn't too hard.  I'm starting to like the darker tan/brown color more than I did originally.

As a bonus, I also finished the seat belt install.  It's ready for seats, and I just got the seat upholstery hardware order yesterday.

Another thing that wasn't very fun was installing my custom dead pedal.  It is bolted to the bulkhead panel with spacers to hold the carpeted side panel in the right spot.  I made a template to use as a drill guide for all of the bolts and screws, and this worked fairly well.  But getting those bolts started while laying twisted on the floor was a back workout.  I think the dead pedal turned out pretty nice and puts the high beam switch just above my toes.

I want to do a little caulking under the car so that less water creeps in, then I can drop it off the jack stands.

You put a high-beam switch on the dead-pedal?  Is this a thing I've just never heard of?

Is there no other accommodation for the high-beams anywhere?

Otherwise, it looks nice.  That color looks like saddle leather.  I think being back West is soaking in on you.

TVR Scott (Forum Supporter) said:

You put a high-beam switch on the dead-pedal?  Is this a thing I've just never heard of?

Is there no other accommodation for the high-beams anywhere?

Otherwise, it looks nice.  That color looks like saddle leather.  I think being back West is soaking in on you.

That's the stock high beam switch position for '72 and earlier.  There's a small box that mounts in the stock position, I just used the dead pedal instead.

I was thinking about the high beam switch being on the floor.  I always thought this was an American car thing up until the late 70s.  Before this TR6, I had no idea that British cars did the same thing.  It does still have the flash-to-pass feature on the column.  It will take some adjustment getting used to the headlight off-flicker-dim control now being on the column.

I shot some caulk around the floor underneath where the center console mounts.  The potential leaks there were made worse by adding the cat heat shield brackets.  While up in the air, a final leak check was done and everything is dry except for a single drip coming out of the diff.  So I dropped the car back down and ran it until warm.  Man, I really need to get a seat in this and register it before the snow starts falling.

My Bugeye has a dead pedal high beam switch too. 1961 (58 design)

Step one of upholstering the seats is to replace the rubber seat bottom diaphragm.  These came with the original pile of parts, so are at least 20 years old.  They appear to be a tad small or unable to stretch enough to fit, at least in width.  I've already managed to tear one, and don't have any hope of the other going any better.

New ones are already on the way because I do not want to tear the seats apart replacing these in a few month or years.  Hopefully the new ones will be more pliable and not complete junk.  Laying them in the sun before installing may also help.  The other option is to add some extra rings on the sides, but the extra hooks shown above left the diaphragm slightly loose and off-center.

In the mean time, I'm going back and sorting some things out.  I noticed that the throttle wasn't returning to idle fully, so checked to make sure the carpet wasn't fouling the pedal.  Some silicone grease on the pedal end of the cable had thickened slightly, but the real reason was that the return springs had popped off of the pedal.  I have since added a return spring directly to the linkage under the throttles, so will beef up that spring instead.  While the intake plumbing is off, I'll also fine tune the throttle synchronization.  Something is causing the AFR at idle to fluctuate between ~14 and 15, so checking for air leaks.  Tuner Studio may tell me more.  The injector dead time also needs to be tweaked, but I can't do that until the AFR is steady.

Yes, the joys of working with replacement parts. I noticed that all the rubber parts on my TR4A are all gone again, so neither the British originals or Chinese copies seem to last very long. I am sure the new replacements will fit fine and your seats will match the rest of the quality of this build!

Hm, trace onto some conveyer belt (or treadmill) material?  then they at least have the fiber internal to assist...  

Apexcarver said:

Hm, trace onto some conveyer belt (or treadmill) material?  then they at least have the fiber internal to assist...  

I believe the material needs to be a little stretchable, otherwise that may work.  There are reinforcement plates cast into where the hooks go, but riveting two plates on either side should work.  I'll wait and see how bad the new diaphragms are before worrying about it.

And to me, the new Chinese rubber seems to actually be worse than British.  But to be fair, these diaphragms were made in jolly old England.

The AFR at idle is still fluctuating too much.  I didn't find any air leaks around the intake, but maybe there's a pinhole on one of my exhaust welds near the O2 sensor.  Doubtful, because I checked for that twice now.  Also, something is also causing RPMs to fluctuate and I haven't enabled Lambda correction yet, so the AFR is a symptom, not a cause.  I'm left with either a poor vacuum signal, injectors, or something else.  To test the MAP signal, I switched to using Alpha-N mode which ignores MAP and uses TPS instead.  Things didn't seem to improve a lot, but the engine load didn't vary as much.  So that leaves me with injectors or a mystery issue.

When I bought the Jenvey throttles, the smallest Pico injectors available from the vendor were 330cc/min.  The vendor convinced me these would idle OK with a hot TR6 motor.  For reference, the bonkers autocross TR6 I used to drive ran with 158cc/min injectors.  I did some Megasquirt logging and noticed that at idle, the injector pulse width is around 2.7ms.  The injector dead time is 1.1ms, so that doesn't leave a lot of room for error.  I have another set of 210cc/min Edelbrock Pico injectors from the other TR6 that would tend to fail when they got hot, but they worked fine for normal driving.  I could pop those in and see if the idle improves.  Reving the engine up to 2000 RPM doesn't settle the AFR very much, but it's still at a fairly low PW.  I could also drop the fuel pressure to get more PW resolution, but how low can I go?  It's currently at 43 psi (3 bar), and the specs for this injector do not list a 2 bar flow rate.

The only other strange thing I see in the log is a fluctuating barometer reading.  I'll disable that to see whether it makes any difference (doubtful).  Also, a McMaster Carr order is inbound that should improve my MAP setup.

I got two parts orders yesterday to address several problems.  The old hazard flasher unit I had was burned out and it's different from the turn signal flasher.  So now I have hazard lights (hoping that they won't be needed).  The new seat diaphragms also worked.  They were stamped as Roadster Factory parts (even though ordered from BPNW) so are probably from Taiwan.

The main event was rigging up a permanent MAP setup.  The existing one was kludged together last January just to get the car running.  It worked fairly well, but wouldn't pull a vacuum much below 50 kPa (with barometric around 80 kPa in CO).  I was low on hose, so joined the adjacent vacuum ports together almost immediately and ran three hoses to a small manifold.  I suspect that this allowed some of the vacuum to pull from the adjacent intake runner.  The new setup runs six individual hoses to a single, larger manifold tucked under the intake.  I also added a hose from an unused port on the balance tube for experimentation.  Here's what it looks like.

The vacuum manifold is under the throttles on the left.  The vacuum hoses now also help support the injector harness.  This setup pulls slightly more vacuum at idle and does really well when the throttle is opened.  I may need to put a restrictor on the larger hose coming from the balance tube, or remove it completely.  Even without it, the vacuum signal is smooth and yet responsive.  Unfortunately, this didn't help my AFR fluctuation issue.  Looks like it's time to try the smaller injectors.

Vacuum reference looks good.  I've no experience with ITB setups but you always hear about bad vacuum-related sensing.

From what you were saying on the injectors, the smaller size sounds like a really good idea.  Especially comparing that 330cc size to your race car.  Any reason to not try going down to a similar 160cc?  Beyond that the 210's are readily available?

TVR Scott (Forum Supporter) said:

From what you were saying on the injectors, the smaller size sounds like a really good idea.  Especially comparing that 330cc size to your race car.  Any reason to not try going down to a similar 160cc?  Beyond that the 210's are readily available?

Yes, that sounded good.  I swapped in the 226cc injectors I had (not 210cc, my bad), and except for creating a couple of fuel leaks at either end of the rail, they worked the same.  I didn't even need to retune, just change flow in the required fuel calculation.  These are a good size injector for this car as the 158cc ones come very close to 90% duty cycle even with a 4 bar fuel rail pressure.  Idle pulse width is now over 4ms, or about 5% duty cycle.  Unfortunately, the revs and AFR are still fluctuating.  I even tried to set all of the fuel map cells being used at idle to the same value, but this didn't help.  Maybe playing around with tuning the fuel map could stablize things.

Next up is checking valve clearances.  There's some noise coming from the valve cover, so maybe one of the intake valves is set too wide.  My valve timing may also be a little too aggressive, but it isn't running like with a lumpy cam.  It's more of a rolling fluctuation.  I also still need to check the barometer setup, but it's only varying by 1-2 kPa and doesn't seem to correlate with AFR or RPMs.

Some forward progress today after some backwards movement.  I pulled the valve cover and replaced the cork seal with a new silicone one.  Hopefully this won't weep oil like the cork ones do over time.  Then it was time to adjust the valves.  That's when I realized that I can't turn the crank from the front without removing the electric cooling fan.  There's just not enough room to get a big wrench in there without risking the aluminum radiator.  It was only four bolts and didn't take too long.  It turns out that four of the valves were a little tight, but one was a few of mils wide.

Next up was checking torque on the manifold nuts.  The upper ones were no problem, but the lower ones required the intake heat shield to be removed.  This is held on by three bolts under the intake that I can just get a small socket wrench onto.  The shield slid out once the bolts were removed, and I found pretty much all of the intake/exhaust clamp nuts were not very tight.  That's fixed, and hopefully any air leaks that were hiding under the intake.

Once buttoned up, I started it with the barometer set to use the initial MAP reading.  This doesn't adjust for barometric pressure as you drive, so won't work great when driving through mountain passes around here.  It ran quite a lot smoother and without the rocker noise.  I'm still seeing some variation in RPMs and MAP, and am beginning to suspect that some of the Megasquirt sensors (barometer, MAP, oil pressure) are picking up electrical noise.  The TPS signal, temperatures, and ignition are solid, at least.  This will be investigated tomorrow, but the car is running well enough now for street tuning.

Crawling around under the dash yesterday looking for electrical noise issues reminded me that I hadn't labeled my fuse block yet.  I did most of the electrical work one year ago and already had to retrace most of the wiring.  It was my purpose to rethink what I had done, so this isn't necessarily a bad thing.  With all circuits labeled, this should go faster the next time.

I'm focusing on why the barometer and oil pressure sensor (and maybe the MAP sensor) vary more when the engine is running.  Both are powered by the Megasquirt VREF wire that also powers the TPS and temperature sensors.  Rather than using the sensor ground wire, the baro sensor uses one of the other Megasquirt grounds.  The oil sensor uses the exact same power/ground wires with the same routing as the temperature sensors, and the signal wires are in the same bundle.  So I'm at a loss to explain why that is picking up noise unless it's simply being in close proximity to the coils and plug wires.

You can see the gold oil pressure sensor connector in the middle of this shot and how close it comes to the coil wiring.  I wrapped that harness and sensor in foil, but it made no difference.  I also separated the baro sensor wiring as much as possible from the coil wiring under the dash to no effect.  Maybe the injector wiring is causing some noise, but it's all bundled together where it come out of the Megasquirt (using the prefab DIYAutoTune harness, BTW).  Following the DIYAutoTune wiring diagram, the Megasquirt is grounded to the block using pins 15-19.  I suppose it's time to review the Megasquirt manual concerning grounding best practices.

Megasquirt Grounding

I found and fixed one thing that made a small difference.  I had the DIYAutoTune Quad Spark (coil control) grounded to the body, and later shared this ground point with both the main relay coil and MS power relay coil ground.  This may have been feeding noise directly into the fuse block.  I rerouted the relay grounds directly to the negative battery cable.  The car is running better with an occasional spike in AFR.  Putting the intake back together and moving on the the seats.

So now I really need a seat to switch this car with the other TR6.  It appears my brother will be buying the mallard TR6.  That works out well as it will stay in the family and I may be able to buy it back in the future.  As typical with selling to friends and family, I won't be making top dollar.  But this will save a lot of work and stress.  The mallard car needs some cleaning up and a few other things, so that will be my priority until it's gone.

I also need to move this project in order to drag down the snow tires for other cars from ceiling racks.

I just binge-read this from post #1. Man, you do nice work. Carry on.

Tight valve clearances will cause a lumpy idle before loose ones do.  I accidently set mine too tight on my Starlet after a head gasket job.  The idle was lumpy but the difference between yours and mine was the aluminum head on the Starlet.  Once the head warmed up and expanded enough, the clearances became more normal.  After a quick check and reset, all was good.

Also, I believe your hypothesis about Megasquirt picking up noise may be correct.  I owned an '83 SAAB 8-valve 900 Turbo, the first year with electronic APC boost control.  After changing spark plug wire brands, the controller went absolutely apeE36 M3 until I installed a set of stock Bougicords. 

AngryCorvair (Forum Supporter) said:

I just binge-read this from post #1. Man, you do nice work. Carry on.

I hope you had some bourbon handy.

Jerry From LA said:

Tight valve clearances will cause a lumpy idle before loose ones do.

I believe I had several small effects adding up.  A few tight valves certainly made the idle worse.  The loose one was surprisingly noisy, so glad that's fixed.  I may still play around with distancing the electrically noisy components from other things.  The QuadSpark could easily be moved to over the tunnel cover, and I can add an elbow to the oil pressure sensor mounting.  The good news is that everything else looks fine.

Good news on the mallard car - I hope your bro gets to enjoy it greatly.  That thing is gorgeous!

TVR Scott (Forum Supporter) said:

Good news on the mallard car - I hope your bro gets to enjoy it greatly.  That thing is gorgeous!

It's in great shape for a restoration I've driven 30k miles.  We need to get the fuel down to 1/4 full for shipping, so what better way than to drive it around. When we got back, I cobbled together a seat for the new/blue TR6 and moved it to the storage side of the garage.  Of course I had to drive it around the neighborhood to get it in the right orientation.   From that short drive, it felt so much more responsive (power wise) and tight than the mallard car.  I'm really looking forward to getting it street tuned now.

I've been checking off things for getting the other TR6 ready to move.  I've cleaned up the interior, reglued the rear wheel arch covers, and replaced the seat back foam on the passenger seat,. The driver's seat is getting the same treatment, but the foam is shipping from farther away.

I should have done this years ago.  The 25 y.o. foam never was quite big enough to fill the side bolsters, and time had taken it's toll.

The grill was removed to reweld a mounting stud that had broken off, so I took a shot at improving the hood and fender alignment.  It is now marginally better, but probably not worth the risk of damaging something.  I replaced the hood release cable to fix it's broken pull handle, and copied the backup hood release I made for the blue car.

Once the driver's seat is done, work will resume on the blue TR6.

Okay, maybe I'lll do a little on this project.  The seat foam for the other car is now back ordered, but I have all of the stuff to put together a driver's seat for the blue car.

I made the mistake of following the instructions provided by Moss Motors.  These have you glue all of the lower seat foam parts together on the frame, then glue just the middle of the seat bottom to the foam ignoring the cloth tabs on the upholstery.  I had just stuck the bolster foam down when I noticed this, so basically tore the foam back apart.  The cloth tabs go between the foam layers, pulling the upholstery down to form the seat bucket.  Complicating things in my case are the seat heating pads glued to the foam.  Everything is lined up and stuck, and we'll see how well it looks when I pull the upholstery over the frame.