Pi - Fuel Accumulator

[michaeldavis39]

If Mercedes with their K Jetronic mechanical fuel system which has injectors that pop under a similar pressure to the Pi Lucas injectors fitted a fuel accumulator to prevent cavitation/vapour lock and help hot starting- why didn't Triumph fit the same to our Tr6 with Pi? Im guessing that perhaps accumulators were not around then? To me Mercedes clearly understood the potential flaws in Mechanical injection and wanted the best for their cars and customers so they made every effort to make their version of mechanical injection reliable. Has anyone in the Triumph world tried fitting a fuel accumulator? If they have what was the result and if not have I discovered something that could improve all the Pi cars which nobody has considered- it makes a lot of sense to my mind. Its also worth mentioning that I read recently of a Tr6 Pi owner in Australia fitting these same Bosch mechanical injectors that the K Jetronic Mercedes system uses and his car drove fine- he used these injectors as they were easier to get a hold of and much cheaper than new Lucas injectors- they pop at around 48psi so very similar to the Lucas Pi injectors- interesting isn't it!

Edited June 9 by michaeldavis39

[trchris]

Trouble is you won’t know the answer until you fit one , did Mercedes use a mechanical pump or 12v one ? Doesn’t the cavitation mainly happen between tank and pump so maybe fitting wouldn’t make any difference?

Chris
 

[john.r.davies]

Possibly because in the '60s, Triumph bought a Lucas Pi system that had been developed over many years since being developed in WW2 for aero engines, then in the 50's for Jaguar and Le Mans.  AFAIK at no stage did it have had an accumulator (high pressure reservoir) between pump and metering unit.  So Lucas didn't think it needed one.

Moreover, Jetronic was very different from Pi.   Single injector (well, there was also cold start injector), continuous flow, working off a simple flow plate in the inlet manifold, no fancy vacuum diaphragms, an electric pump, Chris, and only the "fuel distributor" between pump and injector, so that a major role for the accumulator was to smooth out pump pulses.   Pi, as we know has pulsed injection.

No doubt Mercedes learnt from the Lucas Pi, so that they built a system that was was a lot simpler and cheaper.   Apart from more components, the complex Pi needed precision engineering in the metering unit.   Maybe after Triumph had to spend the equivalent of £1 million to adapt Pi to their cars, there wasn't the budget for it.  There certainly wasn't for dealer staff training.

Anyway, as said, Jetronic was continuous flow, single injector.   The metering unit itself, and the high pressure fuel it contains, must constitute a 'fuel accumulator"!    My belief is that the poor starting of a Pi system is because of poor fuel supply, but not because of pressure lack, but because wear in the M/u allows fuel to drain back on standing, so that it takes a more turns of the engine and M/u for it to get back to the injectors.

But try an accumulator!  A very interesting experiment!   Please report back, Mik!

See: http://www.mbs.id.au/tuning/Fuel_Injection/K/schematic_K.htm 

John

 

Edited June 9 by john.r.davies

[PodOne]

Just my thoughts having spent time with various EFI set ups over the years is that if you can get the PI to run reasonably reliably which I believe you can within the technology of the day. Then with a few mods you have to accept its foibles and that it will never be as good aa a 1970,80 or 90s car with mechanical or early EFI and less so with post 2000 cars. Its part of owning a classic car!

Going for decent EFI/ECU set up at a higher cost than a "sorted" PI will likely get you closer to a modern but still unrefined even with the best of mapping given the millions car manufactures spend getting there. It's an equally time consuming and experimental journey as the Lucas PI.

Cheaper routes would be a full modern donor engine transplant will its ECU and fuel system but then the originality has gone and it brings its own issues but a good project if your not bothered about retaining originality.

That said the Mercedes injectors if there a reasonably straight forward swap is worth a punt as six new ones is the cost of 1.5 Lucas new ones and likely a better design. Who's going to try first 😂. Micheal?

Andy

[michaeldavis39]

Andy it would be an interesting experiment to see how the Bosch injectors perform yes, however I have just bought new Lucas ones last year- ouch, and I have also changed to the Pierburg fuel pump with a coarse guaze Sytec fuel filter with fuel tank pipework insulation and haven't had a day to go for a run out yet with work commitments but we will see- if Abby runs fine with no cavitation then job done me thinks- she ran fine on a run last year before changes just hesitant on stop start traffic on a hot day so hence my changes to improve the running.

[PodOne]

28 minutes ago, michaeldavis39 said:

Andy it would be an interesting experiment to see how the Bosch injectors perform yes, however I have just bought new Lucas ones last year- ouch, and I have also changed to the Pierburg fuel pump with a coarse guaze Sytec fuel filter with fuel tank pipework insulation and haven't had a day to go for a run out yet with work commitments but we will see- if Abby runs fine with no cavitation then job done me thinks- she ran fine on a run last year before changes just hesitant on stop start traffic on a hot day so hence my changes to improve the running.

`Hi Micheal

Agree if she runs fine for you enjoy her. In the same boat regarding injectors but was worth it.

Just moving the Moss Bosch pump/filter kit combo outside. Tight fit., Retaining the cooler and adding a booster pump. After that not much else left to try.

I cant think of a reason not to switch around the MU and PRV returns to the tank so the PRV is the furthest away from the outlet. Anyone see any issue with this?

 

[Mike C]

10 hours ago, PodOne said:

`Hi Micheal

Agree if she runs fine for you enjoy her. In the same boat regarding injectors but was worth it.

Just moving the Moss Bosch pump/filter kit combo outside. Tight fit., Retaining the cooler and adding a booster pump. After that not much else left to try.

I cant think of a reason not to switch around the MU and PRV returns to the tank so the PRV is the furthest away from the outlet. Anyone see any issue with this?

 

No. That's the way my tank is set up- PRV return tank connection as far as possible from the pump suction line tank fitting. The fuel from the PRV return line has the maximum chance of cooling across the tank  before it returns to the pump.

Edited June 10 by Mike C

[multipletriumphsinner]

I used to buy our PI parts (O rings, diaphragms, etc) from an ex Lucas engineer who was involved with the development of the MKII system for use in road cars (he also went on to be the chief Lucas representative engineer at every GP race where their system was being used).

He told me that Lucas were horrified to find Triumph fitting the PRV so far from the metering unit. Originally the PRV was supposed to be screwed straight into the metering unit (there is a brass plug blanking the hole on early TR5's. He was wasn't at all surprised that they suffered from cavitation as a result.

 

[k_raven_smith]

That would have been the late Bob Blurton (a good friend and work colleague). I have modified quite a few units over the years to the metering unit mounted relief valve. The modification works exceptionally well. 

[john.r.davies]

O Expert Gurus!  Please explain why the PRV is better next to the M/u than in the boot and six-odd feet away?

Is it to allow returned fuel.to cool?   Thats how mine is arranged on my Pi Vitesse.    TempTabs on my lines showed worryingly high temps, not helped by my two-stage pumps with a swirl pot between them.   I fitted a fuel-to-air cooler under the boot floor that takes the fuel as it returns from the PRV to the tank.    Seems to work, but not been out on a hot day yet.

John

[Drewmotty]

Any heat taken out the fuel prior to getting back to the tank is a bonus so a nice uninsulated 6ft of copper will help. I’d consider a large diameter return of 15 or even 22mm alongside the chassis worth it to maximise cooling. 
These cars would really benefit from a compensating pump so that the fuel doesn’t get heated up in the first place. 
Is anyone aware of a suitable pump which doesn’t come off an aircraft and cost as much as the car is worth?🥴

[stuart]

I used to run a Volvo 940 with the same Bosch mechanical fuel injection, their setup had a lift pump/fuel gauge unit in tank and that lifted fuel to a Bosch pump/filter setup in a "Sled" bolted below the car under the passenger seat. That was very reliable and never gave a minutes problem in even the hottest weather and would run down to virtually nothing in the tank. If your fitting a new tank anyway why not get a lift pump fitted in the tank and mount the main pump etc under the floor inside the chassis rail? I believe someone from TSSC has done the in tank pump conversion with a Gitfire.

Stuart.

[stuart]

2 hours ago, k_raven_smith said:

That would have been the late Bob Blurton (a good friend and work colleague). I have modified quite a few units over the years to the metering unit mounted relief valve. The modification works exceptionally well. 

If the PRV is mounted direct to the metering unit how would you check the pressure, Presumably via a "T" section mounted between?

Stuart.

[k_raven_smith]

You measure the pressure as normal. The advantage of the metering unit mounted prv is that the system is run in a loop allowing any vapourised fuel to be moved straight back to the tank. Talking to Bob about this method he couldn't remember any race engines suffering from cavitation. 

[k_raven_smith]

The finished item. Having the valve fitted this way makes adjustments a lot simpler. 

[Steve-B]

So the return line from the MU -> PRV  would just see the PRV mounted on the firewall then the return line go straight to the tank then right? If so that'd be a vry eloquent way to also get a lot of air on the PRV so that returned fuel would be cooler by the time it got to the tank.

Anyone done this, and have any pictures of the setup? It seems so simple to implement possibly as the PRV in the boot would be replaced with a join or two?

[stuart]

5 hours ago, multipletriumphsinner said:

 

He told me that Lucas were horrified to find Triumph fitting the PRV so far from the metering unit. Originally the PRV was supposed to be screwed straight into the metering unit (there is a brass plug blanking the hole on early TR5's. He was wasn't at all surprised that they suffered from cavitation as a result.

 

Also would have been made easier with the original fitment of the Lucas pump on the front left inner wheel arch as per CP1/2

Stuart.

[stuart]

1 minute ago, Steve-B said:

So the return line from the MU -> PRV  would just see the PRV mounted on the firewall then the return line go straight to the tank then right? If so that'd be a vry eloquent way to also get a lot of air on the PRV so that returned fuel would be cooler by the time it got to the tank.

Anyone done this, and have any pictures of the setup? It seems so simple to implement possibly as the PRV in the boot would be replaced with a join or two?

See the picture above your post of the PRV mounted directly on the MU.

Stuart.

[Steve-B]

10 minutes ago, stuart said:

See the picture above your post of the PRV mounted directly on the MU.

Stuart.

Ahh, that's what it is. I was refering more to if I already have a modern PRV in the boot, just moving it to the engine firewall.

Edited June 10 by Steve-B

[Spit_2.5PI]

I bet Triumph put the PRV in the boot because the cost engineers insisted on saving the cost of 6' of pipe to the back of the car, sized to take the flow at idle.

As for a variable-flow pump, I bought a fuel flow sensor off Ebay with a view to putting it in the return line and winding down the pump power to maintain a minimum flow. Perhaps a throttle sensor to increase power rapidly if the accelerator is suddenly pressed. But this has turned into another of those ideas I never get round to actually putting into practice.

Cheers, Richard

[michaeldavis39]

So am I right in thinking that on a modified MU with prv the fuel from pump connects to the MU as normal and then the excess fuel from prv goes straight back to the fuel tank?

[john.r.davies]

Please anyone?  WHY is it better to have the PRV next to the M/u?

John

[TR NIALL]

53 minutes ago, john.r.davies said:

Please anyone?  WHY is it better to have the PRV next to the M/u?

John

Maybe because the Fuel has further to travel back to the Tank and stands a better chance of cooling.

[michaeldavis39]

Is that right then on modified mu then as I asked above or would main fuel line go to prv and the return go to the original fuel inlet on mu?

[TR NIALL]

14 hours ago, michaeldavis39 said:

So am I right in thinking that on a modified MU with prv the fuel from pump connects to the MU as normal and then the excess fuel from prv goes straight back to the fuel tank?

I’m not sure but think so,maybe some of our PI experts can explain better.