Upgrading to throttle bodies on a Fiat/Lancia 16v turbo

Competition engines and ancillaries - general discussion
Post Reply
Evodelta

Upgrading to throttle bodies on a Fiat/Lancia 16v turbo

Post by Evodelta » August 10th, 2006, 10:38 pm

Hello all,

I'd like to set up a thread here to discuss converting a single throttle Fiat/Lancia 16v turbo engine into a four single throttle body set up, it's just for me and you to learn from and for ideas and theories to be discussed, I'll start if off here by tellling you what I have learned and what I would like to achieve, please feel free to comment, advise and tell me if you think i am wrong in any particular areas.
The car is a 16v integrale that I use as a test bed for my ideas and as a trackday car in between, it isn't used on the road and when it is finished I would like to compete in it, quite what in I don't know! But I do enjoy driving it hard and have a competitive streak in me.

First of all why Throttle bodies?

Well, the standard 56mm throttle plate on this engine is ideal for up to about 320bhp, after that it becomes restrictive, sure you can get more power whilst still retaining it, but still the fact remains it is causing a restriction. Some people bore out the housing on the Lancia one and fit a larger butterfly (or throttle plate) this is a good move as far as it goes, but go too large and low speed drivability can be difficult, as you can imagine, cracking open a large throttle plate even a tad can let a large amount of air through suddenly, resulting in setting off a bit quicker than you intended (!) and a jerky response. The Fiat (16v turbo) gets around this problem by shrouding the top half of the plate, this means that on part throttle openings only the bottom crescent is used up to a certain angle, then after that the full circle is exposed bit by bit, it's a very clever but simple design and works well:

Image

To sum up: I am looking for more power with good response and controllability. First of all I found a suitable donor for the manifold, it is perfect in a few ways:

1. The bore size is 40mm, this equates well with the inlet valve size of 34.2mm and puts it firmly in ideal circuit/drag car territory, very slightly too large for street/hillclimb/rally, but nothing to worry about, not when I can adjust the total length of the inlet tract to compensate.

2. The flange fits the head spot on. (Almost as if it were made for it :wink: )

3. The injector boss is very close to the base, this is perfect for low speed pick up, for higher engine speed the second set of injectors mounted higher up in the Tbs will come into play.

4. It has circular/tubular ends with a recess for the rubber which joins onto the TB, seperate Jenvey SS bodies fit perfectly.

5. The total length of the whole inlet tract also measures up well with recommendations for circuit/drag use, maybe a little short for street/hillclimb/rally but again, it can be lengthened if needed with longer bellmouths.

The downside: It presents itself to the head at completely the wrong angle. I worked out geometrically what I needed, cut it up and had a local fabricator weld it back together again. The angle isn't perfect, but to get it perfectly in line with the ports would mean the Tbs poking through the bonnet. Here's a comparison: Before (lower) and after (upper)

Image

And here's a shot down the bore;

Image

And then up it:

Image

Here's a shot of the Jenvey throttle body:

Image

And finally, a pic of me having a trial fitting to see what it is like for clearance, the timber dummies are dimensionally correct to help see any future problems, I can see one looming, but nothing my saw won't sort out.

Image

Ok, now some awkward questions:

I need to have vacuum take off points fitted for brake servo, Map sensor and idle control valve, these are going to have to come from the Tbs and the manifold, the ICV is pretty much catered for here:

Image

1st question: How exactly does the ICV work, what does it do?

Will the map sensor work ok from 4 joined together points? I have been told it may not be happy at the pulses hitting it, I can dampen the pulses with small bore hose and a small fuel filter, but then this would delay the signal perhaps and make it useless..... (this is my main worry)

Can I take off the vacuum point for the brake servo from just one runner or will this cause an imbalance? (I think maybe not as it is only working on deceleration and will 'lock up' under boost due to the fitment of a one way valve)

Do you think my efforts at improving the inlet manifold angle have made any benefits and could they be improved?

Martin.

P.S. I'd like to perhaps discuss intake plenum design next if that is ok?

SteveNZ

Re: Upgrading to throttle bodies on a Fiat/Lancia 16v turbo

Post by SteveNZ » August 11th, 2006, 12:16 pm

Evodelta wrote:


1st question: How exactly does the ICV work, what does it do?

Will the map sensor work ok from 4 joined together points? I have been told it may not be happy at the pulses hitting it, I can dampen the pulses with small bore hose and a small fuel filter, but then this would delay the signal perhaps and make it useless..... (this is my main worry)

Can I take off the vacuum point for the brake servo from just one runner or will this cause an imbalance? (I think maybe not as it is only working on deceleration and will 'lock up' under boost due to the fitment of a one way valve)

Do you think my efforts at improving the inlet manifold angle have made any benefits and could they be improved? :wink:

Martin.

P.S. I'd like to perhaps discuss intake plenum design next if that is ok?
ICV, Idle control valve is more correctly stated as an ISC, Idle speed control valve. It is just like a small throttle body operated by the ecu to control the idle speed. It is held closed by a spring and is opened with an electronic solenoid controlled via PWM. This is a simple system and has limited control over the idle speed. It is poor at maintaining the idle speed so is only used to "step up" the idle speed for cold start and A/C etc. This system needs an IAC, Idle air valve to set the base idle speed. This is part of the throttle body. Both need to be clean and adjusted correctly for the system to work correctly.

MAP measurement can work ok setup like you have described. However it can give poor results when MAP is low. It is better practice to exchange MAP input for throttle position. Either can be used as a major load sensor for tuning an ECU. Throttle position does not give enough information for a turbocharged engine though, it depends how good the tuner is but MAP input makes it a lot better. Most aftermarket ECU's can be setup with both.

Factory cars with multi throttles usually have an AFM which is the best. Most people dislike this option though.

Vacuum for the booster can be taken from 1 runner only. The pulses do not matter as the booster "stores" up the vacuum for when it is needed.

Evodelta

Post by Evodelta » August 13th, 2006, 8:56 am

Ok, thanks for the reply and the explanations, very usefull.
Now you've left me wondering if I can't just get rid of the ISCV when swapping onto Tbs? I guess really, it will be just a case of trying it to see. I am running an aftermarket ECU.

What is PWM?

What do people have against MAF sensors? Is it to do with flow loss or slow reaction time?

Again, it may be a bit of trial and error to get this to work properly, the problem is errors can often be expensive and I'd like to be aware of all the pitfalls before going too far, or before they occur anyhow. The problem is finding someone who has been there, done it and willing to part with the knowledge.

Martin.

SteveNZ

Post by SteveNZ » August 13th, 2006, 9:10 am

You maybe able to ditch the ISC but it will probably stall depending how you have it tuned.

An ISC valve does not really work with multi throttles but an ISC actuator would. A mechanical actuator that operates the throttle to controll the idle speed. These are found on some cars from the factory, usually with single point injection like an Uno70 or Ford Falcon EA.

AFM's are ugly and annoying, they need extra plumbing etc. most people think they are a huge restriction too. Modern AFM's (VAF) have virtually no restriction at all. They work well for slow running but a MAP sensor is better, or easier, for tuning with boost.

PWM is pulse width modulation.
Last edited by SteveNZ on August 13th, 2006, 10:21 am, edited 1 time in total.

Guy Croft
Site Admin
Posts: 5031
Joined: June 18th, 2006, 9:31 am
Location: Lincoln, UK
Contact:

Post by Guy Croft » August 13th, 2006, 9:20 am

Good thread!

Re: servo - you can take it off one inlet runner but put a non-return valve on it or you may get erratic idle. These were used on Fiat 124 Sport and 131 models, put a post in sales and wants. I hope that is one small issue identified.

As for load monitoring via throttle position and manifold vacuum vs speed, surely the ECU manufacturer is going to tell you what is best? Who is it?

I don't want to spoil the party but (see my cautionary thread on FI) unless you extremely detailed load site mapping it won't matter what you use - or do. I would want throttle and vacuum feedback and a user friendly ECU like DTA or Weber that can interpret and make use of them. (I am not at all saying other ones are no good but I have no experience of them). I am not criticising anyone here and I am prepared to stand corrected on this (if an only when I have detailed sight of maps of ignition and fuelling and before and after dyno curves for a 16v Fiat/Lancia from any rolling road) - that it is a big job to replicate the accuracy of bench dyno calibration by that set-up method when you are starting from an empty chip, because of the extreme difficulty of holding the load and speed at the precise throttle angle engine rpm load site positions for long enough and often enough to stabilise the torque and adjust the setting of fuel and ignition without overheating the engine and its oil. FWIW I think I'd want to copy the parameters of say - a comparable bhp Vauxhall XE 2 liter 16v as a base map.

I know of people who have come away from RR sessions with an engine like yours with what I would call very poor power results compared with what I would expect and I am certain that this is because too few load sites were optimised. I mean 60 bhp down. And I would be very cautious who I'd choose for this as I know, eg: from bench dyno step tests at Warrior (I'll publish the curves later) that there can be a huge difference in the torque vs speed graphs plotted at 15 second stabilised 500 rpm steps and then closer 250rpm steps. If a rolling road owner wants to step forward here and offer do this, great.

If there is a torque difference then there is a fuel-injection calibration demand difference too, you have to go looking for obscure load site conditions and those 'holes' in the torque curve or you're going to get a fraction of the power you expect. Don't forget the output at the flywheel is the 'mean' torque - the true sum of all the parts (sites), not just a few here and there. OE manufacturers spend weeks doing this and I've watched it take over a week on an F1 engine too.

GC
Last edited by Guy Croft on August 13th, 2006, 10:03 am, edited 2 times in total.

Testament
Posts: 101
Joined: June 22nd, 2006, 7:47 pm
Location: Taupo, New Zealand
Contact:

Post by Testament » August 13th, 2006, 9:21 am

PWM: Pulse Width Modulation
I guess when people think of MAF sensors they think of old flapper vane types which were quite restrictive. The Hot wire types are not so bad, although they have some limitation in that they are going to measure flow well over a certain range.

Rich Ellingham
Posts: 118
Joined: June 23rd, 2006, 6:54 am
Location: Glasgow, UK
Contact:

Post by Rich Ellingham » August 13th, 2006, 4:13 pm

Don't worry about the Idle control valve this is an added complication. An after market ECu can control the idle with the throttle stop being adjusted and equal airflow through the bodies. The ECU then control spark advance to maintain idle speeds. This is how mine runs at the moment and it idles ok. I would imagine that idle condition requires setting up without the inlet box that would be required for the forced induction. So a removable construction could be advisable.

Incidently, jenvey bodies are not tested for pressurised conditions, thus it would be prudent to devise a pressure testing method off the car to see if pressure is lost through the spindles. I have heard of a number of DIY solutions to leaking bodies on forced induction cars.

Rich
book 38

ronan
Posts: 3
Joined: August 12th, 2006, 11:46 pm
Location: Wells, England
Contact:

Post by ronan » August 13th, 2006, 4:56 pm

After reading this post I thought I might mention my experience with hooking up the brake servo vacuum to only one throttle. Under heavy and prolonged braking I noticed a definite pulsing through the brake pedal as the cylinder produced vacuum. It was not noticable under normal driving however I did not like it.

Guy Croft
Site Admin
Posts: 5031
Joined: June 18th, 2006, 9:31 am
Location: Lincoln, UK
Contact:

Post by Guy Croft » August 13th, 2006, 5:16 pm

A fair point Ronan,

no-one can argue with your particular experience, though I will add that every inlet manifold I sold under my name since 1988 (hundreds) carried one servo take-off from one inlet tract. This pulsing is a problem that can come from a worn servo diaphram.

With respect, this being an engine forum and in order to keep on topic, I would simply say that in my view a car of this type would be better served by having twin master cylinders and adjustable brake bias, AP Racing brakes, not a servo in sight.

Any views to the contrary feel free to start a NEW thread in this section. Don't have a problem with non-engine issues at all, as the race seat discussion shows.

GC

Evodelta

Post by Evodelta » August 13th, 2006, 5:44 pm

Thanks again for the replies, it is now looking like ditching the ICV is probably the way to go and it seems that the servo take off point is sorted out too, if one take off point is not enough then another can be added relatively easily. I would have thought that the car already has a one way valve fitted, but again, a point worthy of note.
Still not 100% satisfied on the map sensor problem, but like I mentioned earlier, it may be a case of trying this one out and seeing how it reacts.

Rich: Jenveys can be used with boost pressures of up to 6 bar (!), although you need to check with them if you are using this on some models, otherwise 2.5 bar is not a problem for them.

The car will be running MoTeC, I already have it and a base map which runs the car, although it hasn't been fully mapped yet in it's current state of tune which I expect will be 300-350 bhp.

AP brakes and bias adjuster? Yep, had them in place for a while, no point in going fast if you can't stop! The pedal box and twin cylinders will have to wait for a while yet.
I'll get round to doing a full write up on the car in the relevant section (which I should have done already :oops: ) So you can answer the questions more accurately.

Martin.

Rich Ellingham
Posts: 118
Joined: June 23rd, 2006, 6:54 am
Location: Glasgow, UK
Contact:

Post by Rich Ellingham » August 13th, 2006, 9:13 pm

Martin, the chaps at mech repairs just mentioned it to me as we were talking about things, they said they had experienced blow by if you like on certain bodies. If you go back a few PPC's you'll see a chap who had to seal the jenvey badge holes to stop pressure leaking - not an obvious problem. Not hard to test with a bit of plate and an air line I'd guess.

rich
book 38

Evodelta

Post by Evodelta » August 21st, 2006, 6:04 pm

Dynodave wrote:
I wouldn't have bothered trying to fit multiple throttle bodies , I would have fitted a larger single std type (personal opinion).

Dave
Thanks for the answer Dave, can you explain a little more about why you would opt for the larger single type?
The problems with the set up already fitted is that it is an all in one piece item with the inlet runners and plenum all part of the same casting - impossible to get in to tidy it up and improve the flow, to do this would mean sending it off to Australia or the US for 'Extrude Honing'.
The OE design is also shaped like a bunch of bananas, which doesn't look too good from a flow perspective.

Sadly the only way to know if I am to get any gains (airflow ) is to build the thing and then test it against the original one.

Thanks for your comments also Rich, I just went by what I read on their website, of course, practical experience is very valuable.

Martin.

ACE-GREG
Posts: 8
Joined: June 24th, 2006, 8:30 am
Location: Piraeus GREECE
Contact:

Post by ACE-GREG » August 22nd, 2006, 4:15 pm

Evodelta,

Using multiple throttles at any engine combination forget the MAP based calibration of your efi. You have to use TPS only at atmo engines or TPS/MAP based calibration for forced induction. TPS is used for fueling and TPS plus MAP for ignition setup. My personal advice is to use a larger single throttle body and not a multi throttle setup.
Also, as mentioned before there is no need for the idle valve.

Greg

Guy Croft
Site Admin
Posts: 5031
Joined: June 18th, 2006, 9:31 am
Location: Lincoln, UK
Contact:

Post by Guy Croft » August 22nd, 2006, 6:19 pm

Good to see you again Greg.

GC

Post Reply

Who is online

Users browsing this forum: No registered users and 2 guests