Stock Delorean throttle bodies - why so big?

[Spittybug]

I've been EFI'd for several years now and would never go back. When I built my setup I got rid of the stock intake manifold and throttle body and replaced them with a Peugeot one with an old Camaro throttle body mounted to it. Without much thought I made sure that I had a restrictor plate in place to ensure that the larger throttle opening got sized down to be equal with the stock one. Well, the car has never liked to be tuned in the 'conventional' manner that Megasquirt EFI likes to use; manifold vacuum levels. I have used the throttle position sensor instead with great success. (That's alpha-n versus speed density for those of you with a similar setup). I assumed it was me doing something wrong and just went with what worked well.

While interpreting some of the log files I created while re-tuning things after some post Harvey work I did (new fuel pump, new O2 sensor, etc..) I discovered that my manifold vacuum disappears very quickly - like 15% throttle and boom I go to near atmospheric in the manifold! There are no leaks or other issues at play here, just the amount of air coming in.

So rather than rely on JZ's throttle size, I pulled up an online calculator and plugged in our engine parameters. It called for a throttle size of 53mm (2185 sqmm). Our stock throttle body is DUAL 47 mms (3470 sqmm). That is a very large opening for a 2.85 liter engine. IIRC there was no restrictor plate of any sort downsteam of the throttle body in the stock setup, right? I can only assume that the huge intake manifold and its convoluted air path through the horns, into the upper chambers and then crossing over to the opposite side intakes must have something to do with the need for such a large bore. Or maybe it doesn't matter because of the use of the air flow meter for fueling rather than manifold vacuum to meter fuel. Maybe some of you old school engine guys can opine? This may well explain why some people have had difficulty getting their cars to run well with EFI when using the stock manifold and throttle body.

In any case the cure is now visible to me; a much more restrictive plate behind my throttle body will help maintain the vacuum levels in the engine properly and allow me to tune with the preferred method.

[dn010]

I would think it doesn't matter because the air is already being metered but I could be wrong. This is why I always thought it was nonsense when I read about the 'new' plastic throttle body spacers "improving performance" by being less restrictive and letting more air pass. I'm running the B280F, maybe I can see how large the throttle of that is tonight for comparison sake.

[David T]

A lot of factors come into play. Are you using a throttle body with fuel injection or are you injecting at each intake valve? What is the velocity of the air you want? The throttle body in the stock set-up was sized for an already mixed air-fuel mixture. What is the temperature (density) of the air you are controlling? If you are going to make changes you must re-engineer the whole system to work, not just a few pieces of it. You have to size the throttle body and the mass air flow sensor to the amount of air you intend to control to maintain the proper amount of control.

[dn010]

That's right, I forgot the fact that the B280Fs use MAF sensors so I guess the measurement won't matter. I think I'm the only one running MAF based EFI and it was a PITA trying to find a sensor that would be a good replacement for the B280Fs hot wire bosch maf.

[Spittybug]

A lot of factors come into play. Are you using a throttle body with fuel injection or are you injecting at each intake valve? What is the velocity of the air you want? The throttle body in the stock set-up was sized for an already mixed air-fuel mixture. What is the temperature (density) of the air you are controlling? If you are going to make changes you must re-engineer the whole system to work, not just a few pieces of it. You have to size the throttle body and the mass air flow sensor to the amount of air you intend to control to maintain the proper amount of control.

David, what I think you meant to say to not sound condescending was: "The Delorean was engineered for the whole system to work, not just a few pieces of it. They had to size the throttle body and the mass air flow sensor to the amount of air they intended to control to maintain the proper amount of control."

I'm sure you didn't intend to sound condescending about how my EFI setup, which doesn't use a MAF sensor and automatically adjusts for incoming air temperature, needs to be setup. It is running very well (as good as stock) using throttle position to control fueling. The fact remains that the engine bore, stroke and max power RPM level determine the necessary air flow when using modern day fuel injection when basing fueling on engine load not throttle position.

The stock setup indeed mixed the air with fuel prior to going through the throttle body. The Delorean engineering that you refer to takes into account such items as resistance of the metering plate, the long manifold lengths and the fact that the engine vacuum levels were indirectly used in controlling the fuel injection quantity (via the fueling plate, not directly off the manifold). Your lesson on engine design is best left for another day, thanks.

[David T]

I don't mean my posts to sound condescending but sometimes someone can take it that way I guess. When I post, I post not only to the OP but to the general audience. I am not familiar with each person's Delorean especially if they have modified it. Not everyone is familiar with what it takes to install an EFI set up and all of the considerations it takes to get it to work well. Sometimes I restate the obvious. It is obvious to those familiar with the topic, not everyone reading these post is as familiar with the topic as the OP may be.

[FABombjoy]

The decel valves probably reduce the effective TB size a little.

What issues have people had tuning with the stock intake?

[DARCOM]

I just finished my new setup with bigger throttle bodys. I'm liking them. They may not be practical for my car but i love the throttle response.

[Spittybug]

Don't you guys with stock intakes and big throttles have problems with vacuum drop off very quickly when you open the throttle? I had a restrictor plate in place to choke down to the equivalent of ~54.3 mm throttle body (stock being dual 47 mm) and I was having the issue that even a 12% throttle opening was making me go to just about atmospheric. Very hard to get fueling/timing right that was using speed density. I've been using alpha-n which listens to the throttle position and it has worked very well. Yesterday I made an incrementally smaller plate using an online calculator and our engine parameters as a guide. It is barely smaller; 53 mm total area, but that was enough for me to get a MUCH better MAP reading at more throttle. I did a quick run in the garage last night, RPM up to our max power range of 5500, and recorded the MAP. Stayed below 60kPa. That was with steady throttle opening, so punching it will of course rise higher.

This will allow me to give speed density tuning another shot. It is better at cruising ranges than alpha-n and since more people use it and are familiar with it I won't be the exception any longer. I don't know if my Peugeot manifold or the adapter on top somehow affect my setup materially differently than the stock setup, but I know other Delorean guys with stock intakes that have struggled with getting the tip in to not bog, especially with an automatic. I suspect too much air is allowing vacuum to totally disappear and making tuning very hard.

[dmc4087]

I just finished my new setup with bigger throttle bodys. I'm liking them. They may not be practical for my car but i love the throttle response.


That looks fantastic, what throttle bodies are you using?

[FABombjoy]

I can see how a larger throttle body can affect the granularity of the throttle input - turning the pedal into more of an on/off switch than a linear input. Unless the TB is comically oversized I would think you could account for that in the tune

I can have MAP >= 100 anywhere from 30% to 100% TPS depending on gear/boost targets/etc. I have speed density working perfectly - alpha-n would probably break my car.

Can you post logs of the stumble when using SD?

[Spittybug]

I'm not using SD so I don't have logs of the stumble.... can you post a log and your .msq to show me what kind of VE values you have in your low RPM, higher load bins? When I was using SD I had to put in numbers >100. That's not what it should take.

I'll send admin a request to move this thread to EFI since the conversation has migrated.

[FABombjoy]

I'm on MS3 and we're plumbed differently enough that my values probably won't be useful. I've intentionally messed with the minfuel to gain more resolution in the VE table so I'm 100+ above 85 kPa.

You were seeing lean stumble?

The numbers in VE should only affect steady state operation but it sounds like you're having trouble with transient response? I would think transient issues would relate almost exclusively to accel enrichment. That's assuming the MAP sample is quality with no issues regarding sample source, tubing diameter or length, latency due to excessive smoothing, etc.

[Spittybug]

I just went out on a run for roses (y'all do remember what day it is tomorrow, right?) and captured a log. MUCH better relationship between MAP reading and throttle. I don't go to barometric until I really give it good throttle. Since my timing is tied to MAP it now behaves better, making driving even better. I have my AE threshold too high to kick in except under the hardest of acceleration, so that is next to work on. MS has me a little confused here though.... the notes say that the AE table is a percentage adder to Reqfuel yet the scale of the graph is in 'added MS'. Looking at my log, at 576.788 sec, the TPSdot goes to 114.5, over my kick in threshold. The resulting accelPW is 7 ms, which on my table is supposed to be the added amount when TPSdot is >1000. So it looks like the adder must be a percentage but is labeled wrong. That 7 ms is ~50 of ReqFuel at that VE bin I think.... 14 ReqFuel + 7 AEFuel. Confusing to get right....

What kind of % adders do you use at different TPSdot or MAPdot values? That should be fairly consistent despite our differences.

[Spittybug]

I've confirmed that my MS2e uses an absolute ms adder for AE, not a percentage of ReqFuel. The tech tip that pops up is incorrect.

I'm therefore doing some scatter graphs to look at TPSdot and PW and AFR..... lots of interesting analysis to be done with Megalogviewer.

Since you have your AE as a % adder in MS3, can you tell me what your values are? Then I can translate into ms and compare.

Thanks.

[DARCOM]

That looks fantastic, what throttle bodies are you using?

There Nissan 300zx stock throttle bodys

[FABombjoy]

I'll admit I haven't spent a lot of time here as even the stock curve worked well enough. I've enriched things just a little bit versus stock but the last session I recorded shows I could use a little more.

https://imgur.com/a/OEM4mnE

Are you running ignition control or fuel only

[FABombjoy]

After finding some free time for a few thoughts and image searching LT1 throttle bodies, I think you're seeing the effects of going to both a larger TB and a linear pulley.

The stock DMC throttle arrangement has a fairly nonlinear relationship to the pedal. I haven't measured it but I'd guess the first 10-20% of travel opens the throttle much more slowly than the remaining travel.

"Please excuse the crudity of this model": https://imgur.com/a/0y18kPr

After looking at a few LT1 TBs they appear to have a linear spool, meaning that cable travel is now directly proportional to throttle opening.

[Spittybug]

The larger body was definitely an issue that is now resolved with a slightly smaller restriction plate than what I was using. The value from the online calculator I referenced has done the trick. My throttle body has an elongated cam and another key to getting the better throttle response has been to reposition the throttle cable to follow more of the cam.

MS2e does indeed use an adder in ms for AE enhancement, not % of ReqFuel. It's also picky in what order you define the table... mine was high to low and it wasn't reading it right. That is now corrected and I add between 2ms and 7ms depending on TPSdot.

Took if for a test drive and am very pleased indeed! Very responsive. Good acceleration. Evaluating the logs shows there are a couple of lean spikes I need to massage, but overall very good AFRs in the 13 - 14 range. I'm still on alpha-n since I don't have a problem with it, but I will independently start working on a speed density tune so that I can be more consistent with what others use to make comparisons easier.

I love being able to completely change the car by just loading up different tunes. I could never go back to mechanical fueling or spark again.

The attached pics show the throttle body cam, front view, and view through the Lucite engine cover (shame to not show off the engine after all).

[DARCOM]

Looks like you have 650 cfm throttle body on there. My v8 Z28 had one. I think that's double what you need for a prv. My Nissan ones can only do 350 cfm. Im not sure what the stock Delorean throttle body flows but i know its less. If your at half throttle or less your probably at max cfm the Delorean can consume and with a restriction plate it would be like your at wide open at less than half throttle.

[Spittybug]

I've totally converted over to speed density rather than alpha-n and have no issues at all. The extra definition afforded by simply reducing the effective area of the throttle body did the trick. It's driving better than ever with no sluggish spots or stumbles. Closed loop idle compensates for A/C kicking on and off nicely. I need to go back in and wire up the buzzer (totally forgot about it) to act as a warning for any overheating should I lose a hose or anything. Easy safety!

While I like the look of Darcom's dual throttle bodies and keeping the long runners of the stock intake (better theoretical power but is it noticeable in actual use?), I have to believe that routing cooler intake air might be better. Looks bad ass!

[nkemp]

I know I'm a bit late to this party but .... I was reading your post http://dmctalk.org/showthread.php?7533-Throttle-body-questions&p=110541&viewfull=1#post110541 shortly after reading:


Thegeneral rule is: the bigger the size, the higher the RPM needed for a kick ofpower. If your new throttle body is too big your vehicle will feel tired atlower RPMs, while really taking off powerfully at higher RPMs (when theengine’s demand of air flow matches up with the the internal bore’s biggersize).


From <https://blog.tdotperformance.ca/guides/how-to-buy-the-best-performance-throttle-body/>


Sounds like the original problem.


But it still begs the original question for this thread. I have three PRV's (last two on engine stands):
- The original in 897: 2 x 46 mm … an area of 3,324 mm2 each, 6,647mm2 total
- '92 Eagle Premier 3.0L: 1x50mm, 150bhp … A≈ 1,964 mm2
- '88 Peugeot 505 2.8L EFI, stock intake manifold same as 3.0L. 145bhp: 1x55mm … A= 2,376 mm2

In the above referenced post you said:
optimal throttle body diameter of 52.6 mm (2176 mm2 area)



Which is pretty close to the 505 2.8L with the newer intake manifold and has a 55 mm TB on it with 2,376mm2

And yet the 3.0L has only 1,964mm2

So why are the originals so big? The big honkin DeLorean intake manifold (area, volume, length, all)? It runs reasonably well theoretically oversized.

What drives OEM sizing? Or TB size in general?

[nkemp]

I think the answer to the original question involves running the equation as a 3 cylinder engine to determine each TB size. In essence, our airflow with the dual TBs is two isolated engines with 3 cylinders each. The air after the TB is isolated from the other. The lower profile intake manifold does not split out the two sides so it gets treated as a 6 cylinder engine. Air is common after the TB in the new intake. And that is likely why the need to run such smaller TBs on the new intake manifold.

If I ran the numbers right, the diameter for each TB in the dual setup is 34.69 mm. That is less overkill than running it as a 6 cylinder.

I was almost asleep when this came to me, and I'm sleepy as I write this. So feel free to give it a dose of reality should it be wrong:).

And it still seems that the 3.0L TB should be bigger than the 2.8L TB even though they are not that way on the two extra engines I have.

Now back to :sleep:

[nkemp]

This post is to correct a math error for any future readers.

Years later I spotted an error in my post. Even with the corrected numbers, the question i presented still stands. Why so big?

Originally I said "3,324 mm2 each" when that should have been the 3,324 mm2 total.


But it still begs the original question for this thread. I have three PRV's (last two on engine stands):
- The original in 897: 2 x 46 mm ? an area of 3,324 mm2 each, 6,647mm2 total
- '92 Eagle Premier 3.0L: 1x50mm, 150bhp ? A≈ 1,964 mm2
- '88 Peugeot 505 2.8L EFI, stock intake manifold same as 3.0L. 145bhp: 1x55mm ? A= 2,376 mm2



It should read as:
- The original in 897: 2 x 46 mm ? an area of 1662 mm2 each, 3,324 mm2 total
- '92 Eagle Premier 3.0L: 1x50mm, 150bhp … A≈ 1,964 mm2
- '88 Peugeot 505 2.8L EFI, stock intake manifold same as 3.0L. 145bhp: 1x55mm … A= 2,376 mm2

[FABombjoy]

My total guess, could be way off. Would be curious if someone knows for sure:

The K-jet metering plate is a 2nd upstream throttle, the opening of which is determined by airflow & fuel pressure.
The primary cable-driven throttle may have to be larger to improve throttle response, so the k-jet plate can react more quickly?

[nkemp]

For the nuts of it, I measured the hole the metering plate sits in, The area of the throat is 5026mm2. The throttle body area is about 1684mm2 less than the metering plate hole.