TPS relay trick (REALLY LONG)

[rjaffey]

How to modify the TPS circuit on a Honda for use with forced induction.

Let me preface this by saying the modification discussed here is purely experimental and there's no way I can account for differences between cars, years, models, or technical skills of the installers. I'm also not certain of all the effects which the modification may cause including changes in exhaust emissions, nor do I claim to be an expert on how the electronic engine management systems used on Hondas work. I can only relay my observations so far (I've driven hundreds of miles since making the modification) and reiterate the fact that it's based on a suggestion from JeffM of ZDYNE (see: http://www.hostboard.com/ubb/Forum40/HTML/002576.html ) and it makes sense that it should work as expected. Note that the information provided here applies to '94 GSRs with JRSCs specifically, but may apply to many other models, years and other forced induction systems as well.

Anyone attempting to use the information that follows should have a Helms manual or some other reliable documentation which includes electrical diagrams that cover the TPS wiring, should feel comfortable making slight modifications to the wiring in their car, and should have at least a basic understanding of electronics as I make no attempt to explain how relays or resistors work, nor do I get into specifics as to where or how to mount components, etc.

I certainly don't believe you can harm your car by making this modification, but I can not be responsible for *ANY* problems arising from the use of the information I'm posting here. I also cannot guarantee that the information below is entirely accurate, but I have made every effort to ensure that it is free of errors. That said, here's the idea:

As JeffM states, the ECU depends on the throttle position to tell it when to switch into or out of closed loop operation with the Oxygen sensor. When the throttle is only partially open, from idle to moderate acceleration, the ECU operates in closed loop with the Oxygen sensor. In this mode the ECU tries to maintain a 14.7:1 air to fuel ratio that provides good fuel economy and minimal emissions, but this mixture is too lean for high engine load when the throttle is wide open or when boost is present. Honda's factory engine management relies on throttle position to determine when load is high enough to switch out of closed loop, but this doesn't work very well when forced induction is added where significant boost is present with only partial throttle opening. The result is a tendancy for audible ping and/or hesitation on tip in, and poor performance with part throttle opening when there's boost but the ECU is still trying to maintain the 14.7:1 air to fuel ratio.

Before you continue reading, check out the two schematics at the URL:

http://www.artic.edu/~rjaffey/tps/relay-trick.gif

The upper schematic represents the TPS sensor on a '94 GSR. It's the same on '95 GSR and probably on many other recent Hondas and Acuras, but check your documentation to be sure.

The lower schematic represents the TPS sensor with the relay "trick" incorporated. I'll get to it after a discussion of the unmodified TPS sensor wiring.

Let's look at how the TPS works in a nutshell. Referring to the top schematic you can see the TPS has 3 terminals:

The bottom terminal is grounded via the green/blue wire.

The top terminal is connected to +5 volts coming from the ECU via the yellow/blue wire.

The center terminal supplies signal as a voltage back to the ECU via the red/black wire. The signal voltage ranges between 0.5 at idle and 4.5 volts at WOT depending on the throttle position. This signal voltage tells the ECU what the throttle position is. For those folks using J&S Safeguards, the signal also tells the J&S when the throttle is pressed down enough that it should "arm" meaning it should start displaying and reacting to knock should it occur.

Now for the modification:

The idea is to make the ECU switch out of closed loop operation and to arm the J&S Safeguard if there's one installed, when there's boost present. This makes it possible, through the use of electronic fuel controllers and/or boost dependant fuel pressure regulators, to obtain a more desirable air to fuel ratio for running with boost even when the throttle is only partially open. Were the ECU to stay in closed loop operation, it would counteract such attempts. Arming the J&S Safeguard at the same time causes it react to knock (if present) as soon as there's boost.

The bottom schematic from the URL mentioned above includes the addition of a relay and two resistors to the TPS circuit. At the point between the two resistors there is roughly 4.5 volts; what the ECU would see were the throttle wide open. The "trick" is to route the real TPS signal to the ECU and J&S only when there isn't boost and the 4.5 volts to the ECU and J&S when there is boost. That's where the relay (and pressure switch) comes in.

The details:

The wiring and component layout is beyond the scope of what I'll cover here. Just be sure the connections are good, secure, and protected, the wiring has enough slack to allow for engine movement, everything is well insulated, and the components are securely mounted.

- The required components are all available from Radio Shack at http://www.radioshack.com . The relay I used is not available in their stores, only through their online catalog:

(1) SPDT (Single Pole Double Throw) automotive relay and socket.
(1) 22K Ohm resistor, 5% or better tolerance.
(1) 2.2K Ohm resistor, 5% or better tolerance.
18 gauge stranded wire
electrical tape
assorted crimp on wiring taps and connectors for making connections

- Required tools:

Wire cutters
Wire strippers
Pliers

- Optional, but highly recommended components:

zip ties
wire loom (that plastic tubing like stuff used for sheathing wiring under the hood.)

- Optional, but highly recommended tools:

volt meter



- If you've got an anti-theft radio, be sure you have your radio code.

- Disconnect the negative battery terminal.

Using the schematic as a guide:

- The 22K Ohm resistor taps into the green/blue TPS wire (ground).

- The 2.2K Ohm resistor taps into the yellow/blue TPS wire (+5 volts).

- The free ends of the two resistors are connected together.

- The red/black TPS wire (TPS signal to ECU) is cut. If you're using a J&S Safeguard, I recommend you cut the wire between the TPS and the J&S (folks who connected their J&S units to the wiring by the ECU need not worry about this). This way you'll have one end of the red/black wire connecting to the TPS alone and the other end of the red/black wire running to the J&S and ECU, so the J&S will be tricked into arming as soon as there's boost in the same way the ECU will be tricked into switching out of closed loop operation. The end of the red/black wire that runs to the ECU (and J&S if so equipped) is connected to the center (common) pole on the relay, as it will be switched to either the TPS itself or the constant 4.5 volt source (between the two resistors) depending on whether or not there's significant boost. The end of the red/black wire which runs to the center terminal on the TPS is connected to the pole on the relay, to which the center relay pole connects when there's current flowing through the relay coil. This way the ECU sees the real TPS signal when there's current running through the relay coil (when the pressure switch is closed as there is less than 1psi of boost).

- The 4.5 volt source between the two resistors is connected to the remaining pole on the relay, to which the center relay pole connects when there's no current flowing through the relay coil, so that the ECU (and J&S if so equipped) sees a 4.5 volt TPS signal when there's no current running through the relay coil (when the pressure switch is open as there is 1psi or more of boost).

- One side of the relay coil is connected to a +12 volt source controlled by the ignition switch. The PCS (Purge Cut Solenoid), where the JR enrichment relay coil is connected, is a perfectly good source.

- The other side of the relay coil is connected to the pressure switch, where the other lead from the JR enrichment relay coil is also connected.

- Before you insulate everything, you might want to use a volt meter to verify that things work as expected. Be sure no exposed wires, contacts, etc. are shorting or grounded. Reconnect the negative battery terminal. Switch the ignition to the on position and measure the voltage between the yellow/blue (+5) and the green/blue (ground) wires. It should be about 5 volts. Measure the voltage between the green/blue (ground) wire and the red/black (TPS signal) wire that runs to the ECU (and J&S). It should be about 0.5 volts. Leaving the volt meter connected to ground and TPS signal to ECU (and J&S) wires, move the throttle by hand from closed (idle) to wide open. The voltage should climb from 0.5 volts up to about 4.5 volts. Let the throttle close and the voltage reading should drop back to 0.5 volts. Disconnect the relay from the pressure switch and the voltage should instantly jump to about 4.5 volts. Reconnect the relay to the pressure switch and the voltage should instantly drop down to about 0.5 volts. If you don't see this behavior, something's wrong with your installation or your car's different. Check/correct your work and test again.

- If you've done the testing in the previous step, shut off the ignition. You might also want to disconnect the negative battery terminal again if you want to be extra cautious.

- Insulate/wrap the wiring, etc.

- After everything is wrapped, secured, insulated, etc. reconnect the negative battery terminal and you're good to go. The car should run normally, the air fuel gauge should hunt back and forth while idling, cruising on flat road, or even under very light acceleration, but the instant you put your foot down enough for the boost to kick in, the air fuel gauge should go rich. The tip in performance should be very crisp as the lean spot that previously lasted 1/2 second or more should be gone as can be seen with the air fuel guage.

Notes:

- The values of the two resistors were selected based on measurements I took on my car. The TPS has a 5K Ohm resistance between the top and bottom terminals. The 2.2K and 22K Ohm resistors supply 4.5 volts without increasing the load on the +5 volt supply enough to effect the supply voltage, while still supplying enough current to the ECU and J&S TPS signal inputs. I assume (and it is just that, an assumption) that these values will likely work on a range of years and models, but you may want to take your own measurements to be sure the same is true on your car.

- I have no idea how this will work on an OBD II car. I've only tried this on my car and it's OBD I. *** See update at the end of this post ***

- The benefits of this modification are greater for folks running higher boost.

- I've got a JRSC, but there's no reason to believe this can't work with a turbo. However, I haven't heard of folks with turbos describing the problem which this mod is intended to solve.

- Folks with adjustable rate regulators such as a Cartech 20005 might try lowering the baseline fuel pressure and increasing the rise rate, so fuel pressure up to zero vaccuum is lower, while the peak pressure at full boost is where it was before. This is possible as not as big of a head start is required to?prevent a lean condition when the transition from off boost to boost is made. I'd be interested in hearing what you folks find out. I've tried it and seem to be able to run lower baseline pressure without ill effects.

- This modification requires a pressure switch tapped into the intake system between the blower (or turbo) and the engine. The JRSC systems I've installed and the ones I've inspected all have pressure switches coupled with relays to trick the ECU into seeing very cold intake temps resulting in more fuel delivery, however, some older Civic systems may not use a pressure switch.

- The pressure switch has to open at or around 1psi or less for this to work as expected. 1psi is not very significant and I've seen no problem resulting from closed loop operation between 0 and 1psi of boost. If the switch doesn't open until significant boost is present, this modification won't work. I'm pretty sure all of the "new" style pressure switches JR supplied open at 1psi. Rumor has it that the "old" style pressure switches JR supplied were not consistant as to what pressure was required to open them and that this is why JR sent replacement pressure switches to folks who had the "old" style.


Possible problems that come to mind:

- The modification may affect when and how an automatic transmission shifts as the TPS signal is monitored by the TCM (Transmission Control Module) on automatic transmission equipped cars. I'm not sure whether the affect would be desireable as I haven't tried it. If any of you with automatic transmission equipped cars are thinking about making this modification, you may need to do more extensive wiring mods to make sure the TCM always sees the true TPS signal or it may just be a good idea for you to skip doing this mod altogether.

- As the relay is a break-before-make switch, there's a moment (a matter of milliseconds) when the TPS signal is disconnected altogether while the relay is changing state. So far this has not been a problem on my car, as there have been no MIL codes and no strange behavior. If it turns out to be a problem on my car, the addition of a small capacitor should solve it. I'll likely add it to the circuit and the posted schematic if necessary.

- Some folks have reported that their JR pressure switches flip on and off rapidly under some conditions (I haven't seen this personally and have only heard it reported by two or three folks on this board). I have no idea how this would affect performance, etc., and I'm not seeing it on my car, so I can't comment other than to say it might cause problems for folks with JR pressure switches so afflicted.

Update 12/7/00: Quite a few folks with OBD II cars have made this modification without problems. A search of this forum will yield posts from some of these individuals.


If anyone spots any mistakes in here, please let me know!

[This message has been edited by rjaffey (edited October 24, 2000).]

[This message has been edited by rjaffey (edited December 07, 2000).]

[BlownSOL]

Raff, thank you, a true masterpiece. This post needs to go in the FI board Hall of Fame

My setup is an OBD-1 1.6 sohc, should I go ahead and order those parts(resistors etc.) from the Shack or do I need to confirm my voltage values? I have a voltmeter but the car is not running at the time. Guess I could trust my helms too.

AL

[ringer]

What about doing something similar (modifying the voltage) with the MAP sensor??

CHeck this: http://www.overboost.com/story.asp?id=257

I think most of us already know a lot of the info on that page but one thing struck my eye. In reference to the MAP: "...This sensor has the largest bearing on how much fuel is delivered...".

So how about it? Am I missing something obvious? I know you can't fool it too much or you get a CE code.

P.S. The GIF file came up OK for me.

[This message has been edited by ringer (edited October 23, 2000).]

[rjaffey]

BlownSOL,

I would suggest you verify voltages before you start, after connecting the two resistors, and then after wiring in the relay. This way you'll be sure.

nexusone,

I'd rather not deal with the capacitor issue just now as I'm not seeing any problem with the circuit as described and I don't have convenient access to the equipment I'd need to verify that I've selected a capacitor that addresses the problem correctly.

As to the part numbers, I think these are the right ones:

Relay: 900-2394
Relay socket: 900-2396

ringer,

I think that messing with the MAP signal is something you do after the ECU has switched out of closed loop operation. It's a valid place to tune fuel, in fact that's what some electronic fuel controllers do, but I don't know that it can address the problem I'm solving here, or at least not as simply.

[This message has been edited by rjaffey (edited October 23, 2000).]

[beekayITR]

Just curious...but can't you just arm the J&S constantly by not connecting the J&S to the TPS.

That way you don't have to worry about throttle position.

[Holden]

Is it just me or does Radio Shack's site suck?

That was a painful experience. It took me 10 minutes to figure out the part numbers (this was before Rafe posted them - I got 'em right!).

In case anyone was wondering, I believe

22K Ohm 1/2W 5% Carbon Film Resistor
$0.09 Reg. Price
Cat.#: 900-0411

2k2 Ohm 1/2W 5% Carbon Film Resistor
$0.09 Reg. Price
Cat.#: 900-0387

are the correct resistors.

Greg

[Batoutahell]

Bravissimo! Not that it would stop me from trying out this trick, but does it affect gas mileage? Time to visit Radio Smack.

[nexusone]

Would you mind adjusting your schematic to include the capacitor? If I'm going to make one of these, I might as well include the capacitor now also. Seems like it cant hurt to maintain that line voltage during relay switching.

nice work!

------------------
'98 GSR JRSC with all sorts of high-performance, high-tech, overpriced stuff packed in all over the place.
http://www.donia.org/clandestine

[nexusone]

Also: can you include part numbers? there are several automotive SPDT relays on the radioshack site. I don't want to order the wrong one.



------------------
'98 GSR JRSC with all sorts of high-performance, high-tech, overpriced stuff packed in all over the place.
http://www.donia.org/clandestine

[Holden]

Thanks Rafe.

I will do this ASAP and will report on any OBD-II problems.

BTW, Am I the only one who can't get that GIF file?

Greg