Chasing engine hesitation

[PeachyMomo]

Hi all, I know there are a few threads on engine hesitation here but I couldn't find the exact right issue.

The gist of it is after car comes to temp it occasionally finds itself lacking power around 2-3k. You can get the pedal at like 30-50% throttle and the revs/power just won't come for a few seconds. I typically let off and then jab it past to like 70ish% throttle and then it responds again. The problem is intermittent though and before I go spending money I was hoping to get a second opinion.

Things I was initially considering:
Put in a fuel pressure gauge and check the fuel pump. -- Im the 4th owner not sure if the pump has ever been replaced or not yet
Check the main relay, I've heard some people just slap the area the relays in and have had some luck getting it to work. I haven't personally tried it but was considering poking around with it and checking to see if anything was awry there.

Before I do those two things I did pull some data out of the OBD sensor and did see that my bank 1 voltage was being ridiculous at some points
o2 bank 1 sensor 1 was stuck pretty high and o2 bank 2 sensor 2 was stuck pretty low during one of the occurrences. Now I don't think my catalytic convertor is as good as these charts make it out to be. So I'm planning on changing these o2 sensors first.

But my main question is can a bad fuel pump result in these voltages I'm seeing? (Bank 2 was not behaving this way, so I am leaning toward o2 sensor replacement first)
Would rather not spend the money on parts that don't need a change, thanks in advance

 

[MotorMouth93]

Assuming properly functioning sensors, a narrowband reading of sustained ~1 volt would indicate a rich condition which is rather unlikely in the case of a bad fuel pump, which would typically create a lean condition when it can't maintain rail pressure.

If the sensor is going bad and pegging rich then the ECU will try to pull fuel to get it back to stoich which will result in hesitation as the fuel mix goes very lean. Stomping the throttle at that point would put the ECU into open loop fueling mode and would cause it to seemingly recover.

Without additional context such as tps, aps, fuel trims, or load though its difficult to really know whats going on based solely on narrowband voltage readings.

 

[PeachyMomo]

Thanks for quick reply.
Fuel mix going lean sounds right. I have those other data points collected too for the same time frame.

Does it make sense for the o2 issue to be intermittent though? Another thing of note is even when the car seems to respond well (subjective, response puts smile on my face) the long term fuel trim still sits around -13 (that data isn't in these charts)

 

[Honcho]

How many miles/years on the O2 sensors? What year NSX?

 

[PeachyMomo]

95'
o2 sensors I am not 100% sure. I replaced two of them in 2019 because I was actually getting CEL and codes for them. I cant remember which ones they were but I found part numbers in an old email . Denso 234-4093 and 234-4065, it looks like they were both the upstream sensors.

Not sure on how old the downstream sensors are, again didn't jump the gun on replacing since this hesitation isn't throwing any codes to suggest they needed to be replaced

Also spark was checked every which way, plugs clean, new coil packs 2022 new igniter 2022, no water sitting under the cover either

 

[MotorMouth93]

Your fuel trim logs are pretty far in the negative, meaning the ECU is reducing fuel quite a bit which is pretty strange. Are both banks like this or just one? If its only one then I might start to suspect an o2 sensor issue.

 

[Bram]

Have a look at the IAT sensor.....these can cause drivable issues

easy to clean too

 

[Old Guy]

The fuel pressure test ( gauge installed at the fuel filter) as described in the service manual is useful for diagnosing a serios problem with the fuel pump or the fuel pressure regulator. If it is a problem with inadequate delivery under high load the service manual test will likely not catch that. You would need a fuel pressure sensor and a logger to identify that as a problem.

If the engine is running and you are not getting complete power interruption (electrical interruption) it is probably not (I no longer say never to anything) the main EFI relay.

Because it is relatively easy to do, examine the fuel pump resistor circuit. The wiring and the resistor can get cooked leading to erratic problems. There can also be heat damage to the resistor by-pass relay which results in erratic operation. The by-pass relay is called the fuel pump relay by Honda even though it does not control the fuel pump - unfortunate terminology. As a test, completely by-pass the fuel pump resistor (the paper clip in the connector plug test). If this resolves your problem you could have:
- a fuel pump that is getting weak
- a failed resistor by-pass relay
-a flakey fuel pump resistor
- plugged fuel filter

That said, as noted, negative long term trim is not consistent with under delivery of fuel. Having both front and rear bank values would be useful to confirm that you have a consistent problem. Right now, you are just looking at 1/2 of the engine.

How are your injectors? An injector that is stuck open or has serious drool could generate a lot of negative fuel trim and explain a front - back discrepancy.

A flakey IAT sensor will cause fueling errors. However, the density correction for air temperature is relatively small unless the IAT has completely crapped out which would normally generate an OBD error. A flakey IAT will cause consistent problems on both the front and back banks. That said, removing and cleaning the IAT will not be a bad thing. Heads up - I think it is retained by a JIS screw not a phillips screw so use the right screwdriver to remove it.

Secondary O2 sensor failure will generate OBD error codes; but, will not affect engine operation. The secondary sensor is there for emission monitoring purposes. Your bank 1 sensor 1 voltage appears to be swinging full range which suggests that it might be just fine. When narrow band sensors die they typically read lean / low volts all the time. What is your bank 2 sensor 1 doing? Is it the same or significantly different?

It is possible that a sensor is getting slow to respond without failing. It is normal for the sensor 1 to swing lean to rich when in closed loop operation. Without knowing the operating conditions of the engine, my gut reaction is that your rich - lean swing rate on bank 1 sensor 1 might be slow which 'might' be a sensor ageing problem. Do a Google search on Diagnosing O2 Sensor Voltage Fluctuation and you can get some hits describing what is normal for that front sesnor voltage.

 

[asylum]

Had same symptoms. It was the O2 sensors.

Disconnect them. You will get a code and your car will run in loop mode (richer for safety). But if problem disapears, O2 sensors are the problem.

 

[PeachyMomo]

Sorry for going dark on this. Attached are the other bank at the same time points (look a-okay to me) and the IAT sensor temps(also look okay but Ill probably give it a quick clean next time I can anyways). Injectors I cleaned once a few years back and then a shop I had to do water pump, timing belt and usual things also cleaned (problem was still happening then)

The cars sitting in storage right now, but the North east has been getting some nicer weather so might take it out soon enough and get to wrenching.

Think current plan of attack is

1. Drive and when I get the issue attempt to unplug o2 without burning myself
2. paper clip trick for fuel pump resistor circuit and see if issue stops
3. fuel pressure gauge and try to pass a meter into the cabin so I can check it as I drive

 

[PeachyMomo]

Also I recognize all the heavy hitter names replying, just wanted to say thank you for taking the time. First car I bought with my own money and its been great working on it with my dad. Wouldn't be possible without all the knowledge ya'll have dropped on this forum over the years.

 

[Old Guy]

What really sticks out for me is that the long term fuel trim on bank 1 looks quite high while the bank 2 long term fuel trim looks closer to what I would characterize as normal. Bank 1 short term trims were also running really negative. If the bank 1 and 2 fuel trim curves represent the same operating snap shot I am inclined to say that if something has gone wrong, it has gone wrong on bank 1.

Fuel pressure / delivery problems should affect both banks equally. That does not appear to be happening so I don't think you have a fuel pressure problem. Also, as has been noted your O2 sensors say its running rich on bank 1, not lean. Based upon that empirical evidence problems with the fuel pump, fuel pressure regulator or fuel pump resistor is not where I would start my search. Problems with the IAT will also affect both banks so I don't think the IAT is the problem.

If the car is in storage with the battery disconnected the long term fuel trims will be erased. Pay attention to how the car drives immediately after coming out of storage with what should be close to zero trim values. Once it is up to operating temperature, if you can, monitor the fuel trims as you drive . If the car initially drives better; but, issues appear as the long term fuel trim starts to accumulate the problem may be the bank 1 sensor 1 giving you a bad reading, although sensors reading rich is not how they normally fail. If the car initially drives poorly and improves as the fuel trim accumulates then I am going to come back to a fuel over delivery problem on bank 1.

If after driving for a while the car is back to accumulating high long term trim on bank 1, I suggest that you go 'old school' and pull the spark plugs. If the bank 1 spark plugs look brown compared to the bank 2 plugs then you have a fuel over delivery problem that is just on bank 1. If you have a single bank 1 plug looking really dirty and the other two are really white the I will bet a beer that you have a leaking injector on that cylinder. You can also do the smell the plug test or if it is visibly wet that pretty much says leaking injector

If you can, monitor the fuel trims under two conditions after the engine has gone into closed loop operation. Monitor the trim as you allow the engine to idle for a long time. Then monitor the trim as you go for a trip at a steady fast highway speed. Leaking injectors cause a larger error at idle because the injector should be closed for longer periods. At higher engine loads the injectors are naturally open longer and the leak will be less significant in terms of how it affects fuel trim. If trim is high at idle and then drops under engine load, that tends to point you in the direction of an injector problem.

 

[PeachyMomo]

Sorry wanted to update - I have no idea why but the cars been running great with no investigation done by me after I took it out of winter storage. I'm just enjoying it for the time being instead of tinkering and chasing perfection. If the problem does surface again I will try and post whatever I may find out the cause to be for posterity

 

[nkgosselin]

I just want to say I had the same symptoms on my 91 with 132k miles. Replaced 02 sensors (that's what blink code said), IAT, cleaned throttle body, tested TPS. And nothing made a difference. I replaced the fuel pump and wow. Night and day. I think the pump has been failing for like 3 years. The symptoms came and went. Some rides good. Some bad. But immediately the car idled better, responded better. I imagine if you have an OE pump in a 3 decade old car, many will need replacing.

 

[Chris_Lum]

I just want to say I had the same symptoms on my 91 with 132k miles. Replaced 02 sensors (that's what blink code said), IAT, cleaned throttle body, tested TPS. And nothing made a difference. I replaced the fuel pump and wow. Night and day. I think the pump has been failing for like 3 years. The symptoms came and went. Some rides good. Some bad. But immediately the car idled better, responded better. I imagine if you have an OE pump in a 3 decade old car, many will need replacing.

i think i need to replace my fuel pump too. which pump did you go with?

 

[MotorMouth93]

Unless you want to retune the ECU then go with a stock pump.

 

[nkgosselin]

i think i need to replace my fuel pump too. which pump did you go with?

The Walbro 255. I replaced all the fuel hoses while tank out too.

 

[MotorMouth93]

The Walbro 255 will substantially change your fuel pressure, which changes injector flow rates and dead times, resulting in the car not running properly unless you retune it, it can learn to compensate for this to some extent but the open loop behavior will not be correct. It also won't behave the same way as the stock pump when switching between the low and high speed modes. It's just kind of a crappy pump in general and is very noisy. Unless you want to make it a huge project, just go OEM.

 

[Honcho]

Listen to John. Go with the Denso unless you need something big for like a turbo system.

 

[Chris_Lum]

The Denso 951-0011 OE replacement pump appears to be discontinued. Is my only option buying the expensive OEM full fuel pump assembly?

 

[Chris_Lum]

Looks like Rock auto has a delphi unit i can buy..

 

[Honcho]

Looks like Rock auto has a delphi unit i can buy..

As long as it is built to the OEM spec, you should be ok with a mainstream brand like Delphi. I'd avoid the "economy" knockoffs- production quality could be an issue there.

Edit: BTW Chris, it looks like the Denso is still in stock at Amazon:

Amazon.com

 

[Chris_Lum]

Amazon is the first place i checked and it says it won't ship to my location, no matter what local address i put. I'm betting it's just not in stock. Any site i check says it's out of stock and the one seller on eBay who lists it has poor ratings due to not having stock on hand lol. I ordered the Delphi and will see how it goes!

 

[Old Guy]

Amazon.ca shows the Denso pump as being available while Amazon.com shows it as unavailable.

https://www.amazon.ca/Denso-951-0011-Fuel-Pump/dp/B000C5WFS8

If you are in the US and use a virtual IP you might be able to order it without Amazon.ca diverting you to the US site, or it may kick you out when you try to enter a US address. It is also possible that it may turn into a 'vapour-ware' pump as soon as you try to order it.

Denso de-listed the NSX specific pump package about 3 years ago. I picked up the pump from Rock Auto pre pandemic just to have as a spare. Shortly after I bought it Rock Auto listed it as unavailable. The actual number on the pump is 195130-1251. Lots of 195130 pumps around; but, the 1251 NSX variant truly appears to be history.

If you want OEM I think you are going to have to buy OEM which is not cheap because it includes the complete fuel pump set, not just the pump.

Got to pay to play!

 

[PeachyMomo]

I got a p1201 p1202 p1203 and p1300 today . I've changed igniters, coil packs, injectors, and fuel filter over the years so it looks like its fuel pump time for me too.

@Chris_Lum any luck getting yours? I'm about to place my order for the denso through amazon and looks like it should get delivered in a few weeks, DM me if you're still having trouble getting yours and I might be able to just order 2 of them