Holley Sniper issues

[Allan Robinson]

Hi guys,
I’ve not been on this forum for a while so thought I would share my experience with the Holley Sniper EFI system which I installed in March 2021 and have run with over the following 3 summers.

I have a Dax Rush with a 4.6 RV8 and at that time an Edlebrock 1404 Performer carb. With the assistance of a vacuum gauge, AFR gauge and input from this forum I had it running perfectly but then I got bored and was temped to the dark side by a bright shiny new Holley Sniper system which I installed in March 2021.

Initially I was very impressed, it was so nice just to press the start button and it ran perfectly without me having to tease it into life. The improved throttle response and ease of tuning was also very impressive. Unfortunately after 1k miles my troubles started that continued to plagued me intermittently over the following 3 summers.

Now here’s the thing. If you have 1 fault you have a good chance of solving it. If you are experiencing 3 intermittent faults that present as one then you are in the lap of the gods.

Fault 1 – Intermittent high idle. 1200 rpm plus.
Fault 2 – Intermittent Loss of throttle control and rough running.
Fault 3 – Lambda sensors going faulty or becoming unstable within 1K miles.

Out of desperation I resorted to searching YouTube to find that many others were experiencing the same issues; unfortunately, in most cases gave no credible solution. Phoning Holley in the USA for a lengthy phone call was not really an option. Now after 3 years of wrestling with these issues I have resolved faults 1 and 2. Fault 3 is still ongoing.

Fault 1. The intermittent high idle was due to the secondary butterfly sticking slightly open. Whether this is due to some form of heat distortion of the throttle body or vacuum issue I really don’t know. The solution is to wind another turn on the secondary butterfly return spring. The down side is that the accelerator pedal becomes harder to press.

Fault 2. This fault was due to the Injector / injectors intermittently stop firing. Within the EFI unit there are 4 inline electrical connectors that have a dogtooth clasp holding them together. The problem is that the connector material is very soft and the retaining latch becomes ineffective. Subsequently they slowly come apart and initially cause all sorts of intermittent running problems until eventually its bad enough to witness no fuel spraying in one or more of the throttle bodies. The solution was to secure each connector together with a zip tie.

Fault 3 – I’m still struggling with this one.. I replaced 5 lambda sensors within 3500 mile and at circa £150 each is not sustainable. I have checked all the documented reasons as to why a sensor should fail but all appear good.

So to sum up

When the Sniper runs well it is very impressive but a nightmare when it doesn’t.

I eventually found the solutions to faults 1 and 2 on You Tube, They are issues that others have also experienced and as far as I can see are due to either poor design, poor material choice or poor assembly. Based on my experience I am very disappointed with Holley being such a well respected company. I can only hope that Holley have now rectified these issues, however, I remain skeptical.

I didn’t put the Dax on the road this year so resolving the lambda sensor failure is going to be the main focus for the summer of 2025. If I can’t resolve the sensor issue I will consider running in OPEN LOOP although this is not ideal or I will be reinstalling an Edlebrock carburetor.


Has anyone else on this forum installed a Holley Sniper system? How was your experience?

[GDCobra]

Nice write-up Allan, sorry to here of your issues but seems like you are on your way to a solution, also seems like most of the issues are due to product quality at Holley. Did I understand correctly that the connector issues you had were inside the ECU? Strange to have mechanical connectors inside a box particularly without a more secure means of maintaining contact, that's more like what you'd find in a DIY system (like something I'd do!).

Failing Lambda sensors is a bit odd, I'm assuming these are not Holley specific and is third party, something like Bosch or similar? Is it the actual sensor that is failing or the controller device? If the controller is this a Holley item or third party.

I know you say you've checked all possible causes but if it is the sensors themselves failing and you've potted 5 so far there must be an underlying problem such as:
Poor brand - Quality sensors shouldn't fail in this time-span (or at all), if they are un-branded or Chinese copies all bets are off!
Poor location - Most obvious would be if the sensor is in the lower half of the pipe, from memory the sensor needs to be in the upper half, between 9 & 3 O'clock.
Electrical issues - Are the sensors being fed with the wrong voltage? Too high or low could be a problem, presumably they have a heater circuit as well as the sensor circuit, if I've understoold correctly if the heater is not working correctly this can cause sensor issues.
Mounting issues - IIRC when I got my sensor there were notes about not touching the sensor area or letting foreign (oils, grease ETC) matter contaminate it. I also seem to remember something about not using certain anti-seize products such as copper-slip.
Mixture issues - Possibly 'something' to do with your combustion process which is resulting in 'something' in your exhaust gas which is spiking the lambdas. I can't even begin to think what the 'something' is though (well I could begin but probably wouldn't get far!).

There are probably other possible culprits but those are the ones which spring to mind.


You mention not wanting to run open-loop. What is the reason behind this, running cat's for instance?
I know it will (possibly) give you better behaviour (assuming wideband sensor) but even closed loop would be better than a carburettor, particularly on a car which may be idle for prolonged periods. If the EFi is correctly mapped why would it change?
What about if you ran open loop but used a lambda sensor to feed a gauge to monitor the AFR? This way you'd have a working car, you'd know if there were any mixture issues and if that sensor failed you'd know it was nothing to do with the Holley system and it wouldn't stop you in your tracks.

I'm running the 14CUX system and, as the closed loop capability of this system only runs narrow band sensors, I run open loop and monitor with an AFR gauge. To be honest if my AFR suddenlty changed I'd want to know about it rather than let the system adjust until it can adjust no further. But that's just my logic.

[Allan Robinson]

Thank you GDCobra for your comprehensive response. I started to list all the tests I have carried out but clearly I need to do them again with a better understanding of what I am looking for.

In answer to your question, with the Holley sniper the throttle body is the hub that everything is connected into. The failing connectors sit within this throttle body. There are two casings, one on each side that provides access to the back of the four injectors and their respective cable connectors. It is really tight and congested in there that makes it quite difficult to reassemble without disturbing the injectors or pinching the wires.

Re my O2 sensor problem.
I agree that if I had bought £50 sensors from E-bay then possibly I guess you get what you pay for, but I have tried to buy genuine Bosh sensors. In my attempts to achieve this I bought from well known online companies and paid top dollar; however, even then I’m not confident I was getting the genuine product. On returning one faulty sensor the sales operative said “I keep telling them not to buy this cheep rubbish”!

Anyway, moving on, any help or guidance would be very much appreciated. The issue is that the sensors become erratic or appear to go out of tolerance in a very short space of time. The sensor is fitted in a collector of 4 into 1 header pipes. The shortest header pipe to the collector is 600mm. The sensor is installed in the side pipe at approx. 15 degrees and external to the car. It is a new V8 developments engine with now less that 10k miles running.

To assist me in resolving this issue I constructed a test box with its own AFR gauge that enabled me to access sensor wires for measurement while the engine is running. The sniper ‘Touch Display Unit’ also displays voltage and AFR for comparison. I have no idea what values I should be seeing so using a new sensor as the benchmark I compared 5 used sensors and have ended up with 1 new and 1 good used sensor. The failed sensors were unstable with one having a failed heater.

I did wonder if the unit was receiving over voltage. Does anyone know the voltage range of a Lambda LSU 4.9 sensor? (I would have thought up to circa 14.4v from the alternator would be within tolerance range)?

On examining of the used sensors, my only observation was that they looks slightly black, oily or perhaps too rich.

So my approach this winter will to be:
1. Re check valve preload.
2. Do a Compression test.
3. See how it runs on the road (summer 2025) and monitor for any contamination on the sensor.


Regarding Open loop / Closed loop running, now this is where I could really do with some assistance. My understanding is (please feel free to correct me):
With Open loop the engine runs from the base map but this is not ideal as no real time tune feedback.
With Closed loop the ECU receives real time AFR feedback and adjusts a closed loop map (or does it adjust the base map)?

So, if I run the car for say 100 miles in ‘Closed loop’ to allow the ‘Self Tune’ to take place and then switch to ‘Open Loop’ will I still benefit from the tune data already gleaned over the 100 miles or would I just be reverting to the Base map (pre tune)?
I could then, as you suggested, monitor the AFR on system performance on the ‘Touch Screen Controller”.

Many thanks
Allan

[GDCobra]

Anyway, moving on, any help or guidance would be very much appreciated. The issue is that the sensors become erratic or appear to go out of tolerance in a very short space of time. The sensor is fitted in a collector of 4 into 1 header pipes. The shortest header pipe to the collector is 600mm. The sensor is installed in the side pipe at approx. 15 degrees and external to the car. It is a new V8 developments engine with now less that 10k miles running.

To assist me in resolving this issue I constructed a test box with its own AFR gauge that enabled me to access sensor wires for measurement while the engine is running. The sniper ‘Touch Display Unit’ also displays voltage and AFR for comparison. I have no idea what values I should be seeing so using a new sensor as the benchmark I compared 5 used sensors and have ended up with 1 new and 1 good used sensor. The failed sensors were unstable with one having a failed heater.

I did wonder if the unit was receiving over voltage. Does anyone know the voltage range of a Lambda LSU 4.9 sensor? (I would have thought up to circa 14.4v from the alternator would be within tolerance range)?

On examining of the used sensors, my only observation was that they looks slightly black, oily or perhaps too rich.

I wish you luck on finding decent replacement sensors, I'm assuming it is only the sensor you are having to replace rather than the controller circuit also?
It also sounds like you are using the Bosch 4.9 which is a later version than the 4.2 which I have from an Innovate sensor I bought some years ago when I was aiming to fit a Megasquirt system, I never actually got around to that but have fitted the sensor and it worked fine.

I would think the heater circuit would handle typical alternator output voltages aof around the value you mention as I don't reacall them having to be fitted with a stabilised supply, unless the controller circuit handles that (I'm talking about the heater here, I'm sure the sensor circuit will be stabilised ), it may be worth checking that your alternator is not outputting a higher voltage however.

Regarding Open loop / Closed loop running, now this is where I could really do with some assistance. My understanding is (please feel free to correct me):
With Open loop the engine runs from the base map but this is not ideal as no real time tune feedback.
With Closed loop the ECU receives real time AFR feedback and adjusts a closed loop map (or does it adjust the base map)?

So, if I run the car for say 100 miles in ‘Closed loop’ to allow the ‘Self Tune’ to take place and then switch to ‘Open Loop’ will I still benefit from the tune data already gleaned over the 100 miles or would I just be reverting to the Base map (pre tune)?
I could then, as you suggested, monitor the AFR on system performance on the ‘Touch Screen Controller”.

When you mention 15° I'm assuming you are indicating this to be above the horizontal, that is the upper half of a horizontal pipe?

You mention the sensor looked oily? I guess one thing which could contaminate sensors is if the engine was burning oil to an exessive degree or if the oil which was being burned had some constituant which would be problematic to the sensor, unfortunately I don't know enough in this area to say what that may be. I'm possibly clutching at straws but just putting it out there as something to consider.

In terms of open/closed loop I suppose various systems may have detail differences but in general you will have a map or matrix of values which can be accessed based on the operating conditions such as engine load and RPM and this will determine the injector opening time to give the required air/fuel ratio for those conditions (there will be some adjustment for coolant temperature and other environmental conditions but let's ignore that for now), in an open loop system (such as the 14CUX I'm using on mine) this is all the system uses so is no different to a carburettor in that respect. As long as the engine was correctly mapped originally and nothing changes this will be perfectly adequate (possibly not if running catalysts).

If we compare that to a closed loop system, these will use the base map as a starting point for the fuelling but, by measuring the AFR output from the engine, they will 'tweak' the injector time to correct the mixture, the values used are known as trim values. I've not had much dealing with closed loop as my car doesn't use it but my understanding is that the system works out a proportional value to trim the system up or down and then applies this across the board while constantly re-learning it. There is typically a long term and a short term trim, at least the long term values should be remembered over a shut-down/start-up in which case even if the closed loop was stopped the system should still maintain the latest trim.

Your sytem may be different however, you mention a "Self tune" capability. The devil is really in the detail with this and please bear in mind I'm not au-fait with the detail of your system. I've heard of this capability before (may have been mentioned in the context of Megasquirt, can't remember). As I understand it, in this case, what happens is you'd start with a 'one size fits all' type of map which has values for required AFR under the load-vs-RPM conditions and the system learns what it needs to supply to the injectors to achieve this, effectively assembling a corresponding 'map' of injector timings to deliver the required fuel, this is a little like the system writing the map as would normally be done on a dyno. Another way of looking at it would be the system having multiple trim values, one for each set of load/RPM conditions. I don't know the details of your system so I've no way to tell if this 'Self Tune' feature is something you turn on initially to tune the system (akin to running a dyno session or sessions) and then turn off once all is working correctly, at that point is could still work in closed-loop as described above just maintaining 'simple' trim values. If it is 'turn offable' then I'd say the learned values should maintain.

The advantages with closed loop are the ability to adapt to short term changes in envrionmental conditions - external variables such as ambient temperature and pressure and possibly different fuel qualities - and also to changes in the sysem components such as injector wear which will most likely be a continual process working in one direction over long periods of time.
I can see the advantage of that in a daily driver but for my 'fun' car I'm happy with the open loop style. If something is misbehaving I want to know about it rather than simply have the system adapt to it (there is obviously a limit to this adaptation in any case) but we all have different ideas in this regard. I also like to keep things simple and I don't have to run catalysts so I have the freedom to do this.

Keep us posted on how you get on, I'm sure there's plenty of people here who will help out.

[Allan Robinson]

Thank you for your very comprehensive response. I have just started investigating the sensor issue starting from scratch with a cylinder leak down test and compression test.

All appears good except I have found a badly leaking No3 cylinder exhaust valve so I am now in the process of removing the head. Unfortunately this is not in the same bank as the sensor exhaust.

I will update if I find the cause of the sensor failure but might not be until the summer of 2025 when the car is back on the road.

Many Thanks for your help with this
Allan

[stevieturbo]

I've no idea what you're trying to do with the sensors, but there is nothing useful you can gain from one, by you doing anything with the sensor wires, other than leaving them connected exclusively to the controller they're supposed to be connected to.

You should only be getting readings, voltages from the controller, nothing else.

Generally they are fairly reliable, they are an OEM part after all fitted to millions of VW's and other cars. If you're using an Innovate controller, they're renowned for eating sensors.

Leaded fuel can kill them, wrong controller can kill them, wrong mounting can kill them, too much heat can kill them, any water/condensation getting spewed onto a hot sensor can kill them, Innovate just kill them lol.

If you really want an OEM part, you could buy from a VW dealer. wire lengths and connecting plugs can vary. But it really shouldn't be necessary when there are plenty of people in the aftermarket using sensors from all sorts without major issue ( apart from Innovate )

[Allan Robinson]

Thanks for your replay Sevieturbo.

As you have stated there are many things that could cause an O2 sensor to fail although i didn't know exactly what. I am now looking at any potential contamination. Compression tests and leak down tests were good except I found a leaking exhaust valve but this was not in the same bank as the sensor.

What I have found is that all head chambers have an oily deposit and the inlet valved have a tar like coating on the flow side. It is my opinion that this oily tar like contamination must have come via the PCV system as valve stem seals are good and this engine has only done 10K miles. Cylinders still have the cross hatch marks so together with the compression and leak down test are looking good.

To that end I am firstly trying to find out what the oil level should be in relation to the crank or bottom of the block and have asked on this forum but no response as yet. If I don't find anything conclusive I will just vent the PCV to atmosphere via the oil catch tank and monitor the condition of the chambers.

I don't think it is water contamination as the coolant system looks sound and the sensor is installed at circa 15 degree angle.

So at the moment I am cleaning up the heads and lapping the valves before reassembly.
I'm happy to consider any other suggestions

Many Thanks
Allan
[attachment=0]20241030_101912.jpg[/attachment]

[stevieturbo]

Water contamination tends to be from condensation from cold starts

Some places even suggest to delay sensor heating for a short period to try and reduce water droplets getting onto a hot sensor, instead having the engine warm up a little first ( although obviously the sensor will not be reading during this time if that matters )

Other common guidelines are sensor mounted high in the pipe, not at the bottom, and close enough to the valves it will get heated quickly naturally, but not so close as to cause overheating ( not likely on a n/a engine anyway )

But as said before, if you're using Innovate as a controller, they do their own thing and IME it kills sensors far faster than using normal control methods pretty much every other wideband controller uses.

But generally in normal use you should get a long time out of them

[GDCobra]

20241030_101912.jpg

Is that photo of a valve from your engine? If so that looks in a bad way for 10k miles. I'd say that's indicating a problem with the engine or possibly an incorrect oil level as you have suggested.
This can be a bit difficult on the RV8 as it came with different combinations of dipstick and dipstick tubes, some coming higher out of the block than others, If you have a long tube and a short stick you will have a high oil level, the other possible combination is worse!

I took a measurement on mine a few years back when had the engine out, this was from the sump mount face to the level on the stick unfortunately I don't have that value to hand right now although I do seem to recall putting something on a thread on this site about it as someone else had issues at the time. IIRC it turned out their sump was non-standard with very much reduced capacity so for the correct quantity of oil the level in the sump was rising too high. Is your sump standard.

The other problem with the stick is it doesn't go straight down into the sump, it goes at an angle so the distance between min-max is not that same as the difference in levels measured vertically.

I think you are on the right track breathing the engine into a catch tank, at least for a while so you can check what is coming out and how much of it.