Valve float due to exhaust back pressure?

[Merc Man]

Ok, time to follow up. The engine has been sitting over the weekend, today we continued testing. To answer some of the questions, please bare with me if I miss one or two ...

- Air flow is unfortunately not measured.
- EGT unfortunately not measured.
- AFR is measured and drops clearly when power drops.
- Regarding coil bind, at 1.900"/145 lb seat preassure there was still .135" left before coil bind. Springs are now shimmed .050" which leaves .085" to coil bind and increases the seat pressure by approx. 21 lbs so we're at 166 lb seat pressure.
- Lifters are short travel Johnson/Isky so with the springs used I simply don't expect any lifter issues.
- Oil is per Isky rec's for the lifters and also according to bearing clearances (15W-40). Pressure is almost constant at 68-70 psi from 3600-6000 rpm.
- The curve drops at 6200 very precisely every time, power and torque under that is very consistent and the engine actually runs very good with over 600 lb-ft from 4150 to 5850 rpm and 671 hp at 6000 rpm.
- Exhaust backpressure has not been measured. We found out today that new mufflers payed back 10 hp from a drop of 40 from open headers. However we discovered that even a 3" diameter 3 feet pipe and no muffler still robbed the engine of 30 hp so that must be more a pulse tuning issue than simple restriction in the exhaust. Headers are 2" primaries ...
- Ignition is a MSD6A, i.e. no rev limiter
- Also unfortunate but when running open headers we did not pull further than to just before where we 'think' we did see valve float.

Decision today was taken to replace the pushrods, currently .130" wall thickness, with .080" thickness. Maybe they were a little bit heavy for this combo however it was a balancing act together with increasing both seat load and spring rates from recommendations and today we went even further on seat load.

Please feel free to comment!

[Merc Man]

and the right lobe

Imo, that’s where it starts.
If the lobe is too fast, it can be a big challenge to figure out a way to have the motor turn higher rpms with “regular” HR lifters and “normal” spring loads.

Howards recommend 130 lb at seat and spring rate of 348 lb/in, we're now at ~165 lb at seat with 431 lb/in spring rate

[mt-engines]

- Air flow is unfortunately not measured.
- EGT unfortunately not measured.
- AFR is measured and drops clearly when power drops.

does air flow start to drop before power drops? what s the afr before and after it starts to drop?

[PRH]

We had one with a solid roller do a similar thing years ago, but at 6500.

It ended up being the nylon thrust button someone installed got mushroomed right over, and the cam would walk forward at 6500, retarding the timing a bunch.

Replaced the nylon with a roller button and reset the thrust clearance, which cured it for about 3-4 pulls.......then it was back to the same thing.
The timing cover got pushed right to the back of the water pump with the roller button in there instead of the soft nylon one.
Knocked the timing cover back in....... attached a spacer to the back of the water pump...... problem with high rpm power dropping off cured.

The lesson there was...... check the timing at the rpm where the power drops off.
We dicked around with a lot of stuff before we found the timing retarding way up high.

[PRH]

- Also unfortunate but when running open headers we did not pull further than to just before where we 'think' we did see valve float.

Why can’t you just open the headers up and pull it higher?

Howards recommend 130 lb at seat and spring rate of 348 lb/in, we're now at ~165 lb at seat with 431 lb/in spring rate

Some of the mercruiser race series BBC engines with hyd rollers use springs that are over 520lb/in.

We chased this situation around on my friends Stocker a few years prior to this article......
https://www.hotrod.com/articles/mopp-05 ... dyno-test/

[Merc Man]

We had one with a solid roller do a similar thing years ago, but at 6500.

It ended up being the nylon thrust button someone installed got mushroomed right over, and the cam would walk forward at 6500, retarding the timing a bunch.

Replaced the nylon with a roller button and reset the thrust clearance, which cured it for about 3-4 pulls.......then it was back to the same thing.
The timing cover got pushed right to the back of the water pump with the roller button in there instead of the soft nylon one.
Knocked the timing cover back in....... attached a spacer to the back of the water pump...... problem with high rpm power dropping off cured.

The lesson there was...... check the timing at the rpm where the power drops off.

That was interesting!
Timing gear has a Torrington bearing and we set the cam end play to ~.005" with a dial indicator however the limiter is a nylon button. The thing is though the symptom hasn't changed at all, it was there from start and was still there even with increased spring load.

Regarding our next question, we simply didn't get there as time ran out. Dyno guy is 110 miles away ...

[MadBill]

..Decision today was taken to replace the pushrods, currently .130" wall thickness, with .080" thickness. Maybe they were a little bit heavy for this combo however it was a balancing act together with increasing both seat load and spring rates from recommendations and today we went even further on seat load.

Please feel free to comment!

In the vast majority of cases, heavier/stiffer pushrods help or at least, don't hurt power. Weight on the lifter side is far less important than on the valve side.

GM went to 7/16" PR's for the late sixties solid lifter L88/ZL1 427".

[Merc Man]

..Decision today was taken to replace the pushrods, currently .130" wall thickness, with .080" thickness. Maybe they were a little bit heavy for this combo however it was a balancing act together with increasing both seat load and spring rates from recommendations and today we went even further on seat load.

Please feel free to comment!

In the vast majority of cases, heavier/stiffer pushrods help or at least, don't hurt power. Weight on the lifter side is far less important than on the valve side.

GM went to 7/16" PR's for the late sixties solid lifter L88/ZL1 427".

That was also my thoughts and intentions.

[Merc Man]

- Air flow is unfortunately not measured.
- EGT unfortunately not measured.
- AFR is measured and drops clearly when power drops.

does air flow start to drop before power drops? what s the afr before and after it starts to drop?

The owner of the engine chose to go to this dyno guy because he's experienced with the Holley EFI.
Unfortunately he doesn't use the old school tools as air and fuel flow measurement.
When the supposed valve float occurred the afr dropped at least one point from being stable during the entire run.

[PRH]

If you’re running it with EFI, there is probably some vacuum reading you can look at?
Does anything strange happen with it when the power drops off?

If it were doable with a minimum of hassle, I’d probably try a carb on it...... see if the same thing happened.

[mt-engines]

- Air flow is unfortunately not measured.
- EGT unfortunately not measured.
- AFR is measured and drops clearly when power drops.

does air flow start to drop before power drops? what s the afr before and after it starts to drop?

The owner of the engine chose to go to this dyno guy because he's experienced with the Holley EFI.
Unfortunately he doesn't use the old school tools as air and fuel flow measurement.
When the supposed valve float occurred the afr dropped at least one point from being stable during the entire run.

try getting the datalog. thanks

[lefty o]

- Air flow is unfortunately not measured.
- EGT unfortunately not measured.
- AFR is measured and drops clearly when power drops.

does air flow start to drop before power drops? what s the afr before and after it starts to drop?

The owner of the engine chose to go to this dyno guy because he's experienced with the Holley EFI.
Unfortunately he doesn't use the old school tools as air and fuel flow measurement.
When the supposed valve float occurred the afr dropped at least one point from being stable during the entire run.

as in during the entire run, including before supposed valve float? i think your problem is elsewhere.

[Rick!]

AFR drops during the poor performance? Usually poor running is associated with excess oxygen due to incomplete burn and AFR spikes upwards, in my experience.

I wouldn't be surprised that injector dead time is messed up or the end angle table is activated and way out to lunch.

One would also need to make sure that it's not hitting a rev limiter. There are 3 rev limiters in the Dominator EFI software and 5 different ways to limit rpm with fuel or spark or both.

Is this port injected or one of the throttle body units? What version software are you using? Holley is worse than Microsoft for updates and letting customers beta test their fixes (or fails).

What type of ignition system is on this thing? I've had "Dual-Sync" issues where it would shut down at 3000rpm with a hall effect crank signal and at 6000rpm with an analog crank trigger. Updating the software version and using a dedicated hall effect cam trigger and going back to the hall effect crank trigger fixed that issue.

Like others have said, you need to create datalogs of the pulls and look at them to get a better feel for what is going on.

[Merc Man]

Guys, thanks for all responses!

Unfortunately I could attend today but the dyno guy arranged a pushrod swap to lighter .080" rods and also a bit shorter.
The length puzzles me a bit since I'm only using mid lift method for pushrod length checking. However the engine is now running well pumping out 663 hp at 6000-6100 rpm. That is with 3" pipes and mufflers that robs the engine on 30 hp.

496_afr300.jpg

[PRH]

If lighter pushrods were indeed the cure...... that sure would seem to point towards the upper rpm problem being some sort of valvetrain control issue.

That’s a good looking curve.

I would have pulled it a little higher to see that it didn’t run into the wall right where it’s in the meat of the big power(and if it did, that would help determine where the rev limit needs to be set).