I am about to install an Accusump as I have been seeing oil pressure drops in high G corners when trail braking. The control options for the Accusump verge on the ridiculously crude (e.g. hand operated lever valve), to slightly better with a pressure activated solenoid. The issue I have is there have no pressure switches compatible with the idle pressures of my engine. My current plan is to use the digital out on my AiM EVO 5 to control a relay for the solenoid and enable the oil from the Accusump to be released based on logic built in RS3. Does anyone else have experience of this? I already have the parts so will build it anyway, but would I be interested to know if anyone has tried similar before.

 

I agree with Peter's assessment on the control. Accusump with an electric solenoid, triggered by a digital output. DO on when RPM over 500, off when RPM under 500 (or there abouts).

The Accusump is a pressure reservoir. I don't agree with Peter's statement of "it pushes excess back to the tank but through a calibrated, slowing orifice that ALWAYS makes sure the main oil galley gets priority, thus ensuring safety for the lubrications system of the engine." The Accusump puts out oil when the pressure in the line it is connected to is less than that of the air in the Accusump. You need to make sure the Accusump is plumbed to an oil galley or supply hose that feeds the main oil galley. If you have it plumbed to a general feed or an oil filter, the oil gets distributed as it normally does (mains, top end, squirters, etc). The Accusump has no control of where the oil goes after it leaves the solenoid/ball valve.

 

I did exactly this and "seems" to be working well. I say seems as I would defer to Peter and Matts expertise.

One of the key things I learned was to use the latest "EPC" valves. The valve controls the flow, the logger just "enables" the valve. The latest valve is a one way valve that flip's direction when enabled.
Normally it is set to accumulate pressure, whenever the engine is making more pressure than is stored in the tank, the tank will slowly add pressure.
When the logger "flips" the valve, the valve will no longer receive oil, and wont discharge any until the engine pressure drops below what is stored. So if there is no issue, the tank stays full.
So the trick is to flip the valve and prepare for trouble "just in case"

I set mine pretty high, RPM>2500 (I really don't care what happens below that) and Engine OilP<50psi (as long as there is 50, I "think" it will survive), this keeps the oil in the tank full, and ensures only in emergency does it discharge.

One thing I have been meaning to do is trigger based on a math channel, so I can chart the math in RS3. With it set as only a trigger, you cant see the logic in action, only the resulting oil pressure gains/drops.

You can see in the graph below oil pressure stays above 30psi. While not ideal, i was dropping into the single digits before the accusump.

 

jscott82, appreciate the response. This was the type of feedback I was looking for.

I am going to be honest, ProCoach your condescending know it all commentary had put me off contributing to this thread. The benefit of community forums is for the discussion of ideas. Comments like the below, with no further information discovery, kill the discussion by claiming to have ultimate knowledge.
Because of this I will post information in this thread as I implement my Accusump integrated with an AiM system. As a start, I want to call out the ridiculousness of

Software and electronics can respond way faster than mechanical systems and their behavior can be more easily adapted and integrated with other logic. The Accusump 'electronic' system comes with a pre-calibrated GEMS PS61 Series Pressure Switch The kit I purchased comes with a "35-40 PSI setting" and the switch is laser etched as being configured for 37 PSI. Let's put the response rate, accuracy, and precision of this to the test. The claimed specification are Repeatability of +/-1.5 psi +3% of setting and Deadband of 12% of setting.

To do so I will use the AiM EVO 5 and two of the analog channels. I have these super handy debug cables I made with Wago 224 connectors, Tefzel, DR-25, and Binder 719 female connectors. Makes it very quick to prototype wiring configurations.



The test rig uses a 1/4 NPT Schrader valve for pressurization on a Tee with a 1/8 NPT female and 1/4 NPT female connection. This then has the GEMS pressure switch and an AiM 0-150 PSI (X05PSA00150P18K) pressure sensor.



And this is what the data look likes. Blue is the pressure while red shows the switch activation through voltage measurement change on the analog channel. Even with pressure changes happening pretty slowly (time on the X axis) the GEMS sensor misses it's deadband specification by transitioning at 29.7 PSI and 30.8 PSI when pressure is falling. It does better when pressure is rising and meets specification.


The point of the above is the Accusump solenoid control system as supplied is better than a lever, but a low precision system. It is perfectly fine for a $40 switch. There is no magic to it, either the switch is on or off and the solenoid is activated or not activated. The performance of this system can be easily surpassed with almost any digital control system, especially one as expensive and feature rich as the AiM EVO 5. More importantly, it can take multiple inputs for the logic used to control the solenoid on the Accusump. My data is showing that Yaw rate is a leading indicator of a soon to occur low pressure event when combined with heavy braking. Rather than wait for oil pressure to drop and respond, pre-empt the pressure drop based on Yaw rate changes from my front axel IMU and long_acc.

 

I almost went with an accusump but couldn't quite get my head around a good way to control it.
I think you're on to a great solution. Although, I'd have to feel real comfortable with the consistency of gating the accusump via yaw and long_acc.
Can't you start with just oil pressure and rpm, and look at post event data to see if that's enough? Or, I'm not familiar with the EVO5, but if you could install a math channel to watch rate of change of dropping oil pressure, that might create a early warning also.

 

I realize I'm not addressing your question, but I have a suggestion: If you are able to log whether the Accusump is open or closed while out on track, it can be helpful to gather this data. I had an older Accusump installation which had a different style EPC valve which made this easy. I created track maps comparing where the oil pressure went below a certain value (35 psi in the example below) vs when the Accusump valve opened. Below is how I plotted it in Racepak's DatalinkII.



You could also plot oil pressure vs Accusump valve position to see the scatter of the opening/closing pressure from a session I presume.

 

@sometime, fair enough. Rereading that, can certainly see that. I apologize.

Thanks for your share. Very good information and well done. Again, will try to do better.

I learned something.

 

In reguard to the "speed" that the accusump can respond, I've heard that argument before and there is merit.

In its original configuration, the accusump was a simple hydraulic pressure accumulator, there was no valve separating it from the engine. In this configuration the accusump responds in real time to every fluctuation. This is the ideal scenario.

Is does create other issues,, like flooding the sump and exhausting its supply too early. So the valved solution was added as a work around. This by itself slows the response way down.

now adding a logger/controller to mix adds more delays. 1) where are you sampling pressure from? Depending on where you install the pressure sensor, can add delay as it takes time for the drop to propagate through the system. 2) Aim logger is not a real time CPU, there is no guarantee the aim system is not delayed just at the moment you need it.

Risk/Reward... its not a perfect solution, its not guaranteed to save your engine, but I think it does increases the odds back in my favor.

 

True. I've seen windage issues from high levels in the sump, but that is usually due to not allowing galley and accumulator pressures equalize before checking the static oil level.

My experience with the included electric solenoid is that the discharge from the accumulator is much quicker than the return rate allowed back to the accumulator. I like increasing odds!

 

Thanks for the info dasing. Seems like your developed part of what I was looking for e.g. log of usage. As the valve control will be done directly by the EVO 5 data logger in my scenario, logging of it's activation will be automatically accounted for. The other area I want to know is how much of the Accusump oil reserve is being used. Having spoken to Canton it seems the best way to do this is with another pressure sensor to replace the needle gauge to know the air charge. As the Accusump fills, the air in the back of the chamber will be compressed and the pressure will rise. I have an AiM 0-300 PSI (X05PSA00300P18K) pressure sensor which is replacing the basic gauge to measure this.

 

Appreciate the comment. Let's try something new and see if it works

 

One point I want to make clear is that with prior data of real events, a control system can be given parameters to to preempt the bad system state e.g. low oil pressure; rather than reacting to the system entering that bad state. This doesn't require the system response time to be ultra fast as it is not responding to a trailing indicator of negative system state.

As a side note, my car is heavily tracked, but driven to the track. Having the Accusump dumping it's contents into my sump when the car is idling at some lights is not desirable. My tying the control of the Accusump to the data logger, it won't be freaking out at low idle pressures.