slo

iTrader: (1)

The parts consist of a cooling mist S-HSV cooling mist pump, and cooling mist pressure bypass valve, which effectively regulates the pressure put out by the pump.

Impressions, the S-HSV is very well designed, and it works.

It works fine controlled off the haltech, but I found it sensitive to PWM freq, and PWM duty cycle percentage, the higher the freq the more pump pulsing could be seen in the stream, a lower freq had the made the pump pulses invisible even at very low flow.

When I say pump pulses let me be clear, the pressure bypass switch is an electrical switch that turns the pump in at about 125PSI and off at about 130PSI. The pump tuns on and off fairly rapidly, when the valve is at 100% and and flowing 10GPH (a 10 GPH nozzle) the pump is pulsing about 10 times per second. This pump pulsing is only controlling pressure and not flow. At high flow pulsing can't be seen at all in the stream. At some point as the flow is turned down the pulsing can be seen in the stream, this seems to happen only when the flow goes well below a usable level. At what point the flow begins pulsing with the pump is strongly effected by the PWM frequency a lower frequency seems to provide better low speed control of the flow.

The PWM freq I settled on was 30 HZ, which is very similar to what a boost control solenoid switches at.

This seemed to give the most progressive flow rate of all testing.

The minimum usable flow seemed to be at about 35% duty cycle at 30 HZ, which gave a very nicely atomized stream with no visible pulsing. Steadily increasing the duty cycle up to 100% gave a very progressive increase in flow.

raising the freq also had the effect of changing where the minimum usable flow was available. With a freq of 400 HZ the nozzle wouldn't even produce a drop until over 50%.

If there was a pulsation dampener in the system (like the one in an EFI system) then a higher frequency would probably work better.

What I really need to do now is some tuning to see what I can do with this thing.

What I may do is create a chart by injecting water into a graduated cylinder at various duty cycles, (to determine the actual flow rate) I am guessing that the flow rate goes up with a curve as opposed to a strait line, such that doubling the percentage may not double the flow rate.

I think that this an all water injection systems that are designed to be progressive could benefit from an EFI style FPR, with a pulsation dampener, that would allow the pump to stay on as opposed to rapidly pulsed to regulate pressure. Perhaps a pulsation dampener alone would be enough.

I may adapt a bypass style FPR to this system to see how it works, I know of at least one that is designed to work with methanol.

I am very happy with the cooling mist products, especially the S-HSV and would certainly buy again.

I would also like to say the customer support given by coolingmist is excellent, he has been there and available every single time I called.

coolingmist

Im glad to see the valve is so far living up to the hype I know this valve is very fast, very reliable. When this was patented, one of results of the design is the superior repeatability. It puts out a great curve.



I dont know if this will help you but I have a hysteresis curve for you.

slo

iTrader: (1)

Thanks, than info may well help...

For my initial test, tune I will be starting at 35% duty cycle at 2PSI and increasing in a linear fashion up to 80% duty cycle at 15PSI, for those not farmiliar the haltech generic PWM load/duty cycle output is a 32 x 32 3D map controlling duty cycle percentage on one axis based on RPM and load (boost pressure) on the other 2 axis's. Since a rotary engine has a fairly flat VE curve which varies pretty much only on load after about 3000 RPM this should give a pretty good initial tune. I am keeping the map nearly flat across the RPM axis with just a little more water/meth near the peak torque rpm.

slo

iTrader: (1)

Everything is working great so far, the transition turning the water injection on is invisible (no stumble etc) my initial water injection map (see attached picture). Seems to be working pretty well, I am not initially wanting to run past 18PSI which is why its maxing at 18 PSI.

I reduced fuel in boost by 5% and that appears to be a good amount.

The meth/water combo reduces intake temps by about 50F to levels at or below what I had with an intercooler.

The lack of an intercooler greatly increases throttle response.

Gohan3rdrift

iTrader: (11)

Cool beans,
keep us updated bro.

RE-catch-22

Any updates.....good information.

I am about to do the same thing.....Haltech and cooling mist.....but I have an older model.

slo

iTrader: (1)

Its very stable, it works great.

I changed the wiring a bit, I am now powering the pump with a relay off the haltech fuelpump relay, also using a static boost switch to start powering the pump relay at 1-2 psi through the pumps pressure bypass.

I have a second pressure bypass wired into the system if line with the boost switch that sends a signal to the haltech if the line pressure goes below 105 PSI.

This is on an input that is configured for hard ign cut at 3K rpm.

the Idea is if the line pressure is not there because the pump has failed, and or because the tank ran dry it then the car can't go into boost

coolingmist

Here is a picture of our Dual S-HSV multi injector setup. The supercharger for the Elise is not going to be ready until June or so, in the mean time we are upgrading our Rx-7 with our latest. The S-HSV's on this manifold can turn horizontal instead of vertical which is how I have it installed. It fits perfectly along the firewall. For those that want to use a single S-HSV you can do that as well. Initially I am going to use 3 nozzles in the elbow and one in the Blitz Intercooler coldside core. I plan to remove the intercooler completely and just use water injection.

I will post a thread at somepoint detailing our install and what we did. In the meantime for those that want to see what the S-HSV looks like, here you go.

Xtatic

would coolingmist care to elaborate on this phenomenae and tell us whether the dampener solution would help the above "problem"?

Thanks
Alex

coolingmist

With Vari-Cool we run a fixed PWM and adjust the current from low to high. If you do it that way you can run high frequency. We do over 400 and the valve is rated to 15,000 HZ.

In terms of how the Haltech is doing this or why the valve is working at the said frequency I am not fully sure. Without having the Haltech hooked up and without an Oscilliscope I could not answer that question fairly.

David

Xtatic

I was just trying to verify my conclusion, which would be that this is a consequence of an "open time" of the valve. (which calculates to about 1.5ms at the voltage the OP used, given that the valve stops dribbling at 60% duty cycle at 400Hz).
The "open time" tends to shorten as current is increased, so perhaps that's why it's possible to regulate the real duty cycle by regulating current while keeping PWM frequency and duration the same.

and if i am right, the pulsation dampener won't do anything for this.

coolingmist

Its a proportional valve, so its designed to give a proportional increase in flow with current. Look at the chart in post #2. its by design.

David

Xtatic

going from the description of you product here:
http://coolingmist.com/pagedisplay.a...key=s_hsv_kits

... surprisingly nowhere it says that it is a proportional valve. It does say that it goes on/off quickly, which is what a PWM valve usually does (much like regular fuel injectors).
Also, from what it says, S-HSV means super high speed valve. Can it also be a "proportional valve"? i.e, both PWM AND proportional valve as described in this guide:
http://forums.evolutionm.net/showthread.php?t=292953 )

coolingmist

I am not trying to confuse anyone with terminology. How is this, its a PWM proportional valve.

Vari-Cool is a PWM device. When you connect vari-cool to the PUMP it makes the pump PWM. WHen you connect Vari-Cool to the valve it makes the valve PWM. Same thing. In otherwords S-HSV will PWM the same way the pump does.

yes, its also a positional valve which can be attained by using a driver board and an amplified 0-5V to make it move to a certain position. Also, If you want to open the valve completely for a number of milliseconds just give it 12V for that period of time. It has a very fast r/t.

We designed the valve to work with Vari-Cool, however there are alot of different applications. Someone like Slo that has a great curiousity and is willing to pioneer will have an exceptional application because the performance for the valve is there.

David

slo

iTrader: (1)

No you missed the point, there is no problem.

and the pulsation dampener isn't trying to stem pulsations created by opening and closing the valve.

These aren't fuel injectors which are opened for a set on time every revolution, they are being continuously pulled.

The pulsation dampener would be to stem pulsations from the pump turning on and off. Those pulsations are created by the design of the system which regulates the water pressure, however this is not a problem, because pulsation isn't even visible in the stream until the duty cycle gets below a usable level. What I mean by a usable level, is there is no point in turning the stream down until its injecting 10 CC per minute.

A pulsation dampener while it might absorb this pulsation but it would be a waste of money in this application. In-fact I wouldn't be surprised if pulsation was beneficial in this case and improved atomization of the liquid.

The system works very good the way it is.

I would also like to add that I could make it perfectly usable PWM'ing at 450 HZ, but the duty cycle range would be 60-100 instead of 30-100, it would still work as well.

slo

iTrader: (1)

Bad ***, I have the S-HSV completely hidden, tapped into the y pipe with no intercooler.

Also, pure meth which I have just started playing with works better at cooling the intake charge than water/meth mix.

Injecting fluid is more effective than intercooling at cooling the intake charge, I wouldn't be supprised if it works better without an intercooler with a stronger mix (as in more alcohol) because with a hotter intake charge the alcohol will more easily vaporise. Allowing you to use more, and get more benefit from the greatly increase octane, this would also work for water alcohol. I also wouldn't be supprised if you could get the intake charge with AI and no intercooler just as cold as you could with AI + intercooler.

kev-yorks

Is the haltec capable of controlling the twin HSV set up. Wiring the valves in parallel from a single pwm output or is there going to be too much load on the pwm output line? Or have cooling mist got plans for a higher flow rate HSV?

Kevin

wackaloo13

iTrader: (3)

this is all being done on e8s and e11s. just run a second pwm output. i bet there is a way to do staged injection if you want to run 2 valves as well.

coolingmist

Each assembly that you see draws approx .5 amps, so if your circuit can run 1 amp, no problem. If not, you can use a 2nd PWM circuit for the 2nd assembly.

We could get a larger valve for more flow, it will be slower and harder to mount. Each assembly can run a max of 10 GPH with great atomization.

On my Rx-7 I am using 2 assemblies and running about 6 GPH. 1 assembly can easily do this. I am running the 2nd for redundancy more than anything.

David

slo

iTrader: (1)

In the chart posted by cooling mist above it clearly shows that the valve draws less than .5 amps so to valves wired in parallel should be no problem.

The PWM's have an advertised load rating of 1 amp.

Trionic

iTrader: (4)

What? You're saying a straight analog 0-5V signal can be used to control these? No PWM needed at all; you've tested this? It would be awesome if they can work this way too.

Xtatic

I don't think it'd be all that awesome. Non-fullswing-PWM operation would mean only partial valve opening, which in turn could mean bad atomization due to high pressure drop over the valve (at the low side of the proportional opening spectrum).
This valve seems to be fast enough to be operated like a regular injector at full swing. It would be nice to see the "open time" vs "current" curve for that valve.

once again, refer to this well written guide for difference between PWM and proportional valve operation:
http://forums.evolutionm.net/showthread.php?t=292953

coolingmist

Not a standard 0-5V. Most 0-5V signals cannot carry any load at all. It needs to be supported by a driver board.

Its actually one of the best way to run the system as it doesn't PWM, it opens to a certain position (based on the voltage). As pointed out, it requires a special driver board or custom signal.

David

Trionic

iTrader: (4)

Thanks. I think I mentioned to you before that I'm integrating your WI/AI equipment into a PLC-based control/logging system... with EGT-based injection into the front and rear rotors independently.

If the valve actuation/atomization works well for Voltage control it'll be easier for me to implement, control-wise, compared to managing somthing like a high speed counter or CW/CCW daughter card and pwm driver.

Should I be concerned about poor control and atmization at low flow rates like Xtatic writes?

coolingmist

It depends on the size of the nozzle and the number of valves you have. Its widely accepted you need 40 PSI at the nozzle to get acceptable atomization. You would start the current/voltage at whatever gets you 40 PSI.

We dont have a low end atomization problem with Vari-Cool and S-HSV because the starting current (as well as full current) is adjustable within the software.

David