NITROUS OXIDE ( nos / n2o ) advice forum

Over the past few months I've been carrying out a lot of testing on the dyno with a pulsed system (single Pulsoid jetted at the outlet, into a distribution block and a single dry nozzle fitted directly to each port). Control of the Pulsoid is via an auxiliary PWM output from a programmable ECU. Fuel enrichment is through the injectors. Amount of additional fuel is taken care of by switching to a separate fuel map when the nitrous is activated and is calculated to provide a safe rich nitrous/fuel ratio. The system is set up to increase nitrous delivery based on engine rpm.

All the test figures I'm about to refer to were taken at 5000rpm. Normally aspirated the engine makes 164bhp @ 5000rpm. To start with I used an 80 jet which in theory delivers an increase of 40bhp. The actual power gain was 46bhp (210bhp @ 5000rpm). This was done on a full hit basis i.e. no pulsing. Next I took the jet out to 0.95mm (est 65bhp increase) and to start with pulsed the solenoid at 50%, expecting to see an increase of approx 30 to 40bhp. Instead it made an extra 62bhp! Figuring I'd maybe gone too big with the jet and assuming a 100% pulse width would see power jump by more than 100bhp I stepped right back down to the 80 and decided to fully test it at various % pulses to see how the actual power delivered compared to percentage pulse width.

As mentioned above, simply switching the solenoid on (100% pulse) picked up 46bhp. Switching to a 50% pulse width, power was up by 45bhp and at 25% it was up by 44bhp. Not a lot of difference! I've checked the Pulsoid with an oscilloscope while under bottle pressure and the figures agree exactly with the numbers entered for PWM output in the ECU i.e. the Pulsoid is working as it should.

From the testing I've carried out, I have to come to the conclusion that varying the % pulse width of the solenoid does NOT have a corresponding linear relationship to the nitrous power increase. A 50% pulse width most definitely does not deliver 50% of the possible total power increase. I wasn't necessarily expecting to see exactly half the max power at 50% PWM but I was certainly hoping to see at least something close. Using a 25% pulse width sees a power increase that is 96% of the increase achieved by not pulsing at all i.e. a negligible difference! A 50% pulse width delivered 98% of the total increase!

This bears out what happened with the 0.95mm jet - at 50% PWM it added 62bhp. Not far short of the predicted maximum of 65bhp.

As things stand at the moment I don't feel it's safe to continue using the larger 70bhp jet unless there's some way of reducing the increase at lower rpm and my results show there is absolutely no point in pulsing the solenoid anyway.

All testing was carried out with a bottle pressure of 900psi and for consistency, the bottle wasn't allowed to get any lower than 2/3 full. Pulse frequency was 25Hz and the seat in the Pulsoid is adjusted to approx 0.6mm. I've done back to back tests and repeated them several times to rule out fluke results so these figures are based on hard facts (and a lot of time!).

Does anyone else have real world data to show what effect pulsing the solenoid has in relation to nitrous power increase? Does anyone have real evidence of pulse widths having a linear relationship to power increase?

How can the power increase be so high when the solenoid is only open for half or quarter the time it would when operated in a simple on/off situation???

Looking forward to any input/suggestions/theories etc.

Hi KNS,

I haven't read this in detail 'yet' but as I know my Pulsoid respond PERFECTLY to the PWM output from our own progressive controlles, I can only assume the signal from your ECU is inadequate to control the Pulsoid, which isn't surprising.

The fact that you've checked the correct signal frequency and duration with a scope at the Pulsoids does NOT prove that the signal is suitable to drive the Pulsoid and the fact that you see little to no variation in power PROVES that the signal is inadequate. The minor variations you've measured on the dyno could be due to changes in pressure alone and indicates that the Pulsoids are ONLY being switched ON and not ON & OFF.

I'd drop the ECU idea and buy a Minimax or a Max Extreme, then you can be CERTAIN to get the variation that is shown on the display.

We have NUMEROUS INDEPENDENT reports (even from critics) to prove that even with some ECU set ups (like the Motec) the results are EXACTLY as required.

Your questions answer themselves, it's IMPOSSIBLE for a solenoid that is switched off 1/2 the time to flow the same as one that's open all the time (all other things remaining constant) , therefore the solenoids can NOT be closing 1/2 the time and since the Pulsoids can only do what they are told, the can't be getting told correctly.

Have you looked at the outlet plume from the Pulsoid to check the gas flow is being interupted?

Regards

I've just checked the outlet plume as suggested - at 25 Hz it's difficult to see any obvious difference between 50% and 100% (no distinct pulsing) so I reset to 16Hz (the lowest the ECU permits) and you can plainly see the pulse effect. Altering pulse width at this frequency clearly shows the ECU is controlling the Pulsoid. I also had the engine running during the test to see if electrical interference was having any effect (it wasn't).

Out of interest, what's the lowest pulse width you can use and still expect the Pulsoid to operate correctly. I'm not looking at using anything less than say 40% on the engine but some indication of the lowest it will operate may help diagnose things further.

I just did another pull on the dyno with the pulse width set to 50% (16Hz) and the result this time was better but still higher than I'd like at 81% of the total available.

Assuming the PWM output from the ECU is sufficient to control the Pulsoid, and at this stage I haven't found any evidence to the contrary (plume test, oscilloscope readings), I see no need to add a stand alone controller. BTW the oscilloscope readings were taken with nitrous being fired through the Pulsoid (vented to atmosphere) with the engine running, as opposed to a dry bench test.

The ECU configuration I'm using allows RPM based progressive control - AFAIK the Max Extreme doesn't. If ultimately the ECU proves incapable of controlling the Pulsoid correctly, I'll happily admit defeat and look at the alternatives.

The question I'd really like an answer to: what is the relationship between pulse width and power increase? Are they linear? I suspect not but I'm willing to be proved wrong.

With a correct controller the output is perfectly linear.

When matched correctly my Pulsoids will responde at up to 50 Hz (the higher the frequency the bigger the achievement), compared with 11 Hz for some large US solenoids and they can work over a range of 10 to 90% under the right conditions.

Numerous factors influence the ability for a solenoid to pulse in response to PWM (including bottle pressure, available Voltage and a number of Pulsoid design features) but it's much easier for ALL solenoids to work at the high end of the scale as you wish.

There are ONLY 2 explanations for your situation;
1) The output from your ECU is inadequate for the job
2) Someone has messed with the Pulsoid.

If the spring has been changed, or the seat is not set as you believe it to be or the magnetic circuit has been removed, then anyone or all of these would cause the problem you're having.

How old is the Pulsoid?

Did you buy it new?

Have you ever taken it apart?

Unfortunately seeing an 'electrical' signal at the Pulsoid does NOT prove that the signal is of adequate strength to control the Pulsoid, at 16 Hz the coil has more time to reach the desired saturation level to energise the coil and more time to decay to release the plunger, so it sounds more like it's failing to release adequately.

ALL the parts of my Pulsoids are designed to enhance the pulsing performance and if ANY of the parts has been changed or messed with, it will not perform as designed.

Regards

3) A third explanation would be the solenoid is not pulsing exactly as requested by the PWM signal
4) or the frequency of pulse doesn't allow a linear relationship between pulse width percentage and percentage power gain.

In answer to 1) & 2), the output from the ECU is more than adequate and the Pulsoid is new.

I've experimented further today by reducing seat height to as little as 0.15mm and testing at various frequencies (up to 50Hz). Pulsed performance is still inconsistent. I can see that a wet system will be more forgiving as any inconsistencies will be reflected equally on both the fuel and nitrous side. Because this is a dry system, and fuel enrichment is handled independently by the ECU rather than through a matched fuel solenoid, lack of linearity makes it impossible to accurately set additional fuel requirement for open loop running.

As with boost control solenoids there must be an optimum frequency at which response is close to linear but due to the high pressures involved in a nitrous system, maybe it isn't possible to operate in that range? And no doubt the optimum frequency will vary from engine to engine?

Maybe a Revo would be better suited to this application?

Does anyone have a dyno graph that clearly illustrates a linear relationship between pulse width and N2O power increase and if so, what frequency was used?

I'll take you up on that - thanks.

All my products have a lifetime warranty (to the original purchaser) to perform as I describe them.

If we find a problem with the build it will be re-built correctly and failing that if we find for some reason it doesn't perform as it should, then it will be replaced.

However if my test results prove the Pulsoid is performing as it should (which will be determine by measuring the weight of nitrous flowed over a range of PWM and frequencies), then there will be only one conclusion I can come to.

Regards

That sounds fair enough
I'll carry out a nitrous/PWM/Hz flow test here before sending it back. Assuming that all checks out OK, it's still no guarantee I'll be able to achieve a linear response - that's not a slight on your Pulsoids - just one of the characteristics of pulse width modulation. As I said before, if the response isn't linear, it isn't possible in this particular application to configure the ECU to provide the correct fuel addition - unless I map the actual bhp increase at each PWM percentage I plan on using (not gonna happen ).


Definitely interested in hearing their experiences.

Just out of interest KNS - what injector and distribution setup are you using?

DTA S60 ECU. Single Pulsoid centrally mounted (flat 4 layout) and jetted at outlet into NOS Pro Race 4 way distribution block. Single dry nozzle into each port, mounted just below each throttle body (one butterfly per cylinder). Distance from Pulsoid outlet to distribution block approx 4cm, from block to manifold approx 30cm. Fuel injectors are capable of delivering sufficient fuel for 300+bhp total (flow tested). Max power normally aspirated: 220bhp. Intended N20 increase: 70bhp. The ECU is configured to operate the Pulsoid at throttle openings over 85% and start PWM of the Pulsoid at 4500rpm (40%).

I agree PWM on most solenoids does NOT produce anything like a decent result but we've never had anyone have a problem like this before.

If they didn't work as I claim, then there would be no gain when people use our progressives, because going on your figures, the reduction in flow is too small to achieve the desired reductions.

A quote from Brad (my top UK bike agent) was "I set it at 20% I get 20% of the jets, I set it at 30% I get 30% of the jets, it does what it says on the tin" and he's so satisfied with the PRECISE delivery/control that he thinks there is going to be little benefit in performance from switching to the Revo.
However he doesn't yet appreciate ALL the benefits I know the Revo offers and I'm CERTAIN there will be performance gains.

Regards

Here's how they work for everyone else, when driven my a Maximiser;

Click here to watch Pulsoids

If you look closely at the maximiser screen you can see the percentage rate increasing and if you listen to the Pulsoids you can hear the note changing and you tell the flow rate is increasing by the increasing noise from the nitrous flow.

Regards

I've carried out a similar vented test here - pulsing the solenoid at various percentages and observing the flow increase accordingly. With a stethoscope hooked up to the Pulsoid you could clearly hear the pulse increase, so all looked promising. But a simple test of that kind falls a long way short of proving the system will deliver a linear reponse when injecting nitrous into a running engine.

Conclusive proof would be seeing real world (not theoretical) dyno graphs of an engine tested firstly without nitrous and then again with the max hit of nitrous to establish minimum and maximum power levels. Further testing at different pulse widths eg 25%, 50%, 75% would then reveal exactly how close the theoretical power increase is to the actual numbers.

The test described above is exactly what I carried out here but the real world numbers didn't match up to the theory, despite the visual/aural test matching the results you describe above. I started this post assuming you (or someone else) would have conclusive dyno figures to add to the debate but so far it doesn't appear anyone has set out to conclusively prove the concept with genuine dyno data.

I'm aware there are numerous people running pulsed systems and they report a softer initial hit, so obviously power output is being reduced to some degree but by how much? For most people the exact amount probably isn't too important - as long as the initial hit is softened enough to prevent loss of traction etc they're happy and with a wet system they most likely maintain a relatively consistent nitrous/fuel ratio. But for my application (dry) there's no way I can assume the nitrous power increase is directly in proportion to pulse width (because it isn't) and therefore cannot guarantee correct fuel addition. I suspect there may well be a pulse width frequency that could come close but doubt it's physically possible for the Pulsoid to operate in that region (I've tried).

As I've already stated, I don't have an issue with the Pulsoid design (it's superior to anything else available) but in this one particular application it cannot deliver what I'm looking for. This may be influenced by the type of engine/induction system/camshaft etc etc and factors associated with PWM but I need a lot more than anecdotal evidence and simple plume tests to convince me it can deliver power in a linear fashion. i.e. 50% pulse width = 50% power increase.

The REVO concept has to be a major step forward, as a continuous variable injection of nitrous must surely eliminate most, if not all, of the PWM/linearity issues I've mentioned. Real world dyno tests along the lines described above would demonstrate this clearly, hint
[/i]

Yes there is... if you test it on a dyno exactly as I've been doing.



In the course of this thread I've asked for those contact details TWICE and still haven't received them. I would be genuinely interested in talking with others who have carried out tests to hear their experience. I like to think I'm open minded enough to believe I'll probably learn something



I agree 100%. However, my injectors are matched to the fuel requirements of the engine and max duty cycle on nitrous is 76% i.e they are NOT maxing out. All nitrous dyno pulls have been run at a safe rich A/F ratio of 0.82, therefore there is absolute consistency with the A/F ratio at all points tested. I go to some lengths to achieve good consistency and repeatability during all testing. Also I totally agree there will be a bhp variation due to less than optimum timing but there's no way the conservative retard amounts I've used can account for the HUGE linear discrepancies observed. This was confirmed by back to back testing with differing amounts of advance.



Like I said earlier, I've checked all this. It's easy enough to verify: test the engine on the dyno WITHOUT gas, then test it ON gas and you'll soon find the REAL bhp increase provided for that particular engine. Testing at various pulse widths will then reveal how linear the system really is (for that particular engine).


Mine do.



NOW I've got the ANSWER I've been looking for!!! In an earlier telephone conversation you said I could expect a near perfect linear response i.e nitrous BHP increase would be DIRECTLY proportional to pulse width.... hence all the testing I've carried out. Now you're of the opinion there are too many variables and linearity may NOT be possible. Wish that had come out earlier - but no harm done

Fortunately I'm in a positon to test and verify the results of modifications such as this. If I hadn't thoroughly tested all this on the dyno and had opted to go ahead with the current setup, on the understanding nitrous power delivery would be directly proportional to pulse width and had set additional fuelling via the injectors accordingly, I could be looking at a very bad situation right now. I appreciate this particular application probably isn't too common so the likelihood of others experiencing engine damage going this route is slim but I think it's fair to let people know they CANNOT rely on a linear relationship between solenoid pulse width and power output at the flywheel in a dry application, even if the Pulsoid(s) are totally innocent and it's all the fault of "ENGINE PARAMETERS". At the end of the day these systems are being fitted to ENGINES and NOT BENCHES. True precision response from a Pulsoid during a bench test (I stated earlier my own bench test agrees with yours), does NOT guarantee similar performance when fitted to an ENGINE (well, at least the engine under discussion here) - as you've already explained above.

I'd just like to point out I've had no engine problems as a result of this whole test routine (thanks to precision A/F control via the ECU) and I'm happy with the overall power increase achieved. I didn't start this thread to complain about any of your products. I've tried to make my comments constructive and share honest test results that may save others some time and money.

For this project I set out with a particular aim I thought may be achievable. My test results have shown otherwise and I now have confirmation you accept that's entirely possible.

I have detuned motors on the dyno to try to get them to hook up to the track.I have seen huge decrease in mid range torque with small decrease in peak power.What were the mid range torque numbers?I would like to see you put a jet in it run it at a 100% measure how much nitrous was used, then do the same test at 50% and measure how much was used.

Another thought that just passed my way, what nitrous injector/s are you using and where is/are the metering jets???

Can you post a pic of the system fitted to the engine?

Regards

I've just discovered who KNS is and what components he's using which has proved that my thoughts in my last post were bang on.

It's also a HUGE coincidence that the results KNS has independently produced PROVE CONCLUSIVELY one of the FACTS that I posted in my thread;

http://www.forum.nitrous-advice.com/why ... t2811.html

The vehicle KNS has is a Beetle with a US kit fitted which he upgraded with a set of my Pulsoids. When Beetles are fitted with a single set of solenoids they end up with very long pipe lengths. Not only that but US kits use pipe that is FAR TOO BIG for such low power systems.
To make the pipes seal to the US nozzles, the flare jets have to be fitted at the nozzle, which creates a bottle neck adding to the reservoir effect. Now these were bigger than normal jets but they were still smaller than the pipe bore.
The combined effect of long, large bore pipes with jets at the end has resulted in a SUBSTANTIAL RESERVOIR EFFECT.

Despite the Pulsoids pulsing ACCURATELY as I've repeatedly stated and proved, the RESERVOIR effect almost completely neutralised the pulsing, resulting in power figures that were negligible changed over the entire pulse range.

CASE PROVEN!!!

I've said this many times before but it applies even more with this thread: My job is hard enough when I have all the facts and the problem relates to my own products/concepts, without having spurious factors, like badly designed US parts thrown in.

Regards

i got suspicious when the dist block was said to be nos... the WON system, is after all a 'system'!