[jeho]

A lil background….

I’ve been lurking on the forums since march 08, when I sold my N/A built k20 RSX-S that put ~220-230 whp on a lil 2.0Litre 4 banger with Hondata Kpro ECU Management (fully user tune-able; a/f, ignition, low speed cam angle & high speed cam angles, vtec engagement points, datalogging, user maps uploadable and/or dyno tune able)…For the last 8 months ive been lurking and asking myself… “JB or Vishnu”… Luckily, I had a friend with both JB3 and JB2H. The first time I drove em, I was blown away by how F*CKING fast it was and the power delivery. For the price, everyone would say it’s the best bang for buck tune BUT there were a few ‘flaws’ that kept me from ordering this tune. I didn’t want to get a tune that was far less adjustable compared to my previous Honda haha, it wouldn’t make sense.?

[BEFORE….my V3 came…]
Heres why I chose the V3 > any other tune (late FEB)…
- ability to update maps instantly (who LIKES to wait things that make ur car faster/smoother..?)
- datalog (how else are you gonna know if ur tune is doing your engine well, butt dyno’s don’t mean jack!!!)
- user adjustability in torque settings for 2 different maps. It was important to be able to adjust 1st and 2nd gear torque settings. It rains more than ½ the year here, therefore I don’t have my max summer tires on (Dunlop Direzza SS) on until the months of summer so for ¾ of the year, I am on just high perf summers that don’t compare to my Direzzas. The V3 would allow me to lower the 1st and 2nd torq settings so I could still full throttle since the tires don’t grip as good as my max summers. Now I can 100% throttle with both sets of tires, obviously 1 making less boost.
AND, if for some reason, the car is running too much boost (yes I have boost gauge) on HOT track (circuit) days, I would have the ability to lower the torq settings.
- pulls timing, compare to some other tunes that don’t?

[After… V3 is here and installed (march)…]
Heres why I LOVE the V3….and don’t regret getting it a single bit…
- Shiv releases the news of CANBUS adjustability in rev 2
WHO WOULD HAVE EVER EXPECTED THIS? Now this gives every reason for the extra $ for the V3 Rev II. *no need to explain the features for this is the next big thing :P
- The low-end power delivers sooo smoothly and linear, and it pulls so hard!
- The engine revs much much free-er and feels lighter in addition to my minor weight reductions… cf hood/trunk/18lb wheels/lighter BBK/coils
- the car pulls from start to finish, 3rd gear pull feels like 2nd, and 4th feels like 3rd…
- boost is very consistent and doesn’t spike (Recorded a run via Defi Boost Gauge to show how strong it pulls)

Final Verdict---

V3 is the best bang for the buck and I can’t see how the other piggyback tunes can be even comparable….Yes, I’ve tried JB3 and the power delivery does not feel IMO as good as the V3, even tho they both make similar power. The power delivery for daily driving on the V3 is much much more fun than the JB3. I have no say on the ¼ cuz I don’t drag. I can’t wait to test out V3 on the track(circuit) when I gets warmer here against the E90/E92 M3 and stg 2/3 STi/Evo.. No, I don’t want to elaborate more on V3 vs JB3 because we have enough threads on tuner wars here.

Anyways, THANKS to SHIV and Vishnu team for having a product that makes the 335i drive better than I expected…

**AR Catless DP’s on the way…I can’t wait til even more power >_<

Here is the recording from my Defi Control unit of 2nd, 3rd, 4th gear pull (very short 4th)..
The boost is very stable...I know I should have datalogged via comp...

[marty_mcfly]

thanks for the review

how come in that video though, theres no engine sound? lol

[jeho]

with the Defi Control unit...

U can record from the data from the boost gauge and then play it back to watch it...

[Soni]

nice write up Jeff

[sflgator]

Nice review, J86...enjoy your PROcede v3.2 335i!!!!

[E82tt6]

Good writeup, but one minor technical correction, the procede does not control timing. It offsets crank position sensor, tricking the ecu into thinking the pistons are somewhere they are not, which provides some limited timing retard until the dme learns around it.

Not bashing your choice in tune at all, I'm very glad you're happy, this is just a common misconception (primarily the result of marketing), though I'm not sure if it would have effected your choice.

No piggy will control timing until we see one that directly drives the ignition.

[OpenFlash]

Originally Posted by E82tt6

Good writeup, but one minor technical correction, the procede does not control timing. It offsets crank position sensor, tricking the ecu into thinking the pistons are somewhere they are not, which provides some limited timing retard until the dme learns around it.

Not bashing your choice in tune at all, I'm very glad you're happy, this is just a common misconception (primarily the result of marketing), though I'm not sure if it would have effected your choice.

No piggy will control timing until we see one that directly drives the ignition.

Please explain what is involved in order to control timing.

Because the spark timing effects of retarding timing 1 deg through CPS offsetting is exactly the same as subtracting 1 deg of timing through the ignition map via ECU reflash. With your logic, neither approach controls timing since they are both subject to DME learning. Both approaches dictate the timing advance target that, depending on mapping, will dictate the extent of additional DME learner you refer to.

If you don't agree with that, could you tell me why?

Shiv

[E82tt6]

Originally Posted by shiv@vishnu

Please explain what is involved in order to control timing.

Because the spark timing effects of retarding timing 1 deg through CPS offsetting is exactly the same as subtracting 1 deg of timing through the ignition map via ECU reflash. With your logic, neither approach controls timing since they are both subject to DME learning. Both approaches dictate the timing advance target that, depending on mapping, will dictate the extent of additional DME learner you refer to.

If you don't agree with that, could you tell me why?

Shiv

The difference is, with a DME flash, you can set timing to 8* of advance at a certain point, and keep it there if you so choose. You could also make sure that the ECU is more cautious in terms of how much additional advance it will learn in, and reduce the rate at which it will advance timing on its own. You cannot do that with CPS offset.

With your CANBUS integration, you can add more dynamic CPS offset based on current ignition advance, however there is a limit to the amount you can offset the crank position sensor before it begins to have unintended and deltrious effects on other systems that rely upon crank position.

And unless I'm unaware, it has not been clearly demonstrated that there is a ceiling on timing advance that can be reached by throwing in a reasonable amount of CPS offset, correct me if I'm wrong.

CPS offset is a tried and true method of modifying spark advance on many cars. This ECU, however, will learn CPS offset out, and based on how quickly ignition timing seems to advance back after a knock event (per Scalbert's testing) I wouldn't be surprised if the CPS offest is learned out within three WOT pulls. That doesn't it's useless for perhaps reducing knock during adaptation a bit (as I have always conceded), however, I don't think it can legitimately be called timing control.

IMHO, real timing control would entail directly driving the ignition. This removes all limits CPS control has, and lets you decide exactly how much advance/adaptation/etc you would like under a given set of circumstances. For example, at 13psi/55*C IAT on a map designed for 91 octane, you could decide that you NEVER want the ECU to add more than 7* of advance at 4600 RPMs, and if it tries to add more, you can modify the feedback the ECU recieves about the current level of timing and knock sensor input, to ensure it doesn't try to modify spark advance, and since you're directly driving the igniton, you can have the spark happen whenever you like, say exactly 7* of advance.

This would allow you to program in more "safety margin" than what the stock ECU has, since we are dealing with higher stakes. You could even modify the amount of noise it had to get from the knock sensor to retard/hold stable adaptive advance, by simply adding more noise to the signal as desired.

Correct me if I'm wrong, but I don't believe current piggyback technology allows for that, or you'd already be doing it.

Also, if you see a major flaw with that solution that I don't that would make it somehow inferior to CPS offset, point it out to me.

My point is simply that I don't believe CPS offseting can be realistically called "timing control". Perhaps it helps prevent SOME knock during adaptation. Frankly we haven't seen evidence of that yet, but I know Scalbert has some more testing to do along.

Don't taking personally, I just thinking saying CPS offset is timing control is a bit like throwing a "good enough" label on it, when it isn't good enough. I think it's an imperfect solution at best, and I think trying to make it sound like a better solution than it is discourages people from pushing for progress, and a better solution.

[OpenFlash]

Originally Posted by E82tt6

The difference is, with a DME flash, you can set timing to 8* of advance at a certain point, and keep it there if you so choose.

I think this is where the confusion comes from. No, you can't. You can map your 8 degrees of timing at any particular load/rpm point but it is still subject to additional (not to be confused with "additive") timing correction from the knock control system.

The same applies to the CPS offset approach. You can dictate the same 8 degree of timing at the same load/rpm point. But you would do so by inducing an offset. If the stock timing value at that point is 10 deg, you just apply a -2 deg correction at that point. When it comes to actual spark occurrence in the compression stroke, the result isn't "similar to" or "almost good enough" the first approach. It is the exact same result. No difference.

By absolute control of timing (ie, timing map input = output), you would have to disable all secondary timing corrections which isn't feasible or advisable.

You could also make sure that the ECU is more cautious in terms of how much additional advance it will learn in, and reduce the rate at which it will advance timing on its own. You cannot do that with CPS offset.

With the exception of cold start mapping, the DME does not advance timing above the set point. So the idea of additive timing offsets is imaginary. All sensor feedback induced corrections are negative.

With your CANBUS integration, you can add more dynamic CPS offset based on current ignition advance, however there is a limit to the amount you can offset the crank position sensor before it begins to have unintended and deltrious effects on other systems that rely upon crank position.

Have you determined the measurability of these "deleterious" side effects caused by +/- 1% of timing offset? And where exactly do these side effects manifest themselves?

And unless I'm unaware, it has not been clearly demonstrated that there is a ceiling on timing advance that can be reached by throwing in a reasonable amount of CPS offset, correct me if I'm wrong.

If by "ceiling of timing advane" you mean maximum timing value/curve, then yes, there are maximum values. That is, with excessive octane or sufficient negative CPS correction, the DME will assume a maximum timing curve. It will not keep on advancing to the point of knock.

CPS offset is a tried and true method of modifying spark advance on many cars. This ECU, however, will learn CPS offset out, and based on how quickly ignition timing seems to advance back after a knock event (per Scalbert's testing) I wouldn't be surprised if the CPS offest is learned out within three WOT pulls. That doesn't it's useless for perhaps reducing knock during adaptation a bit (as I have always conceded), however, I don't think it can legitimately be called timing control.

As explained above, these DME is not inherently different from other DMEs as far as adaptive knock control logic. There is no more "learning of CPS offsets" and there is of learning of ignition timing mapping. Once again, CPS offsetting has the same exact effect on spark timing as remapping of the active ignition advance map.

IMHO, real timing control would entail directly driving the ignition. This removes all limits CPS control has, and lets you decide exactly how much advance/adaptation/etc you would like under a given set of circumstances. For example, at 13psi/55*C IAT on a map designed for 91 octane, you could decide that you NEVER want the ECU to add more than 7* of advance at 4600 RPMs, and if it tries to add more, you can modify the feedback the ECU recieves about the current level of timing and knock sensor input, to ensure it doesn't try to modify spark advance, and since you're directly driving the igniton, you can have the spark happen whenever you like, say exactly 7* of advance.

The same effect is accomplished by finding the max advance value of the factory DME at that given load/rpm point. If it is 10deg and you decide you want to limit it to 7deg, you simply add a CPS offset of -3 deg. Simple as that. If you want to apply additional trims based upon IAT, you can do that to with this table using the same logic:

This would allow you to program in more "safety margin" than what the stock ECU has, since we are dealing with higher stakes. You could even modify the amount of noise it had to get from the knock sensor to retard/hold stable adaptive advance, by simply adding more noise to the signal as desired.

Correct me if I'm wrong, but I don't believe current piggyback technology allows for that, or you'd already be doing it.

Also, if you see a major flaw with that solution that I don't that would make it somehow inferior to CPS offset, point it out to me.

My point is simply that I don't believe CPS offseting can be realistically called "timing control". Perhaps it helps prevent SOME knock during adaptation. Frankly we haven't seen evidence of that yet, but I know Scalbert has some more testing to do along.

Don't taking personally, I just thinking saying CPS offset is timing control is a bit like throwing a "good enough" label on it, when it isn't good enough. I think it's an imperfect solution at best, and I think trying to make it sound like a better solution than it is discourages people from pushing for progress, and a better solution.

I respect your opinion but it would be nice if we can bring all the facts to the table before making statements about what each approach can and cannot do.

Shiv

[runningman]

Originally Posted by shiv@vishnu

I respect your opinion but it would be nice if we can bring all the facts to the table before making statements about what each approach can and cannot do.

Shiv

Looking at the timing logs posted the stock values ("ceiling") is 11-13 degrees depending on RPM. Maybe more depending on IAT. The "tuned" values with CPS offset are hovering around 6-9 degrees, lower with no CPS offset. Since 6-9 is less than 11-13 the tunes are all bumping against the knock threshold. Correct me if I am wrong but I believe that is E82tt6's point.

[OpenFlash]

Originally Posted by runningman

Looking at the timing logs posted the stock values ("ceiling") is 11-13 degrees depending on RPM. Maybe more depending on IAT. The "tuned" values with CPS offset are hovering around 6-9 degrees, lower with no CPS offset. Since 6-9 is less than 11-13 the tunes are all bumping against the knock threshold. Correct me if I am wrong but I believe that is E82tt6's point.

That's not how I read it. But then again, english isn't my native language. And unless I'm mistaken, he was referring to inherent differences between the two approaches. Nothing to do with the current tunes bumping against the knock threshold or not.

BTW, most of the PROcede logs I've seen have shown logged timing look just like that of a stock car after reasonable amount of adaption.

The good news is that there is now a tool that we can all use to further refine maps so that everyone gets the performance and repeatability they want.

It would also be interesting to get logs of those running the currently available reflashed ECUs. I'd be curious to see if they exhibit similar behavior at the lower power levels they run. They seem to be as repeatable on the dyno as the Procede results posted recently.

Shiv

[Septro]

Good review and it looks like you found a tune that matched your needs. The best bang for the buck statement is pushing it imo.. you may mean 'the most features I like for the buck', but the difference in price is stark for the performance or the true bang of the product.

The datalogging, downloadable maps, and user adjustability features are solid differentiators well worth the price of admission to some customers. CANBus is not until the feature is actually out and in production.

Also with full disclosure - you bought your own tune and received no discount correct? <this may be required for all review posters from here on out>

[jeho]

Originally Posted by Septro

Also with full disclosure - you bought your own tune and received no discount correct? <this may be required for all review posters from here on out>

Yes, i didnt get the discount cuz since this was my first tune and did not have JB or previous Vishnu to get the upgrade price,
BUT I DO NOT REGRET IT whatsoever...
I thought the price might of been a tad bit higher than I was expecting to pay BUT now i realize and understand why it ISNT, also with these new features that are planned to come out as shiv posted.

I bought it when it was 3.1, then 3.2 came out and had more user torq. adjusting....Soo gooood

[mmmotornutz]

I am confused, if the V3 controls timing and the JB3 doesn't, then why are the timing logs from each tune from Scalbert's test look similar??

[OpenFlash]

Originally Posted by mmmotornutz

I am confused, if the V3 controls timing and the JB3 doesn't, then why are the timing logs from each tune from Scalbert's test look similar??

They don't. The jb3 showed twice as much knock retard activity during the midrange roll-on tests and significantly lower timing curves during the single gear full sweep test.

Shiv

[mmmotornutz]

Originally Posted by shiv@vishnu

They don't. The jb3 showed twice as much knock retard activity during the midrange roll-on tests and significantly lower timing curves during the single gear full sweep test.

Shiv

The knocks are in two different places but still show both tunes experience knock retard. Looking at the timing curves, the overall look the same.

[OpenFlash]

The PROcede logs show more than 10 deg of timing everywhere except for ~500rpm band. The Jb3 everywhere except for a ~2000rpm band. Not what I would call "similar."

And the fully adapted runs are even more different with the jb3 logs in the single digits everywhere, IIRC. It's good that you are happy with your jb3. But data is data.

Shiv

ps. That drop of timing you circled in green is no more knock retard than the drop in timing the jb3 sees at 3400rpm (which you didn't circle).

[runningman]

Originally Posted by mmmotornutz

The knocks are in two different places but still show both tunes experience knock retard. Looking at the timing curves, the overall look the same.

Both appear to have the same amount and severity of knock to me. Based on the data presented both are relying on the ECU to "learn out" timing.

Shiv does bring up an interesting question with how the flash tuners are handling timing. Are they really making changes to the timing tables, or just letting the knock system sort it out like the piggybacks?

Originally Posted by Shiv

ps. That drop of timing you circled in green is no more knock retard than the drop in timing the jb3 sees at 3400rpm (which you didn't circle).

We're looking at timing under WOT. In one chart you can see boost has stabilized for a few hundred RPM and in the other appears to be in the middle of an off boost to on boost transition.