Toyotec's 2.0 16v MK2 dyno comparison. ABF (Digi vs Kjet vs MS) vs 9A MS. P3 14-11-10

Discussion in '16-valve' started by Toyotec, Oct 6, 2010.

  1. Toyotec

    Toyotec CGTI Committee - Happy helper at large Admin

    Joined:
    Jul 26, 2006
    Likes Received:
    3,097
    Location:
    Creating Pfredstarke
    You cannot conclusively say the 9A car in this comparison is running "low" with this bunch of vehicles as you can argue they all ran "low". Unless you have seen a vehicle with this same spec of engine on these rolls produce a better number curve then that is the data that is presented. The reality is, relative to the rest of ABF MK2 vehicles, it achieved less torque yes, drove well but still matches that fact at WOT.
    Ignoring absolutes, if this test were to take place on another brand of rolls, once the dyno procedure is heeded and the process is controlled as in this exercise, I would expect to see the same type of delta even if all the vehicles where reading "high".

    Your plots has a questionable rpm range for a 16v engine particularly if it has performance cams. That range is more inline with an 8V engine though the plot does look 16v in profile if you imagine a typical 0-7500rpm range compressed. You statement regarding the MK2 9A in this thread would have more validity if all cars where tested on this dyno in your plot. And what initially appears as an operational error would affect all equally. It is for this reason I requested that plots other than what is being discussed on here not to be put up as it introduces a non relevant reference to what is being discussed here. This plot cannot be used to compare what was done here to a specific powertrain as it is not transparent.

    Hope you understand what I am trying to say.
     
    Last edited: Oct 15, 2010
  2. Toyotec

    Toyotec CGTI Committee - Happy helper at large Admin

    Joined:
    Jul 26, 2006
    Likes Received:
    3,097
    Location:
    Creating Pfredstarke
    I certainly agree with what you are saying.
    It would be nice to perform a KR K-jet vs a ABF K-jet. As that is a step that many take to replace worn out a worn out KR. But with these these things it takes a lot of time just organising and validating vehicles prior to controlled testing. The window as to be right.
    I will remember your offer though.
     
  3. 3hirty8ight Forum Member

    Joined:
    Oct 28, 2005
    Likes Received:
    109
    Location:
    United Kingdom
    The ABF is a 'develpoment' of the KR/9a, so by rights it should be more efficient in the way it makes power.

    Whats really under scrutiny here is the engine management side of things - Where K-jet is outclassed hand over fist. Its just the calibration of the equiptment that is the overriding factor.
     
  4. Ben S

    Ben S Forum Junkie

    Joined:
    Aug 1, 2005
    Likes Received:
    345
    Location:
    Essex
    Worth remembering that both Megasquirt cars had only had a couple of hours of tuning on the street. No WOT dyno calibration whatsoever yet.
     
  5. A.N. Other Banned after significant club disruption Dec 5th 2

    Joined:
    Oct 23, 2003
    Likes Received:
    395
    General posts about 9A-ABF differences now split off this thread, a far better home being a continuation of the same topic here from earlier in the year.
     
  6. Toyotec

    Toyotec CGTI Committee - Happy helper at large Admin

    Joined:
    Jul 26, 2006
    Likes Received:
    3,097
    Location:
    Creating Pfredstarke
    ABF digi 3.x vs ABF KJet. Results walked back to original comparison.

    Introduction and baseline method

    The original comparison, performed on the 5th or Oct should have included a vehicle fitted with an ABF engine and controlled by a K-Jet system. This is the original vehicle that was used in the How to ABF your Mark 2 Golf thread.
    Finally this vehicle became available and to ensure the results generated were comparative, the red Digifant 3.2 vehicle was sourced as a control to the exercise.
    First the Digifant vehicle was run up and the results in terms of atmospheric conditions were accounted for so that the result could be compared to the original comparison. The result which was within 2% of the original test and was considered to be transparent. With a reliable baseline established, the K jet car was then strapped on the rolls and ramped up.
    As the two vehicles had different methods for picking up rpm, dyno overlays were not possible. Instead a spreadsheet used to display differences captured.

    My thoughts on Digifant vs K-Jet results .

    Digi 3.2
    [​IMG]
    Dav's ABF digi car is a very well kept and serviced example of the this conversion. It was good to see after so many years after the conversion, this powerplant was still pulling very strong. As the vehicle was left with me over night I can report it had excellent cold starting performance and in traffic the car behaved as any modern Golf.
    We have seen on rare occasions more powerful "std" Digifant ABF engines on this rolls for various reasons, but this example is typical of your average good working ABF Digifant unit. Having driven the vehicle to the dyno I could confirm it felt the same as the last comparison run. With the throttle set to WOT this vehicle pulled very urgently from 2500rpm and gathered plenty of pace when the revs rose to 4000rpm. In top gears this urgency would dampen slightly as the 6950 cut out was approached. So how did this factory example fair when re-baselined?
    [​IMG]
    This is what I would expect nominally, from an ABF engine w/VW Digifant on these rolls.

    K-JET
    [​IMG]
    Don's ABF KJet car surprised me. I knew it was aggressive when the throttle was flattened but did not expect the engine to achieve these strong torque readings. In these WOT conditions, it had a more obvious torque dip at 3500rpm which was not as noticeable on the Digifant vehicle. It should be pointed out that this "dip" never dropped under 138lbft in this vehicle. Surprisingly this engine did not rev to 7200rpm as first thought but to 6800rpm. This was double checked with the dynometer pickup. This 6.8K rev limit threshold was later associated with the early 16v ignition modules in MY 86-89, mentioned in this forum before. Not that it had to rev any harder. Still defiantly this was much stronger in terms of max torque 156lbft compared to the max 142odd lbft from the Digifant vehicle and trust me you would feel that! In gear acceleration was pretty fierce with the engine really pulled the car urgently was the revs got to 4500rpm, then it seemed to want to accelerate at the same pace right to the soft rev cut out similar to a KR engine just with much stronger acceleration. No doubt in top gears, a tail off in acceleration may be felt, but rather than try to attempt on public roads, lets see how the dyno measured this sensation.

    [​IMG]

    Not bad for an old head and block purchased from Steven's VW dismantlers years ago and limited to many iterations of 5mm Allen key and 10mm ring spanner adjustment!

    By capturing the WOT performance for two engines hardware types installed in conversions, we now have and upper and lower torque profiles that can be used to assess other similar engines performance.

    ABF Digifant 3.x vs ABF K-Jet plot
    [​IMG]

    Applying these dyno numbers to real life and using Tesla's GTECH accelometer, it can be demonstrated, these two MK2 vehicles can accelerate to 60mph in the very low 7secs range for the digi 3.x or slightly less than 6.8secs for the K-Jet, which on OE 020 gearing and tyres is much faster then when 16v MK2 cars left the factory.

    What custom mapping can do.

    The ABF Digifant 3.X vehicle in the above example has good part load performance. The mapping strategy is much more improved than the old Digifant 2 GTI 8v cars. There are no nasty torque transitions between idle loads, part load and full load conditions, a topic that many an enthusiast has complained about, as there are many bodges and workarounds for the 8v Digifant 2. Idle accessory loads are compensated for via fast path ignition scatter as well as slow path ISCV. The transition from a point of part load to full load is very progressive too. Unlike what has been witnessed and experimented with in Digifant 2, the ABF 3.x fuel calibration tends to maintain a robust stoich up to a rpm and part load threshold and then runs a richer mixture. With work already on the way to remap one of these modules, but with the understanding of what is happening with the ECU logic, the results as displayed when and aftermarket controllers are installed can be repeated with the OE controls. I plan to demonstrate this soon in another thread using a MK3 16v and another vehicle with an aftermarket controller both mapped for maximum performance and response dyno and vehicle time dependant. The research into the STD Digifant 3.x ECU shows it can even be used in turbo applications.
    Ben_S MS 16v car, when tested on it's first dyno baseline already started to show typical characteristics an ABF engine that has accessible engine controls with a good enough road cal. In his case the two established baselines shared in this thread show what limited (K-Jet) or no easy adjustment (Digifant 3.x) to fuel and spark controls can achieve. With his car already displaying a good trace albeit with less torque than the K-Jet vehicle, I am sure with further mapping, the traits of the K-jet and the Digifant can be further combined. After sharing the above information, Ben has been working on his cal to increase and flatten is WOT torque, response to various throttle and load positions and drivability in part load and we will see how it all come tougher in future updates. Once the new road cal is re-evaluated for drive improvements, dyno time will be purchased to perform more work on the WOT curve. What the final product will measure we will wait to see.
    In the case of Mike and the 9A on MS again with more mapping time on the dyno I am sure the torque profile can be improved to superseded that of the Digifant ABF vehicle, but not match the torque profile of either the MS vehicle or the K-Jet vehicle. Again dyno time dependant we will wait to see what happens.

    So there we have it.
    4 types of EA827 2.0 16v engines.
    4 very different traits.
    3 different methods of controlling them.
    All MK2 vehicles.
    Tested using a very controlled process.
    Dyno Dynamics 450DS machine used to measure power runs.

    My conclusion based on this exercise and testing many more 16v engines.

    9A/6A/ACE engines are very good upgrades to replace your worn out KR lump. As in my How thread you can get them to perform better than std, but it involves more work due to replacing intake cams, getting the cam to crank phasing right was well as transferring components, for 6A/ACE, that make the head and block a transverse application. It is more involved and in the end you will end up with a combination that is quicker than a well fettled KR engine or just slightly short or about even with performance of a factory transplanted ABF on VW Digifant. My data and experience so far suggests for a 9A or otherwise to be significantly quicker, component modifications will have to be undertaken at increasing cost.
    When an ABF engine is used as a swap engine in a MK2, the vehicle acceleration characteristics seem to change significantly right from the onset if for example the old engine was a KR 16v. Why does one ask. Plus simple, well thought out improvements can extract further power and torque in a way that could be quite expensive if the base engine was a 9A or otherwise.
    Let us review some facts we know about the base ABF engine code.
    The ABF engine is already a transverse application found in a host of pre 97 Seats, 3A chassis Passats and The 1H0 Golf MK3. In the search for more performance for little outlay, the base long engine does not have to be opened to swap internal components. NVH improvements can also be noted as seen by the use of an aluminium oil pan and various brackets to link to the block to the gearbox in addition to lower block reinforcements. There is the use of pistons with thinner rings and a shorter skirt, for reducing internal friction and when coupled to longer 159mm rods, transmits more piston force to the 92.8mm crank pin for longer durations. The block is capped off with an improved cylinder head with less valve shrouding, around the 27mm exhaust valves as well as more flow around the exhaust valves helped by higher lifting cams. It is my belief that these combined improvements are what gives this engine the edge as a factory unit when swapped in a conversion vs a 9A/6A or ACE variant. This view becomes stronger as more and more 16v engines are tested and validated on the dyno and show similar patterns to what is being described. Unless this data trend changes the conversion to NASP ABF power will be one I will always recommend over the other 16v options when you are about to freshen up your older Golf from that aging 16v and even 8v lump.

    Acknowledgements

    Many thanks to Garage Streamline for use of their DD 450 2400BHP machine.
    Davinder for the red ABF Digi at short notice again
    Chandrin for GerTIe the ABF K-Jet,
    Mike for his 9A MS vehicle and
    Ben for his cherished 1.6 driver w/ABF MS fury.
    And to those who perfer to remain unnamed for making this comparison happen.

    Regards
    Toyotec.
     
    Last edited: Feb 26, 2016
  7. chrismc Forum Junkie

    Joined:
    Oct 22, 2003
    Likes Received:
    0
    Eddie,

    Great write up- I agree as a direct bolt-in upgrade to a mk2, the ABF is top of the pile. Such a shame an Eibach strut brace is not compatible with the extra block height though.

    Is a great deal of variation required from stock KR values when using K-Jet as a base on a 2L 16v in your experience?

    I have set my idle CO to 2% and now have 8-9degs base ignition timing. With fuel control pressure set to 48psi (richer end of factory tolerances) the car feels to be performing fairly well.

    Allowing for all engines being different do you recommend any other tweaks prior to a dyno session?

    Tempted to try yet more ignition advance as I can't hear any pinking as yet? Engine runs nice and cool with it's mocal cooler which obviously helps...

    With a 1990 KR ECU can a tiny bit of low rpm/high load pinking be tolerated as the benefit is more top end advance and power? I know early/late ECUs have different curves but not sure how this affects how each type of car may be set up?

    I'm quite happy to avoid high load/low rpm conditions if the extra advance is beneficial....

    Any advice gratefully received

    Chris
     
  8. Toyotec

    Toyotec CGTI Committee - Happy helper at large Admin

    Joined:
    Jul 26, 2006
    Likes Received:
    3,097
    Location:
    Creating Pfredstarke
     
  9. chrismc Forum Junkie

    Joined:
    Oct 22, 2003
    Likes Received:
    0
    :thumbup:

    Thanks for the swift response!

    Il try a tad more advance and see how it drives on Wednesday (got a 3hr round trip)

    I have always found the mfa readings are a good indicator to how the car is running. A good mpg return indicates the engine is running well.

    Will be in contact to maybe sort out a day for a setup if that is ok with you?

    Cheers

    Chris
     
  10. A.N. Other Banned after significant club disruption Dec 5th 2

    Joined:
    Oct 23, 2003
    Likes Received:
    395
    Nice update, and even more interesting to see the K-Jet ABF socking it to the Digifant.

    It is interesting to see the ABF outperforming the 6A/9A/ACE at least at this state of tune, and to hear about the piston ring size. That, the rod lengths and the head diffs really help focus the mind on how these units differ, ever so subtly.

    vw_singh's K-Jet ABF was tested at 160bhp with a KR head in Nov 2009. Is that a reasonably like-for like comparison with the above?
     
  11. danster Forum Addict

    Joined:
    Apr 2, 2008
    Likes Received:
    7
    The 051 103 373D heads were fitted to late 6A engines remember.
    Installing an ABF inlet cam to one of them would make the top ends all but the same for an accurate comparison.
     
  12. bens_cab Forum Junkie

    Joined:
    Dec 30, 2005
    Likes Received:
    0
    Location:
    Ashford kent
    paul_c has an abf mk1 on k-jet and he whoops my **** on the strip pulling 14 second dead times
     
  13. vw_singh Events Team Paid Member

    Joined:
    Oct 29, 2003
    Likes Received:
    781
    Location:
    United Kingdom
    Mine did 150bhp on K-Jet. [:D]

    Gurds
     
  14. A.N. Other Banned after significant club disruption Dec 5th 2

    Joined:
    Oct 23, 2003
    Likes Received:
    395
    Damn! Getting mixed up with TrackCab, but that's fettled!
     
  15. tshirt2k

    tshirt2k Forum Junkie

    Joined:
    Oct 21, 2004
    Likes Received:
    325
    Location:
    Herts
    Very good write up as usual.:thumbup:
     
  16. chrismc Forum Junkie

    Joined:
    Oct 22, 2003
    Likes Received:
    0
    My engine is as such- short block 2L 16v with 051 head and now with ABF cams;)

    Only addition is a 4 branch and the obligatory intake/exhaust tweaks:)

    I'd be glad to offer it up for an optimisation/comparison as it's in a pure home-brew tune at the moment:thumbup: As much as we debate it- I remember a few years back with GVK and Tubthumped- Gary with the 9A and Craig with ABF. Their cars always performed very very similarly despite the different motors....both on k-jet

    Makes for interesting reading that the crude old k-jet can match the modern efi setup on a basic engine with good old fashioned tweaking.

    Always had a soft spot for k jet- for the sound alone I think. How sad[:$]
     
    Last edited: Nov 15, 2010
  17. Toyotec

    Toyotec CGTI Committee - Happy helper at large Admin

    Joined:
    Jul 26, 2006
    Likes Received:
    3,097
    Location:
    Creating Pfredstarke
    The 9A in this comparision had ABF cams but not the head. I would like to test the above combinaton in K jet form albeit optimised to the max, but with STD dress kit, to see if the result could match the full ABF engine. As with all these things it takes a lot of time to set up and pre check vehicles.
     
    Last edited: Nov 15, 2010
  18. Toyotec

    Toyotec CGTI Committee - Happy helper at large Admin

    Joined:
    Jul 26, 2006
    Likes Received:
    3,097
    Location:
    Creating Pfredstarke
    No worries Chris. Once I am free.
    Keep tweaking intelligently. It does take time.
    You may be surprised what can be achieved if you make wise tweaks to these things.
     
    Last edited: Nov 15, 2010
  19. R.b!n_16v Forum Member

    Joined:
    Nov 3, 2010
    Likes Received:
    1
    Nice job toyotec! are they all runing stock exhaust system? with a cat?
     
  20. Toyotec

    Toyotec CGTI Committee - Happy helper at large Admin

    Joined:
    Jul 26, 2006
    Likes Received:
    3,097
    Location:
    Creating Pfredstarke
    All on STD exhaust systems, digifant has a full 8v system and the k jet runs a 16v system.
     

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice