Since BMW turned to petrol turbo motors, the interest of the “chipsers” in these engines exploded. One of the options is how to “upgrade” the engine’s performance – change its software. The second option – install the outer “black-box” without intervening in the DME itself. This second option is incredibly luring because it is simpler and does not put the guarantee in danger (when you have to visit the dealer, simply disconnect the equipment, and nothing indicates possible intervention).
The basic principles described below apply to other BMW turbo engines: N20, B38, B46/48, N55, and further, according to the specifics of each engine.
The most popular product in the industry of the other “power boxes” – JB4. Price of the unit is around EUR 500 .. 800 (depends on the taxes paid and assembly).

What does JB4 do?
JB4 cheats DME by telling it that the air pressure in the inlet and the air mass consumed by the engine is less than in reality (measured by the car’s sensors). As a result, DME increases air pressure and flow until necessary/required, but the airflow (as it turns out) is more significant. More air = more power. At least, that’s the idea of JB4. But – how is it in truth?

JB4 connections
JB4 is connected to the MAP (pressure sensor of the inlet manifold) and MAF (air flow sensor). Plugs of both these sensors are conducted to the JB4, signals (in the range of positive boost pressure) are modified, and already revised readings are sent to DME. JB4 is also connected to the diagnostics socket. From the diagnostics socket, JB4 receives the supply voltage (12V) for its needs and communicates via the CAN data bus. Communication via CAN bus allows switching the management modes “on the fly”. There are different modes – starting with “stock”, continuing with several maps of increased power, and finishing with the “custom” option.

JB4 problems
One of the problems – if JB4 is connected, the OBD socket is occupied. It is not possible to connect and read EDIABAS data. Yes, newer JB4 releases have their own OBD socket (or BT connection, by choice), but the amount of data via it is very modest (these are OBD data; the amount – is incomparably narrower than EDIABAS). Why is this a problem?
First – if the engine starts to experience problems, it is impossible to connect INPA/ISTA. Even the OBD Mode 6 option to see the misfires is missing.
Second – analysis of these EDIABAS data is critically necessary to analyze the engine’s performance (ensure that everything is OK when any of the increased power maps is switched on, not even talking on a custom map).
The second significant problem is that if the inlet manifold parameters are cheated in such a way, all other management maps (fuel, ignition, etc.) will work in the segment, which is inappropriate for the actual situation. Yes, the difference is as if not so large – 10 .. 15% (if the maps with +50hp addition are used, I recommend not even trying to test “more powerful” versions), but actually, the consequences can be critical. “Feature” of JB4 (and other similar products) – these other management maps will “adapt” using adaptation maps. For a perfectly complete engine (all components without/close to 0 parameter deviances), there should not be any troubles with a difference of 10 .. 15%. The problems hide in another place – depending on the peculiarities of the exact engine, already with the stock management maps, separate adaptation maps could be pretty close to their min/max allowed corridor. Trying to “move” these adaptation maps by these/even 10 .. 15%, there could be (and are) situations that/when some of the adaptation maps will reach the limit of the corridor. As a result – error messages in the DME error message memory; reduced power (limp mode is activated – any of its types), and often even EML (Check Engine) lights up in the panel.
A perilous situation, in this case, appears with B58 and its TwinScroll turbo aggregate. This turbo aggregate has a different load for cylinder groups of 1st to 3rd and 4th to 6th cylinders. Using such power increase boxes, nobody knows how significant the difference of the fuel mixture will be in precisely these groups in the mode of ample torque (exactly in this mode, the inlet manifold data are faked). DME will adapt and try to maintain the average fuel mixture (because the B58 series engine is equipped with only one Lambda probe, joint for all cylinders). Accordingly, in one of the groups, the fuel mixture can become lean; in second one – rich. And nobody will know it (including DME itself).
The third is the main problem. JB4 encourages – switching the management maps “on the fly”. And there is no mention of readaptation of the engine (what is that; when, and why should it be done). Both are crucial nuances. Switching the maps without readaptation of the engine – is the fastest way to damage the engine. Why?
As some know, DME creates multidimensional adaptation maps of each hub (throttle; wastegate valve management; fuel pressure sensor; each injector, etc.) for a longer time. These adaptation maps contain temperature; fuel quality, and many other dimensions. If JB4 switches the maps without warning, the consequences are harsh. Let’s see the transition “from stock to JB4”. The user increases power (previously, the vehicle has worked with stock maps). What happens?
a. JB4 cheats DME in indicates reduced inlet air pressure and airflow;
b. DME increases boost pressure for it (and the airflow) to correspond to the planned. Now the actual air amount is more enormous than DME thinks;
c. the increased amount of air means a lean fuel mixture. DME sees it and uses STFT to fix the situation urgently.
The impact of the STFT is not momentous/without inertia. During the first seconds, the fuel mixture will be lean! A lean fuel mixture is hazardous to the engine! If the engine is allowed to adapt correctly to the new conditions, the situation would possibly be more or less acceptable. But, by switching these boost maps, the user completely paralyzed this principle of adaptation creation. If the STFT of any current engine reaches the max value (+30%), the fuel mixture will stay lean till the moment when DME will understand that “everything is bad”. At this moment, DME will reduce the boost pressure or switch the turbo mode. But during these seconds, which DME will need to react, a lean fuel mixture will do anything possible to damage your engine!
The second aspect – in swiftly changing modes, DME works in open loop mode, which means – readings of the Lambda probes are not taken into account; DME works only by maps (management maps + adaptation maps). DME switches to the Closed-loop mode 2 .. 3 seconds after the stable driving conditions (required torque and RPMs) is reached. And here, JB4 has two options:
a. do not modify the boost data in these modes to avoid the crash of DME (“falling out with error messages” or switching to the limp mode);
b. fake data from the first second without any consideration of the consequences.
In the first case, in the mode of swift spurt (0 .. 100 km/h), there will be no gain in the 1st and 2nd gear; in the second scenario… Such “upgrade” is already out of control.
Finally – this reserve of 10 .. 15% BMW has intended in case of fuel of poor quality; for mountain areas (the oxygen level in the air is reduced), and for technological allowances. There is a 50/50 chance that your turbo aggregate is not able to “pull” the most potent JB4 maps, and then… EML in the panel, loss of power, and desperate questions in internet forums “my car lost power after installing JB4, what to do? ” are guaranteed.
+80 .. 100 hp, indicated by the manufacturer of JB4, is not realistic to reach. By “realistic”, I mean – without error messages in different driving modes and at least partially safe for the engine. Yes, for newer releases, using programs of max power, JB4 connects to the electric wastegate and, obviously, checks – if the DME keeps it completely closed for a long time (obviously, the turbo aggregate “can’t handle”), reduces the fake amount of the pressure (for DME not to “fall out” with the error messages regarding boost pressure). This partially solves the problem with DME “fall out” with issues of the boost pressure, but everything previously mentioned – remain. It is not even worth starting a topic regarding ignition maps and knock sensor data when switching the maps – “everything is bad”.
The obstacle is that developers of JB4 are not even mentioning the readaptation of the engine and, at the same time, recommend switching maps “whenever you want”, indicates that these people, unfortunately, have minimal knowledge and understanding of how DME works nowadays. Yes, BWM manufactures sturdy engines, but I will not challenge destiny by trying to connect the JB4!