Review: PMDG 737 NGX

This review was originally published in German on our sister site and has been kindly translated by Martin Boehme. All references in the article lead to the original articles in German.

Precision Manuals Development Group (PMDG) has followed suit and recently released their interpretation of the Boeing 737 NG. This was a release that the community has spent a long time waiting for – after all, the airliner was initially announced for 2010. “Good things take time“, as they say – but has a “good thing” indeed come of this?

Our reviewers Oskar Wagner, Dr. Stefan Benzinger und Holger Kistermann have taken a closer look at the Boeing 737-800 / -900. To read their verdict, click “read more”…

By the way: This is the second part of our three-part 737 NG review special. We have already reviewed the product from iFly. After this review, we will write a comparison between PMDG and iFly. Stay tuned!

Introducing the reviewers

Oskar Wagner was a Captain with a large Swiss airline for many years, flying a large variety of types, and he also worked as a Fokker company test pilot flying new aircraft in preparation for delivery. Today, he consults for and tests various Flight Simulator projects. He will tell us how realistic PMDG has managed to make the systems simulation of the 737 NG.

Dr. Stefan Benzinger is a physician in Dachau and has been a member of the simFlight review team for half a year. He is now our “man for the fine details”, and the focus of his assessment will lie on visuals, haptics and sound.

Finally, Holger Kistermann. A marketing and sales director in „real life“, he is also a beta tester and project consultant and has been with simFlight for ages. His task will be to chair this review and formulate the final verdict. In 2003, by the way, he reviewed PMDG’s Boeing 737 NG for FS2002 and FS2004, and he later went on to review the various Boeing 737 NG from Ariane, too.

Short characterization of the product

Before we get down to the meat of the review, we briefly want to characterize the PMDG Boeing 737 NGX in the wider context of the current add-on market. After all, this is an add-on that has been awaited for a very long time, with the focus of interest lying on the desire for a realistic implementation of the real aircraft. Three companies are currently competing for the implementation of the best Boeing 737 NG simulation: iFly and PMDG – and strictly speaking Ariane, too, though the latter product has been on the market for a few years and can no longer keep up with the other two. iFly released their Boeing 737 NG fleet only a few weeks ago, and we reviewed it here. Though iFly has won the race with PMDG to be the first to release, many enthusiasts see PMDG’s product as being far out in front – after all, PMDG’s developers have already delivered impressive proof of their skill in the shape of other complex airliner simulations. For PMDG, this is already the second time they have created a Boeing 737 NG: In 2003, they released the first version, which was implemented for FS2002 and FS2004.

PMDG’s Boeing NGX product philosophy

It has already been decided that, in addition the Boeing 737-800 and -900 reviewed here, there will be an update at a later date containing the -600 and -700, the two shorter variants. If there is sufficient customer demand, the ER variants of the -700 and -900 may follow, but a final decision on this has not been taken. All of these models will only be available for FSX; an FS2004 version is not planned – after all, such a product already exists in the shape of version 1.

PMDG’s product philosophy aims not only to deliver a realistic airliner implementation but also to offer corresponding accessories. Hardcore simmers will surely be delighted to note that printed manuals are available covering everything from Crew Training to the Quick Reference Handbook. All of this at unbelievably high prices – but this is an indulgence die-hard 737 simmers will grant themselves.
PMDG enable their customers to equip their aircraft as they please – the real aircraft is available in different configurations, and the same is possible in the simulator. With winglets or without. With the characteristic eyebrow windows or without. Honeywell or Collins MCP? The customer has the choice. Each configuration can be tuned further to one’s heart’s content. Should the autopilot control be realistic or simplified? Should steel brakes or carbon brakes be simulated? You can probably tell that many of these questions will not interest most virtual Boeing captains, who follow the motto: “Can I use this to recreate my last summer holiday flight or not?” But those who have knowledge of the subject matter and know what the different configuration options mean will find great satisfaction in them. By the way, practically all adjustments are made via the Flight Management Computer and take immediate effect without having to restart the simulator. Wonderful!

A PMDG product is always complex?

True. The Boeing 737 NGX is certainly not an airplane that can be recommended in good faith to an absolute novice. Instead, it is an add-on for intermediate or even (very) experienced desk pilots. And this is a good thing – after all, there are already more than enough other freeware and payware options for novices .This is not a “game”, this is a simulation of a working environment. The deeper you are prepared to involve yourself with this simulation, the more demanding it becomes to fly the beast realistically. This is fun! Once you start to achieve your first successes with this, you should be ready to hazard a step into a real Level D simulator and try your hand at the real thing.

Holger: Oski, why do you recommend that even experienced Boeing 737 NG pilots should use the documentation (at the very least parts of it) to familiarize themselves with this add-on?

Oski: I would indeed make this recommendation in those cases where this experience has been gained with older FS aircraft. The reason is simple: The PMDG B737 NGX has really been simulated almost completely in all systems – including Flight Guidance, of course. This means it is certainly worth your time to take a look at an up-to-date manual. You will become aware of many details you would not previously have paid attention to.

Holger: „Complexity“ is a term that is hard to define precisely in virtual aviation. Even more so if it takes place at your PC at home – with a system that may have cost 1.700,- € in a generous case and can “only” deliver a corresponding performance. Is it even possible that this PMDG software can simulate a Boeing 737 realistically?

Oski: To a certain degree: yes! You will never be able to learn to fly using a computer simulation, but you can certainly operate a B737 using correct systems procedures – this add-on proves that beyond a doubt. You can familiarize yourself with all of the complexity of correct system operation to a very large degree – if you put in the necessary effort. And of course this entails that you need to obtain the necessary documents and work them through. The commitment that this requires in terms of cost and time should not be underestimated. I doubt, however, whether the majority of simmers will do this – and in my view, it isn’t necessary. To me, at least, part of the fun of simming consists of flying different aircraft types. Doing this at the full operational system depth would, in my opinion, definitely require too large an effort – just think of the objections we hear that add-ons should be suitable for casual “after-work” flights. But this is in fact where one of the great strengths of the NGX may lie. In principle, it does not take particular effort to fly. With a bit of practice, the FMC can be programmed in 10 minutes at most. 5–10 minutes are enough to configure the rest of the systems for „home use“. Even though the systems simulation is extremely comprehensive, the aircraft itself, apart from the FMC, is not particularly complex, as long as you don’t punish yourself with systems failures.

Holger: Can you give an example of a particularly well-implemented complex workflow that is possible with the PMDG aircraft if you exploit it to its full potential?

Oski: I don’t even have to go that far. It’s enough to consider an engine failure at or shortly after V1 and the correct response to this problem. The first steps are PPAA -> Power – Performance – Analysis – Action to achieve safe single-engine flight; next, we want to restore system redundancy to the degree possible – while of course not forgetting to navigate and– under coordination with ATC – to manoeuvre to a position from which, after re-programming the FMC, a single engine approach can be made to the departure airport, if this is possible from a meteorological and operational point of view. If not, we will of course need to divert to the takeoff alternate. This is already complex enough but can be carried out flawlessly with the present add-on. This also implies that single engine flight is simulated quite well – at least in terms of performance. The only weakness I have found in this regard is the way the engine loses thrust when the fuel lever is closed. In this situation, the thrust should drop away almost immediately. But it doesn’t do us this favour; instead, the engine spools down to zero quite calmly, which may be much more pleasant from a flyability point of view but does not correspond to reality. It makes the transition to single-engine flight much too harmless.

Holger: In spite of everything, we find ourselves in a two-dimensional world: We sit at a computer screen and try to operate a simulated three-dimensional world using a mouse. Stefan, can this complexity actually be operated well? The way Oski describes it you probably don’t have much time for wild mouse acrobatics.

Stefan: To me, personally, the interaction concept of a panel is one of the topmost quality aspects of an aircraft add on. What use is the most detailed systems simulation if I have to click through three sub-panels in critical phases of flight with high workload? Or if the FMC crashes completely because sloppily programmed clickspots cause me to flip the wrong switch on the overhead and power down the flight computers? Or if I adjust the bank limiter instead of the heading bug? In such conditions, a “flow experience” will never come about. In the real world, after all, engineers go to incredible lengths to create a human-machine interface that is as optimal as possible – this is true in many fields, but particularly in aviation.

It is clear that PMDG has given a lot of thought to this area, and this has resulted in a first-class interaction concept. Even in critical situations I can make all necessary inputs using my mouse hand without having to let go of the yoke, for instance to open a panel (I have a Saitek yoke/throttle unit and rudder pedals). The basic concept takes some getting used to at first but should for the most part be familiar from previous PMDG products (particularly the MD-11). Once you have internalized the concept, you can conveniently operate the NGX to its full extent without any major investments in home cockpit hardware. The concept is explained extensively in the documentation and has even been improved over previous products (MD-11, Jetstream).

Another point that definitely deserves a positive mention: All of the load management – passengers, cargo and fuel, including the correct distribution across the aircraft – is carried out from the cockpit via the CDU. There is no external tool for this, everything can be configured directly from the Flight Simulator session. I’ve had to wait a long time for this convenience – the last payware product that offered this luxury was the Super 80 Pro by Coolsky/Flight1. All of the various equipment and setup variants are also controlled via the CDU – this is intuitive, and changes are made at once. A reload of the aircraft is not required, which can save a huge amount of time.

Holger: The „complexity“ is of course not restricted to the systems simulation but continues with the visual model. Was the design team able to live up to the high standard in the area of graphics?

Stefan: As a longtime model maker I can answer this with a clear and – I admit – slightly surprised „yes“. To date, PMDG has not necessarily been known for eye candy; an accurate rating for the external models of previous years might be “solid work”. The aircraft they have delivered this time round, however, is really quite beautiful and modelled in great detail; it even features a cabin that is visible from the external view, which is not really typical for PMDG. Here at the Benzinger home, it is customary that I build a Revell model kit for every new FS add-on I buy – and in this case, the PMDG model clearly beats it!

Holger: When the BAe Jetstream was released, Robert Randazzo, the head of PMDG, wrote that there would be no more add-ons with 2D panels. I was a little surprised that the 737 NGX now has one after all. Are there any operations that can only be performed in the 2D panel?

Stefan: No, on the contrary: To start the engines, you have to switch to the VC to open the fuel valves. This is because, strangely enough, the throttle quadrant does not have its own 2D panel. Also, flaps and spoilers have to be assigned to a key combination or joystick axis if you want to fly in 2D mode only, or you have to switch to the VC to operate them with the mouse. The 2D panel concept has thus not been implemented completely, which I don’t quite understand. Who knows, maybe an update will address this.

Holger: Oski, you fly with a multi-monitor setup, right? You should welcome the existence of a 2D panel, I assume? Has PMDG thought of pilots like you?

Oski: Stefan has already mostly answered this question. It has also been discussed in the PMDG forum. Obviously, at least I’m not the only one with a preference for 2D panels. It is, after all, more realistic in the end if all necessary instruments and systems are in plain view.

I hope no one is going to claim that they use the VC to carry out the complete initialization of the FMC, for instance. A 2D panel is simply more convenient for that. But the thing that really amused me slightly was RSR’s explanation for why the throttle quadrant was left out: He said it was due to lack of time. But of course it is true that you should equip yourself with enough hardware that the mouse has to be used as little as possible – no matter whether you use the VC or the 2D panels. My solution to this has been the VRInsight MCP combo on one hand and a touch screen on the other. And there is an almost unlimited range of solutions to this problem. The usability of the FMC on multi-monitor systems is bad – to put it mildly. I say this because if the 2D FMC is on a secondary screen, you cannot use keyboard entry – you can neither use the tab key nor activate the input line in the scratch pad. It just doesn’t work. You are left with only two options: Either you move the FMC back to the main screen, or you use the mouse in the old-fashioned way and click around on the input pad until your fingers are sore.

Holger: …but the virtual cockpit is of course completely sufficient! I fly using TrackIR, and I don’t need the 2D panel at all. But let’s get back to the core topic of complexity for a moment: Do I have to be an “expert” to fly the 737 NGX within the boundaries of the “Normal Procedures”?

Oski: In simming, „expert“ is always a relative term. When I browse various forums and see inexperienced simmers asking questions on the level of “what button do I need to press to take off”, I do have to say that a certain level of “expertise” is required. And I also believe that you will quickly lose interest if you operate at the level of merely pressing buttons. You should bring a bit of systems knowledge to the table. But – that usually takes a bit of effort, and many simmers still shy away from that. And this is something I don’t really understand, because if you do have a little systems knowledge, you can quickly prepare the airplane for flight, and you will definitely have a more profound experience. But an expert in the sense of the question – you certainly don’t have to be that. You simply need to know the standard procedures.

Holger: So the complexity of this add-on can definitely be learned and experienced – without needing a Level D simulator to do it. A home PC in a standard configuration is enough. It’s great that this is possible today. Even a few years ago, flight instructors in airliner flight schools would have shaken their heads at the mere thought. What is also great is that PMDG attaches equal importance to the visual aspect of the flight experience. The Boeing 737-800 and -900 simply look incredibly good, too!

Exterior values: Model, performance, and sound

„First impressions count“, as they say. And PMDG has done a good job at leaving a good first impression. To guarantee some “ahhhs” and “ohhhs” from the FSX community, PMDG has collaborated closely with Boeing. The exterior and interior model are based on the original blueprints from Seattle. What this means in practice is evident to the customer the first time they load the 737. Provided, that is, that they have suitably powerful hardware…

Holger: Stefan, if you look at photos of a real 737 NG and compare them with the PMDG copy, you conclude that it comes pretty close to the original. You are the connoisseur among us, with an eye for correct design. Is the first impression misleading?

Stefan: Well, in direct comparison with other 737 NG implementations I have to say that overall, the VC in those implementations is noticeably lower, making the view through the forward window less high and wider. This really applies to all of the working space of the captain. As a result of this, the “A pillar” in these airplanes is more inclined, and the bulge does not remind me as much of the nose of my former Latin teacher as the one in the PMDG. In the PMDG, the whole panel is somehow less wide, everything in the VC seems more confined. I have to say, though, that I don’t have any real aircraft experience to compare to, nor do I have the time to climb into one of the fixed-based simulators in my vicinity. The image galleries on don’t really help either, because many of the images have been post-processed and stretched. Which of the implementations conveys the most “realistic” feeling of space? I don’t know.

As I have already indicated, the visuals of the 737 NG exterior model are as a whole correct and feature many details. In my view, PMDG has made clear advances here. But it is only natural that the cooperation with Boeing – which has been stressed time and again – and the use of original plans and blueprints should also have its effects here, and not only on systems depth.
And after all, the community is made up not only of pilots with virtual type ratings and hardcore simmers who plan and execute their flights with great dedication and knowledge, but also of many with an aesthetic inclination who frown at sludgy textures or mushy pixels and complain if exterior features such as pitot tubes or antennas are missing. PMDG obviously did not want to expose themselves on this flank, which is why the add-on contains a balanced mixture of detail and functionality, without losing itself in resource-hungry bling-bling a la CaptainSim.

Holger: Frank, who recently wrote the review for the DA-20 Katana, pointed out the cockpit crew animations to me. They’re wonderful, don’t you think? The animations of the crew are great – down to the smallest detail. The other animations are also executed very well. From the extension of the flaps to the wing flex on takeoff! All of this, in combination with the high-resolution textures and the well-executed model, gives the aircraft an incredibly realistic effect.

Stefan: Another example: The elevators have trim tabs with their own animations – at extreme deflections, they move in opposite directions. The aerodynamic purpose of this is not quite clear to me, but it is definitely beautiful to watch. The hydraulic cylinders of the slaps and flaps are also modelled in 3D. The engines don’t look entirely “new from the factory”, and the numerous official PMDG repaints are also a joy to behold. Special mention should be given to the lighting of the exterior model: The position lights that reflect of the ground just look great!

Of course, visual treats like these have their price. After all, every animated object has to be calculated dozens of times per second – light incidence, reflections, shadows. Add to this the effects on flight characteristics, systems, gauges, and so on and so on – to tell the truth, I am amazed that a simulation of this depth and visual quality is even possible on a single PC. This, too, is a credit to the developers: Consciously crossing the boundary of what had been thought possible to date and pushing it out – true to the American pioneer spirit of “bigger, faster, further”. In this respect, the NGX is a new benchmark by which many future developers – as well as future alternatives to FSX as a platform – will be measured. For it is an interesting question whether there will ever be a comparable product for the coming X-Plane 10, for instance – and if so, when? Things will remain interesting, my friends.

Holger: “Performance” is a good key word. Oski, how does the NGX run for you?

Oski: My current system consists of an i7-960 @3.2 GHz, 6GB triple channel memory and two GTX285 graphics cards with 3 displays in total. All of this runs under Win7 64-bit. I have good frame rates (20+ for instance at Mega-Airport Zürich) when in the VC and even better frame rates (30+) with the 2D panels. I have no problems whatsoever with CTDs – at least not so far, despite using various Mega- and Orbx airports and ASE. What I observe time and again is that there are suitable and less suitable configurations. There is one thing you simply have to realize today: Given the high memory requirements of complex simulations, you’re simply better off with a 64-bit system because, as a rule, the address space is large enough – in contrast to 32-bit systems, where you quickly reach the limits. And this is why PMDG recommend a 64-bit system in their requirements.

There is an amusing aspect to the image update of the NGX in the external view. As the model is obviously very detailed and consists of very many polygons, less powerful systems display a “skeletal” airplane for a shorter or longer period of time as the image updates.

Stefan: A few months ago I treated myself to a new computer because I finally wanted to max out FSX and be prepared for products like the PMDG 737NG. The investment has been worthwhile; with very high display settings, MyTraffic, ActiveSkyEvolution and REX2 clouds, the PMDG flies absolutely fluidly at a Mega airport with a good 20+x fps; at the gate, frames may dip to 12-15 on occasion. My specs are: Intel Core i5-2400 CPU 3.1GHz, 8 GB RAM, Windows 7 64-bit and nVidia GeForce GTX470 with 2 displays or one display for the 2D panel and a projector for the external view.

I probably won’t even install the PMDG on my good old dual-core Vista 32-bit PC from Aldi (though other PMDG aircraft and even the FlightSimLabs Concorde run quite acceptably on it).

Holger: I bought a new computer only three weeks ago. Frank Schmidt’s review of the Diamond Katana might have something to do with that… I now have an i7 Sandy Bridge processor (2600K) that runs at up to 3.8 GHz. Also in the case are eight GB of memory and an Nvidia GeForce 570. For FSX, this is a real quantum leap – and the system is perfect for the PMDG 737 NGX. No jerkiness whatsoever! My old computer, a first-generation quad core with 2.93 GHz (QX6800) should probably be able to handle the PMDG, too, but a mixture of complex sceneries and the 737 NGX would definitely be pushing it! In the system requirements, PMDG even writes something about i5 processors .In my view, that should be considered a warning. In the final evaluation, though, my opinion is that the NGX is still comparable to other complex aircraft add-ons.

Holger: Sound is an aspect that should not be neglected – after all, it accompanies us from the first click on the panel to the completion of the flight – though we are usually unconscious of it. And this is why the better the sound, the better the overall impression. What is your take on the sound, Stefan?

Stefan: In PMDG’s forum, Robert S. Randazzo, the head of PMDG, confers the highest accolades on his sound designer, and I would like to echo this praise. To me, the acoustic environment in the cockpit is tightly woven and convincing. In the beginning, the acoustic backdrop seemed almost too loud to me, but if I hark back to the pre-9/11 era I recall that various airliner cockpit visits at the time really left a similar impression on me. I suppose a cockpit is anything but a quiet place. There is always something that hums, buzzes or beeps. No doubt there will always be some room for discussion whether the trim sound, say, is too quiet or whether the spoiler extension on landing sounds quite so or a bit different. But these are minor points, and moreover they depend on subjective perception.

What is essential to me as a lay person is the overall atmosphere, and that is definitely believable. By the way, the setup option in the CDU allows the acoustic backdrop to be adjusted to personal taste – again, this is a real innovation and one of the strengths of the product.

Holger: Oski will be more familiar with the acoustic environment than us – after all, he is much more familiar with the sound from the „front left seat“…

Oski: Well, it’s always a bit difficult to judge the sound on a PC. On the one hand, the specific sound system has a strong influence. I use “only” a sub-woofer and two tweeters – all in all, a pretty standard PC sound system. On the other hand, it is not a simple matter to mix the various sounds correctly. As a user, you don’t have total control over all sound levels – except, that is, if you have some knowledge of the sound.cfg file and fiddle around with it. But this is something I generally don’t do – after all, I want to assess the sound of the model I bought, not the model I fiddled around with. For my taste, the air conditioning is a bit too loud – but unfortunately I don’t know the B737 from practice. The engine sound, on the other hand, comes across as quite convincing. A comparison would have to be made, though, across the whole range from the -600 to the -900. I can imagine that on the B737 the difference in engine sound from the shortest to the longest variant is quite noticeable – I know it is like this between the A319 and the A321. But nothing conclusive can be said yet in this respect since the shorter variants have not been released. A negative point that I noticed – and again this applies to my multi-screen 2D environment – is the fact that various sounds drop out when the focus switches from the main screen to one of the auxiliary screens. For example, this applies to the air conditioning sound and – more seriously – to all of the altitude callouts including the altitude warning. This is very inconvenient and detracts from the overall impression. Better solutions to this are possible, as A2A, for example, have demonstrated strikingly with their sound system. There, I don’t have these problems.

How the PMDG Boeing 737 NGX flies

High complexity, good design and great sound do not necessarily guarantee that flying the PMDG will provide hours of fun with lots of „ahhhs“ and „ohhhs“. After all, an “uh-oh” is something no one wants to experience. But they do occur: In the PMDG forums, more than a few customers have been puzzled by sudden crashes – not crashes from the cruise flight levels but ones of a technical nature. Some of these crashes are even accompanied by a blue screen. PMDG recommends the use of a 64-bit operating system. Where high complexity and exact design meet, even a capable family PC can reach its limits. Don’t forget – in Flight Simulator, we move in a three-dimensional world where the position of the sun, the view of objects and the world, the depiction of weather as well as animations, systems logic and high-resolution textures have to be computed anew with every frame! This is an extreme feat, and to us lay people, it is hard to grasp that it is even possible to execute such a procedure many times a second without a hitch. It seems all the more understandable, then, that such a calculation can, on occasion, go awry… To be quite clear: The absolute majority of customers do not have any problems with the product, and PMDG has already released a hotfix to help customers that are suffering crashes.

So we should really assume that everything runs normally. The PMDG flies great – whether by hand or by autopilot. The systems simulation adds substantially to the hours of enjoyable flying.

Holger: Oski, we really expect an exact systems simulation from PMDG. But let’s put the 737 to the test, Captain! Can the PMDG be flown „by the numbers“? What is your impression of flying the Bobby?

Oski: Well, it would be easy to hand out superlatives. But those who know me know that I generally don’t do this, as no further improvement is possible beyond a superlative. And at this point, I again need to make an important digression: As usual, I have read the whole spectrum of assessments in the PMDG forums from “too sensitive” to “too sluggish”. This assessment depends on important factors – particularly the personal controller settings. This is why here, too, I have my personal “standard setting”, which I never change. In this way, I can at least compare between similar aircraft. For my taste, the NGX feels very good. Not too lazy and not too agile. An important indicator is always the transition to single-engine flight. This is accomplished without overly large rudder deflections and is easy to stabilize – my feeling, as I have already mentioned, is that it is a little too easy! If I consider how an A320 with FBW behaves – which, after all, supports you extremely well during this transition – the NGX seems to me to be a bit too harmless in this respect. The assessment of the flight envelope as a whole, however, delivers a very good result in my opinion. You never have the feeling that something might not be right. Here, again, the fact is that we are not sitting in a Level D simulator but in front of a screen – and those are two principally different worlds. But the general impression is very good.

Holger: I recently talked to a flightsim-crazy airline pilot – it might even have been you, Oski. We got to discussing the behaviour of the autopilot in good Flight Simulator add-ons and real aircraft. I learned that many airline pilots would sometimes welcome it if real autopilots flew as precisely as they do in the simulation. Obviously, “George” also has problems from time to time with leveling off exactly at a given altitude or predicting a descent point precisely. Does the PMDG fly precisely? Maybe even too precisely?

Oski: I hope it wasn’t me who made that claim. My experience with “real world” autopilots has been consistently good and definitely on the level of simulator autopilots – if I may make the comparison in the reverse direction for once. Let me put it like this: The NGX autopilot makes about the same errors as a “real” autopilot – and to a degree that is definitely justifiable and does not to be corrected. For example, at low speeds it has the same difficulties in holding altitude precisely when rolling into or out of a turn as some of its “big” colleagues. Whether this particular behaviour is seen specifically on the NG is something I don’t know. But I do know a number of autopilots of other airliners that have precisely this “problem”. That this is also the case here is an absolutely positive indication for me that the aerodynamic fundamentals have been implemented realistically.

If there is something to find fault with in comparison with a “real” autopilot it would be that turns are sometimes initiated a bit harshly in normal LNAV cruise. If we want to count all of the lateral and vertical flight guidance by the FMC with LNAV and VNAV as part of automatic flight, we have to begin noting the first imperfections. I suppose it is simply impossible to program a perfectly flawless FMC. And so also the FMC of the NGX has a “hiccup” from time to time. It repeatedly happened to me that the flight path displayed on the ND contained inconsistencies. The following image shows an example of this that occurred with the SID RWY 28 ZUE 1V in LSZH.

To what degree the data conversion performed by Navigraph is responsible for this cannot be determined. After all, they convert the data according to the add-on developer’s specifications. What is clear, though, is that even the NGX is not immune to such errors. In cases such as these, this also applies to vertical navigation, where a target speed sometimes does not behave the way it should. It can also happen that you experience lockups if you “tweak” the routing too much in flight. But I don’t want to overstress this. I really took the system to its extremes here. If you consider that the database consists of several tens of thousands of procedures, you can certainly imagine that there can always be a flaw somewhere. Maybe I was just unlucky and happened upon one of these very few flaws. In general I do have to say that the FMC runs quite stably and without errors. It really is a joy to work with. I did not find any function that displayed any obvious errors. At this point I do not want to enter into the tiresome discussion about different V-speeds, for instance between TOPCAT and FMC. They use entirely different methods of calculation and therefore cannot yield the same results.

However, I have to deduct more than a few points for one of my “favourite procedures”, the automatic go-around. Any modern flight guidance system can handle it – though only of course if the initial situation is correct. The NG, too, can perform a go-around from a fail-operational dual autoland configuration, but the NGX can’t. Here, the developers blundered! Instead of raising the pitch to 15° and accelerating with a climb rate of at least 1000 fpm, it tiptoes away with 5–6° ANU and 500–600 fpm climb rate. This is not what the FCOM describes. Honestly speaking I would feel very uncomfortable about this in “real life”, particularly if you consider that one of the main purposes of a go-around is to regain altitude quickly. Holding entries are another sticking point for a precise FMC simulation. Here, at least, there are no wrong entries, but the procedures themselves are flown very imprecisely.

Holger: Are there systems that are particularly innovative or that deserve particular attention?

Oski: Particularly innovative is not something you can really call the NG – and here I mean the real aircraft – even if NG stands for „New Generation“. This New Generation has a few years under its belt now, and the overhead panel really doesn’t look exactly super-modern – to put it mildly. So in that respect there is not really a lot that is innovative. But the EGPWS, for instance, definitely deserves a mention. The depiction is clear and has very good performance within the limits of the available data – after all these do come from the FSX terrain model. And if the nitpickers now go searching for non-functional switches or panels, they will quickly find something in this area: The radar image, which can also be overlaid on the ND, does not exist. The radar panel is dead! Surprising, but it’s a fact.

Holger: Nick Collett from Angle of Attack has published a walkaround through the virtual cockpit on Vimeo that shows not only design-specific details but also introduces some of the systems. We had already presented it on, but we would rather mention it again in case someone hasn’t seen it yet.

Holger: I always enjoy it when an add-on surprises me. In the case of the PMDG this applies to the convincing flight dynamics. On one of my test flights from Munich to Palma de Mallorca I had made an error in the fuel calculation and still had quite a lot of fuel in the tanks during the approach. I could really feel the large mass „pushing“: The aircraft was quite unwilling to lose airspeed and altitude. On final approach, I was only able to reduce to Vref right before the threshold. And during the subsequent autobrake 1 deceleration I had to intervene manually to leave the runway in a reasonably “sensible” fashion. An impressive experience!

To me, the PMDG is a perfect add-on… but in view of the great add-ons that have appeared in the last three months alone, you have to be really careful not to overuse the superlative. What do you think Stefan?

Stefan: You’re completely right. Whereas only a few months ago there was talk in forums and magazines about of a crisis of flight simulation, new developments are now coming thick and fast in the shape of high-quality releases. The present 737 interpretation from PMDG, to me, is the climax of this development (for the present?). But it is also the add-on that has had to contend with the highest level of expectations. In the first few days after the release, there were also some slightly disappointed comments from time to time. In addition to the performance problems on older systems, which were to be expected, there were also reports of inexplicable crashes – CTDs and “freezes”. I, too, was affected by this on several online test flights. But I have every trust in the expertise of PMDG’s team – and in the community’s “swarm intelligence”. To me, the 737 NGX is by far the best and most innovative – and so the most significant – aircraft add-on in a long time. However, a few isolated but relevant problems will definitely still have to be fixed. But please give PMDG a bit of time – it’s not even two weeks since the release yet.

Oski: Objection, your honour! There is no perfect add-on! To be sure, the NGX sets new standards in a few areas that all future aircraft will need to be measured by – but it also does have its flaws, even if they only marginally impact the overall impression. It is very nicely done and deserves a very good rating. Why don’t I say it deserves the maximum rating? Quite simple: The better is always the enemy of the good. I don’t think much of the superlatives that are so commonly used today. After all, if something is perfect, it can no longer be outdone. You can’t improve on “perfect” – syntactically maybe, but not semantically. So let’s leave it at “very good” or “excellent”. There is always space for improvement, and there are also a few noticeable flaws that remain to be corrected.


PMDG has managed to implement a very exact simulation of the Boeing 737 NG for FSX that does not leave much to be desired. There is no question: PMDG has set a milestone in the history of Microsoft Flight Simulator! The PMDG Boeing 737 NGX does have a few flaws, but they do not detract from the general impression. PMDG gives their customers the good feeling of having bought a package that is very complete in terms of a serious Boeing 737 NG simulation. And this is indeed the case!

…If it weren’t for the CTDs! PMDG has currently released the third hotfix for this issue, but some users continue to experience problems. It is reassuring, though, that the developers are hard at work on a solution, and we are sure that PMDG will soon be able to offer such a solution. The CTDs occur sporadically, and the degree to which they occur depends on the PC configuration and any third-party add-ons that are installed. And of course there are also pilots such as Oski who have not experienced a single CTD to date. Holger, on the other hand, has already been unlucky several times.

Quite apart from the CTDs, the general principle of course always applies that „where there is light, there is also shadow“. With this present product, however, the shadows are very few and far between. There are a few isolated flaws, but in view of the huge complexity of the add-on they should not be overstressed. The bottom line is that PMDG has released a truly excellent product!

The one question that remains is about the „reviewers’ choice“. PMDG should really have deserved this distinction, which – unlike the simFlight awards – is awarded by the reviewers. But before we can award the “reviewers’ choice” to this add-on, PMDG needs to fix the CTDs for good – and then we will follow through.


  • Good, extensive documentation
  • Configuration of the aircraft from the cockpit without a restart
  • Realistic flight characteristics
  • Correct systems simulation
  • Collimated HUD
  • Top-notch sound
  • Very cleanly designed models
  • Very high-rate virtual cockpit


  • Crash to desktop (CTD) on some systems (see PMDG support forum)
  • FMC sometimes has “hiccups”
  • Flawed go-around procedure

Developer: Precision Manuals Development Group, PMDG
Distributor: PMDG
Home page:
Price: 69,99 US $, available through PMDG

Test system
Multiple different systems

Oskar “Oski” Wagner
Dr. Stefan Benzinger
Holger Kistermann

0 Responses

  1. Very informative review. Almost makes me want to run out and make the purchase, but my i5 and 3gb ram and windows xp may not be able to keep up. I think my Nvidea GeForce 9800 GT would give it a run for it’s money.

    In the case of cockpit views, I am of the same mind frame as Captain Oski, I prefer 2D panels and was pleased to hear of the inclusion of a 2D panel but disappointed that some things were left out of it.

    Thanks, great review,


  2. I hope PMDG for those who use multiple set up monitors and for more realistic the situation add the First Officer 2D Panel as well. After all not all use VC or single monitor.

Toggle Dark Mode