From this point on, unless otherwise stated or completely obvious from the context, all abbreviations are written in Cyrillic.

Today, I’m continuing with the assembly of the M198 model.

Last time, I completed steps 2, 3, 4, and 5, so today I’m naturally moving on to steps 6 and 7:

Step 6 is fairly straightforward and easy to complete, so I won’t go into much detail about it. Well, a bit of cleanup is needed to remove some leftover molding marks from a few parts — mainly B42 and B22 — but nothing too serious.

Step 7, however, is a different story.

Let’s start with part B33. As you can see in the video, the locating holes don’t line up with the locating pins, so you’ll need to open them up a bit before the part will fit properly.

But the real “fun” starts when you try to fit the subassembly made up of parts A43 and A28 to what you’ve built so far — it’s simply physically impossible as instructed. Just watch the last 4–5 minutes of the video and you’ll see what I mean. That’s why I called it a day at this point — I need some time to figure out how to tackle it.

So, after about 35 minutes of work, this is where things stand:

And, as usual, here’s the real-time build of these steps:

I’m starting to think that following the instructions quite so literally may not have been the best idea. As you’ve probably seen, I’ve run into a fairly serious problem with the assembly.
So, I bought another copy of the kit and will most likely start over from scratch. I’ll keep this one as a source of spare parts in case something goes wrong again.

From this point on, unless otherwise stated or completely obvious, all abbreviations are written using the Latin alphabet.

Today I’m continuing with the assembly of the M198 model.

Last time I completed only step 1, so today I’m moving on to steps 2, 3, 4 and 5:

First, however, I had to go back a little and “finish off” step 1 – part B16. If you look at the instructions, it isn’t really clear exactly how it is supposed to be fitted, so I took a slightly creative approach and glued it in the position that seemed most logical to me.

The remaining parts are described clearly enough in the instructions, so I don’t think they need any additional explanation – just watch the video.

So, after about 55 minutes of work, this is the result:

And, as you might expect, here is the assembly of these steps in real time:

From this point on, unless otherwise stated or completely obvious, all abbreviations are written using the Latin alphabet.

The M198 is a 155 mm medium towed howitzer, developed between 1968 and 1977 and manufactured between 1978 and 1992 by US Rock Island Arsenal. It has been retired from US service and replaced by the much newer M777, but continues to be used by the armed forces of Bahrain, Iraq, Lebanon, Morocco, Saudi Arabia, Pakistan and other countries.

Kit 02306 is made by Trumpeter and contains four fairly small sprues – one A, one B and two C sprues – along with two large plastic parts, a small photo-etched fret, four tyres, cord and decals. It is a relatively small kit and, at least at first glance, looks easy to assemble.

As is typical of Trumpeter, two painting schemes are provided – camouflage and green – but with no additional information about which military units they represent.

Well, let’s start assembling step 1:

The step itself is fairly straightforward, but, as it turned out, there is one small catch.

It is clear how the howitzer’s breech – parts B13 and B12 – should be glued to the barrel. The other parts, however, can rotate around it (you can see what I mean in the video), and, at least to me, it is not clear how they should be positioned relative to the breech. So I made the wise decision to leave them unglued to the barrel, and I strongly recommend that you do the same, at least at this stage.

The instructions also do not clearly show how part B16 attaches to B9, so I glued it in the position that seemed logical to me. And I have to admit that I forgot part B16 while assembling this step, so you’ll be able to see exactly how I fitted it in the next video

That’s about it. After around 30 minutes of work, this is the result:

And, as tradition dictates, here is the assembly of this step in real time:

From this point on, unless otherwise stated or completely obvious, all abbreviations are written using the Latin alphabet.

Why am I winding the clock back, you might reasonably ask? Well, because I actually completed this stage on 2 May 2025

I built it, filmed it, brought the footage home, sat down to edit the video and… Somehow I couldn’t be bothered. Then life swept me along: I filmed (and edited videos of) Elena and Katerina’s ballet concert, organised the Thracian Ultra ultramarathon in the Rhodope Mountains, worked and… Only today did I finally get around to this post and its accompanying video.

But back to the model. I needed to carry on with step 11:

And I say “needed to” because it wouldn’t be me if I hadn’t forgotten something – in this case, the rather enormous parts A39 Usually I forget tiny little pieces, but apparently I’m making progress in this department.

Anyway, I glued them on, with one small mishap – you can see one of the parts coming loose in the video.

I continued with step 13:

The step itself is easy, with no technical challenges. It is one of the monotonous ones, though: assembling eight wheel rims is straightforward, but repeating the same actions over and over takes a lot of time.

That’s about it. After around 40 minutes of work, this is the result:

And, as tradition dictates, here is the assembly of these steps in real time:

I decided to add a little variety to my videos. I changed the opening, and this is how I intend them to look for the foreseeable future. I’d love to hear what you think of the new format in the comments.

From this point on, unless otherwise stated or completely obvious, all abbreviations are written using the Latin alphabet.

Once again, I find myself facing a paradox Everything in the workshop was supposed to be easy: sit down, film and build… “Yes, but no!” as Bulgarian journalist Petko Bocharov used to say. It turns out this business isn’t quite that simple. As the number of cameras increased – now four – so did the time needed to set them up. Positioning all four cameras (fortunately, this time I didn’t have to move them or adjust the tripod heights), connecting their SSDs, formatting them, setting each camera’s exposure, filming the colour chart with each one… There go 40 minutes. First world problems, as the wise people say

Anyway, back to the model. I had decided to at least start step 11:

The step is not particularly difficult. Fitting the parts to the chassis can be a little awkward, mainly because the locating holes in the chassis are too small for the pins on the parts that go into them. Nothing that a little extra cement – meaning plastic solvent – can’t sort out.

Of course, it wouldn’t be me if I hadn’t forgotten something – in this case, the rather enormous parts A39 I don’t think it’s a disaster, though. I’ll add them in the next session…

And since I still had plenty of time, I continued with step 12:

No problems here. I was actually surprised by how nicely and easily the photo-etched parts W7 and W8 folded. There was a small workshop mishap when one of the stabilisers came loose, but nothing serious. I glued it back on.

But I’m proud of myself for at least remembering the fuel tank cap this time

That’s it. After around an hour and a half of work, here is the progress:

It looks a little like an AT-AT

And, as usual, here is the assembly of these steps in real time:

From this point on, unless otherwise stated or completely obvious, all abbreviations are written using the Latin alphabet.

I had left it looking as though a bomb had gone off in there the other day Not that I left it much tidier this time, but…

It took me an hour and 20 minutes in total to finish arranging the cameras and everything else and… I still have more rearranging to do. But that’s another story.

Back to the model. I had decided to at least start step 10:

At first glance, it didn’t look like much work, and that proved to be the case. Nothing particularly difficult either – only fitting parts E32 and E27 together gave me a little trouble. All the other parts fitted together without any problems. I was even pleasantly surprised that, unlike on previous occasions, the air filter practically found its correct position by itself. In fact, if you watch towards the end of the video, you’ll see that I had forgotten to glue the air filter and radiator to the chassis, and they stayed there for quite a while without falling off

That’s it. After around 45 minutes of work, here is the progress:

I’m not happy with how the video turned out, but if you’d like to, you can still watch the build in real time:

From this point on, unless otherwise stated or completely obvious, all abbreviations are written using the Latin alphabet.

My mood improved on the way, though. I knew there would be quite a bit of work arranging the cameras and everything else, but somehow I was looking forward to spending that time on my own… It took around 45 minutes to set everything up, including two minutes to realise that my optical HDMI cables are directional, and the end marked Source has to be plugged into the camera. Otherwise, there’s no picture on the monitor

I had decided to finish step 8:

With other similar models I’ve built that have steering linkages like these – D1 and D3 in the picture above – things never end well: the wheels always end up crooked. This time, I was pleased to find that MENG had done an excellent job, and this particular step was surprisingly easy. Except for part C28. Not that assembling it is difficult; quite the opposite, it’s simple. Once you’ve found part C28, that is

I must have spent three minutes staring at sprue C, blinking and counting the parts – no part 28 anywhere. Finally, I thought to check the other sprues in the box. Of course, it turned out that C28 is a separate little sprue of its own, packed with sprue E – naturally, right at the bottom of the box

And since I finished fairly quickly, I moved on to step 9:

I ran into three problems here. But let’s take them in order…

First, part C53. MENG let themselves down a little here. It is supposedly possible to glue it in only one orientation, but the locating groove is rather… loose, so the part can rotate through quite a large angle. I positioned it by eye…

Second, parts A54. I’m not quite sure how I managed to fit these. It is clear how they should be assembled – the instructions even draw special attention to it – but… They have to be tucked underneath another part, which is an awkward process involving a fair bit of force. Looking at it, that other part could easily have been glued on after installing A54…

Last in order, but first in importance: attaching part D32 to the assembly built so far. A difficult job. Apparently, I had introduced a slight misalignment between the two large parts during assembly – something like 0.05–0.1 mm – but… It made D32 very difficult to fit over the rest. Watch the video and you’ll see what I mean. I’m not happy with how I handled it, but there we are.

So, after around 45 minutes of work, here is the progress:

I got home, sat down to edit the video and… Hit another snag

I imported the footage from the three cameras into DaVinci Resolve, selected Create Multicam Clip, chose synchronisation by timecode and… Got a 20-hour clip. Errr?! It turned out that two cameras (and the audio) were using time-of-day timecode, but camera B was on free run – meaning it had started at 00:00:00… Luckily, I had clapped my hands just before the usable part of the recordings began, so I managed to synchronise cameras A and B by audio. Then, thanks to my brilliant mathematical abilities, I calculated the offset I needed to apply to camera B’s timecode

So, here is what the build looked like in real time:

I won’t be building today – Saturday, when this goes live. We’ll be cleaning the studio, assembling shelves and doing that sort of thing

From this point on, unless otherwise stated or completely obvious, all abbreviations are written using the Latin alphabet.

Nadya and the girls went to visit the grandmothers, while Preslav and I stayed in Sofia because he has classes today.

So around lunchtime we headed to the studio: I was going to build, and he was supposedly going to work on some exercises. Supposedly, because he “accidentally” left his workbooks at home. While I was building, he watched matches or some football programmes. At least he helped me move the last model still at home – the Topol-M.

I had decided to finish steps 6 and 7:

Nothing particularly difficult here. Fitting the two drive shafts, D66 and D68, to their respective transmission components is a little fiddly, but nothing too painful.

I moved on to step 7:

The first slightly trickier step. For one thing, it’s monotonous: those eight brake drums involve quite a bit of repetition, but that’s manageable. The problem is that the instructions aren’t entirely clear about the relative positions in which parts A18 and A10 should be glued, and likewise A19 and A11, A29 and A3, and A28 and A4. So I went a little by eye. Of course, it wasn’t until the third rear drum that I worked out the logical orientation, which meant redoing the first two I had already glued… Have a look at the video. I was pleasantly surprised by how easily the four movable front drums went together, though. And since I have an unpleasant feeling that I’ve messed those up too, at least I’ll be able to take them apart, correct them and reassemble them easily…

So, after around an hour and a half of work, here is the progress:

And here is what the build looked like in real time:

Another small pleasure: I left the cameras where they were and simply cleaned the worktop. On Tuesday, when I plan to continue, I’ll only have to adjust the camera exposures before sitting down to work

From this point on, unless otherwise stated or completely obvious, all abbreviations are written using the Latin alphabet.

After work, I stopped at the workshop – which, incidentally, is quite close to my office – intending to carry on with the model.

First, though, I had to clear the decks a little. We had played Vecna: Eve of Ruin the previous evening, and since it got late, we left everything as it was. So I had to rearrange some chairs and tables and set up the cameras and lighting. There went 45 minutes… Not that I’m complaining!

Anyway, on to the model. Of course, it wouldn’t be me if I hadn’t forgotten to glue on parts C22 and C23 last time

I glued them on and moved to step 5, determined that this time I would pay attention to the picture in the instructions and not forget any parts.

The step itself is very straightforward. The instructions are clear, and the parts fit together and onto the existing assembly without any problems, so there’s nothing particular to point out. For more detail, watch the video

Of course, there were still a couple of interruptions: my colleagues in Argentina and Brazil interrupted me twice. But what can you do? These models cost money…

So, after around 45 minutes of work, I have made some progress:

And here is what the build looked like in real time – well, almost:

I left the workshop as it was and simply cleaned the table. I plan to continue tomorrow, and I hope the only preparation needed will be switching on the lights and cameras…

From this point on, unless otherwise stated or completely obvious, all abbreviations are written using the Latin alphabet.

For the last two or three NASAMS building sessions, I’ve been meaning to finish step 3 properly because, as you may already have guessed, I had forgotten to glue on a couple of parts: the two C3 pieces.

It’s so easy working this way! I had left the tools on the table yesterday, and the lighting was already set up. I simply took out the box containing the model’s “progress”, picked up the box of remaining parts from the shelf, and sat down. I took out the two C sprues, cut part C3 from each, a couple of strokes with the file, a couple of brushfuls of glue and… Ivan and Martin called to say they were downstairs with the chairs and other equipment, and asked me to help carry everything up.

So, after around five minutes of work, this is where I got to:

In fact, taking the photos and putting everything away took longer than the assembly itself…