Farish 37 Lowering, DCC Sound, and Detailing Project

[justin1985]

I've been building up quite a stock of new generation Farish Class 37s, which are beautiful models. The main problem with them is the excessive body-bogie gap. I've also acquired a set of the exquisite Shawplan Extreme Etchings detailing parts for this model - finer windscreens and exceptionally fine roof grilles and fans. 

 

And I've also been getting more curious about trying DCC sound ... so this project is meant to be about doing all of these things to a Farish 37 - currently 37 239 "The Coal Merchant's Association of Scotland" (although I might consider changing it to a Construction sector loco, if I can summon the courage to try and remove the sector branding with T-Cut etc ...) 

 

Several people have documented methods of lowering the Farish 37 - some raising the bogies, some lowering the body. It seems like when it first came out people went for a very "belt and braces" approach, but as time has gone on, people have gone for lower impact changes. I've largely followed Jonas' blog for this. As I'm adding DCC sound at the same time, I've adapting the method a bit, and also took inspiration from the YouChoos DCC sound tutorial.

 

The first step is to unsolder the lighting board connection wires. I found that once the iron was applied the wire popped out easily. I then used a solder sucker to tidy the PCB up and leave the hole open to re-use.

 

The next job was to cut down the notch in the lighting connection PCB that seems to be the critical part in defining the height of the body. Like Jonas, I used a white paint pen to mark along the side, but then used the point of a pre-set digital calliper to scribe a line exactly 1.5mm up from the existing bottom of the notch. Then I used a fine blade in a fretting saw to cut the notch out, and tided up with a file. My logic was to start with 1.5mm and then file further to get to the "about 2mm" that others have recommended. I might actually have taken it a bit too far, and in future I'd keep it to 1.5mm. 

 

The cab interior units did definitely also hold the body up higher than I would have liked, so especially as I would need the space for DCC sound, I snapped these out. I remember seeing someone saying they had filed the bottom surfaces of these units down by 0.5mm or so to retain them, which I might consider doing in future if I move the speaker down into the underframe tanks - but for now I've left them out. The other job was snapping off the lugs on the underframe tanks (one had fallen off already) and reattaching it temporarily with bluetack. Then the ride height was looking pretty good! The bogies do just about swing freely, but are at risk of catching if the track isn't perfectly level etc - hence thinking I would not cut the boards down quite so far in future.

 

I hadn't noticed until I posed this photo that Farish had changed the bogie moulding to use a much more discrete coupling pocket. I'll probably fit B&Bs anyway. Also, I only just noticed that the pair of 37s here are exactly the same pair Jonas has in his pictures, just the opposite one converted!

 

Rather than chop up the existing main PCB to retain the pickups and lighting functions, as per the YouChoos DCC sound tutorial, I worked out which lighting connection was which, measured the existing resistors, and made up a new board from offcuts of veroboard. I'm yet to test this properly, so not certain the resistance is correct, but I used two 670R SMD resistors I had in stock for a signal project in series to give c.1.4k resistance. The central contact from the Farish lighting connection board is the common, so I connected a blue wire to this, and both ends connect in to one end of the resistors. I wired the other connections with yellow and white decoder wire and paired one of each on isolated sections of the veroboard. So the decoder function wires will connect blue to the other end of the resistors, and yellow and white to the two paired connections - not sure which way around - trial and error I suspect.

 

The only other thing I've done so far is to slightly drill out a hole on two 2 x 1 offcuts of veroboard to sit on the screw in pickup connections, replacing the ones on the original PCB. Face down, with the wire feeding down through the hole to the bottom, should do the trick.

 

On the roof grille end of the chassis (the notices only allow it to fit one way around) there is a bit of a recess under the original decoder socket, so I decided to saw off the "wings" alongside this to make a wider gap, which nicely fits a YouChoos "SugarCurve" speaker. This is the area at the right hand end covered with Kapton tape in the picture above. This crosses over into the cab space, but I don't think its that visible. 

 

The other complication is that if I wanted to fit the Shawplan roof grille, it will also sit over the speaker, and almost certainly prevent the fan blades fitting properly underneath. So, once I've established it all works, I'll think about going back a few steps and removing the smaller of the two underframe tank chassis lumps (the larger one has the main isolated split chassis joining screw) and fitting a YouChoos "Minnow 5" as per this other YouChoos tutorial I've just found. This would let me fit the roof grille, fans, and think about reinstating the cab interiors (if I can cut them down)

 

So, next step is actually hard-wiring in the decoder - I'm waiting until I've got a good chunk of time to sit down and do this. Slightly more apprehensive about soldering in a £100 sound decoder than the usual £30 ones ...

 

Justin

 

[jonas]

Hi Justin, I’m glad you found the blog useful! I think the Farish 37 is dramatically improved by reducing the gap and it’s good to see someone else having a go. I look forward to seeing the whole project come together! I did pick up some 37 bits (Frost grills and a pair of noses with no high intensity light) that might be suitable for your 37239 depending on the time period you are modelling. Drop me a PM if they are of use!

[justin1985]

So this ended up as another of my projects that sat half finished in a drawer for a few months! 

 

The stumbling point was getting the body to sit down evenly and make contact reliably with the lighting contacts on the ends of the chassis. It turned out that the decoder (ZIMO MX648R, supplied by digitrains) was a bit too "lumpy" when placed on top of the motor, causing the body to pivot back and forth on it. The lump looks like it might actually be some of the square type capacitors? So the lumpy bit needed to hang off the end of the chassis lump into the cab space.

 

However I'd already used the lowered notch in the chassis lump intended for a normal decoder for the YouChoos Minnow 5 speaker, so the decoder had to move to the opposite end and hang off into the cab space. The gotcha here was the fact that the pillars for the original PCB screws are in opposite corners of the chassis block, so stuck up under where I wanted to put the decoder. I bit the bullet and carefully cut off one of the pillars (and carefully filed it smooth and covered with Kapton tape).

 

This meant that I had to create a new contact for the split chassis pick up for the decoder on that side of the chassis. I just drilled a hole using a 0.7mm bit and then used a cut down 14BA brass screw, simply screwed in to the soft metal to tap itself. I gave the decoder red and black wires a little square of Veroboard on the end and screwed them into the two sides of the chassis, one into the original pillar, the other using my new brass scew. both press the copper board\solder lump into paint free parts of the chassis by the screws.

 

If I was starting again I'd put the decoder in the provided "dip", crossing over into the cab, and the speaker in the fuel tanks - which would mean milling off the extensions of the chassis that fill them. I didn't fancy dismantling absolutely everything to do this safely without getting metal filings into the mechanism or decoder once I'd already started though. I have left plenty of slack on most of the decoder cables, especially the speakers, making a bit of a cat's cradle, so I could still move things around a bit more if I ever did feel like it.

 

 

This also meant I had to move the improvised Veroboard LED resistor board - on top of the motor now (with kapton tape between). After the last picture I worked out the contacts for the lighting:

 

Contact A - red LEDs - I used yellow cable

Contact B - common - blue

Contact C - white\yellow LEDs - I used white cable

 

The veroboard is now quite well hidden, between two layers of Kapton, and under a bundle of cable, so no picture. The basic layout is 

 

decoder yellow - 670R resistor - 670R resistor - one yellow, one white 

decoder white - 670R resistor - 670R resistor - one yellow, one white 

decoder blue - 2 x blue 

 

Obviously I'd use a 1.4k resistor if I had some handy!

 

I was still left with the problem that the lighting PCB contacts weren't reliable - the whole body was now sitting too low! I'd cut 2mm off the contact PCBs that define the height, and this was clearly too much. I ended up fitting some offcuts from Evergreen strip, 0.75mm thick to the inside of the metal channels that the PCB slots into. So this gives a total drop of 1.25mm - this is the amount I'd cut off the PCBs next time around. It now seems to sit nicely and firmly at a respectable height and with reliable contacts to the LEDs.

 

 

Compared with another unmodified 37, this time a renumbered Dutch 37/0 (the pictures show up how wonky the numbering is - might have to redo this!) you can see the drop quite nicely - looks just right, I'd say. Oh and the sound is pretty good too! (sorry no video yet)

 

 

The final glitch is that in the fiddling with the lighting PCBs, I seem to have managed to blow the white (headlight) and red tail lights on one of the end PCBs. The yellow LEDs still work, and if I swap this PCB with the one from the other end, that works at either end, so the problem is clearly the lighting PCB itself, rather than the resistor board or the decoder etc. I asked Bachmann if I could get a replacement lighting board, but they don't have them - their best offer was a whole new body. So I've ordered some new 1206 surface mount LEDs, and I'll try replacing them myself. If no luck, then I might go back to the Bachmann offer ...

 

J

 

[Steven B]

Does the decoder you've used have extra function outputs? If so you can use them to give independent control of the red lights. This means you can turn them off when pulling a train.

 

 

Steven B.

[justin1985]

Does the decoder you've used have extra function outputs? If so you can use them to give independent control of the red lights. This means you can turn them off when pulling a train.

 

 

Steven B.

 

That occurred to me after I'd chopped off the green and brown function wires! (doh!). Something to think about for the next hard-wired conversion, I think.

 

I've not used a decoder with extra functions like this before, so am I right in thinking the tail lights would need to be separately on F1 and F2, so switched independently? I suspect it would also mean re-mapping all of the function codes in order to make it reasonable to access via the controller handset. I'm using a Roco Lokmaus2, so its a bit of a faff to access functions beyond 9.  

 

I was planning to just dial down the brightness of all of the lights using CV60, which would make brake lights shining mid-train less conspicuous. In general I'm really pleased with the ZIMO decoder - lovely smooth control and great sound. This and the CT ones I've used in shunters seem much better than the Digitrax wired ones I'd previously used in "DCC friendly" Farish diesels, and Bachmann own brand 6 pin decoders.

 

Justin

[Steven B]

I've only done it with a Loksound Micro v4 from Olivia's Trains (fitted to a class 37). That sound project had a couple of spaces in the function list that could be used for the red lights at each end. Remapping was done via JMRI using a Sprog.

 

I'm planning to do something similar using the Zimo MX622N, using F1 to control the red lights on the No1 end, and F2 for the red lights on No2 end. In this case I'll need to attach wires to the decoder itself as the extra function outputs aren't wired to the 6 pin plug or as pre-fitted flying leads.

 

Steven B.

[Mike140]

The lowered 37 looks so much better. Very well done. How well does it run? Wouldn't the lower body catch on the wheels as the bogies turn?

It's a shame the lights blew. I personally have taped off the connections to the lights on one end of every loco I own as I cannot bare the tail lights being on when the loco is coupled to a train. It really makes me wonder why they don't simply fit a switch to the newer models for the tail lights. They have done in the past.

[justin1985]

I haven't noticed any problem with the running at the current height. I've tried testing it on both 1st and 2nd radius PECO set track, and it does still go around the 1st radius curve, albeit with a bit of a growl - although I'm not sure that is actually the bogie or wheels rubbing the body. It seems totally fine on the 2nd radius curves.

 

The SMD LEDs arrived, but I'd mistaken the sizes - 1206 was much too large. I did try fitting one of them anyway, and it did work, but was much too bright. I've now re-ordered some 0603 LEDs to try again.

 

I made a short video of testing on second radius curves with some of the sound functions.

 

 

Justin

 

Edit - video added. The board was on the floor and my cat wanders in to the shot half way through, thankfully he didn't attack the loco!

Edited January 22, 2018 by justin1985

[Mike140]

I haven't noticed any problem with the running at the current height. I've tried testing it on both 1st and 2nd radius PECO set track, and it does still go around the 1st radius curve, albeit with a bit of a growl - although I'm not sure that is actually the bogie or wheels rubbing the body. It seems totally fine on the 2nd radius curves. 

 

The SMD LEDs arrived, but I'd mistaken the sizes - 1206 was much too large. I did try fitting one of them anyway, and it did work, but was much too bright. I've now re-ordered some 0603 LEDs to try again.

 

I made a short video of testing on second radius curves with some of the sound functions.

 

(upload to Youtube much slower than I was expecting - will edit this post once its live)

 

Justin

 

I always thought the entire reason for the high body position was for clearance of the wheels. If they still run on set track I wonder why they can't lower it themselves. Very odd.

I'm also extremely envious of your Coal Sector 37! I can't find any!!