Quarter Scale Merlin V-12

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
After the holidays and when the house went quiet again, I decided it was time to finally replace the motor in my lathe. Replacing it required a couple afternoons that included hashing-up some fixtures to help me maneuver the old motor out and shoehorn the new one into place. The close-fitting enclosure around this lathe helps to keep the shop floor clean, but it really sucks when maintenance or modifications are required on the machine. Most of this is my own doing, though, since the location I chose for it in my shop made the enclosure unremovable.
Even though I was assured the replacement motor would be the same physical size as the original, it isn't. The new motor protrudes outside the rear envelope of the enclosure by a quarter inch or so. This required a large hole to be cut in the rear of the enclosure and a cover plate to hide the 'bump'. The motor is also about an inch longer than the original, and so I had to modify the fresh air intake that I recently plumbed inside the enclosure for the motor's cooling fan. After working through a few errors in the updated wiring diagram supplied with the motor, I finally had the lathe running.
The new motor has a few new 'features' that took some getting used to. Most obvious is that its integral VFD and controller remain active for some 30 seconds after its power is removed. If the Mach session is terminated before the energy storage devices inside the VFD have time to discharge, the motor will briefly spin up in some random direction before discharging them. Also, if the lathe is used in its manual mode, almost the same amount of time is required for the motor's internal controller to boot up before the motor can be run. This had me running around in circles the first few times I tried to test the new installation.
I disassembled the original motor hoping to find something simple that I could repair. The integral VFD in this motor is an assembly of four complex circuit boards that fit together in Chinese puzzle box fashion inside a cast aluminum finned enclosure. This enclosure is normally hidden by the cooling fan shroud. There is also a large internal potted inductor assembly as well as a shaft encoder, and so this VFD is not a typical sensor-less type. One of the boards contains four 350uF 400V capacitors sitting on the dc bus. The shrink-wrapped tops of three of them appear to be bubbled which isn't a good sign. The power connector to the motor windings was badly overheated (melted), and indicates the motor has drawn some excessive currents. A temperature-related insulation failure would fit the symptoms I've been seeing, but the failed/failing caps could be also be involved. The issue with replacing them, though, is that they are soldered to one side of the same circuit board that contains all the heat-sinked semiconductors on its opposite side. It seems that a heat conductive epoxy was used to permanently attach them to the interior surface of the finned enclosure, and so this board may not have been designed to be repairable. This is disappointing since aluminum electrolytics are known ticking time bombs in any product especially when they're inside a close-fitting enclosure with heated stagnant air. In frustration I boxed up all the pieces for another day.
Getting back to the Merlin, I thought I had completed all the machining on the crankcase so I could logically start on the liners. However, I had forgotten about the tapped mounting holes on the rear of the crankcase used to attach the wheel case. Five of these 20 holes were drilled earlier using coordinates from the wheel case drawing in order to attach a temporary rear fixture for line-boring the crankcase. It's critical that the centerlines of the crankcase and the wheel case are identical, but only a few of the wheel case holes are accessible for match-drilling to the crankcase. So, I decided to drill the crankcase and wheel case holes in two separate operations using the coordinates provided on the wheel case drawing. I had also planned to add a couple dowel pins for good measure, but I couldn't find suitable locations for them. Instead, I reamed the wheel case holes for close (.002") screw clearances and hoped their sheer numbers would be sufficient to register the two assemblies.
I stripped down the crankcase (again) and indicated it along all three axes on the mill in order to verify the coordinates of the previously drilled holes and to drill and tap the remaining holes. I also scraped away some more investment that I found hiding in some of the corners of the casting.
Before drilling the holes in the wheel case I first had to do some foundational machining on the casting. The wheel case is an overwhelming casting that will eventually require a lot of precise machining since it supports a number of geared take-off shafts and countershafts in addition to the supercharger. The gear spacings associated with some of these shafts will not adjustable, and so there will be some non-forgiving machining ahead.
The first step was to face the front mounting flange to the crankcase. This had to be done iteratively and in small steps with the rear flange to which the supercharger will mount. My particular casting turned out to be significantly warped, and Ihad a lot of difficulty distributing the casting errors between the front and rear flanges so circumferential variations in their thicknesses were not so obvious. The notes warned that this casting might be problematic but cannot be straightened due to its complex shape. The mounting flange for the timing chain cover was then faced just enough to get it flat. This surface defines the horizontal axis of the wheel case and was used as one of the references for the mounting hole coordinate system.
The coordinates are referred to the center-line of the crankcase which, in turn, must correspond to the centerline of the supercharger. Therefore the wheel case was moved to the lathe where its crankcase-side flange was mounted to a faceplate. After compromise-centering the casting, the mounting flanges for the supercharger and its bearing plate as well as the opening for the crankshaft on the front flange were all concentrically bored. These operations established the finished centerline of the wheel case.
A confusing note in the documentation also called out the machining of a concentric recess on the front flange of the wheel case for use as a register to the crankcase. I did this without understanding why since the rear of my crankcase is flat, and there is nothing for this recess to register. This note may have pertained to an earlier casting version, or it may make more sense later when I better understand some of the remaining operations on the wheel case.
After spotting, drilling, and reaming the wheel case holes I turned a snug-fitting crankshaft plug to locate the wheel case to the crankcase. All 20 screws freely went in as hoped and were snugged down. I then removed the plug and after loosening each screw one full turn I could measure only a few tenths movement of the wheel case with respect to the crankcase. This gave me some confidence in the alignment of the two sections even though I don't like leaving such things up to screws. Four additional holes were transfer-drilled between the wheel case and the oil pan. Finally, with the crankcase/wheel case assembly sitting flat on the mill table, the timing chain cover flange was indicated. It turned out to be horizontal as expected but was also serendipitously at its finished height above the crankshaft centerline. - Terry

559.jpg


560.jpg


561.jpg


562.jpg


563.jpg


564.jpg


565.jpg


566.jpg


567.jpg


568.jpg


569.jpg


570.jpg


571.jpg
 
Hello Terry,
It's me again. The other Varicon guy. When I read your comment about the 30 second delay in the motor speeding up and the 30 second delay when stopping the motor, it looked to me as if your motor has a programmed ramp up and ramp down parameter that was programmed ex factory as extremely long or wrong or your motor might be wired incorrectly.

I looked up my Varicon manual and there are two parameters that govern the ramp up / ramp down speed on start up and after shut down, expressed in either acceleration / deceleration (min-1 / sec) or HZ per second.

I would suggest to contact your USA Wabeco representative and question him about your situation. Hopefully he can provide you with the answer to your problem. I am not sure whether you already contacted him since you indicated you found mistakes in the wiring diagram that was provided with the new motor. Hope it all will work out, eventually.

Peter J.
 
First off.. I must say wow.. To all of this..

Secondly, about your old VFD. Those caps don't look that bad actually.. Usually a cap REALLY buldges when it goes bad or it blows up completely..

The burnt connector probably has more to do with the trouble you were experiencing.. It has been my own experience in the past that it isn't excess current that caused the heating, but just a loose connection with perhaps a touch of oxidation thrown in.. With all the vibrations in the lathe over the years that connection has probably been worked loose just enough to cause this.

I have fixed similiar issues just be de-soldering the plug, and then cutting the wires and soldering them directly to the board. But as you mentioned it sounds like the heat sink covers the solder side making this difficult..
 
Peter,
I did emailed him but received no reply.
It isn't a ramp up or ramp down time that I'm seeing. It is a 'boot up' time required by the controller before it will even accept commands. After waiting that period of time just after power is first applied, the motor responds normally. The wiring discrepancies were related to one of the wires in their diagram being a different color from the one actually used in the wiring harness in my machine. It involved one of the 24volt direction-select wires, and so I felt on safe ground using the wire I thought was correct for my machine. It turned out that the new motor actually ran in reverse to the old one and so I ended up having to reverse the pair of wires in addition. The second discrepancy was a shield ground for the signal cable. The manufacturer originally grounded this shield to the machine at both ends of the cable, and there was a terminal provided for grounding it in the original controller housing. The new controller had no provision for this ground. It isn't good practice to ground a shield at both ends, and so I didn't have a problem just leaving the shield open at the controller end, but I asked my dealer to check with the factory anyway. He often just ignores what he feels are PITA emails. - Terry
 
First off.. I must say wow.. To all of this..

Secondly, about your old VFD. Those caps don't look that bad actually.. Usually a cap REALLY buldges when it goes bad or it blows up completely..

The burnt connector probably has more to do with the trouble you were experiencing.. It has been my own experience in the past that it isn't excess current that caused the heating, but just a loose connection with perhaps a touch of oxidation thrown in.. With all the vibrations in the lathe over the years that connection has probably been worked loose just enough to cause this.

I have fixed similiar issues just be de-soldering the plug, and then cutting the wires and soldering them directly to the board. But as you mentioned it sounds like the heat sink covers the solder side making this difficult..

You might be right about the intermittent connector. That could explain my symptoms as well. If you look at the photo I posted, the burned connector is around The rightmost blackened blade inside the white header on the pcb. I think you've inspired me to repair that connection and put it all back together.
 
Terry,
Seems I didn't read your original problem description thoroughly and correctly. I do understand now that your problem is not a ramp up / ramp down matter but a power up / down situation.

Having a similar Varicon motor of about the same size on my milling machine, I looked and tried to see if this kind of behavior is found on my motor as well but must say that it does not behave as you described. When I power up, the motor is instantly ready for use and on powering down, the motor does not turn in any way.

If your motor turns briefly on power down, I could imagine that this could lead to an injury if one is somewhat distracted and may get hurt. If the motor behaves as you describe, I would assume there is something wrong with the motor and its internal controls / set up and may need replacement.

I am sure your Wabeco representative (Assuming its MDA Precision out in CA) will resolve this matter to your satisfaction. Them, being of Swiss heritage, do take care of their customers.

Peter J.
 
I don't know that particular controller but I've had similar problems.

The boot delay being on power up - thereafter the motor is switched using the "run" or "enable" switch points - if you simply switch the controller on and off at the mains you will keep getting the boot up delay.

Those "Hot Joints" are bad news and need to be repaired / bypassed (by soldering directly to the blade if necessary).

I agree with the prior response by brendonf that the caps are probably O.K. - they normally bulge significantly or blow completely (those indents in the can are deliberate weak points to "control" the explosion of a detonating capacitor).

As regards your build, I've been watching in fascinated silence - there aren't enough superlatives (as said by others) this is just fabulous, mythological levels of workmanship.

The amount of effort you put into documenting what you are doing is a great service to all of us following this build and is greatly appreciated.

I can't wait to see this finished.

Being British I have a special place in my heart for the Merlin engine - without which I might have been writing this in German.

Regards,
Ken
 
Brendanf,
Your comment about the burnt connector being the source of my problem made so much sense that I spent the day repairing the connections and reassembling the motor/controller. I set it up on the bench fully expecting the problem to finally be solved, but after an hour or so it began intermittently acting up again. There is a 'controller ready' led on the controller circuit board that should be ON when the controller is powered up and all is well, and I noticed that this led is OFF when the motor refuses to start. The 10 volt analog voltage that is generated internally so the motor speed can be controlled with an external pot drops to about 6 volts when the controller is acting up. There is so little else that can be done without a schematic that I think the problem is going to be nearly impossible to troubleshoot without a schematic. Probably the best I will be able salvage is the motor, itself, with an external VFD. Again, thanks all for your insightful comments. - Terry
 
Last edited:
Intermittent faults are a sod to dope out.

Yours might be temperature related given that it happens after an hour - try a can of electro-freeze (or similar) and try selectively cooling suspect parts - this is a fairly standard bodge for this kind of problem.

The next place to go looking is bad joints or cracks in the circuit board / tracks.

Remove and re-plug anything that has friction connections and examine the connections. Plug in and out several times to clean up. This includes plug in I.C.s etc.

As regards the board / tracks, examine with a high power loupe or magnifying glass - anything suspicious solder a jumper across any two convenient terminals either side the "crack".

It gets worse with multilayer boards where you need to take resistance readings from point to point (make a list - particularly those near zero) check again when the fault recurs. Again solder bridges over the suspect.

I once did this over several weeks before I finally nailed the sucker.

Hope this helps.

Regards,
Ken
 
Hi Terry:

I agree with everyone else' analysis of your controller issues. I've been in the electronic maintenance game for 40 years now and I've been through this type of problem many times. The freeze mist is a really good idea and should get you in the area of the problem components. Spray a small area of board and see if the problem goes away. In this respect you are lucky that it is intermittent as you can duplicate the problem. I'm confident you'll find it in this fashion.
You didn't say what you did to rehabilitate the burned connector. As others have said it is a loose connection that causes it to burn up and it will be difficult to make it tighter once it's damaged. So soldering the wire directly to the board would be a good idea. Or somehow bypass the bad pin(s).
In addition I have found many times that the solder joints on the back side of the board where the connector pins go through can have hairline cracks. This comes from years of vibration of the wiring of the connectors. This may be why the pin was heating up as much as a bad connection in the connector itself. I think you said you can't get to the back side easily. But if you can, re-solder ALL of the connector pins on the PC board and any high(er) power device leads all over the board. (power resistors and transistors etc.) You can usually see the crack(s) around the pins under high magnifications but it easier to just re-flow and add a bit of solder as it is to look for them.

Oh. And amazing work on the engine too. :)

Good luck

Sage
 
Last edited:
Dave,
Thanks for the suggestions. It turned out that the male-side of the burned connector on the power pcb was just four standard terminal blades soldered through a plastic header. I had a number of female crimp terminals that fit snugly on the blades, and so I cut the melted connector off the harness and replaced it with four individual terminals. Without access to the rear of this board I was afraid to do any soldering on it. Before inserting the terminals over the blades, though, I coated the blades with Nyogel. This is a contact grease that I purchased back in the early eighties to solve oxidation and fretting corrosion problems with the tinned Molex connectors that were used in some PC busses at the time. This particular grease not only keeps air away from electrical contacts to eliminate oxidation, but it also enhances conductivity by becoming electrically conductive under the influence of an electric field. In addition, it doesn't harden from heat like some silicone-based dielectric greases. I still have the original jar that I bought and still use it to this day in landscape lighting fixtures and various motorcycle and automotive electrical contacts that I've had to deal with.
Although the Freeze-mist is a good idea for troubleshooting heat sensitive problems, it wouldn't very practical in this particular case. Three of the circuit boards use multi-pin pcb connectors between them for various interconnections. With the power pcb essentially being epoxied to the interior of the enclosure for its semiconductors' heat-sinking, the whole assembly has to be inside the enclosure to be functional. This leaves only a small opening for troubleshooting and pretty much eliminates any directed access for the Freeze-mist nozzle.
While it was disassembled I should have examined the solder connections on the pcbs that I had access to under a microscope, but I was too focused on the burned connector. I'm including a few photos that I took during reassembly so you might get a better idea of what I'm trying to describe. - Terry

572.jpg


573.jpg


574.jpg


575.jpg


576.jpg


577.jpg
 
After seeing those pictures I'll have to agree that's a tough one to deal with. Certainly designed to be a FRU (field replaceable unit) not something designed to be repaired for sure. I hope you lucked out on replacing the slip on connectors since, as mentioned, the problem could just as easily be a bad solder joint on the back side of the board heating up the pin. I guess you can only do so much with that. A shame they made it that way really.

Oh, BTW. We used to dig through epoxy coating quite easily with an electrically heated tool sort of like a small knife but a soldering iron with a wide tip would do as well. A bit stinky but epoxy decomposes easily with extreme heat and you can dig down to the board to get at connections. But you'd want to limit that exercise to ones you know are bad because it's a bit of smelly work.

Sorry to digress from your wonderful engine build.

Sage
 
Last edited:
While the crankcase/wheel case assembly was still set up on the mill, I transfer-drilled and tapped the fourteen 1-72 mounting holes in the timing chain cover flange to match those previously drilled in the cover. In a second set-up, the through-holes for the three idler sprocket shafts were also drilled and reamed. The center hole was slotted since its sprocket will be used as the tensioner for the chain. I placed the tensioner sprocket on the crankshaft centerline even though its boss is a bit off-center.
Flanged ball bearings were then pressed into the three idler sprockets that I machined back in September. Shafts for these bearings were cut from drill rod, and setscrew flats were milled on their ends. I temporarily assembled the cylinder blocks and heads onto the crankcase so I could check the fits and clearances of all the chain cover components. Since I haven't machined the rear driveshaft with its integral cam drive sprocket, the final chain length can't yet be determined. Fortunately, the chain appeared to ride in the center of its housing. This was something that I'd been concerned about since a lot of parts need to line up properly to make that happen. The clearances between the chain and the i.d.'s of the two inside cover tubes, though, will ultimately end up being a function of the position of the tensioner sprocket in its slot.
I could already tell that final installation of the chain is going to be difficult especially with the cover tubes in place. So, I wanted to make sure that all the cover components fit properly at this point so the chain won't become an even bigger issue later.
I was initially going to let the three sprockets find their own running positions on their shafts since it would simplify their installation. After feeding the chain down through the cover tubes a few times, though, I decided to turn three pairs of centering bushings to insure the sprockets could not move off the tubes' centerlines.
When I machined the chain cover components a few months ago I could only rough cut some temporary cover tubes and test their fits since I had no overall dimensions to work with. Fabricating the actual tubes turned out to be a bit tricky since the outside ones with their mitered ends required a lot of trial and error fitting that resulted in several scrapped parts. When I finally had a working recipe, though, I made a few spares. The ends were shaped on a sander using a miter gage since the soft thin wall aluminum tubing didn't play nicely with my mill. Using a pull-string I was eventually able to snake the timing chain through all four cover tubes and around the five sprockets and verify it didn't rub anywhere. My hat is off to the designer of that chain cover. I was sure it was going to need help from a Dremel tool.
I'm still trying to decipher the wheel case drawings and understand how the supercharger castings are to be machined and fitted to the wheel case. Details of the diffuser ring seem to be missing, and currently it seems that so much material will have to be removed from the impeller that I'm not at all confident in my understanding of the single and very busy supercharger drawing that was supplied. It seems it was here that Zapjack's build (http://www.usinages.com/threads/rolls-royce-merlin-v12-echelle-1-4.42350/) became stalled a few years ago. The documentation and notes for this portion of the build are very sparse, and so I expect my own progress will slow some as well. I've emailed John (http://www.homemodelenginemachinist.com/showthread.php?t=24717) with some questions since he's successfully navigated his way through all this. My plan is to start working on the (billet) bearing plate for the supercharger which is a less risky part but one that I think I do understand. Hopefully, making it will shed some light on the machining required on the remaining irreplaceable castings. Terry

577.jpg


578.jpg


579.jpg


580.jpg


581.jpg


582.jpg


583.jpg


584.jpg


585.jpg
 
Terry is there no provision for sealing the 3 shafts to the block that the sprockets ride on? Would there not be oil leaking out? Or do you have something planned?

Your work is absolutely fantastic.. Wish I could do this!
 
Terry: It was great talking to you, your wife, and your son at Cabin Fever. I am in total awe of what you are doing on this build. Your work is something above what I would expect to be possible for us mortal beings. OMG. How have you accomplished all of this in such a short time? I now believe in alien life forms on this planet. I have only touched the surface of what you have written, but I have come away shaking my head. Such in depth coverage and your ability to come up with fixtures and tooling to do such accurate work is mind blowing. You deserve some type of medal or an award for your abilities. I am in awe!
I think i said that before but it needed repeating! Keep your sanity if possible.

Ron Colonna
 
Ron,
It was great getting to talk with you again also. It's such a shame we grew up in the same town without ever meeting one another, and now that we have these unique and common interests we live half the country away. Thanks much for your very generous compliments. Looks like I'm really going to have to finish this thing now. - Terry
 
While at Cabin Fever I ran across this gem. I wanted so much to meet its 94 year old builder, but the engine was displayed for only a short period of time during the first day by a friend of his. It's a 1/8" scale Merlin with a two stage supercharger, and it has been run. The entire engine was built from scratch (no castings) over a ten year period by, obviously, a master craftsman. Even the scale fasteners were machined. I took a lot of close-up photos of this beauty, and now I finally understand the coolant flow path through the engine. This one sets the bar at a height that's really difficult to see over. - Terry

image.jpg
 

Latest posts

Back
Top