Puma / Dorman 6LE / MacFarlane B46B / 110 kVA Generator Resurrection

[awemawson]

Joe, the paint was from Bitec who feature on eBay:

http://www.ebaystores.co.uk/MDA-Bitec/Hamer-Finish-/_i.html?_fsub=25448221011&_sid=210466991&_trksid=p4634.c0.m322

Pete, sorry but no, no bits picked out in different colours, it's not a Fairground Ride 

I decided to investigate the wiring for the upper instrument panel and associated current transformers that feed the main ammeter. As I altered the selector switch I realised something was badly amiss with it. It was so stiff that the plastic shaft was literally doing a 90 degree rotation before it clicked into it's next position. I think I've mentioned before that I'm not keen on these cam switches but obviously I was going to have to dig into it and see if it could be saved. I'd expected a squirt of switch cleaner / lubricant to fix it but it made no difference.

So I carefully documented all the wires to it, and multiple links fitted to it, before gingerly dismantling it. As it came apart it rapidly became obvious that getting it back together wasn't going to be simple as there were several forms of operating cam, and their phasing was individually set by positions on a spline that of course had been lost in the process of dismantling

I found the stiffness to be a general swelling of several of the plastic parts where operating cams passed though locators, but the main issue was the 'four click' main detente mechanism which resisted all attempts at freeing it up to a workable state.

So what to do

Well I could ignore it and not fit the panel - most of it's features are incorporated in the Deep Sea control - but critically not the ammeter, and the Deep Sea unit only derives it's voltage display by calculation from one phase and neutral, whereas the panel measures phase to phase.

So I tested each bit of the panel, the ammeter and volt meter seem to work fine, as does the vibrating reed frequency meter, but the hours meter is making a dreadful racket although it is counting up.

I think the way forward is to make a new panel with just the Voltmeter and Ammeter and devise an alternative switching method. The ammeter is simple just needing a single pole four way switch. (3 phases and off.) The voltmeter is slightly more complicated, switching both sides of the meter to one of three combinations of phases without shorting phase to phase in the process.  Presumably break before make switches are essential. I've not quite got my head around what is needed so any clever chaps are welcome to make suggestions!

[John Rudd]

As you already have one voltmeter, why not just buy 2 more, and then directly wire them across all three phases? Maybe a bit of a faff having to cut a panel....

You can then see at a glance what each ph is doing without having to mess with a switch ( and the issues with the wiring?  )

[awemawson]

John that certainly is an option and might well be cheaper.

I've worked out that I need a two pole four position 'break before make' switch rated at 500 volts. (Switches far less than an amp)

BTW panel cut outs are a breeze with the CNC plasma table 

[John Rudd]

You could go even cheaper and just have a 240v lamp ph-n  x3.....as a simple go-no-go phase indicator....all the voltmeters are going to show is any slight variance in phase voltage....

Depends on how much work you want.....

The 3.6MW sets we had ( when I was working..), their control panels were not complicated but had enough instrumentation on them for monitoring.... The basics, Volts, Amps,  Frequency.....Turbine Outlet Temp, then a nest of modules for control/sequencing on start-up and a seperate panel for excitation/ syncing....when not in island mode, we would sync them to the incoming site supply for excess power management or if there was a chance of supply failure( quite common )

[awemawson]

Turns out Schneider do a pair of specific switches for these applications, and RS Components have them in stock so hopefully arriving in the morning 

I'll now go and sweep the remains of the original switch into the dustbin and clear my desk


(not cheap but at least 'the right thing' ! )

[AdeV]

but the hours meter is making a dreadful racket although it is counting up.

Pfff, like you'll hear that over the noise of the engine  It's probably been making that racket for years!

[mc]

I've had those style of switches apart.
Some cherry pickers use them as a function selection switch, and once the original grease starts to dry out, they get a bit hard to turn, which is usually then a prompt for either the operator (or more commonly one of the idiots I work beside  ) to flood them with WD40/release oil/dewatering spray to try 'freeing it off', which then washes out all remaining grease, and they become impossible to turn, as the plastics used just bind against each.

Once you know how they work, you can normally work in some fresh silicon grease using a small screwdriver, provided you can get it to turn. Failing that, you have to strip it down to get grease between the sticking surfaces.

[awemawson]

MC that switch is now history as it's in my dustbin - I already wasted too much time on it yesterday 

However Parcelforce delivered the new switches from RSComponents and also a full set of documentation that I won for the Dorman 6LE engine - always best to grab these things as they flit past as opportunities are rare - these were on eBay (naturally).

So I started putting things back together to clear a bit of space in the tractor shed. The main air intake filter housings, starter motor, main alternator end dome and ventilation screen, and the lid to the main alternator connections all went on easily.

[awemawson]

BUT - when it came to mounting the fuel solenoid plate and solenoid it was a different matter. The axis of the solenoid was a good inch out of line with the axis of the ball joint on the diesel injector pump.

Then I remembered - prior to painting I'd flattened out what appeared to be a random bend in the plate that certainly didn't look as though it was supposed to be there. I suspect that the solenoid has been replaced at some time and this one sits with it's axis higher than the original. I thought just bending the plate and letting the actuator work at an odd angle wasn't very good,  so I cut the plate and inserted a 1" joggle piece to get the heights nearer kilter.

A quick coat of primer and matt black, and while it was drying I sat down and drew up the new meter panel in Autocad. Dropped it as a DXF into SheetCAM, ported it to the CNC plasma table and cut it out in 2 mm steel plate. The original was 1.2 mm but very flimsy as the square meter apertures are nearly as wide as the panel itself.

I left the mounting holes off as I wanted to replace the (previous) self tapping screws with Rivinuts and they had obviously been set out by eye rather than from a drawing so to match the cabinet I had to use the original panel as a drilling jig. Drilled 3.1 mm, opened up to 5.5 mm for a bit of clearance on M4 screws, and fitted Rivinuts in the cabinet face.

A quick grit blast and a coat of silver hammer finish paint left over from the Versatool Cabinet refurbish, and while the paint was drying I returned to the fuel solenoid plate and got it mounted satisfactorily with it's new joggle.

Once the fuel solenoid was fitted I returned to the new meter panel and populated it. All fits rather nicely

[John Rudd]

The meter panel looks familiar....


Nice job..

[awemawson]

The meter panel looks familiar....


Nice job..

Yes it's lost the Elapsed Hours meter, and the switches are now separate for voltage and current whereas previously one (faulty!) switch did both.

Two more panels to make - one to mount the Deep Sea controller with an eStop button, and one to control the cabinet and water jacket heaters off domestic mains.

[vintageandclassicrepairs]

Hi Andrew,
The new ammeter selector switch hopefully came with a proper wiring diagram?
The switch should keep the feed from the two CT's that are not feeding the meter shorted out
It was drilled into us as apprentices on the danger of open circuiting a current transformer !!
Where I worked there were identical looking selector switches and meters on many panels but usually reading kV and kA
From memory (rapidly failing ) the switch that operated the frequency meter is spring loaded to the off position as the meter is not meant to read continuously ??

John

[awemawson]

John,

The ammeter switch wiring 'as given' is in the picture below.

I can understand why a vibrating reed frequency meter perhaps should only be used intermittantly, but I think that previously it was permanently connected

[John Rudd]

Ordinarily, ct's would have a burden resistor across the reading side to prevent dangerous voltages being developed across them....when not in circuit...

Cant really remember about the Hz meters,

[awemawson]

John, no sign of 'burden resistors' in the original implementation

As an order for some short M5 hex bolts had arrived it was time to re-assemble the engine instrument panel. When taken off I was amused to note that it was only retained in place by the copper oil feed pipe, the three anti-vibration mounts having completely collapsed !

New mounts were not quite the same as the originals so a bit of ingenuity with custom spacers and some bending were called for ! Still it all went back together OK and was re-fixed to the engine and the oil feed pipe re-connected.

Then I turned my attention to the box  that houses the relays for the Fuel Solenoid and the Starter - I'm re-using the original relays as they are in good condition, but I had to pilfer a replacement base for one from the defunct relay based controller, as one was cracked.

Investigating the box brought me to what I assume to be a bi-metallic over temperature cut out that had been mounted in the box, I think it was in series with the power feed to both relays. Well it's open circuit, and on heating it, it fails to change state, so this may well be the reason that the generator failed to start for the previous owner last Christmas!

I then spent some time working out what large crimp lugs I will need to wire the starter, alternator, ammeter etc, and how the wiring conduits would run. The original conduit had been retained by 'P Clips' held with small self tapping screws driven into any near place, including the heads of bolts! I've decided to use 'Girder Clips' and fabricate some suitable tabs to go under bolt heads for them to clip to.

The box is a main junction for many things on the engine, and was originally all joined together with 'chocolate block' connectors. I'm not keen on them and have decided to use proper DIN rail terminals.

So several orders placed for bits and pieces !

[awemawson]

Today's first job - make a battery tray: up until now the two large batteries have sat on the floor. The machine came with an angle iron stand for them, but I wanted to integrate them onto the machine itself.

I bent up some 2 mm Zintec sheet, and drilled a series of strategic holes. Three 'just in case' drain holes, and a series of through holes to fix it on the right hand 'floor bearer' RSJ. The batteries are very heavy and it cantilevers towards the engine off the 3" RSJ but seems pretty sturdy. Holes in the tray are sized to be an interference fit with the square 'under head' of coach bolts, with the holes in the  RSJ being a clearance fit for these squares. So when finally assembled after painting, each coach bolt will get a thump with a hammer to set it in the sheet metal of the tray, then the nut below the RSJ tightened and the surplus length chopped off.

Then I made a start on the 'Heater Panel' - two heaters both powered from domestic mains - one in the oil / cooling system and one in the cabinet so a new panel was needed to make room for an extra switch and also 'tell tale' indicators to show at a distance that the heaters are on.

One more panel to go and then this lot can have a coat of paint all in one session.

While this was happening several deliveries arrived:

 - a 'Special' eStop button with THREE normally closed contacts -(these are rare and I wanted to avoid having to use a relay)

- a box full of DIN terminals from RS Components

- two 10 metre lengths of 20 mm flexible conduit and one 25 mm length

So now I have the bits to start on the 'relay box' and the Deep Sea Controller panel, but we have an influx of grand children over the weekend so things might get a bit fraught 

[awemawson]

Had a slight change of plans this morning, and decided to incorporate the on / off switch for the battery charger / maintainer in the same panel that controls the two heaters. After all, all three devices derive their power from the external domestic supply, so it is logical. So a quick redraw and re-plasma-cut to make a new panel.

I also drew up and cut the control panel that will hold the Deep Sea Controller and the eStop switch.

All panels and the alternator mounting bracket got grit blasted and sprayed in silver hammer finish and are currently drying.

That I think is the last replacement panel made up so perhaps I can soon start wiring them up !

[awemawson]

Not much done today due to family commitments, but I did manage to drain a fair bit of the coolant allowing me to unscrew the engine block heater and water temperature sender for identification and replacement.

I've temporarily put a 1" BSP plug in place of the heater to stop drips.

The sender : I don't know if it is an analog variable resistance, or a thermal switch, but it is 1/2" BSP and bears the following on the hex nut:

DCT
9-9
ZJ

The block heater is 1" BSP and protrudes 2 1/4" into the heat exchanger and bears this on the hex flats:

Bray JJ
ZA2
200/250V
250W
K8

Both are currently open circuit - if you have one in your back pocket please let me know 

[RussellT]

I can't help with the details.  I remember being told that Smiths marked everything they ever made with a part number so I surmise it's not one of theirs.  If it's a thermal switch then being open circuit doesn't mean it's dead.  Did it connect to a temperature gauge?

Russell

[awemawson]

Russell Temperature Sender:

 it's cable disappeared into a maze of wiring within the original 'relay controller' There was minimal electronics to it but it's possible that it had an analog sender with something looking at it and detecting a particular level. It doesn't alter it's resistance when heated to charring temperature (of my bench) so I reckon it's dead !

I've just ordered an analogue sender (50K ohm at 21 degrees C) so eventually I'll probably build a little Schmidt trigger circuit to give a digital 'over temperature' input to the Deep Sea Controller.

[awemawson]

Late (workshop) start today due to Porcine Processing, three dismantled fattened pigs having been delivered at 7:00 this morning as IKEA self assembly kits. The commercial butcher that I use cocked up so when I went the two hour round trip yesterday morning to collect, they hadn't been done  Steam issuing  from my ears convinced him that him delivering today was a 'good idea' 

So, back to work - still waiting delivery on some vital bits for the engine block heater (which I'm having to fabricate) so . . .

I refitted the charging alternator mounting bracket, the alternator itself, and the three new fan / drive belts which I'm glad to say seem to be correct.

Having done this it was time to investigate alternator regulators. The machine came with a CAV440A , the replacement charging alternator came with a supposedly working CAV440D and I also obtained (again supposedly) another CAV440D as 'new old stock' in a CAV cardboard box.

All were in pretty awful condition with loads of muck, paint and corrosion. I decided to lightly sand blast  the CAV440A which at least allowed me to unscrew it's screws to open it up.

I'd not usually take such drastic action against innocent bits of electronics, but I think that it was justified in this instance! I've read somewhere that the 440D version is 'improved' in that it clamps one of the inputs in the case of faults but I'm struggling to find the refeerence.

So, the 440A had one Lucar connector virtually corroded away, the others have cleaned up well enough that I can re-tin them to stop further corrosion.

All fairly straightforward inside, only really two mystery components - a blue and white block that I think must be a thick film resistor circuit, and one of the three NPN transistors is unmarked. As the other two are 2N3055 and BFY50 I'd be surprised if the unmarked one is not a BC109 or very similar.

Comparing resistance values at the Lucar terminals of all three units they are remarkably similar, so it is actually possible that they all work !

More web hunting to find that reference to the 440D variant - if it proves sensible I may clean one of them up and uses it, otherwise I'll have to replace the riveted male Lucar on the 440A that's so badly eaten away (probably electrolytically)

[awemawson]

Found it - a US web site:


"The type 440 regulator has three setting depending on the rate of charge required in certain circumstances:
HI – Setting for normal ambient temperatures.
MED - Setting for countries with high ambient temperatures where there is a risk of battery overheating.
LO - Setting for countries with high ambient temperatures where there is battery gassing. Once batteries have
cooled the setting must be returned to MED or HI.

Initially the 440 regulator was fitted, but was superseded by the 440A that could work at higher temperatures.
Both types had the disadvantage that should the connection to the LO, MED or HI terminals fail then too high a
voltage will be given out resulting in excessive alternator output. The 440D, which is directly interchangeable
with the earlier regulators, has a fail-safe arrangement. In the event of a similar fault the output voltage is held
down."

Read more at http://www.ybw.com/forums/showthread.php?260872-Battery-Charging-System-Help-Required-To-Understand/page2&s=c9f86ab41675f20586bc5700b1919f2b#O4RPPR0gII7CcFOc.99

[awemawson]

After supper I had a go at re-tinning the Lucar connectors on the CAV440A. Not come out too badly, certainly functional, but there is deep pitting in places stopping full wetting. I left the badly corroded terminal strictly alone.

Then I gave the 'new old stock' CAV440D a light blasting - no significant corrosion on this one - the light blasting didn't even have to remove the tinning so I suspect it's declared 'NOS' status is probably true.

The other CAV440D that came with the new charging alternator has one terminal (MED) in a very similar poor condition to the CAV440A that had the High terminal corroded.

I understand that these regulators switch the field at a reasonably high frequency and this has probably contributed to the terminal corrosion seen.

Comparing the internals there is very little difference between them, a few resistors and an extra diode.

So the 'New Old Stock' regulator wins the beauty competition and is the one I will initially install, holding the others in reserve in case it doesn't work.

[awemawson]

I decided to give the CAV440D regulator a coat of paint before installing it to stop it's casing going all furry like the others - not Bling I promise you 

It actually had three coats of acrylic can paint with accelerated drying under the workshop hot air hand drier - this dryer gets used far more for this than hands - I prefer a towel !

The real job of the day is to make a new water jacket heater. I've not been able to source a replacement for the original so need to make one. I had a 1" copper bar to hand, a 1" BSP male/male nipple that is through bored 26 mm, I'd ordered a 1" BSP Stop End cap, and a 10 mm cartridge heater 50 mm long 250 volts 200 watts and had to hand a 20 mm male conduit nipple to give a wiring connection point so we need to set too.

I wanted to get the 25.4 mm copper bar to be a snug fit in the 26 mm male/male nipple prior to brazing, and had hoped that knurling the copper would squeeze it out enough to swell by the 0.6 mm needed, but it was just too loose, so in the end I made up a brass shim packer with dimples to hold them concentric as heat was applied.

[awemawson]

So I cut a slice off the 1" copper bar, faced it, and bored it out 10 mm leaving a 4 mm wall at the far end.

Then fitting my packing shim I set too with the oxy-acetylene, got a bit over enthusiastic trying to ensure full flow of the braze, and manged to melt the end of the 1" BSP male / male nipple - Oops 

Never mind, probably recoverable - so I mounted it up on the lathe, skimmed it back to good brass and ran a 1" BSP die nut down it - the threads on the nipple were too long anyway so I don't think that this will be a problem.

A good old clean up, then I bored out and tapped the 1" BSP cap to take a 20 mm conduit nipple, which I screwed in and sweated in to stop it unscrewing. The original heater has an integral ceramic terminal block, but the new has flying leads, so a terminal box will fix onto this 20 mm conduit stub.

Altogether not perfect but I think that it will do the job.