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

[hermetic]

Entirely up to your usual standard Andrew! Excellent work!
Phil

[Pete.]

Excellent work, and imagine the hours of entertainment you've provided the next owner when they search catalogues of OEM parts looking for THIS heater's replacement

[awemawson]

Thanks Phil 

The advantage of my re-design is that the heater element can be replaced without draining all the coolant. I'll probably eventually put it in with transistor heat sinks compound, although it's a nice tight sliding fit.

[awemawson]

Pete that did actually occur to me 

There is another Dorman powered generator on ebay at the moment, and close examination of the pictures reveals that it's heater has failed at some time and been replaced by a 1" BSP plug, as of course mine has at the moment !

[Pete.]

Are the heaters cheap? If they are I'd be tempted to buy a spare and bag it up in the cabinet.

[awemawson]

£10 including postage, and yes that’s not a bad idea 

[awemawson]

Today the plan was to measure the Temperature Sender Characteristic Curve to aid me designing a small circuit to trip at a set upper temperature (Overheat Shutdown)

But first I needed to source some 'Jumper Bars' for my DIN terminals - these things are simply a copper bar with holes at a suitable spacing that when installed parallel up adjacent DIN terminal blocks. I'd ordered some up from RSComponents with the terminal block, and despite being listed as correct for these 8 mm wide 6-10 mm csa cable terminal blocks they were in fact the wrong spacing

I've had a 'technical query' logged with their help desk since Monday trying to source the correct part, and yesterday they came back with a manufacturers part number but they don't stock them  And nor does anyone else in the UK unless you want to buy 500 off !!! (I only need four 'two hole' lengths!)

OK change of plan for the day - MAKE SOME. No copper strip to hand, but slightly thicker brass should suffice sliced off a bit of chunky angle that was in the scrap box. No great problem, drill a row of  holes spaced at 8 mm, slice them of on the band saw, cut them into 'two hole' lengths, clean them up and tin them.

Using the screws and spacer collars from the 'wrong' ones proved the collars to be too long - so all had to be cut down to 8 mm l.o.a. Altogether not a complicated job but very fiddly and time consuming.

. . . got there in the end !

[John Rudd]

Shame I wasnt in work, I could have gotten them for you.....
But, hey ho...,such is life..

Looking good as ever Andrew, cant be far off being finished.....?

[awemawson]

So now perhaps I can get on with what I'd planned to do 

I attached the Temperature sender to a 'Type K'  thermo-couple with a hose clip, and wired it to my Comark temperature meter and Fluke V.O.M. on ohms. Firstly I dipped the assembly in a boiling kettle getting 160 ohm and 98.7 degrees Centigrade (It was only later that I noticed that the Comark battery needed changing, but for my purposes this IS OK)

Then I transferred the assembly into a Pyrex flask, poured in the almost boiling water from the kettle, and sat down to plot a 'cooling curve' - this way hopefully the sensor temperature will be a true analogue of it's installed state.

Then I plotted the figures in Excel producing the attached chart.

I've decided to use an LM393 comparator chip and it turns out that eBay abounds with them conveniently mounted on a PCB with a 'set point' twiddle pot. I am going to mount one of these circuits inside a DIN module case along with a miniature relay to give me a closed contact on 'overheat' - the twiddle pot on the PCB will be replaced by another glued in the DIN module case and accessible through a hole for setting.

[John Rudd]

Have you considered the case of an open circuit...? Your detector triggers on a rise in temperature wich correlates with a fall in Ohms....

If you have a wire break then it wont trigger?
The relay wont close and no shutdown....

Better to have a function that is failsafe? Ok we're not talking SIL levels here but some protection is better than none?

[awemawson]

It's a point to consider but it doesn't overly concern me considering the intermittent use this is going to see occasionally driving my induction furnace for maybe an hour or so.

Without managing to source a suitable sensor with a curve opposite to this one I don't really see any other option, and this one was the only one I could find on the correct 'root stock' ie 1/2" BSP. My preference would have been an on/off switch operating at about 100 deg C - this would have avoided all the faffing about with electronics, but I couldn't find a suitable  one!

I could get really fancy and use another controller input to sense that the sensor is passing current but that is getting over complicated even for me

[RussellT]

It's too late now but here are a couple of links.  I had to look as I was surprised you couldn't find one.  Mind you at that price I'd probably be mucking about with electronics anyway.

https://www.asap-supplies.com/high-temperature-warning-switch-95-degrees-1-2-bsp

https://pressurevacuumlevel.co.uk/product/ts-tsp-temperature-switch/

Russell

[awemawson]

Russell thanks for the links - yes I admit I'd found one or two at very silly prices!

John's comments set me thinking - I'd thought that the Deep Sea controller would only take a closed contact as an error signal, but closer reading of the manual, if I'm interpreting it correctly, is that the error state can be defined as open or closed, the open obviously being safer as John points out.

Now over temperature switches intended for surface mounting are readily available - in deed I've just bought a 70 deg C one to put in the relay box to cut off the fuel and starter relay feeds in case of fire. If I could somehow use the body of the original sensor to mount one of these all would be well.

I decided to eviscerate the device - I used an old 1/2" end mill naughtily in the tail stock chuck, and chomped out all the filling gunge and eventually the sensor itself emerged on the end of the end mill!.

Then I turned a slug of brass to fit inside, along with a platform to fit on top to take the actual switch - all a bit crude but it should work.

Then ONCE MORE ! I managed to melt it all while brazing it together. I've run out of proper silver solder and have been using Harris Phoson which has a melting point of  732 to 815 degrees C which is obviously too high for the brasses that I've been using. (It's excellent with copper and needs no flux - beloved of the Heating and Ventilating Engineers!)

So - what to do? Well stop being so miserly, dig out that nice big bar of silicon bronze (I think) that's been in my stash for years, and make one from fresh.

So into the Beaver CNC mill to have a hex put on one end (just using the machine as a dividing head) then make the rest in the lathe. Came out OK - no brazing required so no more oops moments.

Now the switch is obviously not in the water but gets it's heat by conduction through the bronze bar which IS immersed in the cooling water, so there may be a degree or so difference actually at the switch, but these switches are available in a whole range of trip temperatures so if necessary I can put a 95 in rather than a 100 for example.

[mc]

I was just going to suggest those temperature switches, as I happened to replace one at work this morning on a driveline compressor, with a new modified sealed switch unit. It's only taken this supplier about 15years to finally come up with something that might last longer than a couple years on the underside of a vehicle!

I'd like to know where they got the new switch, but it had been bonded into a bit brass/bronze (I'd guess about 3/4" thread, but it's an Italian compressor so it could quite easily be metric..), with no obvious information.
I'd guess the switch came with flyleads, as it had very flexible wiring in a high temperature sheath, with a superseal connector heatshrunk on the end.

Whatever it is, they must be available cheap from somewhere, as the whole assembly only cost about £60, whereas the old unit which was a brass/bronze plug with one of those surface mounted switches attached with some heat transfer compound, used to cost about £80.

[mc]

Just done a bit googling, and I'd guess what I fitted today was something like this - https://cpc.farnell.com/microtherm/t11a10005u112l310100/thermal-switch-nc-100-c/dp/SN36221 stuck in the fitting, then held in with resin/epoxy.

The price would certainly explain why it was cheaper than the old surface mount switch.

[awemawson]

Yes there's a version of that type of switch on ebay that caught my eye - but when these things are exposed on vehicles or plant I'm not at all keen on flying leads - once ripped off that's it!

It'll be interesting to see how the surface mount switch survives vibration fixed to the engine - not that there is much, it runs pretty smoothly. Dead easy to change though now, just two M3 screws and no need to drain down

We had almost the exact same devices bolted to the side of the combustion chambers for the gas tumble driers in my launderettes and had remarkably little trouble with them

[hermetic]

LIKE!

[awemawson]

Thanks Phil 

I spent today wiring up the relay box and documenting it. Far easier to do this on the bench than mounted on the machine down close to the floor, especially  with my varifocals!

It's a useful sub assembly that pre-determines various conduit runs, and as at long last I've found a source for metal conduit clips that are wide enough to drill out to mount on existing engine bolts (ordered but not arrived) I can soon get on with the wiring proper.

(I've printed and laminated a copy of the picture below to go in the box lid to avoid future head scratching)

I mounted the mains cabinet tubular heater today, mainly to get it off the bench and out of the way !

Meanwhile I'm giving consideration to how to get power OUT of this box of tricks. Simple I'd thought, just mount a 125 Amp 3 phase, neutral and earth 'Commando' socket on a panel by the terminal block. But it's not actually that simple, as then you can't close the door

Alternatives are to :

A/ mount a socket on the side of the main cabinet, but then it will be vulnerable being the widest part of the machine or

B/ Cut a hole in the door that exposes a suitably mounted socket that slightly protrudes when the door is closed or

C/ mount a socket on the door itself on flexible leads

Option A I think is bound to cause tears in the future as the machine is moved about. Option  C is neatest but at this sort of power level leads are not that flexible !

So at the moment I'm thinking option B but would very much welcome comments from anyone reading this.

[John Rudd]

I cant really add any ideas for power connections, 125Amps for a TPN socket is a bit wieldy....

As this is supposed to power the  furnace, why not hard wire the two....?
BUT, You should consider some protection relays for the genset.....Undervolts, overcurrent...etc..

Our 3.6 MW turbine sets were used in island mode ( there were two originally....) one duty one standby, and auto start....but one day someone decided to start a compressor ( a refrigeration system..)  that was driven by a 6.6kv motor rated at 2.1MW.....suffice to say, the load imposed on the genset caused the shearing of the pins on the shaft coupling as the motor tried to start....

[awemawson]

John, under / over volt monitoring is no problem as the Deep Sea control has that built in and is just a case of setting the right parameter.

I'm not absolutely certain but I think that the output breaker by Merlin Gerin is set to trip at 150 amps. I need to investigate further.

[russ57]

Could you mount it behind the box, so it is accessible from the side, but recessed for protection?

Another little fabrication job - keep the thread going for another few days....

Russ

[awemawson]

Russ thanks for that idea, it’s certainly one to chew on.

Sorry to disappoint, but there’s little prospect of this thread fading out for a while yet (unless I do!) as there’s loads to do still 

[modeng200023]

Andrew, I'm wondering what you do in your spare time 

[awemawson]

Well actually NOT much spare time today as we had guests, all very nice but you trip over them !

I did at least get a bit of time to turn the 'Meter Panel' into a pre-wired module awaiting fitting

[awemawson]

Still awaiting for bits to arrive, but I did manage to move the main Neutral connection to allow me room to mount the modified consumer unit that I'm using as protection for the controller and meters. At least now there is 30 mA protection for user facing controls whereas previously as built by Puma there was none.

I think I've solved the output socket location issue. So far I've been working with the door removed, and it didn't dawn on me that the door is not full width, there is a panel on the right of the door about 8 or 9 inches wide that is still exposed when the door is shut. Ideal place for a socket, but I realised really you want an isolator as disconnecting 125 amp connections without switching the load off makes BIG flashes ! But the main isolator is inside the door 

. . . never mind, looking at my stash of 125 amp plugs and sockets I discovered a humongous switched interlocked 125 amp unit that will just squeeze onto that panel !