99 3462 Build diary

[Roger]

It also looks like there are no cutouts in the wheel casting on the far side on the RH axle. You can see a substantial portion of the centre of the casting, but you can't see any holes in it at all.

[mbrown]

That made me look twice! But I think you can just make out one of the oval holes if you look in the V between the two sets of wheels. There is no daylight showing through because the mis-aligned flycrank on the far side is blocking it ... I think!

I would love to get over to Germany and see her while in pieces. Some constructional details are still obscure to me. But after a curt acknowledgement of my initial query, the DKBM at Muhlenstroth where she is now based didn't repond at all to my follow up email.

Malcolm

[mbrown]

Turbo Generator

Having finally got a satisfactory boiler, the next thing I ought to do is strip the chassis and assemble it properly, re-doing some of the springing etc as I go. But I am enjoying looking at the boiler perched on the frames and can't get enthusiastic for a major strip down now. So I thought I would make progress elsewhere...

Some time ago, I established that the little turbo generator I made years ago will illuminate a number of LEDs easily. The challenge now is to house the generator itself in a body that is as close as possible to the German prototype. My early jenny had the inlet and exhaust at tangents to the turbine wheel, but the German one has both steam and exhaust coming at right angles to the face of the wheel, which makes for complications.

Here is a picture of the standard Deutsche Reichsbahn generator - in this case on 99 3461, another MPSB 0-8-0 by Vulcan in the 1920s.


1451482_441758985935720_364435112_n (2) by malcolm brown, on Flickr


The steam enters on the far side, opposite the four-stud flange facing the camera, travels up the diagonal duct, and the exhaust is through the thick pipe, also with a four-stud flange, at the bottom.

The main body should be straightforward - I made it from brass tube bored out to 3/4" dia for a push fit for the generator itself. The fun and games came with the fan and its housing. In fact, the turbine wheel wasn't too bad to make, although I forgot to take pictures while I was making it. It is 7/8" dia and the shape of the blades is clear in the shot below. I did them on the mill with a small rotary table set vertically. Directing a jet of compressed air at it showed that it will spin brilliantly but the angle of the jet is fairly critical.

The next interesting bit was the trunking to take the steam inlet. This is a piece of 3/8" square brass, with four 1/8" holes drilled through it, turned down to reveal the holes and then four pieces of 1/8" dia brass, drilled through at No.54, silver soldered in. Here it is before soldering.

IMG_20220108_150448 by malcolm brown, on Flickr


The housing for the turbine was straightforward turning from 1.25" bar. The trunking is a push fit, held in place by the end cap, and the diagonal duct is from 3/16" brass, drilled trough and carefully filed to fit. Below, the various bits and bobs are assembled ready for it all to be soldered up, along with the generator and its turbine wheel and the main housing. I am hoping all the internal passage ways line up OK...


IMG_20220109_173203 by malcolm brown, on Flickr


Once again, I was beaten by the loss of daylight for soldering - and there is a lot more to do before it is finished. But I am quite pleased with it so far.

Best wishes

Malcolm

[mbrown]

Turbo Generator continued....

The weekend saw the generator almost finished. Here it is, with the little junction box waiting to be screwed on. The junction box is carved from Tufnol with a brass inset. One wire will come in at the right hand end and is clamped by the 10 BA brass screw at the bottom, the outgoing wire comes through the little white tube and is clamped by the screw at the side.

The real thing has a rather prominent brass plate on top with all the usual details electrical kit has.... I have attempted this using brass shim and punching figures through from the other side to emboss them. I have no idea what was on the original, but it was a 25 volt, 500 Watt generator, so I have put those figures on the plate. Obviously, one can only use digits that work in reverse, but I have fudged it by using 2s for 5s and vice versa. I must do it again as this was the first experimental version to test the idea and it could be a lot neater. 


IMG_20220117_182423 by malcolm brown, on Flickr

At the back, the end cap is turned from black nylon as it is very close to the terminals and could short them out. Slits in the side of the housing will allow the wires to come forward to the junction box rather than tailing out at the back, and these are mostly covered by a collar of Phos Bronze shim which resembles the removable panels which are held at the top by wing nuts on long bolts on the prototype.

Here is the front of the main rotor housing, with the second shot showing the hole for the jet and the exhaust port. The jet is a short piece of 6 BA threaded rod (so that it can be removed) drilled through at No.60. When I get it going, I will experiment with different jet sizes until I find the optimum.


IMG_20220116_124049 by malcolm brown, on Flickr


IMG_20220116_124129 by malcolm brown, on Flickr


The finish looks rough, but remember this is less than 1.25" diameter. Most of the marks are where copper plating from the pickle has not completely been polished away.

And here are the two main halves, showing the rotor in its housing.


IMG_20220117_182708 by malcolm brown, on Flickr


So - does it work? The answer is - not yet! I tried it on air but nothing happened and, on investigation, I found that there is no clearance between the rotor and the front cover. Obviously, I want the clearance to be minimal but I overdid it and the rotor just touches, which is enough to stop it moving. As the front cover is now very difficult to hold, the easiest thing will be to skim a few thou off the rotor. But I ran out of time last night and am away all next weekend, so it will be a couple of weeks before I discover if that will be enough to bring it to life.

Best wishes

Malcolm

[stevep]

Can we have a video too, please?

[mbrown]

When I get it work, and when I have worked out how to post videos!

You'll need the soundtrack too - when that little generator gets up speed it whines like a banshee....

Malcolm.

[suctionhose]

Hi Malcolm,
I have some questions about speed control of the generator which you or others may be able to answer.

I have a working turbo on one of my loco's and made a pressure regulator - or rather and adjustable pressure reducing valve - following age old model engineering practice. It may have originated with the Basil Palmer designs from what, 1970?

Anyway, simple regulators as such step the pressure down reliably from a constant supply pressure (P1) to a manually set output pressure (P2). The problem is that the P1/P2 ratio is not constant for all values of P1. In other words, the output pressure from the regulator fluctuates with fluctuations of boiler pressure - not as much but certainly significantly.

There are 'precision regulators' when you google around but it is also noted that 'drift' as described above is inherent with these simple spring over diaphragm type of controls.
Possibly matching spring rate to diaphragm area may reduce the affect but the exact relationships escape me.

The question therefore relates to speed control of the generator under variable boiler pressures. (One can assume the electrical load remains constant in the circumstances of locomotive lighting).

Is there a control system that works?

[steamer5]

Hi Ross,
Dad has a steam generator on his loco. There’s no pressure regulator, as you rightly point out it’s speed varies. When he got the loco the brightness of the lights went up & down as the boiler pressure changed…..he wasn’t to worried about it as he doesn’t run the generator for long periods, worried about the generator bearings failing as it fair hums when running!
Anyway one of the bright electric gurus in the club built him a voltage regulator solved the issue of the light fluctuations!

Cheers Kerrin

[mbrown]

Thanks Ross (and Kerrin),

Some, but not all, the articles in ME etc over the years advise a pressure regulating valve. I certainly see the point. But my approach is similar to Kerrin's Dad's - the generator is so noisy (very high pitched, too, which challenges the ears!) that I don't expect to use it except for the odd demonstration. I just like making working gadgets... (or rather, I am not keen on dummies).

I don't do any night running, although I guess there's always a first time... and having working lights might be an incentive I suppose.

Also, this is a small loco and I am going to be quite challenged for space to put things, so the fewer un-prototypical bits and bobs the better.

I found, experimenting on air, that the speed of the generator fluctuates more according to the number of LEDs in the circuit than according to pressure. With no load at all, anything over 20 psi spins it excessively fast. I wait to see if the new generator housing, with more convoluted passages, performs similarly. I may make a voltage regulator as Kerrin has already sent me the diagram, although again, I am not sure where I would put it. But I intend to ensure that at least one LED is in the circuit permanently so that the generator is never without some load. I will also make a fairly fine control valve so that I can set it to a reasonable pressure and perhaps turn it up a bit if boiler pressure falls (although if that happens, I suspect all auxiliaries will be turned off to help get pressure back up!)

In the end, it may not be good engineering, but lack of space is the biggest consideration.

Malcolm

[suctionhose]

I have run my generator for hours on end when running on a gloomy afternoon among the trees at the club or on those crazy all nighters we used to do when I was younger!

Now I have a mate retro fitting one with a brushless motor inside the detailed generator casing so the control issue came back to mind.

I did theorise about the pressure reducing valve performance at one time. Probably the answer is to optimise the pressure control plus add a voltage regulator.

At that point I lost interest and manually adjust the PR valve according to sound (if you can hear it over exhaust and other noise).

Interesting thing is that it slows down as the pressure rises and speeds up as it falls! I have to listen to the sound because it over sped once and bent the shaft (which is 1/4" brass with a magnet encased in it).

I wondered whether anyone had used some kind of servo flow control valve driven by looking at voltage. The fullsize turbos have some sort of governor built in ...

[Roger]

Personally, I think I'd use a pressure regulator on the Steam side to guarantee a known input. Then I'd use electronics to vary the load to control the speed. Dumping the small amount of excess power available into a power resistor is easy enough.

Ideally, I'd use a Hall effect sensor on the turbine shaft so the electronics could see what RPM it's actually doing. You could have a simple set point that would progressively load up the output resistor, the higher the RPM is above the set point. An 8pin PIC Microcontroller could both monitor the input tacho and control the amount of resistance across the generator. The viability of this would depend on how much power the generator could provide. Everything could be powered by the output, the Microcontroller would automatically boot up when the voltage reached 3v3. An interesting project. You can't use any of the Arduino type of boards for this kind of thing, they draw too much current, but a single Microcontroller with careful design could certainly do the job.

This sort of thing would do the job. These cost 62p from Mouser Electronics in singles. The chip only draws 48microamps when running with a 32KHz clock, going up to 1mA when flat out at 4MHz. I doubt of you would need to go past the slowest clock rate for a simple application like this. It will run at anything from 1.8V-5.5V, so you could choose to run it at whatever voltage suits the LEDs you're driving. You could also use the LED driver chip I used for the Lamps, and that would make for even outputs with high brightness.

The advantage of doing it this way is the near instant response and there's no physical control of the turbine, other than the pressure regulator.

[mbrown]

I must have been off school with a note from my Mum when they did electricity - all this is way above my knowledge, ability or understanding. The only time I have ever done any electronics it was (a) kindergarten stuff and (b) I had a picture of what it should look like as I couldn't understand the diagram!

This little generator is the furthest I expect to go beyond basic steam technology. Even that may be biting off more than I can chew!

Malcolm

[Roger]

Jan 19, 2022 20:39:32 GMT mbrown said:

I must have been off school with a note from my Mum when they did electricity - all this is way above my knowledge, ability or understanding. The only time I have ever done any electronics it was (a) kindergarten stuff and (b) I had a picture of what it should look like as I couldn't understand the diagram!

This little generator is the furthest I expect to go beyond basic steam technology. Even that may be biting off more than I can chew!

Malcolm

Love it! Personally, I love the interplay between the disciplines, but I can see that it's not for everyone. None of it's difficult, but there is a bit of fundamental electronics and computer science you need to know to pull it off.

One day I might be inclined to do this as a fun project. It's not trivial, but it's not terribly difficult either. The solution I've proposed is physically simple, with just half a dozen components on a tiny circuit board. Of course, it's a hideously inefficient way to regulate the generator, but it might provide a way to run it at less than an ear splitting speed.

[suctionhose]

If you could see the stuff surging along I might be more captivated by electricity! That said, it sure is amazing stuff!

I'm thinking again about the pressure regulator. I think my diaphram might be too large for the spring or visa versa. A redesign is needed and of course making one to prove the the change. That's where I get stuck - having to do it again... We'll see. Finish the TE next.

BTW Malcolm the P2 Locomotive Trust have kicked off "The TurboGen Club". The unit looks to be the same as yours:

[mbrown]

Thanks Ross - how interesting to see the familiar German turbo-generator going onto the P2, but (if I have read correctly) with axle drive rather than the steam turbo. I imagine they are buying a new piece of kit - and the clip you put up cuts out the price tag which would be interesting to know! But these were standard on all DR locos, large and small, and may also have been used by the DB in W Germany. So I imagine there will still be plenty of "used" ones knocking around.

But I still have to get mine to work!

Malcolm

[Roger]

Jan 19, 2022 22:45:09 GMT suctionhose said:

If you could see the stuff surging along I might be more captivated by electricity! That said, it sure is amazing stuff!

That's why we have tools to make visible what's going on, from Osilloscopes to Logic Analysers. It's no different to figuring what's going on in a Boiler using pressure guages and thermometers. I had an Epiphany when I first grappled with Microprocessors and realised that they were way more interesting than anything I'd ever studied before. I switched careers from mostly Mechanical Engineering to Electronics and Computing as a result. If you're ever bored doing the same old things, I'd highly recommend looking into them, they're fascinating!

[mbrown]

Indeed Roger. Mechanical (model) engineering fulfils the same purpose in my life. Trained in philosophy, professionally an ethicist, and doing a lot of teaching to help students grapple with ideas and abstract concepts, the pleasure of being able to handle solid materials, and see a product one can get hold of and manipulate at the end of the day, is the perfect antidote to the day job.

The "bridge" for me recently has been teaching ethics to students doing PhDs in AI and Machine Learning. They include techies, mathematicians, lawyers, engineers, even a historian. Trying to link different disciplines is fascinating.

But at least I know that none of the machines in my workshop or on the track are going to learn enough to pose some of the questions and threats we are examining on the programme!

Malcolm

[simplyloco]

Jan 20, 2022 11:25:14 GMT mbrown said:

SNIP

The "bridge" for me recently has been teaching ethics to students doing PhDs in AI and Machine Learning. They include techies, mathematicians, lawyers, engineers, even a historian. Trying to link different disciplines is fascinating.
SNIP

Question: Am I concerned about the increased use of Artificial Intelligence?

Answer: No. I am much more concerned about the decrease in REAL intelligence!

Apologies from simplyloco...

[steamer5]

Hi John,
A video my son has……

“Oh no our computer is down…. What we have no intelligence!”

Cheers Kerrin

[mbrown]

Working Turbo Generator

After a fortnight with no workshop time, it has taken me most of the weekend to get the generator working properly.

The rotor seemed to revolve freely, the air jet seemed OK - but when the two halves were bolted together, nothing happened except air out of the exhaust. Every time it was taken apart and reassembled involved 12 x 10 BA nuts and bolts, some in quite fiddly positions, so diagnosing the problem has taken time.

In the end, I felt that it perhaps wasn't getting enough air pressure to get the rotor moving, so instead of having the jet impinge on the face of the wheel, I modified the arrangement so that it hit the wheel on the periphery. This was done by replacing the little 6 BA jet with a plug, then drilling into the diagonal passage higher up. The hole now emerges in the flange and seals onto the opposite hole with a 3mm OD O ring. The hole in the main body meets a diagonal passage drilled from the top and plugged with a 10 BA screw (the turned down head is just visible), ending in a No. 60 jet at an angle to the vanes on the rotor. Trying this with the two halves apart, directing a jet of air down the hole, made it spin beautifully.

Here are the two halves showing the new arrangements.

IMG_20220130_152550 by malcolm brown, on Flickr


However, when reassembled, still nothing turning.... In the end, after many experiments, I concluded that there was still not enough air reaching the rotor - the only solution was to drill out the passages to a larger size. This was a delicate matter as the inlet passage took several turns, most of which had been incorporated before the parts were silver soldered together. But by drilling through into the passages from the back of the corners, opening them out, then plugging the redundant drill hole, it was managed. The passages "grew" from a 1.3mm to a 2.3mm, which was a substantial increase.

... and the result was that as soon as the air was turned on the familiar high-pitched whine began! Success at last.

Here it is spinning merrily and illuminating six LEDs - the equivalent of three headlamps, two cab lights and a tail lamp.


IMG_20220130_151919 by malcolm brown, on Flickr

Sorry Steve, I haven't worked out how to post videos. But there are no moving parts to see, although I can play tunes on it by varying the air pressure or switching the LEDs in and out! Here, it is running nicely on 60 psi, but it will go down to 40 psi before the LEDs start to dim.

And here is a view of the finished article. It is good to end the weekend on a note of success.


IMG_20220130_165404 by malcolm brown, on Flickr


Best wishes

Malcolm