Very nicely explained video! Though i would like to add that The clip you chose for the ÖBB 1016/1116 is not the full sequence. There are two more frequency changes happening at ~60 Km/h before fixing the PWM.
I am very happy to see there are people doing this void of knowledge. I'm an EE student myself and find train traction sounds to be very interesting and not talked about enough. One thing I would've liked to see mentioned is a tap-changer. Though not in use on UK railways as far as I am aware, my home country India does use a lot of trains that still run on tap changers (yes even EMUs). The newer ones are IGBT VF Drives, but I still like the old units. I'm not sure how you could explain it in a video though, you have to feel it. The best way I can describe it is that every few seconds, there is a slight tug you feel that pulls you slightly faster.
Apparently the z3 trams here in melbourne use thyristor choppers, but they do feel like they jerk a bit while accelerating. It's probably just the 70s control system. Also, I assume it's changing motor taps, a bit like star/delta starters, which seems like the most obvious way to change impedance without using resistors but also rather annoying to engineer.
@@ArchieHalliwell The difference between what you are describing and a tap changer is that when you're in a tap changer driven, there are periodic, regular, very small jerks. And there are a lot of them. It is particularly noticeable on an EMU (I had the unfortunate experience of riding on one of them which had a damaged arm so every time the driver changed tap the motor would disconnect and reconnect which was awful). It is very small, usually periodic jerks. You can usually hear the tap changer clicking away as the gears move to change taps. Also, I wouldn't expect a tram to have enough space to actually fit a tap changer (depends on power output so who knows maybe)
Thanks so much for this - very well explained. I visited Vancouber, British Columbia for the World's Fair in 1986. They built a commuter rail system for that event. It used a linear motor drive system and articulated trucks. It sounded different from any of these and of course had no geartrain noise. But it did have a "gear change" sound like most of your examplesy. Might be interesting for you to look into. I didn't know anything about VFD then, I just knew enough to understand I was hearing the coils pulsing.
These train are still in operation, the ones that make that noise are the Mark 1 trains. They are scheduled to be all be retired by 2027. The noise is high tended by the linear induction motors. I believe they plan on keeping one heritage set. They have been very reliable. New cars are larger and longer and have AC, but they sound boring.
It's everything that resonate loud enough to be heared. It's a physical phenomenon known as "synchronicity" (that's what make laterally wobbling the Millenium Bridge in London in the early 2000s, everyone was forced by the bridges moves to walk at same time and pace, which act as an wave amplifier, worsening the situation but i digress a bit) If you go under a really high-voltage powerline (at least 200 000V (you can found it often near high voltage substations)), when you hear the current huming, you actually hear the whole conductor "shaking" at the current frequency cause it's so powerful that it act as a speaker...
+1 on this... AC is better at being transported over long distances, so to me, id make sense to build overhead lines with AC, even if I were to have to rectify it on the locomotives themselves. I mean, cars do this, they dont have dynamos, they have alternators. Its in the name, "alternator" as in it creates an alternating current. Its just easier and cheaper to slap 4 diodes on it (and some smoothing caps, etc...) than to put in a heaver and more complex dynamo.
It depends. old trains with DC motors commonly used mechanical converters that converted AC to DC, modern DC trains mainly use electric rectifiers that convert AC to DC. AC motors are split between single phase and 3 phase motors. single phase motors use the overhead power directly altough only with low frequency (mainly 13,7 Hz) 3 phase motors the most modern and common types convert the single phase AC with H-bridges to 3 phase AC. The main advantage of H-Bridges are that they can generate a three-phase current with a variable frequency, which means that the motors can be switched continuously and braking energy can be fed back into the network (regenerative brake) and is not wasted on Braking resistors. They also support an emergency brake mode which short circuits the motor and transfers the energy into heat.
And most mainline trains here in continental Europe use 25 kV AC @50 Hz like many main lines in the UK, or 15 kV AC @16 2/3 Hz (mostly Germany, Sweden, and Norway). The single phase AC is converted by power electronics to a variable three phase supply to AC traction motors on modern stock as described here.
That was a really interesting video thanks so much for making it. A lot of information though and i feel a lot went in one ear and then went out the other so i might have to rewatch to properly digest ig
Nice video, I liked that there was a lot of different real life examples. However, this is something that I have put a lot of research into and have a strong knowledge on, so I have a few things which I would like to comment on, feel free to provide discussion on them: 01:47 Its not really to do with efficiency, AC motors have always been more efficient than DC as they don't use brushes. Its more to do with the fact that they really only generate usable torque when the supplied frequency is similar to the rotation of the motor (something which you touched on afterwards). 02:23 I wouldn't have thought that its to do with inrush current. More being able to control the speed. Generally with DC motors, more voltage means more torque. Surely if it can't handle a certain amount of power to begin with, then it wouldn't be able to sustain that amount of power at higher speeds??? 02:49 Bypass??? The way resistance works is that the more resistors you have in parallel the lower the resistance (ohm's law), this is usually how they work, they increase the number of resistors in the bank until it can then be bypassed. 06:24 Technically you can have a VF drive on any type of motor, works in a similar way with PWM managing the speed (this one is really nit picking). 10:34 Didn't the 323s get their VFD replaced?? They just used the same patterns (not my specialty so likely to be wrong) This ISN'T a dig. The video was great. There's just so much misinformation surrounding this topic I felt the need to add some of my own thoughts. Of course, I could be wrong too, hence why I ask that you reply with why you disagree and I will try provide evidence to support my claims.
On the motor inrush current, it is a problem because a 200 ton train at a stand, sometimes even with a small amount of brake on, is gonna draw a lot of current. The current will decrease over time as the motor speeds up because there's less load on it. And I was talking about series resistors when mentioning bypassing them one by one, I maybe should've cleared that up. Yes, the 323s now have IGBTs but with the original programming (thank god)
@@WildWildWeasel Most modern trains have a DC bus too. The electronics (GTO or IGBT) switch the DC on and off to create a pseudo-waveform that the AC motors use, as explained in the video. Third rail trains (750v DC) take the DC straight from the rail, whereas 25Kv AC trains additionally have a transformer and rectifier to covert the AC to DC. On an AC train, DC can also be directly fed to the PWM/VFD drive, bypassing the transformer and rectifier, this is how the dual voltage units found around Londion work.
@@WildWildWeaselmost modern evs use inverters and ac motors. But their lower system voltages and power levels allow them to use mosfets rather than igbts or gtos, which can practically switch at 25+kHz no problem, so you simply won't hear it. Recently sic-mosfets have made it possible to even do this on megawatt scales, so even a train drive could be almost silent. Even then though, a manufacturer may still opt to switch slower because it saves some power.
I have two questions. 1. How do you get +ve and -ve voltage in PWM mode without the two "halves" like in pattern mode? 2. And why even bother with PWM mode at all, and just use pattern mode from the start? Are the motors wired in star and delta configurations respectively or something like that?
In PWM mode for an AC wave, the Pulsing frequencies on both sides are the same. At the peak of the wave, the width on the positive side is large while the width on the negative side is small. And in the middle of the wave the widths are the same. The reason I believe they use PWM to start is because each pulse has to be calculated by the programmer which makes it more controllable at lower speeds, I think.
You seem to have completely neglected the loudest and most common noise associated with any electric traction : that of the transmission type between motor and wheel. Axle hung, Frame hung, Radial gear, Straight cut gear, Flexible coupling... etc etc.
Was it 'neglected' or was this video's topic designed to look at just the one thing and be only and solely about the electrical system and not about the factors involving mechanical components?
The S-Bahn Berlin 475 class is such an interesting sound and not like any of the others. But there is still resistors and 14 stages but there is a loud click that repeats once or twice. It even switches to a lower resistor automatically. That thing was declared the most modern rail veichle in I think 1930. First trains operated 1927, last regular operation was 1997. It was replaced with some "modern" "faster" and "better" veichle in which pretty much every part had to be replaced a few years on because they worked so well that they had to decrease the max speed from 100km/h to 60km/h (below the pre ww2 trains) and then take them out of service entirely. Still remember how much better that train is?
This pre WW2 railways working better than modern ones problem is apparent also in Poland. Its slowly getting better, but we will never reach the level of perfection as in Japan at this rate lol
@@Weisior actually the new trains have had all off their failing parts (aka brakes, wheels, doors, heating, motors and more) switched and are now working fine. But imagine having bought 1000 new train cars and they just all fail and you wish you still had the now 82 year old trains
Yes, a transition from engineering efficiency towards economical efficiency. Mostly, we didn’t even discuss that transition in the late 80’s - we just called it “progress”… (Analogous stories exist across most industries and sectors, btw…)
@@musiqtee They are today replacing the mechanical security system with a computer based one. Pros: Can also brake the train not to fully stop Cons: Few million different things can go wrong. The mechanical system consisted of 3 parts that could break, the toggle between a stopping and a non stopping state (can also be set by hand), the lever to slam against the plate (has to be tested every morning) and the brakes (if they break idk what went wrong but there is bigger issues)
Are there any UK locomotives specifically designed for AC, rather than the current range of AC locos with onboard rectification to DC (like the Class 88, 91, 92 etc.)
It was just last Saturday while changing trains at Woking that I was wondering why the express train that I was leaving sounded so different to the suburban service I was boarding. Thanks for the information. For a splendid example of "musical" motors may I cite the example of the older trains on the Montreal Metro.
You could probably do a whole video on chopper as there are several types: magnetic field, armature, 4 quadrant and possibly more. Can’t say if it will get many views though. 😂
My confusion is why the whining noise? Pwm systems are used in small electronics, computers, and the like and run silently. Is the sound due to noise (play) in the (traction and drive) system?
Because train motor are enormous and need a lot of power, the GTOs or IGBTs can't switch the many hundreds of amps as fast without getting overloaded. Smaller appliances have much less load on them and they switch at frequencies outside of human hearing
pulsating current in a wire create a magnetic field around it, this magnetic induced magnetic field then resonate in all the drivetrain and to your ears :)
Its interesting to consider that PWM is otherwise known as FM Synthesis, the technology used in 80s synthesisers such as the Yamaha DX series. There's a whole bunch of distinctive 80s tracks which owe their sound to essentially the same technology as these traction methods. It should be noted that much of Europe used low frequency AC OHL systems (16 2/3 Hz from generators, or 16.7 if converted form mains AC) with universal motors, series wound like DC motors. The main advantage at the time was not needing to rectify into DC, this was long before semiconductor diodes, and even mercury arc rectifiers weren't always reliable (see the tale of the Glasgow Blue Trains). Both the London, Brighton and South Coast Railway and the Lancaster, Morecambe & Heysham both used 25 Hz systems at, I think, 6.6 KV. The LBSCR was absorbed into the Southern on Grouping and everted to third rail, while the LMHR joined the LMS and was changed to a 50 Hz system some time in the 1950s.
The sound from the Bombardier (former Adtranz) Regina (Swedish X50 to X55) and Contessa (Swedish X31/X32, Danish ET) series have a distinct sound. The shape of the pulses changes, giving them their distinct sound. Sound: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-ruocbxmzVzs.html It's very similar to DB's Br 442 (Bombardier): ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-CQaJawYB_F4.html And Stockholm's C20 subway (Bombardier): ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-U6V1zI5FaSg.html
Here's some really funky sounds from the dutch IRM trains before they were converted to IGBTs: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-HTiys7ahfV0.html These trains were nicknamed "bagpipes" because of it. The three tones (PWM frequencies I presume) it's switching between are 550 Hz, 600 Hz, and 650 Hz.
That's a really good explanation of traction drives. I was developing a traction drive from scratch for Vivarail until they ran out of money and went bust. I'd got it running on a test rig. The video I'm linking to is the first time I turned a traction motor. This was in V/f mode but would soon run in closed loop vector in torque control. The encoder was crap too and I wasn't sure it'd work. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Xxu61U_0iGA.htmlsi=HphKcdRUGvw0ICK9
The rotation speed of an AC motor depends on the voltage you apply to it. As a result for trains it would require gigantic amount of power, especially for takeoff. Whereas an AC motor's speed depends on the frequency of the voltage (fixed).
Fascinating info 😀 Y'know I don't remember if 87101 sounded any different with its thyristors compared to other Class 87's. Was only a teenager back in the 80's, but will have to go on a search for footage now!
Non-GTO thyristors at 33-100-300 Hz carrier frequency: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-XdvMqAEuNgg.html Otherwise, I missed the explanation of what "pattern mode" is. Fixed width pulses with varying amount of them skipped?
Immediately after DC PWM, I said that it uses a pattern starting with a short pulse, then they get progressively longer, then progressively shorter again, then the same for the negative half of the cycle
@@zfrailways I swear I watched that part multiple times before asking! However then I have another question: what does then "PWM mode" as opposed to pattern mode mean for 3-phase motors? The explanation around 5:00 fits only brushed DC motors, for 3-phase motors, once you do PWM, you do it in way to approximate the sine wave, so I fail to see the distinction...
@@JasaDavid what distinguishes PWM specifically is that the pulses are at a fixed frequency, regardless of the waveform you're trying to produce (hence the fixed pitch sound it makes). It looks like what he calls pattern mode is a fixed pattern of pulses per cycle of the generated waveform, which means the pulse frequency varies directly with the frequency of the generated waveform (and the sudden jumps are changes in the pattern).
@@MatthijsvanDuin this makes sense. I'd expect than that sound irregularities at "pattern mode" may be caused by avoiding frequencies used by signalling (e.g. track circuits at 75 and 275 Hz or indusi magnets that IIRC use 500, 1000 and 2000 Hz).
Excellent video! It is crazy how little there is online and especially on RU-vid about these drives and modulation strategies. I am in the process of modifying a Renault Zoe to make these sounds and the lack of info is not helpful, unlike this vid!
@@zfrailways Of course! It's slow progress but eventually I will definitely make several videos about it. Unfortunately there doesn't seem to be much appetite online for EV content :(
Some further notes I gleaned from down the rabbit hole: The higher-speed pulsing patterns (and most fixed-pattern regimes on IGBT inverters) are overmodulated (ie they target a higher amplitude than they can physically deliver, in this case "cutting off" 1/3 of the cycle at 100%) to artificially boost moror winding voltage, reducing resistive losses. This isn't done at low speeds due to the vibration and reverb it produces, but it makes for heroic beat drops (Toei 12-600, MTR TKL K-Trains, WMATA 7000, perhaps SMRT C830 if you don't like your ears). As for the fluttery carrier frequency on the Desiros, that's a feature most newer inverters have now for resonance reduction, but on the Desiros' inverters vary their frequency over a much longer period than most. (Not sure why they do that.) Oddly enough, the cl. 700s and the other dual-power Desiros only do the frequency fluttering on third-rail DC power. (Maybe this is due to electrical noise being less of a problem on AC lines? I know some older ATO systems have problems with third rail noise, kinetically in the case of WMATA.) A bunch of ABB and some CAF trains have a "howling" mid-low speed fixed-frequency phase which (I'm pretty sure) is a square wave rather than sinusoidal. I'm not sure why they're programmed like that, but I think it sounds great (CAF Inneo Napoli, DB class 474s in Hamburg, probably the Civity sets too). The DB cl. 421s on most major German S-Bahn networks run a very low carrier frequency, something between 80 and 120 Hz I think, and those have the same square-wave phases as the cl. 474s.
I'd be interested to hear your thoughts on the JR(West) class 221-2000 series EMUs, I believe that they have Mitsubishi VVVF control systems. They are the most melodious units I have ever heard (on video ony. I've never been to Japan).
Fun Fact: the sound the ÖBB 1016/1116 also known as DB BR 182 or Taurus makes as it speeds up from 0 was indeed an Easteregg placed by the Engineers at Siemens who programmed the GTO's to sound like that. However with new Software versions this iconic sound is being phased out and some locomotives don't do it anymore.
judging by your explanation i'd wager you know more than dovetail about half these trains - the 323 especially, where it seems to just have random gaps in audio where some of the pulse drops are
Most diesel locomotives have electric motors and use diesel to generate electric current. I think the reason we rarely hear those sounds coming from diesel locomotives is that the diesel generator is just much louder? Not sure.
@@Pethrenne It's also because the generators just feed the motors directly, with the speed of the engine controlling how much power is going to the motors. At least, that's how it works for DC motors. I'll need to do more research about why you can hear the motors on 800s on diesel
Overall a very good explanation of the different train sounds Though just a few nitpicks: I feel like the 320 wasn't a very good example as the chopper hum sounds basically the same as the transformer hum The 319 has a much more distinctive hum with the 300ish Hz chopper and the 100Hz transformer Also the 323 and the 465 sounds from TSW are inaccurate as the game uses incorrect sounds (the 465 hum is pitched incorrectly and misses pulses and the 323 is missing pulses too)
Thanks for the advice, I was sadly unable to even ride a 319 before they were retired. I also wanted to try and give examples that are still in service today
@@zfrailways That's quite understandable, there are now very few chopper trains left in the world, the 92 stock is being retractioned with new AC motors and VF drive and the 319s are gone. Though I think there are still a few 769s in service with Northern which still have the original chopper equipment and motors since they are basically a 319 with diesel generators.
They still have a whole stack of Tangara trains in Sydney that are chopper trains. If you decide to look into these, would you be able to explain why the Millenium M sets (from Sydney) make the distinctive high pitch noise when they start moving, before then "dropping" pitch abruptly to continue to accelerate. I would imagine this is the change from PWM to pattern mode or something
Why did you copy Beno's video? You bring up the exact same points in the exact same order but change the clips. Why don't you cite your sources/credit appropriately in future?
I found Ben's video (or podcast) difficult to follow at first, so I made one with the intention of anyone being able to understand it. The more people that are talking about it, the more people are informed about these things and the misinformation hopefully dies away.
