The inverters are what is mainly the difference in sound. Simply how the voltage and frequency is controlled to feed the motor is where you get the difference.
As I found out on here a while back, they may look similar but they are completely different units despite being under the Networker family. Even the doors are completely different.
Something I pride myself on was the attention to detail I had aged 4-5 (in the year 2000) in relation to the Networkers. Back then I could link the two different motor sounds to the door buttons, door sounds and to a lesser extent the slight window shape variation. Connex's 365s were a major curveball for me at that age, as they combined the BREL exterior with a MetCam motor sound, while also looking much fancier inside. It wasn't until I was a teenager and had better web access, that I learnt about Class numbering and understood the differences in more detail.
Still would very down to how frequently the GTO inverters of the Brush Alstom switch on and off the voltage going into the motors vs that if the GEC. Its really all down to the frequency.
Nope, just different approaches to cooking the same egg in a nutshell. I doubt that a stock tender would have specified to that degree of intricacy in terms of the frequency the inverter switches on and off.
The frequencies are similar in different GTO drives from the early 1990s, as they are limited by the switching speed of the GTO thyristor as mentioned above. Different manufacturers faced the same challenge of driving the motor effectively across the speed range while avoiding critical frequencies that would interfere with the signalling. They will have adopted different algorithms in the software, so the sound varies in a different way. There may also be differences in the motor designs or the gearing, which changes the frequency they need to be fed with to get the same speed.
Because it's a hard problem without one obvious solution.
It's very exciting that it's been done for British trains now because for a while it was only rail enthusiasts in Japan who had worked out that you could do this.
It is a lot more complicated than that:Because it's a hard problem without one obvious solution.
This video - How electric trains work and why they make interesting sounds - has some great explanations of electric train motor control from the enthusiast perspective, if you've not seen it before.
I'll do my best to explain my layperson's understanding but it's been a long time since I had physics lessons. Corrections are welcome.
This GIF is a good visual aid for getting your head around what the asynchronous AC motor needs to do - you need a constantly varying magnetic field to control it, which you get from supplying three different phases of AC power to separate coils in the motor. Each phase needs to be synchronised with the others, but offset, to keep this field changing the right way. And in order to change the speed - and as the speed changes, because there's a feedback loop - you need to constantly change the frequency of all the phases.
So you need to be able to generate AC power at arbitrary and changing frequencies - this is why this form of motor control is called a variable-frequency drive. And the way this is done is to start with DC power and convert it into the correct shape of AC power, which is all 'inverter' means. It's done with transistors (or thyristors, but the differences don't really matter for understanding the train sounds. "GTO" and "IGBT" etc are implementation details and can be misleading in that regard).
You can think of a transistor as being like a remote-controlled switch that can turn on and off in fractions of a second. If you turn the DC power supply on and off fast enough and with some strategically placed delays, you can basically fake an AC waveform - it's ripply but it's good enough to get the motor to work.
This technique is called pulse-width modulation and it's used all over the place, including in some later trains with DC motors where it's just used to vary the voltage. The inverter generates the pulses and they induce audible noise in the motor itself.
The transistors are switched by a microcontroller in order to make the pulses and the pulsing strategy is just part of a computer program that runs on it. Where the pulses should be in order to generate the necessary waveforms is up to the designer of that system. Think about all the diverse software you use on a PC, and how different programs can be even if they are solving much the same problem - programmers are reinventing the wheel all the time.
The Wikipedia page for PWM shows some of the different strategies just for generating waves in the first place. Your program may well have inputs that it has to make complex decisions with, like the current speed or state of the magnetic field in the motor. You're switching very high power motors so you'll generate a lot of electromagnetic noise, in an environment where there is sensitive signalling equipment all over the place. Then there are economic and physical limits to microprocessor power and the viable switching frequencies of the transistors/thyristors. And one of the factors is how smart the programmer was feeling!
All that affects which pulse modes you need and where, and how complicated the strategy can be in the first place. A more complicated strategy might produce less noise, be more efficient, or prolong the life of the motor, but may well be beyond the capabilities of the components you have to work with.
Add on intellectual property and closely guarded trade secrets and it's no wonder that there's all these different ways of solving the same general problem! It's just complicated.
There are families of these systems. This stuff tends to be developed by big engineering conglomerates with a long history of mergers and acquisitions; so Bombardier even before they were bought out by Alstom, for example, produced at least two different kinds. Which leads to strange things like how the Electrostars on Gautrain have the same (or similar) motor but a different inverter to the British ones, and therefore sound different.
It is broadly the case that trains from the dawn of this technology sounded more 'interesting' because the lower pulsing frequencies tend to be more noticeable, and they have more and/or earlier changes in pulse modes which are what produce the 'gear changing' effect. But there are lots of modern trains that still have very distinctive pulse changes, like the dramatic switch to synchronous pulsing modes you hear on Electrostars and the Hitachi AT family.
Post #4 demonstrates something really cool, which is that, as I mentioned earlier, this process doesn't just generate audible noise in the motors, it generates electromagnetic noise over a much further distance than just the inside of the motor itself. So you can use an inductive coil to turn this back into electrical signals and see the waves that they make up for yourself and hear them much more distinctly. You can go and look at induction recordings on YouTube and compare by sound and visual representation of the waves what each type of train is doing, and become even more attuned to and obsessed with these sounds, if you're the type to
Neither are Siemens built. Alstom was the manufacturer.
I'm wary of spreading misinformation on a topic where enthusiasts are often left grasping at something technically complicated, but hoped it was worth trying. An intuitive/conceptual approach can only get me so far before I have to find a textbook and learn some maths...There are other frequencies that must be avoided too, such as the Reed track circuits that operate at a range of frequencies in the low hundreds of Hz. DC systems generate hundreds of amps of return current, but if more than an amp or so of this is at one of those frequencies (or 50Hz) and there is a danger of the train disappearing from the signalling system. So this makes the traction software even more complicated, especially if it was previously developed for a different country that didn't have these constraints. Not to mention the problem of proving to very high levels of certainty that there is no credible fault that can produce any of these frequencies.
It is a lot more complicated than that:
All rail three phase drives use the most complicated Vector control* or more precisely Direct Torque Control (DTC) or Field Orientated Control (FOC) , of which there are several several sub flavours and descendants of each depending on whether you are the in the Siemens / Infineon (DTC) or Brown Boveri (Mannheim) / RU Bochum (FOC) [e.g. ABB/Bombardier] camps for IP . The family tree in the wiki article below needs extending to the right with more generations in practice (no mention of DSR or ISR descendants of FOC (your Electrostar/Gautrain examples)).
* a very basic wiki article on Vector control (with lots missing):
Field-oriented control - Wikipedia
en.wikipedia.org
VVVFD is the correct terminology most people get lazy and lose a few Vs (or need a TLA?)
One reason the 345s are quiet is that the motors have external blower motors for cooling so there is much less motor rotational noise escaping and no motor fan noise. Also the electronics are all contained in fewer larger boxes (often covering motors on multiple vehicles) with just 4 boxes overall on 345 so much less radiative potential, not the several boxes (often multiple per vehicle) of older products like Aventra or Desiro, which is much better for EMC and weight (far less box and cabling mass).
Toshiba and the other Japanese electronics firms involved in rail traction are generally philosophically aligned with the ABB (mostly ex BBC Mannheim with elements of Zurich and Vasteras) school of thinking so no surprise they sound similar to the original ABB for Brush 465 drives (also see Cl 92 and EuroTunnel locomotives for similar ABB - Brush partnerships)
Nice theory, but the “Do re mi” sound is from a German company (Siemens) and can/could also be heard on DB ICE power cars (underneath the sound of the blowers) which predate Taurus by several years.
I recall when they were new that there were a lot of problems getting the two types to work together.
I don’t know about whether the Taurus sound was intentional, but I agree that manufacturers can change how trains sounds as I’m sure there used to be a video of a Taurus “singing” the Austrian National anthem or something for some national holiday
Siemens engineers have got “test mode” on ICE3 sets to play the German national anthem.
Entirely deliberate - Done by the Siemens Vienna works (also where the 450/444 and the LU2024 were / are mostly made), some other Siemens Vienna engineered products have a few notes of Mozart. Siemens don't just do stuff in Germany.
"The trains are alive, with the sound of music..."
