ECN Electrical Forum - Discussion Forums for Electricians, Inspectors and Related Professionals
ECN Shout Chat
Recent Posts
Does NEC 551.71 (F) apply to dwellings?
by BigB - 01/20/23 10:46 AM
by HotLine1 - 01/19/23 09:10 AM
Power submeter connections
by HotLine1 - 01/19/23 09:09 AM
AFDD's coming to the UK
by Texas_Ranger - 01/17/23 07:22 PM
Circuit Breaker Lube
by renosteinke - 01/14/23 06:40 PM
New in the Gallery:
Burger King crown sillyness
Burger King crown sillyness
by wa2ise, December 11
240/208 to a house
240/208 to a house
by wa2ise, October 9
Who's Online Now
1 members (Scott35), 32 guests, and 21 robots.
Key: Admin, Global Mod, Mod
Previous Thread
Next Thread
Print Thread
Rate Thread
#167888 08/22/07 05:54 AM
Joined: Aug 2001
Posts: 7,520
pauluk Offline OP
Thanks to Alan Belson and his library for the following.

HVDC: The Transverter, 1922

Announced to the public in 1922, a 2000 Kw Transverter machine was demonstrated at the British Empire Exhibition at Wembley, England in 1924 by the English Electric Co Ltd., where it received considerable attention.

A contemporary exhibition report stated it was designed to convert ac supply to high voltage dc for transmission purposes, was configurable for any commercial frequency and could have been used to link interconnecting power stations operating on different frequencies. With an HVDC line, the power-factor complications of parallel running were avoided and the overall losses were stated to be on a par with those involved using step-up and step-down transformers and frequency changers.

Some observers claim the Transverter is the forerunner of all HVDC transmission systems.


The efficient transmission of electricity over large distances was linked to the growth of National Grids at this time. Large power generating stations, often sited on coal-fields or as hyro-electric schemes, produced electricity far more efficiently and cheaply than a plethora of small local generating plants. These often ran dc supplies to consumers, but were being phased out as uneconomic, though I understand Consolidated Edison still supplied dc to 1600 customers in New York till about 2005. High-voltage transmission was essential to reduce losses for transmission over any great distance and required ac for conversions using transformers. However, many consumers still required dc supplies.

To make use of the cheaper ac supply, many dc consumers opted for conversion machinery. This was more economic than running their own prime movers and dynamos. Solid state devices were not available other than lab curiosities and many British Engineers distrusted contemporary mercury-arc rectifiers, [MARS], such as Peter Cooper-Hewitt's 1902 invention [ see ], as being 'unreliable & unproven technology'. It is true that these early MARS rectifiers could be cantakerous and unpredictable, with the arcs 'going out' for no apparent reason, but more sophisticated MARS rectifiers are still used in some HVDC installations in Sweden.

For small-scale installations, such as private houses and small businesses, what better than a simple rectifying machine, a Motor Generating set? PIC 1 shows a neat robust ‘Crypto’ set of the period, reasonably efficient and available in 16v 5A (80W) up to 200v 100A (2000W). Run your lights, a cinema projector, charge batteries, or supply a welder, but no good for use on inductive loads due to pronounced dc ripple - which causes brush-sparking due to power factor lag.

[Linked Image]

A motor-generator does not make efficient use of materials, particularly of copper, once you start scaling up to serious capacity for traction, factory or mining use. For that, one had the choice between a Motor Converter or a Rotary Converter. Both machines had their vociferous advocates, with not much to choose between their salient advantages and disadvantages.

The Motor Converter is similar to a motor-generator, with 2 stator-frames: A separate Converter armature is driven at 1/2 synchronous speed by an inline induction or synchronous motor. The field coils of both motor and generator are connected to the supply and the motor rotor and generator armature are interconnected electrically. Later machines were 2-bearing designs which eased this interconnection. Near ripple free dc is tapped off via slip-rings and brushes. About 50% of the dc power comes from power converted from mechanical torque arriving from the shaft, with 50% transformed directly in the windings, (when running at half synchronous speed, the usual arrangement then). See PIC 2:

[Linked Image]

Click here for larger image

Capacities up to about 2500kW were available at this time, large enough for most dc purposes. Features in favor of this machine are its ruggednes, good dc voltage control and good sag (brownout) and harmonics/transients characteristics, linked to its mechanical inertia. These machines' motors could be connected directly across an 11000v supply.

The Rotary Converter went further: All the windings, motor, primary and secondary were on one armature in one stator frame, (albeit that the armature usually had to be bought up to operational rpm by a pony motor and exciter circuit). Since both motoring and generating take place in the same armature conductors, we get near 100% direct conversion, as if the ac supply were being freshly generated and with the field windings' flux being merely there to maintain synchronous motion. This is a slightly more efficient, lighter machine, the shaft has to transmit little torque; but its detractors said not lighter if you added in the transformer and its tappings needed to supply the motor and control the output volts, for the supply/output voltages had to agree. It was also susceptible to harmonics and transients, in which the Motor Converter had the edge. Slip-rings and brushes again took off the near pure dc output. See PIC 3. Again, up to about 2500 kw capacity machines were being built at this time. Later machines went into the 5MW range.

[Linked Image]

Click here for larger image

Now, characteristically, Converters have 2 wave-forms in their windings; the sinusoidal waveform of the primary ac delivered to the slip-rings and the waveform of the secondary direct current side which, because of the salient field, is approximately rectangular. A differential current will therefore flow, due to the difference between these waveforms. This current will be absorbed in the closed circuit of the armature windings - it’s a fundamental principle of multi-phase transformers, even those rotating in an armature, that the sum of the ampere turns in each limb shall equal zero at all times.


This machine tackled the conversion problem in a different way, for it was designed for HVDC transmission rather than to make end-use dc. Two Transverters were thus required.

Here's the spec. for the 1924 Exhibition machine: A group of transformers formed the magnetic circuit, a commutating group gave the requisite number of fixed commutators, and a discrete shaft carried revolving, synchronous-motor driven brush-gear. Incoming 3-phase ac was taken to windings on a transformer group with 18 limbs to produce 36-phases, thus performing all voltage changes, phase multiplications and smoothing simultaneously in non-rotating primary windings. The secondary windings consisted of similar coils placed on the same 18 limbs as the primaries, i.e. as many sets of secondary windings as there were commutators. In the Wembley machine there were 10 commutators, so there were 10 different sets of secondary windings, each connected to its own commutator.

At the receiving end the process was reversed and ac re created for distribution. By varying the brush-gear rpm one could change from 50hz to 60hz for instance, and the two stations need not run in sync.

The only moving parts were the brush-gear and its synchronous drive arrangement.

In the magnetic system, the flux rose and fell in unison with the alternating-current frequency, and further, the flux rotated by using delta connections for the primary windings. The secondary coils on the eighteen limbs to the commutator bars were connected so that the voltage of the latter was built up regularly and progressively from one pole to the other. In effect, the opposite arrangement to an ordinary direct-current revolving armature, in which the armature coils are mechanically brought, in succession, under the influence of the fixed magnetic field, so that the voltage builds up around the commutator in a regular progression. In the Transverter the coils were fixed, but their connections were such that the regularly rising, falling and rotating flux successively influenced them, also in a regular progression. The brushes revolved around the commutator, successively short-circuiting the secondary coils that were at zero potential, thus achieving dc conversion.

In this case, as in all poly-phase rectifiers, there were two waveforms to consider, i.e. the sine wave of the incoming ac, and the rectangular waveform of the secondary windings, in which the current is constant for any particular load. These have been plotted in PIC 4 Fig. 242, showing the difference in a separate curve:

[Linked Image]

The magnetic system was essentially a 3-phase transformer, even though it had 18 limbs and 36 phases. As stated earlier, the sum of the energy ampere turns on each limb must equal zero at all times. So, if the primary waveform differs from the secondary waveform, there must be a reaction to fulfil this fundamental principle. In fact, harmonics would be produced in the circuit, and these would flow under the brushes and produce the most violent sparking if unchecked.

To meet these conditions the Transverter was compensated by placing additional windings on the magnetic limbs, connected in opposition to similar windings of an equal voltage and of a known sine wave-form.

The synchronous motor driving the brush-gear formed the source for this sine wave. 18 tappings were brought from the windings of the motor and connected to 18 compensating coils on the Transverter magnetic circuit. The sine waveform of the synchronous motor was thus imposed on the Transverter’s magnetic system. Should the Transverter waveform tend to differ from a sine, then a transfer of ampere turns would occur from the magnetic limbs that were high to the limbs that were low, via the compensating windings. The undesirable harmonics were thus filtered from the main primary and secondary windings, whose waveforms remained truly sinusoidal irrespective of load. Consequently the brushes had only to commutate the ordinary reactance voltage of the coils under short-circuit.

The value of the reactance voltage with coils such as those used on transformers is much less than would be the case with an ordinary slotted rotating armature. With the practically unlimited winding space available on a transformer limb, any number of sets of secondary coils might be used, which would all be undergoing the same induction at the same instant.

In this Transverter there were 10 sets of secondary windings, each connected to 10 commutators, each commutator producing 10,000 volts. All the commutators were in series, and so produced 100,000v dc. A vector diagram, PIC 5 Fig. 243, shows how the component voltages built up. Note that in the outer circle, which represents the resulting dc, there are 36 voltages concerned, and that for any one of these there exists an equal and opposite component. These individual voltages were obtained from the 3-phase voltages, corrected or filtered by inductive coupling with the stator windings of the driving motor.

[Linked Image]

In the actual machine, construction was as follows. The 18-phase transformer consisted of 3 separate units of 6 phases each, construction following ordinary transformer practice of the time. The outer limbs were -removable to facilitate the withdrawal of windings. Primarv windings were carried next to the magnetic core, mounted on insulating cylinders. Next to the primary windings were the compensating windings, separated from the primary by insulating cylinders and spacers. The secondaries formed the outer windings.

The bottom half of the transformer tank was formed by a cast iron bed-plate. There were 3 upper tanks, one for each 6-phase transformer, all of welded steel, bolted to the bed-plates. Connections between the transformers and commutators were carried in racks laid in the bottom of the tank, and were made of insulated cable suitable for oil immersion. Breathers and sludge-tanks were also fitted. At emergence at tank seals, the cabling insulation changed to such suitable for air. Radial commutators were used, all bars being cast solid in a ring of molded insulation and then bolted to a slate frame, insulated from ground by porcelain insulators.

The rotor was simply a straight steel shaft carrying a number of insulating discs and distance pieces. It formed, from end to end, a completely insulated unit, with the brushgear carried from the discs. The brushgear was extremely simple, consisting of a box which carried the carbon and a light flat spring to hold the latter in position. A lever arm was arranged to compensate for centrifugal forces.

PIC 6 Plate No. 5 shows two 2500kw Transverters in the Works, while PIC 7 Fig. 244 shows a rotor:

[Linked Image]

Click here for larger image

[Linked Image]

Click here for larger image

PIC 8 Fig. 245 gives a nearer view of the 36-phase transformer:

[Linked Image]

Click here for larger image

In operation, the machine makers claimed it had generally the same characteristics as a rotary converter, and could be started up from the ac or dc side. It might be expected that starting up from 100Kv dc would present difficulties, but this was claimed to be not so, the dc was switched on to the standing brush-gear through a high resistance and the machine was up and running in a few seconds. The characteristic curve drooped slightly from zero to full load, the drop being due to the resistance of the windings. Overall efficiency was high, in the order of 95% being claimed.

Later HVDC equipment used MARS, or semiconducting rectifiers, to achieve the switching done in the Transverter by its brushgear, with the advantages of cost, longer life and lower maintenance.

The Transverter was patented in the names of J. E. Calverley and W. E. Highfield.


Last edited by pauluk; 08/23/07 02:33 AM. Reason: Typo.
Joined: Mar 2005
Posts: 1,803
Correction- my typo; “Later machines went into the Mw range” , should have read “5Mw range” of course. They were already at 2.5Mw by 1922.
Correction made in main text - Paul.
1942; Westinghouse. Incredible pic., at 5.44 Mw, the biggest Rotary Converter I have yet discovered, with, unbelievably, a 2-storey workspace built into the end-cap, with two doors, guys working on both floors and complete with windows! cool
5 star site! Showing Rotary Converter machines, in NYC subway traction system, with color pics and a superb description of their history, technology, layout, buildings and use. Work your way through all the "click heres" and enjoy! cool cool


Last edited by pauluk; 08/23/07 02:36 AM.

Wood work but can't!
Joined: Mar 2005
Posts: 1,803
Bernard S Greenberg, writing in the link [which I mentioned in my previous post ], poses a question on the Rotary Converter- Who was responsible for its invention? In what looks like its final form, ac to dc for electric traction, a 1902 Converter is shown on this site, [ click you would see what looks for all the world… Manhattan Elevated Railway, exhibit # 8 ]. Mr Greenberg plumbs for Nikola Tesla.

If we go back 9 years earlier to the
[The opening shots in the War of the Currents GEC/ Edison v Westinghouse/ Tesla] :-'s_Columbian_Exposition
See ‘3, Electricity at the fair.’
This exposition was remarkable and could be said to have created the 20th Century as we knew it.
Westinghouse & Tesla got the contract to ‘electrify’ the Chicago Expo, doing it for half the Edison/ GEC’s $1M ‘dc’ bid. In an impossible timeframe they supplied ac generation, circuits for lighting, power and demonstrations. They exhibited, among other things, induction and synchronous motors, Electric meters, phosphorescent lighting, neon tubes & incandescent bulbs. GEC actually banned them from using Edison’s patent bulbs at the Expo, in a fit of pique at not winning the contract. Tesla just side-stepped and broke their patent with a unique double-stopper incandescent bulb.
Also shown were polyphase systems, transformers, and a Rotary Converter for making dc from ac supplies . This exhibit included an operational dc railway motor.
This success was certainly a factor in ac being chosen for the

The final Victory in the Current Wars.
The board of directors changed the specification from dc to 25 hz ac generation during construction of the project.
According to records, Rotary Converters were used in this plant.
If we examine the Westinghouse bronze plaques at Niagara, [ and strangely there appear to be 2 versions, 1897 and 1899 ], and look at Tesla’s patents listed on them, we find that he did indeed file patent on a ‘rotary converter’, Dec 23 1887.

Application serial no 258787 23 Dec. 1887
US Patent No 382282 granted May 1 1888;
Here it is in full:

The patent is unclear as to the output current form, [ac or dc] indeed there is little detail as to function at all. This is a common feature of the ‘perfect’ patent application, written by lawyers – don’t say anything you don’t need to which may lead to someone actually learning something and breaking your patent!
It appears to be some sort of rotating transformer, yet why have a commutator set on an ac machine if not to draw dc output? Further, he claims that the device keeps the poles at their maximum flux during rotation pole to pole- implying a rectangular waveform? . In his file, Tesla mentions schemes which already exist where motor and generator windings are on the same armature, which had been built in lieu of static transformers, so someone seems to have invented that arrangement prior to Dec 1887. However he claims these devices were built with little understanding of the principles involved in correct proportioning of the parts and placement of commutators, ie “it wasn’t me!” He also claims that his converter operated as if it commutated an infinite number of poles. All in all I’d tentatively guess this was indeed an ac-dc converter which does not generate dc from mechanical shaft torque.

But, I’m still not 100% certain of whodunnit!

Home page-
With enough pics to keep you going for months!


Wood work but can't!
Joined: Mar 2005
Posts: 1,803
I think I found out Whodunnit. smile

While Tesla came close with this patent, granted Oct 22 1889, [ refer specifically to Fig 5. ]

I am confident the honor must go to Charles S Bradley of Yonkers, Westchester County, NY., another disaffected ex Edison employee who struck out on his own.

US Patent # 409450
Filed Oct 20 1888. Granted August 20 1889,M1
A full rotary converter scheme, all on one armature, ac supply, dc delivery.

Bradley issued a raft of patents at this time for polyphase motors and related electrical devices.
He sold his company to GEC in 1897, who by now were committed to ac generation and transmission. [Edison seems to have taken a ‘back seat’ at GEC after Niagara.]

Westinghouse Electric are recorded as frantically working on both the Columbia Expo and research into polyphase motors and a rotary converter in 1892, after ceasing research in 1890 due to funds shortages. These devices were both essential to their Niagara proposal, since they had to provide dc supplies for Buffalo as well as effective ac industrial motors if they were to succeed against Edison’s dc bid.


Wood work but can't!
Joined: Nov 2002
Posts: 174
When I was a child, my dad worked in a plywood mill and they had an arrangement similar to this to convert 480 AC to DC for use in machine control.

Link Copied to Clipboard

Tools for Electricians
Tools for Electricians

* * * * * * *
2023 National Electrical Code (NEC)
2023 NEC Now Available!
* * * * * * *

2020 Master Electrician Exam Preparation Combos
2020 NEC Electrician
Exam Prep Combos:
Master / Journeyman


Member Spotlight
lil suzi
lil suzi
Posts: 57
Joined: August 2003
Top Posters(30 Days)
BigB 9
Popular Topics(Views)
302,338 Are you busy
231,740 Re: Forum
216,365 Need opinion
Powered by UBB.threads™ PHP Forum Software 7.7.5