My late father was a clerk on the LIRR from 1963 through 1983 and worked for some time in the Ticket Receiver's office in Penn Station. Often those FL9s would run through the tunnels and into Penn Station on diesel power, the rules be damned. They wouldn't shut the prime mover down for fear that they wouldn't be able to restart it. I personally experienced the stench and the haze of their exhaust, on more than one occasion, as a commuter on the LIRR. The same applied over at Grand Central Terminal. The steam boilers on the GG1s were supposed to be shut down too, from the Bergen Hill portals through Long Island City, but very often weren't.
Electroliner 1935 daveklepper (I suppose you could put a 50 plus 50 6L6 output tubes in push-pull-parallel to drive one 16-ohm loudspeaker.) Nice in winter but the summer air conditioner bill would be a problem.
daveklepper (I suppose you could put a 50 plus 50 6L6 output tubes in push-pull-parallel to drive one 16-ohm loudspeaker.)
Nice in winter but the summer air conditioner bill would be a problem.
I had an uncle who was an EE at Lockheed for a couple decades after working at radio and TV stations for a decade and a half. One of his projects was making a direct coupled tube amp, one probelm with it was that the output would drift close to one of the supply rails until a loud segment of music came on and it would go back to center.
One way to deal with the load impedance issue is to wire a bunch pf 16 ohm speakers in series...
Dave's right about transistors' low impedance making direct coupled outputs possible. It did take a bit of breaking away from tube design with the use of bipolar power supples instead of single ended supplies to get away from output capacitors. That and some nifty tricks in current limiting for the output transistors.
Now for something a bit more on topic for the forum...
I was browsing through the "Wolfspeed" (Cree's RF and Power division) website and came across a paper on three phase inverters for motor applications. They were claiming that an inverter running running at a 5 kHz switching frequency and producing ~60Hz output was doing so with what looked to be 99.2 to 99.3% efficiency when supplying a 50HP motor at full load, where the best standard silicon inverters were barely touching 99% efficincy. If a 0.3% improvement in efficiency doesn't sound like much, consider that a 40KW inverter would pump out 400W of heat with silicon and 280W of heat with SiC. The weight of the electronics is just a few pounds (the module with 6 SiC FET's weighs about 1/2 lb), dominated by the DC link capacitors and the heat produced is probably about what my uncles direct coupled tube amp put out.
- Erik
Power transistors are inherintly low-impedance devices, and so direct-coupling is possible. High-powered output vacuum tubes are (were -- although I understand there are still hi fi fans and manufactureres who swear by tube amplifers, the remaining distortion is supposed to be more tolerable and pleasant to their ears than transistor amps. I confess I don't hear the distoriton in either type.) are inherintly high-empidance amplfiers, and so transformers were esssential. (I suppose you could put a 50 plus 50 6L6 output tubes in push-pull-parallel to drive one 16-ohm loudspeaker.)
In addition to being low-impedance, transistor audio amplifiers have the advanages of greater efficiency, much lower heat production, far more compact and less weight. Greater cleverness is required to control distortion, but this is a solvable design problem. Also, transistors can seem to last indefinitely, whereas vacuum tubes needed replacement, much like incandenscent bulbs. This was particularly true of output tubes. The technlogy has progressed where multi-driver loudspeaker systems, with specific loudspeaker elements to cover specific ranges, and perhaps several for each range to shape directional characteristics, can have one power amplifier for each speaker element, and then frequency response and even microsecond delay opmtimized for the purpose of that particiular driver, with the whole shebang coming from the digital to analogue converter right at the loudspeaker system, and the whole audio system between there and the microphone, inlcuidng condtrol, digital. Not very possible with vacuum tubes!
A couple of more notes on multifrequency transformers.
The upper frequency limit for transformers using laminations is set by eddy current losses in the laminations, with limit efined by how much eddy current loss is tolerable. The approx 500VA transformer in my Crown D-150 amplifier was stated by Crown to work at 50, 60 and 400Hz, though the transformer may be on the upper limit of acceptable operation at 400Hz. High frequency transformers typically use nonconductive ferrites for the magnetic material.
Designing a good High Fidelity Audio transformer is not a trivial undertaking, with about a 1,000:1 range in frequencies (hats of to Dave on that). Tricks used in design include high-mu magnetic material (eg. mu metal, permalloy) and layered windings. One reason that Crown and other manufacturers went to direct coupled amplifiers was to get around the difficulty of dealing with output transformers.
Locomotive transformers were not considered as all manufactured for this country are dual frequency dual voltage motors. Now if you consider overseas the UK 50 Hz transformers would be close but not exact.
Just for grins. If the NEC including SEPTA is ever (?) converted to 60 Hz would replacement transformers be smaller ? Although only if dave and I live to be 200 years would we see it happen..
Remember that the dual-power locomotive transformer handles about ten-to-fifteen times the power of an mu-car transformer. This is in addition to the weight and size required for the 25Hz capabiklity
It would be interesting to compare the weight of say NJT or Septa's 25/60 Hz EMU transformers to the M-8's 60 Hz only transformers.
You and I are in agreement on not needing to change turns ratio if the primary voltage remains the same. My point was that for a constant maximum flux density, a transformer would need the same number of primary turns when running 10KV/25Hz and 24KV/60Hz - which is darn close to 11KV/25Hz & 25KV/60Hz.
I did have to edit my prevous posting to reflect constant Phi_max, not constant dPhi/dt (the latter gives constant voltage).
If you are clever in handling the design and balancing the losses, you should not need to change taps when changing between the two frequencies, only when changing input voltage. In the case of 25Hz, the resistive losses would be highter, and in the case of 60Hz the core losses would be higher, but the total losses would be equal at the two frequencies. Otherwise, your analyis is correct.
daveklepper While a student at MIT, and I have both SB and SM degress in EE, I designed transformers of all types at Mystic Transformers, while also an unpaid employee of Boston and Maine Transportation while working on my SB thesis on Diesel Locomotive Load Regulator Controls, solving my Cambridge - Winchester transportation problem. rather neatly. A transformer at 50 or 60 Hz can be smaller than one for 25 Hz, because during each half cycle, flux build-up in the magnetic structure is half or less. You can use a 60-Hz transformer at 25 Hz, but only at one-half the current, meaning one-quarter the total power. The other way around is also a problem, and in that you are correct. A normal 25 Hz transformer will be much larger than the equivalent 60 Hz transformer, and capacitive coupling short-circuiting between coils and hysterises magentic losses will start to reduce effiency, the power lost will create heat, and the danger of insulation burning or charring with the best of materials is possible. Multi-frequency transformers are possible. Indeed classic tube-type audio power amplifiers have wonderful examples, with reasonable efficiency, say 75-90%. over the hearing range 20-15,000Hz, typical. But the iron is far more expensive (Alnico or better), and often a tiny air-gap is introduced to linearize the response while sacrificing a bit of efficiency. (Think the famouse Williamson Amplifier transformer, and its interpretation in a Heathhkit.) An efficient dual-frequency power transformer that works well at both 60Hz and 25Hz can be designed using regular iron, and indeed Amtrak's electric locomotive fleet and Acela trains would not work without them, but they are much larger and more expensive than regular 60Hz transformers.
While a student at MIT, and I have both SB and SM degress in EE, I designed transformers of all types at Mystic Transformers, while also an unpaid employee of Boston and Maine Transportation while working on my SB thesis on Diesel Locomotive Load Regulator Controls, solving my Cambridge - Winchester transportation problem. rather neatly. A transformer at 50 or 60 Hz can be smaller than one for 25 Hz, because during each half cycle, flux build-up in the magnetic structure is half or less. You can use a 60-Hz transformer at 25 Hz, but only at one-half the current, meaning one-quarter the total power. The other way around is also a problem, and in that you are correct. A normal 25 Hz transformer will be much larger than the equivalent 60 Hz transformer, and capacitive coupling short-circuiting between coils and hysterises magentic losses will start to reduce effiency, the power lost will create heat, and the danger of insulation burning or charring with the best of materials is possible. Multi-frequency transformers are possible. Indeed classic tube-type audio power amplifiers have wonderful examples, with reasonable efficiency, say 75-90%. over the hearing range 20-15,000Hz, typical. But the iron is far more expensive (Alnico or better), and often a tiny air-gap is introduced to linearize the response while sacrificing a bit of efficiency. (Think the famouse Williamson Amplifier transformer, and its interpretation in a Heathhkit.) An efficient dual-frequency power transformer that works well at both 60Hz and 25Hz can be designed using regular iron, and indeed Amtrak's electric locomotive fleet and Acela trains would not work without them, but they are much larger and more expensive than regular 60Hz transformers.
A transformer with a 25KV/60Hz primary should just about work at 11KV/25Hz, the required Phi_max (flux) is about the same provided that the number of primary turns stays the same (the unit for Phi (Flux) is the Weber, not to be confused with B, which is flux density measured in Tesla (Webers/sq meter)). Available flux is limited by the saturation flux density in steel and may be limited by core loss). The bad news is that the secondary windings would have have the change-over taps to maintain the same output voltage. The increase in copper losses would only be partially offset by reduction in core losses.
The rule of thumb for power transformers is that the copper loss should equal the core loss. Since this relationship varies with frequency, a multiple frequency transformer will end up being a compromise. Compromise is sort of relative here, it may mean 98% efficient versus 99% efficient and heavier than something designed solely for operating at the highest frequency.
P.S. The cgs unit for flux is the Maxwell, also known as a "line" of force. The cgs flux density is the Gauss, equal to one Maxwell per sq cm. Engineering flux density was lines per sq inch.
P.P.S. I'm currently working on a unique transformer, hence the familiarity with transformer design rules - had thoughts about Dave when working on it over the last year.
blue streak 1 NEC standards have gradually increase allowable voltage from the pre WW-2 110 V to ~130 V although still referred to as 110 V. For those of you that have electric motors check their plates. Have a pre war listed as 110 V and motor bought last year stating 110 - 130 V / 220 - 260 V My line voltage here is reading 127 V.
Where I worked, the substation transformer voltage regulators were set to provide 127 V leaving the sub, this allowed the first customer to be within the voltage limits and with additional switched capacitors on the line to maintain a voltage to the meters at 113 V or above to meet the 120 V ± 7 volts service requirements. At least that was the plan.
Blue Streak, do you know how close you are to your utility's feeder source? The 127 V looks like you are close to it or perhaps they have a regulator near you.
Amtrak HEP is the industry-standard utility 277/480 "Y" configuration. This allows the use of 480V 3 phase motors (for A/C) and 277v single phase fluorescent bulbs, as well as easy setup of shore power in servicing areas. A standard transformer set makes 120v available for things like laptop chargers and vacuum cleaners, in a 120/208v "Y".
What Daveklepper is describing is a high voltage "Delta" configuration with a center tap on one phase for an intermediate voltage. In a low voltage (120/240) commercial setup a large center-tap 120/240 transformer will be accompanied by a smaller 240v transformer on a second leg of the delta. No actual transformer is required for the third leg, where a "shadow" 240v phase is formed.
There is something about the 25Kv system not mentioned. The feeders on the Amtrak New Haven - BOS system uses the same system as your house. Your house has 110 V / 220 V system where it is 2 - 110 V to ground and 220 V feed to feed. Amtrak and probably NJT uses two feeders that are 25 Kv to ground and 50 Kv feeder to feeder. Remember doubling voltage increases power capability 4 times. V sq / Resistance.
Just as an aside if you put an voltmeter on your house wire you will find the indicated voltage higher. NEC standards have gradually increase allowable voltage from the pre WW-2 110 V to ~130 V although still referred to as 110 V. For those of you that have electric motors check their plates. Have a pre war listed as 110 V and motor bought last year stating 110 - 130 V / 220 - 260 V My line voltage here is reading 127 V.
Note Amtrak HEP is listed as 480 V - 3 phase. so dividing twice give 120 V.
Back to Amtrak. An auto transformer is connected to lineside feeder to feeder and then has a center tap to ground and outside taps giving 25 Kv to the CAT.
The feeders get their power from the commercial grid by way of an auto transformer connected to 2 phases of the grid. There is a balancing circuit to the 3rd phase.
How MNRR does it is unknown . It may be 12.5 / grd / 12.5 giving 25 Kv feeders . Have read that New Haven RR used straight 11Kv 25 Hz but have no confirming information. If so that may have been the reason NYG - New Haven had so many power problems. The rebuilding of Hell gate line probably is the 12.5 - grd - 12.5 60 Hz feeders.
The Amtrak 25 Hz system is similar as each high voltage feeder is 69 Kv with feeder to feeder 138 Kv.
Again the M-8 can not handle 25 hz, to go to nyp it will run on LIRR third rail, the pantograph can not be raised if any of 8 shoes on a pair touches third rail, all Lincoln, nebr M-8's except for first cars made in Japan have new dual purpose third rail mechanisms.
The third rail is there only for the LIRR trains, and as far as I know, the only "foreign" power that ever used it were the Hew Haven FL-9s. If I remember, the eastbound flyover is used by both Port Washington LIRR trains and NEC trains, and has third rail. Just east of that flyover is the switch taking NEC trains off the LIRR, and the 25Hz cat continiues until the track is off LIRR property, where the break to 60Hz occurs. Westbound is similar, but without a flyover. So, to get 60Hz-only M8s into Penn, about 1-1/2 miles of double-trackthird rail would be required. Note the thru NH-NJT thru service for Meadowlands events used NJT locomotive-push-pull trains. NJT locomotives, like Amtrak's, can swith from 25 to 60 Hz and back at speed.
Dave: Below is link to Amtak's 25 Hz system that has a graphic showing the 12.5 Kv 60 Hz of Hell Gate. New Rochelle to Hell gate now is same voltage and freq the trains only pass thru a phase break.
25 Hz begins at Gate CP which is just south of Sunnyside junction ( freight line to LIRR ) but before Harold intersection. Your spectulation is probably correct of no 3rd rail Gate - Harold. Adding 3rd rail would require changing signal system to different frequency see article.
Article does have a few errors. PRR raised 25 HZ CAT from 11Kv to 11.5 KV in 1948. Amtrak then further raised voltage to 12 Kv at some later date unknown. These raises were accomplished by simply raising the High voltage feeds. CAT Transformers did not require changing only changing Safe Harbor, rotary converters, and solid state converters to higher outputs. Since the raises were within normal voltage ( see article ) +/- not a real problem.
https://en.wikipedia.org/wiki/Amtrak%27s_25_Hz_traction_power_system
Thank you both.
Mr. Klepper,what is that about "alnico" in audio transformer iron? I am only familiar with that alloy in permanent magnets which I would NOT think good for limiting hysteresis at AF! So I am missing something again. Do you mean "Armco iron" or something like mu metal?
(I found a historical review reference on transformer iron here (PDF download) for those who may be interested...
blue streak 1A very memorial moment in college was when a electrical engineering professor said to our class that electrical engineering was a mature profession and to only expect minor improvements.
There's a famous similar comment about physics, around the time "Hertzian waves" were recognized. Just more decimal-place improvements... there do seem to have been some changes since then.
With respect to 'never or always', I think Gilbert and Sullivan had the best observation...
An earlier poster noted that the Hell Gate tracks owned by Amtrak have been converted to 60 HZ, and I assumed that was correct. But there is still some LIRR track not converted and without third rail between the end of Amtrak ownership and the LIRR third rail. I am unsure whether the Hell GAte and Bronx Amtack tracks are 12,500 or 25,000 volts, but I think the former.
I am the one with Dislexia, not MIT. The degrees are correct as you state.
daveklepperBut I understand there is still a section of 25Hz bertween LIRR third rail and the start of 60Hz on Amtrak. Possibly LIRR third rail can be extended over this short section, and a mile further to provide some necesary overlap with the 60Hz catenary.
You have me confused here. I presume youmean the catenary AC voltage & frequency and not the third rail which is DC. I think the Hell Gate Cat is still 25 Hz as is Sunnyside's .
daveklepperWhile a student at MIT, and I have both SB and SM degress in EE,
Dave, Did you mean BS and MS as in Bachelor of Science and Master of Science or did MIT have dislexia?
Dave thanks for the transformer primer.
A very memorial moment in college was when a electrical engineering professor said to our class that electrical engineering was a mature profession and to only expect minor improvements.
So now we have much more efficient transformers, IGBTs, variable frequency AC motors, fiber optics, solid state electronics, all the information now on just plain telephone wires, etc. Allways laugh at those who make never or always statements especially pols.
Dave, I had not known of flat wire, but after seeing your reference to it I understand how it reduces the size. I think of how difficult it is to lay an ordinary cord flat when winding it on a reel.
Johnny
But then, at least most, if not all, the NH electrics with adjustable shoes by power, simply retracted them, not using the LIRR 3rd rail. Indeed, I have learned from these forums that before the Trenton - Sunnnyside electrification, New Haven Electrics would be exchanged for DD-1's at Harold, and these of course would run through to Manhattan Transfer.
In general, the FL-9s were used to Penn only when a Jet was not available. After the great EF-3's had been sidelined and Van Ness closed and the freight dieselized until the advent of the Ex-Virginia EF-4's.
A power transformer that operates at both 60 and 25 Hz would be designed with the same size and number of coils as a pure 25 Hz. But it would have twice the number of iron laminations, each half the width, approximately,,with superthin insiulation between laminations, and it would have a segmented coils rather than straight coils. It would perorm even better, marginially, at 25 Hz than the lower-cost regular 25 Hz transformer, and it would be efficient at 60Hz. In any power transformer, more expensive and better iron, and the use of expensive square wire instead of round wire are measures to reduce size and weight.
I was unaware that the new M-8 has a shoe similar to the original EMD-FL9 dual-sprung third rail shoes, that with good shoe and rail maintenance, can work well on both overrunning LIRR and underruning MN third rail. I thought they still had the standard MN shoe which works only on underruning third rail. But I understand there is still a section of 25Hz bertween LIRR third rail and the start of 60Hz on Amtrak. Possibly LIRR third rail can be extended over this short section, and a mile further to provide some necesary overlap with the 60Hz catenary.
Although my memory may be incorrect, what I remember from frequent contact witih FL-9's while at MIT and at BBN, (1949-1954) 1956-1967, is that the orginal FL-9 shoe was double-spring, with the horizontal rest position both lower that the top of the overruning LIRR (and NYCTA and PATH) third rail and also higher than the bottom of the MN underruning third rail. This was different than certain NYNH&H electrics, including EP-5 GE`rectifier "Jets," which did have air-operated cam adjustable third rail shoes. Possibly someone has access to drawings?
Wizlish For once, and hopefully for all, THERE IS NO PROBLEM WITH A MODERN 60Hz TRANSFORMER RUNNING ON 25Hz POWER. The problem is the 'other way'; an older transformer built for 25Hz power (like the ones in GG1s) can't handle the higher frequency. But there aren't any M-8s with that problem, are there?
For once, and hopefully for all, THERE IS NO PROBLEM WITH A MODERN 60Hz TRANSFORMER RUNNING ON 25Hz POWER. The problem is the 'other way'; an older transformer built for 25Hz power (like the ones in GG1s) can't handle the higher frequency. But there aren't any M-8s with that problem, are there?
New Haven's FL9's were initially fitted with air-operated shoes that could be set either way or folded up. In later years they were set one way or another until replaced by fixed shoes in the 1970s. Amtrak operated Turbos, FL9s and P32s were supposed to work either way, but dual capability was dropped completely after 1991. Today's P32s are required to operate with shoes folded in MNRR territory.
DutchrailnutThe 25 Hz transformer has been discussed ad nauseam and is way too heavy ...
... plus if used to NYP they be powered by LIRR third rail, not catenary.
Why in God's name would an M-8 coming from 'east of Stamford' need to switch over to DC third rail at any point coming into NYP? SURELY it is simpler and easier to switch AC power keeping the pans up?
I have a technical question, which perhaps has been answered and I haven't seen it or read it properly. "Springs" won't be enough to assure correct switching from overrunning to underrunning contact on third-rail shoes. There would have to be something else, and I don't think that 'something' would be a mandated by-hand switchover of every shoe on every train before a train goes from one 'system' to another. So isn't there aome other system, perhaps air-operated, that switches the shoe on a car from overrunning position to underrunning, with the spring then providing the necessary action in the 'correct' direction thereafter?
Our community is FREE to join. To participate you must either login or register for an account.