Paralleling of Bridge Rectifiers

[Ingenieur]

For large scale FACTS equipment
how do they rectify xxx MVA 25kv power to DC?
then step it up for hvdc transmission bipolar 500kv

 

[Phil Corso]

For large scale FACTS equipment
how do they rectify xxx MVA 25kv power to DC?
then step it up for hvdc transmission bipolar 500kv

Ask the Chinese how! No other country in the world has more HVDC transmission systems in operation than China! In fact, in South China, 5 HVDC lines carry some 26GW! (Ref: IEEE Spectrum, Dec '16)

Phil Corso

 

[Ingenieur]

Ask the Chinese how! No other country in the world has more HVDC transmission systems in operation than China! In fact, in South China, 5 HVDC lines carry some 26GW! (Ref: IEEE Spectrum, Dec '16)

Phil Corso

that's the advantage of building infrastructure now, using current tech rather than 50-100 years ago

the US has alot more than I imagined
some not for transmission but back to back for grid interconnection

 

[Phil Corso]

that's the advantage of building infrastructure now, using current tech rather than 50-100 years ago. the US has a lot more than I imagined some not for transmission but back to back for grid interconnection

"Current Tech"... Enjoyed the pun ! ! !


Phil

 

[Besoeker]

It sure looks that way.
This 18 pulse bridge depends on adding the absolute values of all three winding sets at once to get the waveform that is to be filtered at the capacitor.
So the cap voltage is higher than the peak value of any one winding.
I had not given any thought on how to assemble an 18 pulse supply. I imagine that the parallel configuration would work too but has disadvantages compared to the serial form.
Each bridge contains diodes that are exposed to the peak value of its own windings' AC, but never the full bus DC.

Most of what we have done is an arrangement for up to 24-pulse. There are normally reactors between the bridges otherwise each element would carry the full DC current but for a shorter time depending on the number of pulses.

For the solenoid application here, nothing as complex is required. I'd use standard bridge modules. Something like a pair of DD89N18K modules would be more than adequate.

 

[Besoeker]

For large scale FACTS equipment
how do they rectify xxx MVA 25kv power to DC?
then step it up for hvdc transmission bipolar 500kv

I think it uses transformers to get to the right voltage before the HVDC converter bridge.
An inverter, sometimes mains commutated, sometimes force commutated, inverts it back to AC and then it is transformed down to whatever voltage is being used for the AC distribution network it is feeding into.

 

[FionaZuppa]

And debatable what it does for increasing the total current capacity.

no concern for flyback control? i can only imagine, a hefty ATS that has a "solenoid". hopefully the src to bridge has a fuse :thumbsup:

 

[ATSman]

no concern for flyback control? i can only imagine, a hefty ATS that has a "solenoid". hopefully the src to bridge has a fuse :thumbsup:

\
The NEC does not require fuses in the operating circuit of an ATS, only on the indicating lights.

 

[Besoeker]

no concern for flyback control? i can only imagine, a hefty ATS that has a "solenoid". hopefully the src to bridge has a fuse :thumbsup:

The OP said he measured the solenoid at about 22 ohms. That determines the current from a 480V supply.
Flyback? I assume you mean the current that flows in the inductance after the power is removed. Sometimes called flywheel current.
A full-wave bridge provides a path for that without any additional components.

 

[FionaZuppa]

The OP said he measured the solenoid at about 22 ohms. That determines the current from a 480V supply.
Flyback? I assume you mean the current that flows in the inductance after the power is removed. Sometimes called flywheel current.
A full-wave bridge provides a path for that without any additional components.

22ohms of wire ohms, the thing is likely a big inductor, with a 80ms pulse on that coil when the power is cut the coil will create a huge voltage that is seen as a "PIV" on the diodes. as-is there is no path for flywheel current, DC+ bus is on the cathode side of all diodes on both bridges. if you dont squash the flyback voltage (which will peak way above ~676v with fast hard-off pulses) you can bust a hole in diode (possibly short it). and depending on the inductance you may not see 480v/22ohm, depends on henries. i recommend getting a L reading of the coil using LCR meter, from here we can see what the flyback voltage may be. i already mentioned how one can handle the flyback usin zener, or diode(s) across the coil. and if the ATS switch is sensitive to bouncing (bouncing is typically not wanted with solenoid controlled switches) then you might also need to squash the ringing that may be there as mag field collapses, a small tuned RC to gnd will handle ringing. w/o RC i can easily see what freq i need to handle, i then squash it with a proper RC.

for some these pics are familiar, others will have no idea. the below is engineered by me (fuel control system), and then validated on the scope.
this is 12vdc @ about 1A on a fuel injector coil of decent henries, after the pulse stops the coil creates nasty flyback and will damage the fet switching. the upper flat as shown is clamped at 33v via a zener (w/o zener the voltage is very high), the width of the flat is ms it takes to dump off most of the mag field energy, the ringing is obvious w/o RC to handle that, RC is used in 2nd pic. the ringing can cause coil to bounce, which is bad for a switch and will shorten life of the switch, also bad for precision controlled fuel injection !

 

[Besoeker]

22ohms of wire ohms, the thing is likely a big inductor, with a 80ms pulse on that coil when the power is cut the coil will create a huge voltage that is seen as a "PIV" on the diodes. as-is there is no path for flywheel current, DC+ bus is on the cathode side of all diodes on both bridges. if you dont squash the flyback voltage (which will peak way above ~676v with fast hard-off pulses) you can bust a hole in diode (possibly short it). and depending on the inductance you may not see 480v/22ohm, depends on henries. i recommend getting a L reading of the coil using LCR meter, from here we can see what the flyback voltage may be.[/IMG]

There are no pulses. The current commutates from one diode to the next when it gets forward biased. At switch off, the current flywheels through the diodes in the bridge.
Both bridges - what both bridges?

I'll try to find an image of the circuit........

 

[Besoeker]

Duplicate.

 

[gar]

161222-0847 EST

Besoeker:

Your drawing did not have a switch in the line from the AC source to the bridge. Put the switch in the circuit in the input side position and draw the load as an inductor and you do have a very large PIV applied to the diodes in the bridge upon opening the switch.

Put the switch on the DC side and there will still be a PIV on the diodes the magnitude of which depends on a number of circuit factors.

.

 

[GoldDigger]

If you open the AC side current will continue to flow through the forward biased diodes of the bridge. There will, as Besoeker said, be no reverse voltage anywhere in the circuit above the 1.4V of the double diode drop.

With the switch on the DC side it is, as you stated, more complicated.

Sent from my XT1585 using Tapatalk

 

[FionaZuppa]

If you open the AC side current will continue to flow through the forward biased diodes of the bridge. There will, as Besoeker said, be no reverse voltage anywhere in the circuit above the 1.4V of the double diode drop.

With the switch on the DC side it is, as you stated, more complicated.

Sent from my XT1585 using Tapatalk

true.... my stuff is switched on the DC side of things by fet.
so, what i said earlier --> not true for the setup as shown if the switch is on AC side. my bad.
as-is, when the current stops because of diode dropout you'll see a jump in voltage, but not enough to be PIV worthy.
as-is you could still see ringing due to the LRC ckt that exists.

there has to be a switch somewhere, OP said it would only be on for ~80ms

switch on DC side, then you will have a PIV to worry about on the switch (whatever that is). remember that DC arc'ing across switch contacts can conduct for a long time vs AC that has a zero point.

so, how is this thing switched on and off ???

 

[ATSman]

true.... my stuff is switched on the DC side of things by fet.
so, what i said earlier --> not true for the setup as shown if the switch is on AC side. my bad.
as-is, when the current stops because of diode dropout you'll see a jump in voltage, but not enough to be PIV worthy.
as-is you could still see ringing due to the LRC ckt that exists.

there has to be a switch somewhere, OP said it would only be on for ~80ms

switch on DC side, then you will have a PIV to worry about on the switch (whatever that is). remember that DC arc'ing across switch contacts can conduct for a long time vs AC that has a zero point.

so, how is this thing switched on and off ???

FZ
go back to post# 23 in the drawing on page 2, plug J5 shows the microprocessor controller
wiring to the coil control relays (CN, CNO, CE, CEO) that send the pulses to the coils.

 

[gar]

161222-1007

It is correct that the diodes are a clamp if there is no substantial capacitance associated with the inductor.

Since switching is on the DC side and an FET is used for switching there is something else on the inductive side to clamp the inductive kick.

.

 

[FionaZuppa]

FZ
go back to post# 23 in the drawing on page 2, plug J5 shows the microprocessor controller
wiring to the coil control relays (CN, CNO, CE, CEO) that send the pulses to the coils.

ok, i read the schematic in detail.
the MCU board has isolation to line, see http://apps.geindustrial.com/publibrary/checkout/Installation and Instruction|50R-1000|generic
MCU controls a line pulse via relays (RT box page 2 of the of the PDF from post #23)
MCU has four control signals, CN/CEO and CE/CNO

what i dont see is, what those four controls actually are on page #1 of PDF from post #23. is it a quad coil 3PDT latching relay ??

and from that PDF on post #23, the bridges are single bridges.

 

[Besoeker]

161222-0847 EST

Besoeker:

Your drawing did not have a switch in the line from the AC source to the bridge. Put the switch in the circuit in the input side position and draw the load as an inductor and you do have a very large PIV applied to the diodes in the bridge upon opening the switch.

Put the switch on the DC side and there will still be a PIV on the diodes the magnitude of which depends on a number of circuit factors.

.

You put the switch or contactor in the AC supply because it avoids transient overvoltage transients for the rectifier elements. Done it dozens if not hundreds of times on the shunt field of a DC motor.

 

[FionaZuppa]

what i dont see is, what those four controls actually are on page #1 of PDF from post #23. is it a quad coil 3PDT latching relay ??

what transfer switch model is it exactly?

ZTS/ZTSD Series Transfer Switches 40-4000 Amps
the 480v solenoid for standard transition is zenith(ge) ZC K-2209, retail $230.00 !!
the delayed transition model (shown in PDF on post#23) is ZC K-2248B, retail $951.00 !!