# Compound-source static exciter analysis

Hi, i am currently analysing a static exciter setup of which i included the following link to the schematic:
http://members.home.nl/damellema/CS4.PNG
It is a compound-source exciter, e.g. power to the generator field is delivered by a voltage source (Power Potential transformer) and a current source (Boost Current Transformer). This type of exciter is very similar of the General Electric SCT/PPT exciter, with the exception that the current transformer is not saturable.
A Basler whitepaper says the following: "The SCT-PPT excitation system uses both a generator stator voltage source (PPT) and a generator stator current source (SCT) as the power source for the main field. When the generator is offline and not supplying current to a load, the excitation field current is supplied by the 3-phase, power potential transformer. When the generator is on-line and supplying current to the system load, a portion of the excitation field current is supplied by the three saturable current transformers. Since there is both a current source and a voltage source used in this compound-type excitation system, linear reactors are utilized to prevent the PPT from being shorted when the SCTs are saturated. These three magnetic components (PPT, SCTs and Linear Reactors) constitute the power magnetics for this compound-type excitation system, all of which have been designed specifically for each generator to ensure the correct amount of excitation field current is available for all load conditions. This makes the SCT-PPT compound-type excitation system very desirable." Why does the SCT short the PPT?
Basler also claims the following: " Under normal steady-state conditions, the phasor summation between the PPT/linear reactors and the SCTs provides the correct compensated voltage and current to the generator field for all loads at any power factor." Can someone explain to me how this works?
The Boost Current Transformer is a pretty weird device. It essentially is a wound CT with 1 primary coil (3 windings) and 2 secondary coils (19 & 3 windings). They secondaries are series connected to eachother in a sort of 'delta way'. Isn't it illegal to connect two current sources in series. Does someone have a explanation of this arrangement?
There is a second 'current transformer' in the schematics, T1, which i do not know the function of. The output of this transformer is also essentially a current source. So you have a rectified current source and a rectified voltage source, wich are added and put in the field. Can someone explain how this thing works?
The field current is controlled by a thyristor setup which shunts current from the middle phase of U2, essentially shorting the Current Transformer for half a cycle. Is there some relation to coil L2?
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

Without having the SCT-PPT schematic, it's hard to say for sure. But if the CT's are wired so their output is parallel to the PT's, then when they saturate they would short the PT's.

As the load is increased on a generator, the required field current to maintain terminal voltage also rises. With a simple PPT regulator, the regulator must have a very wide range of control in order to maintain terminal voltage at all ranges of load. By adding CT boost, the current boost automatically increases the field current.
In most large units, the match is *not* a perfect one and the PPT portion is still regulated. But the overall range of control does not need to be as wide since the CT boost automatically adds to field current with increasing load. So the static regulator and PPT only need to compensate for the imperfect match between CT boost and the exact field current needed for constant terminal voltage.
One simple CT boost method is to use a 3-phase full-wave bridge for the CT's connected directly to the field, and a second bridge for the PPT (or thyrister bridge for controllability).
But the amount of field current needed for a particular current level on the generator also varies with the power factor. Lagging power factor loads cause a sharper drop in terminal voltage than resistive loads. (due to the generator's internal impedance). So ideally, the CT boost is also sensitive to the power factor of the load.
By connecting the CT boost from particular phases of the generator to particular phases of the PPT, this can be achieved pretty well. A unity power factor current in the CT boost would be 90 degrees out of phase with the PPT output and the vector sum of the two increases field current only slightly (hypotenuse of a right triangle). But a lagging current in the CT boost would be 0 degrees out of phase with the PPT output and the vector sum of the two increases field current much more. Exactly what is needed by the generator to maintain terminal voltage.

They are from different phases. The result is a current that is phase-shifted from both CT's. The ratio of the two CT's will determine the magnitude and phase shift of the resultant. As long as the output isn't open-circuited, should be okay.

The connections of T1 suggest that the two outputs are phase-shifted from each other. The comment down by the rectifiers about 'U1' being +30 degrees seem to confirm this.
I *should* know this, but the 'Dzn0' by the potential transformers is one method of describing the primary/secondary connections. But I don't recall which. I think that is delta-wye (judging from a 'n' for secondary neutral and no 'N' for primary). This introduces yet another phase-shift. I think that puts the PPT secondary voltages in-phase with the generator phase currents (before the CT).

Hmmm... so SCR2 is controlled by the regulator (AVR). SCR1 looks like it's triggered from the diode alone (zener??). My guess without studying this all night is that SCR1 is used to commutate SCR2 periodically (i.e. stop the current flow through SCR2 so it goes to a non-conductive state).
I can't be sure, but I notice that L2 is in the same leg as the resistor R3 to SCR2 anode. Perhaps that is important.
Overall, I'd say the big 'mystery' in this regulator would be in understanding the phase-angles of the CT secondaries and the PPT/T1 output. Looking at the voltage applied to SCR2 and SCR1 should reveal how their conduction is controlled.
There have been several variations on CT-boosted regulators, but I hadn't seen this particular variation before.
Don Kelly or Charles Perry might be able to give us more....
daestrom
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

Why are the currents out of phase with each other. Doesn't unitiy power factor mean that the voltages or currents in phase. Or is the CT arrangement causing the currents to be out of phase with each other? Can you supply a phasor diagram?

The PPT in the schematic with vector group Dzn0, is actually a Delta Zigzag (interconnected wye) transformer with a neutral and no phase shift. After looking at a 'phasor' diagram of a zigzag winding, i realized that by arranging the ct's a little differently the ct arrangement could be drawn the same way. --> the compound-boost CT is actually a wye zigzag CT! The problem with this 'interconnection' is that essentially current sources are placed in series. Isn't that illegal according to Kirchhoff? Also is it possible to derive the phase and magnitude characteristics of the boos CT?
http://members.home.nl/damellema/zigzagct.PNG

When simulating this circuit in Simulink, it seems the currents in U1 & U2 are in antiphase. Do you have an explanation for this?

There are 2 zigzag transformers in the circuit. I've read somewhere that these transformers attenuate triplen harmonics. I presume these transformers are used to attenuate the harmonics caused by the diode bridges.

SCR1 is some kind of transient protection for the diode bridges i think. When a big back-emf occurs by the rotor windings, then the breakover diode is activated, and the SCR's start to conduct.
The currents in U2 are sines; the SCR2 commutates every half cycle. This also means the field current can be regulated only for half a cycle (red path). Can L2 help in no-load situations?

<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Got some help from a PDF:
http://members.home.nl/damellema/zigzagct2.PNG
Does this also mean that the current in the secondary is 150 degrees leading in comparison with the primary?
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
I immediately dit a patent search on 'Yz5' and found this patent: http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&p=1&u=%2Fnetahtml%2FPTO%2Fsearch-adv.htm&r=2&f=G&lP&d=PTXT&S1=yz5&OS=yz5&RS=yz5
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
I immediately dd a patent search on 'Yz5' and found this patent: http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&p=1&u=%2Fnetahtml%2FPTO%2Fsearch-adv.htm&r=2&f=G&lP&d=PTXT&S1=yz5&OS=yz5&RS=yz5
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

The currents in the individual CT secondaries will be in phase with the *generator* load current. So when the generator is supplying a unity pf load, the individual CT secondary current is in phase with the phase-neutral voltage for the same leg. When the generator is supplying a lagging pf load, the individual CT secondary current will also lag behind the phase-neutral voltage for the same leg.
So by using the CT's in this manner for a boost, and adding the CT currents to the PPT currents, a given number of ampere generator load will have different phase angle between the CT boost and PPT whenever the generator load pf changes.

Sorry, no. Do you have the correct ratios? I thought there was some note about +30 degrees on one of them?
But I don't see how the difference in angle between U1 and U2 can make any difference. They are only connected through the two bridges. But the current through the U2 bridge includes the CT boost, so the currents flowing in each leg of the U2 bridge are probably not matched/synched with the currents in the U1 bridge.
And I still don't see a purpose for L2.

Nah, now that I think about it, I think I was totally wrong here.

The rotor winding is a very large inductive load. The L/R time constant of rotor windings on large machines can be several seconds. Without SCR2, the two 6-pulse rectifiers would apply full power to the rotor winding. This and the rectifier means the currents in each leg of U2 are very *non* sine. But yes, I can see where the current in SCR2 drops to zero and commutates each cycle when the voltage of the middle leg goes negative compared to the other legs of U2.
I don't know if its a mistake on the drawing, but if the two 'break-over diodes' are supposed to be zenor diodes, they look like they're connected wrong. Their anodes are connected to the positive supply/output of the 6-pulse rectifiers. In this connection, they would be forward biased and conduct almost all the time. They would only be in 'zenor' mode (reverse biased) if the polarity of the rotor winding reversed. This can't be the case because then SCR1 would trigger shortly after SCR2 triggers each cycle. That would effectively short the two 6-pulse rectifiers and never shut off.
If those diodes are zenor diodes but drawn backwards, then the two thyristors SCR1 and SCR2 would fire only when very high CT boost occurs. This might be necessary if the output of the generator is faulted (say a three-phase 'bolted' short). The extremely high generator load current would provide a very high field current (via the CT boost transformers). The extreme field current would generate higher fault current. So under generator fault conditions, these zenor diodes would conduct, short the CT boost current away from the rotor winding and effectively 'shutdown' the generator. (PPT output is often ignored during short-circuit faults since the terminal voltage is shorted).
Many regulator designs I've worked with got around this issue by ensuring CT boost supplied only a fraction of the field current needed. So a fault would remove the PPT portion of the supply, and the rotor current would drop. The lower rotor current is not enough to maintain rated current into a bolted fault so the power supplied from the CT boost would also decay over time and the generator would 'shutdown'.
(of course, all this happens only until a protective device operates, but most generators are still designed for 'bolted fault' conditions at least for a short time, and the particular response of a given regulator is often considered in system fault analysis).
daestrom
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

## Site Timeline

• - the site's newest thread. Posted in
• Share To

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.