Transformers

I was trained on a voltage regulator for a diesel generator that used a DC transformer (DC in the primary) for the derivative feedback loop. The manufacturer even called it a DC transformer. ARM

Reply to
Alan McClure
Loading thread data ...

in article snipped-for-privacy@corp.supernews.com, Alan McClure at snipped-for-privacy@gwis.com wrote on 9/18/04 12:38 PM:

I was trained on a voltage regulator for a diesel generator that used a DC transformer (DC in the primary) for the derivative feedback loop. The manufacturer even called it a DC transformer. ARM

Please explain more about the dc transformer.

Bill

Reply to
Repeating Rifle

IIRC - The primary of the transformer was in series with the field winding (DC) of the generator. (450v/60hz/3ph) If the regulator changed the field current the secondary would develop a voltage proportional to the rate of change of the current, thus providing a dI/dt signal by using only one component.

ARM

Reply to
Alan McClure

in article snipped-for-privacy@corp.supernews.com, Alan McClure at snipped-for-privacy@gwis.com wrote on 9/20/04 12:25 PM:

From your description, the device did just what an ac transformer should do. That is, transfer a potential difference to a secondary winding proportional to dI/dt. It is just that there is a primary bias current that may adjust the coupling coeffecient by the amount of current flowing in the primary.

That said, I have no idea what the purpose behind such an arrangement would be. In steady state, the output would be zero irrespective of the current in the field winding.

Bill

Reply to
Repeating Rifle

We also used this type of feedback. It is just a rate-sensitive feedback to the regulator. The output is zero in the steady-state, but when the field current is changing (as it would when a large load is applied to the generator), the signal from this transformer would be non-zero and proportional to the rate of field current change. Signal was used to dampen oscillations in the system caused by a somewhat high-gain regulator and a long time-constant in the field current (due to field winding's inductance).

We also used another form of 'DC' transformer that was really a form of mag-amp. An SSR (self-saturating reactor) would be placed in the half-wave rectified supply to the generator field. The control winding of the SSR would carry a DC signal from the regulator. After just a couple of cycles with no control-winding current, the rectified DC of the power circuit would saturate the reactor (i.e. self-saturating) and the reactor would present low impedance to field current. As generator voltage rose, the DC current in the control-winding would increase (fed from regulator). DC control-winding current would 'de-saturate' the reactor so that the reactor was a high-impedance on the subsequent rising half-cycle. At some point during cycle, the half-wave current would again saturate the reactor and the impedance would suddenly drop again. End result is a 'chopped' half-sine wave current to the field, much like you see today with phase-controlled SCR bridges. The exact point of 'turn-on' was a function of the DC control current.

'Bushbadee' could probably give a much better explanation of mag-amps, but they were small, controlled a lot of power, handled extremes in temperature with little trouble and pre-dated SCR's and power-electronic devices.

daestrom

Reply to
daestrom

in article CiJ3d.6007$ snipped-for-privacy@twister.nyroc.rr.com, daestrom at daestrom@NO_SPAM_HEREtwcny.rr.com wrote on 9/20/04 4:00 PM:

Why will a large change of load on a (dc?) have a big effect on field current. You did say that the time constant of the field was large.

I remain confused, but that is not unusual. Why was this kind of control necessary?

Reply to
Repeating Rifle

No your description is pretty good. I put two saturable reactors in the control circuit of the Vela satelite. They worked at about 7Kz. They had the input voltage on them so that if the input voltage rose, they saturated sooner and turned off the power transistors sooner. Fantastic regulation with just 6 db gain in the loop. That allowed a 12 db per octave roll off and the circuit could regulate way out in frequency. Even with 12 db per octave roll off we had a fixed 6 degree margin which was set by the turns ratio so the cirucit was unconditionally stable.

inductance).

temperature

Reply to
bushbadee

The particular diesel generator I was in reference to was the emergency diesel on a nuclear submarine. (since decommissioned, the whole sub. class for that matter) The speed and voltage droop characteristics of this generator were designed to be nearly the same as the characteristics of the sub's (steam) turbine generators so that the generators could be paralleled easily. The voltage regulators on all the above mentioned generators were built using magnetic amplifiers, since at that time in history solid state devices were not considered reliable/tough enough and there were ground isolation concerns. To answer your question, as load increased, voltage output would decrease, the voltage regulator would sense the error and change the field current to compensate. The "dc transformer" would send a negative feedback signal to the regulator to prevent "hunting" (oscillation) of the voltage. This was the "D" in the PID control loops of the voltage regulator.

As I said above voltage and speed droop is needed to get generators to "play well with others".

ARM

Reply to
Alan McClure

Repeating Rifle wrote:

in article CiJ3d.6007$ snipped-for-privacy@twister.nyroc.rr.com, daestrom at daestrom@NO_SPAM_HEREtwcny.rr.com wrote on 9/20/04 4:00 PM:

"Repeating Rifle" wrote in message news:BD749CFD.23D21% snipped-for-privacy@sbcglobal.net... in article snipped-for-privacy@corp.supernews.com, Alan McClure at snipped-for-privacy@gwis.com wrote on 9/20/04 12:25 PM:

IIRC - The primary of the transformer was in series with the field winding (DC) of the generator. (450v/60hz/3ph) If the regulator changed the field current the secondary would develop a voltage proportional to the rate of change of the current, thus providing a dI/dt signal by using only one component. From your description, the device did just what an ac transformer should do. That is, transfer a potential difference to a secondary winding proportional to dI/dt. It is just that there is a primary bias current that may adjust the coupling coeffecient by the amount of current flowing in the primary.

That said, I have no idea what the purpose behind such an arrangement would be. In steady state, the output would be zero irrespective of the current in the field winding.

We also used this type of feedback. It is just a rate-sensitive feedback to the regulator. The output is zero in the steady-state, but when the field current is changing (as it would when a large load is applied to the generator), the signal from this transformer would be non-zero and proportional to the rate of field current change. Signal was used to dampen oscillations in the system caused by a somewhat high-gain regulator and a long time-constant in the field current (due to field winding's inductance). Why will a large change of load on a (dc?) have a big effect on field current. You did say that the time constant of the field was large. The particular diesel generator I was in reference to was the emergency diesel on a nuclear submarine. (since decommissioned, the whole sub. class for that matter) The speed and voltage droop characteristics of this generator were designed to be nearly the same as the characteristics of the sub's (steam) turbine generators so that the generators could be paralleled easily. The voltage regulators on all the above mentioned generators were built using magnetic amplifiers, since at that time in history solid state devices were not considered reliable/tough enough and there were ground isolation concerns. To answer your question, as load increased, voltage output would decrease, the voltage regulator would sense the error and change the field current to compensate. The "dc transformer" would send a negative feedback signal to the regulator to prevent "hunting" (oscillation) of the voltage. This was the "D" in the PID control loops of the voltage regulator.

We also used another form of 'DC' transformer that was really a form of mag-amp. An SSR (self-saturating reactor) would be placed in the half-wave rectified supply to the generator field. The control winding of the SSR would carry a DC signal from the regulator. After just a couple of cycles with no control-winding current, the rectified DC of the power circuit would saturate the reactor (i.e. self-saturating) and the reactor would present low impedance to field current. As generator voltage rose, the DC current in the control-winding would increase (fed from regulator). DC control-winding current would 'de-saturate' the reactor so that the reactor was a high-impedance on the subsequent rising half-cycle. At some point during cycle, the half-wave current would again saturate the reactor and the impedance would suddenly drop again. End result is a 'chopped' half-sine wave current to the field, much like you see today with phase-controlled SCR bridges. The exact point of 'turn-on' was a function of the DC control current. I remain confused, but that is not unusual. Why was this kind of control necessary? As I said above voltage and speed droop is needed to get generators to "play well with others".

ARM

Reply to
bushbadee

A large change in the load on the generator output (DC or AC) will change the voltage output of the generator. Hence the need for a voltage regulator in the first place. But to get the desireable regulation, the regulator unit must have a pretty high gain (field current must more than double when the generator goes from no-load to full-load).

Such high gain comes at a price. A high gain coupled with a long time-constant on the field winding of the generator (DC current through a massively inductive load) can result in sustained oscillations. Of course there are several ways to counteract such oscillations. One might be to sense the rate of change of the generator voltage and use that as a second input to the regulator, giving you PD (proportional and derivative) control. Another method (the one chosen in this case) is to sense the regulator output and its rate of change. Use the regulator's output rate of change as a form of feedback.

So, one could sense the regulator's output any number of ways. A simple, rugged method chosen was to use the dI/dt of the field current. But the current in the power stage of the output needs to be electrically isolated from the regulator. A 'DC transformer' is used since it's secondary voltage is proportional to the dI/dt of the 'primary' (the generator field current). It also provides the electrical isolation all in one package.

Any number of control mechanisms *could* have been used, just have to meet the requirements. The power needed to supply the field ranges from 30-40 amps at no-load to ~100 amps at full load. The entire unit must be self-contained and capable of starting up with no external electric source. It must be operable in some extreme temperature ranges (~20F to ~150F) and must withstand the severe vibration of the diesel engine and, believe it or not, being depth-charged (severe shocks of several g's).

Nowadays, it could be done with solid state devices, but in the '60's and '70's, mag-amps and SSR's were the way to go. Also used in motor-generator sets used for 'vital busses'.

daestrom

Reply to
daestrom

Hi Bushbadee,

Yes, that was one of the requirements of many pieces of submarine machinery!!! (even the main feed pumps :-) I worked on several different

400 hz sets, perhaps the oldest was 400HZ, 120V 3phase 42kW motor-generator sets used on the old Permit class (renamed after the Thresher accident in '63). IIRC, these were all mag-amp units. The only 'electronic' thing in them was the silicon rectifiers.

daestrom

Reply to
daestrom

OUr converters were to replace motor generators. Lots of 2N174's the output was 120 volts, 3 phase. I wonder if there was any relationship.

motor-generator

Reply to
bushbadee

I did also work on some smaller power 60hz to 400hz converters. IIRC, they had two 400hz inverters feeding a Scott-T for the 3 phase output. One 'master' that fed main leg, and the 'slave' feeding the tee-leg. Phase A-B voltage controlled by master, phase B-C and C-A controlled by not only magnitude of 'slave' inverter, but phase shift delay from 'master'. But these were smaller units, only a couple of kW.

daestrom

Reply to
daestrom

(shudder!!!) I just had a flashback of re-brushing Amplidynes used for some old excitation systems ;-) Twice as many brushes as similar sized DC machinery. Considering the occasional failures I say of SCR systems, always thought mag-amps were the best.

daestrom

Reply to
daestrom

At least the amplidyne had a reasonable response in comparison to the pilot exciter-exciter-field set up and used less space -with only one commutator.

Reply to
Don Kelly

We built them also. Did you by chance work for ElectroSolids or a competitor.?

I worked on an interesting one. It was a pulse width modulated inverter. ac amplifier. An Input ac signal was amplified and rectified. That signal then acted as the reference for a pulse width modulated converter which chopped at double 40 Kc. The amplified half wave rectified signal was then fed to the center tap of a transformer which the then derectified the imput wave by haveing a pair of transistors which were tied to the input and detected 0 crossing of the input wave and kept the out put transistors in sync with the input.

The thing was pretty flat to 4KC and down to about 20 cycles.

The output had about 5% distortion. One of my ideas was to feedd the amplified signal to a class B or even a Class A amplifier as the DC input so the dc on the amplifer would track the output and achieve higher efficiencies than the theoretical efficiencies of class A and B amplifiers.

I think Sony took out patents on the idea, but their patents were useless after I had presented papers on the theory and device we had built to the IEEE about 20 years before their patent.

DR Middle brook also used the same idea to come up with a similer amplifer, but he used a bridge output and not a transformer which has quite a few advantages. I do not know if David ever built t he device, but we built, demonstrated and packaged our device and tried to market it as a high efficiency shake table driver.

Reply to
bushbadee

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.