Signal Mixer 2

[KhalDirth]

Hello All,

I am analyzing a signal mixing circuit that belongs inside of a generator voltage regulator. The mixing circuit is part of the feedback loop for controlling generator output voltage onto the system. A part of the circuit is a bit confusing to me. It involves diodes and transistors. While I am used to working with both seperately, together I am not sure as to how they work together.

In the schematic, the output voltage is going into an inverted op-amp (type 741) with a gain of -1. The schematic information book reads as follows:

"Inputs C and D are "auctioneering" inputs, in which the most positive of the two signals is given control.

A negative signal at Input A will override any signals present at the other inputs (note: Input A is considered a "limiting" circuit)."

To make things easier, Input B is usually riding around 14 Vdc.

I guess my questions are:

1) How does Input A override all other inputs?
2) I believe that if Input B were to change, it would override. Is this true?
3) Are the transistor amplifier circuits used to prevent overloading of the diodes? They seem to always be gated on, so I'm not sure of their purpose.

I am aware of how an auctioneering circuit works, but usually just in a diode-OR function, not in parallel with "limiting" circuitry and NPN transistors.

Any help is appreciated!!

Sincerely,
Mario Capuozzo
Associate Engineer

http://files.engineering.com/getfil...-483e-82da-492ed0b1942e&file=Signal_Mixer.vsd

 

[VEBill]

The only thing that affects the output is Input B which is wired directly through (lacking a junction dot). All the other inputs short each other out and emit smoke, but have no effect on the output.

 

[KhalDirth]

I will correct the print. There should be a junction dot there.

Any ideas now?

 

[IRstuff]

Is this for school?

TTFN

FAQ731-376

 

[VEBill]

On its face, the schematic still makes no sense. You wrote that it was part of a generator voltage regulator circuit. If so, then the various inputs (why so many?) might be connected to low impedance voltage sources. And if they vary so that a diode conducts, then the current will be unlimited and the diode will burn up.

Is there any chance that the schematic you've posted is wrong, or incomplete? Is it a 3-phase generator and you're missing the components to complete the diode array rectifier and filter needed to make DC from AC?

 

[ScottyUK]

VE1BLL,

This style of circuit is very common in the older Westinghouse AVR designs: I am half-wondering if this is part of the schematic from a WTA-300 series AVR. I've spent a few highly entertaining callouts trying to work out what the heck that design was doing.

The inputs are typically DC levels from various limiters and from the error amplifier output. Normally the error amp output drives the next stage, typically a gain stage and firing angle controller, but it can be overridden by supervisory inputs such as the excitation limiter and the Volt/Hz limiter for example.

I suspect the output should be taken from the collector of the 2N1711 and not from input B as shown.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[KhalDirth]

To IRstuff:
No, this is not for school. I am not sure why you ask that.

To ScottyUK:

You are correct, sir. This is a WTA-300B. You are right about the inputs. I tried not to be too specific about their functions: not everyone has seen the WTA-300B voltage regulator. I thought that I would boil it down into a basic circuits question. The output signal is actually coming from an inverting op-amp whose input is Vout. If you would like, I can make a scan of the original westinghouse diagram. I was initially hesitant in doing so, as the print from westinghouse is actually an "information only" style print, and not an "as build." Therefore, some of the inputs on the drawing are actually not being used inside the power plant.

As to my initial questions, I'm not sure what those transistor circuits are for. I've been on a trouble call recently where the AC error detector card was failing. While I was troubleshooting I came across this circuit, and so I'm just trying to make sure that I understand its operation completely.

Thanks!
~Mario

 

[ScottyUK]

The transistor circuits appear to be a pair of current sources, with different current outputs. Exactly what it does is difficult to say from a fragment of the circuit. By all means scan it and post it - it's nearly two years since I even saw one of these regulators.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[KhalDirth]

Here is the first page of the instruction leaflet, and also signal mixer schematic. Keep in mind, the summing inputs are not being used.

 

[KhalDirth]

I guess you can only upload one at a time. Here is the second document.

 

[IRstuff]

1) How does Input A override all other inputs?
All other outputs except for B are tied with a reverse-biased diode.

2) I believe that if Input B were to change, it would override. Is this true?
Only if the source impedance of B is lower than A's

3) Are the transistor amplifier circuits used to prevent overloading of the diodes? They seem to always be gated on, so I'm not sure of their purpose.

The diodes bias the transistors on. The emitter resistor of each transistor determines the current output of each transistor. This appears to be basic freshman EE course material fodder.

TTFN

FAQ731-376

 

[amptramp]

I agree with ScottyUK, transistors Q1 and Q2 are configured as classic current mirrors.

 

[VEBill]

Scotty gets a star. The only way to top Scotty's reply would be if the actual designer of this nightmare shows up to explain himself.

 

[KhalDirth]

IRstuff:

I don't believe that current mirrors in series with diodes is a normally covered topic in freshman EE. As I was unsure whether or not the transistors would have some sort of loading effect preventing the diode from become reverse-biased, I had to ask the question.

That being said, I don't believe it is the purpose of this forum to leave acidic comments in lieu of constructive statements (at least, the FAQ in your signature reads that way). I could point to several threads on eng-tips that are simpler in nature than my questions. I might also add that no one felt the need to jump on those threads and note how "basic" they were. As this is the first thread I've created here, I hope that I don't need to expect the same reaction to any future questions.

That's all I have to say about that.

amptramp and ScottyUK:

Thank you for that verification on the transistors behaving as current mirrors!

Sincerely,
Mario

 

[IRstuff]

Duly noted; nonetheless, freshman EE is where I first saw that type of circuit.

And, note that this site does not tolerate student posting, so postings that resemble homework are particularly scrutinized.

TTFN

FAQ731-376

 

[ScottyUK]

Hi Mario - I didn't think I would see one of these things again. Thanks for bringing back some memories of happier times in a place I used to love working.

Are you still looking for an explanation of the circuit operation? Here's my take on it - but I'm unwinding with a very large whisky after driving back from Sheffield in rush hour traffic so my thought process might go awry


Assume for now that the signals on 13 & 14 are high and thus D2 and D3 are reverse biased, and whatever is on 12 is high impedance.

Current from the current source Q1 and associated components flows via D4 into current sink Q2 and associated components. The voltage at the collector of Q2 is determined by the highest of inputs 9, 10, & 11, less the forward drop of the respective diode. The voltage at 12 is equal to the voltage at Q2 collector plus one diode drop, so it is effectively the highest of the voltages on 9, 10 & 11. The current sink Q2 draws roughly double the current delivered by the Q1 current source, and the balance of current is drawn from the highest of 9, 10, & 11.

If the limiter signal on either 13 or 14 goes lower than the voltage at 12 minus one diode drop, then it becomes the dominant signal and overrides the signals at 9, 10, & 11. The function of the signal at 12 is difficult to determine without knowledge of the upstream circuit driving it.



Westinghouse made a couple of test sets for this regulator which are worth looking for - Buffalo Power bought out this part of Westinghouse during the break-up and they might be worth speaking to about them. SWPC in Orlando will charge you a fortune if you have to buy them through their aftermarket support.



VE1BLL - it just shows how much time I have spent fighting with these damned controllers that I can recognise its circuit from a sketch a couple of years after leaving the plant. I'd like to say it is a masterpiece of analogue design, but in reality it is a horrible piece of equipment comprising of large plug-in modules containing next-to-nothing and which cost a bloody fortune from Westinghouse. Some of the circuit design is kinda cute in a very old-fashioned way, but the little discrete blocks make it a pig of a system to fault-find on. On the positive side they are remarkably easy to repair in the field.



----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Skogsgurra]

Slight (and completely unnecessary)correction: They are not current mirrors, but current sources and (perhaps) sinks.

The sink (lower 'source') sucks twice the current of the source. Not sure why.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[KhalDirth]

12 is actually an input from the Maximum excitation limiter (MXL). It will ride around 12 to 14 Vdc in normal operation (non-limiting). 12 is kind of what's confusing. If the signal is 14 volts, and the signal at 13 or 14 is 2 volts, then what is the true voltage level at the collector of D4? Is the input to the op-amp more or less being determined by the currents? Strange how I can't wrap my head around this one. Perhaps I should try throwing it into spice until I get it down. Haven't done anything with electronics in years - mostly been working on computer-engineering related projects. Your help is very much appreciated.

You are 100% right on the equipment. I was at the Basler factory in Illinois last week. They took over all of the ECS 2100 voltage regulator line from Cutler-Hammer. Now, they also make the DECS, along with WTA and MGR voltage regulators (pretty much cornering the market on reliable voltage regulation). I got to see their circuit fab operation, first person. It is interesting to me that they still charge $5k for a board with a few 741 op-amps and a few resistors!

Sincerely,
Mario

 

[ScottyUK]

Hi Gunnar,

Agree on terminology - am I guilty?

My explanation for the difference between source and sink currents is that diode D4 acts to restore the voltage at 12 to the value of the highest of the signals on 9, 10, & 11. The current in D4 is determined by the Q1 source, and the current in the other diodes is determined by the difference between the Q1 source and the Q2 sink. By making the currents roughly equal the diode drops should be quite accurately matched.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[KhalDirth]

Skogsgurra,

That makes sense. The current mirrors I've seen involve transistors whose gates are tied together in some way or another (like in the 741 opamp). Thanks!