SCR Heating Control

C

Thread Starter

Condern1

I have a circuit that I need help in troubleshooting. The circuit is a 3 phase 480 volt supply feeding an SCR assembly that is controlled by a 4-20ma signal from a temp controller. The output goes to a 110kva transformer where the primary is 480v and the secondary is 107v used to heat a heating configuration in delta where each heater is 36.7kw at 107volts. This is a phase fired SCR controller with RMS Current limit and an over current trip with a relay output. This includes a auto trip reset module when fault disappears. adjustment range is about 30 to 125% of units current rating limit and 30 to 200% for the over current trip. system is protected with MOV RC snubber and fuses.

The system is started and heats properly and then for an unknown reason the scrs stop firing and the unit trips on over current. I have monitored the line current and have seen nothing to cause this. The auto reset does not work and manual startup is required. I have 3 systems here so I took this unit and exchanged it with one that is working and the problem did not follow the hard ware. I am perplexed and wondering if someone has some ideas on what is going on.

Thanks
 
"the problem did not follow the hardware"

please explain what you mean, basically did the problem go away?

Over-current trips without load changes, has been associated with transients in the power feeder. It could be the result of SCR/Triac switching transients produced in other systems in the plant. Snubbers prevent SCR turn-on are the usual fix.
 
The problem did not follow the hardware means. When I put a unit from another system into the one I was having problems with, it did not work. And the unit I was having problems with working, I put in another system and it works properly. The problem did not follow the hardware or better said the SCR power controller. I have talked to the manufacturer, and they really don't know why this is happening.

 
J

John Correa T.

Could be a problem with the transformer that feeds the load. Some heather units have parameters to adapt to the transformer inductance. May be inrush currents or saturation.
 
Check the manufacturers voltage spec, then measure what is actually delivered to your unit.

For example, the EU voltage rating was deemed adequate the cover the nominal on-site rating, but the mains transformer taps were set at 575 vac. and considerably higher than the unit could take long term. The water cooled model could handle the voltage, but not the standard model. The corporate gurus had no clue, probably because they wanted no liability, but the shop tech nailed the problem and had the transformer taps adjusted.
 
C

Curt Wuollet

I think you should examine the heaters to see if a sag might be happening if the load doesn't move with the unit. And check the SCRs/heatsink temperature. But the fastest way would probably be a DSO set to stop recording on the fault so you could see what happened before it, as well as any transients that could cause false triggering. One failure to commutate or a false trigger is all it takes to trip. Some power quality monitors have this capability as well. But renting the instrumentation often costs far less than the downtime, etc. My DSO is paid for many, many, times over. But people continue to guess.

You can get a decent, long record, DSO for 3 or 4 hundred bucks brand new from OWON, Hantek, or Rigol. Add a couple 100:1 probes and you will find uses where you know in seconds what you could take days to guess.

Regards
cww
 
C

Curt Wuollet

No problem!
But, do watch the voltage rating on the probes and think carefully about your grounding. The probe grounds are tied together and to the 110 V ground. The FFT function is great for tracking down harmonic problems.

Regards
cww
 
I appreciate your input about the transformer as this is where I think the problem is also, but could you give me a little more details on your comments. How would you proceed to troubleshoot the possibility of the transformer being the problem?

Thank you
 
curt,
I am wondering if you could give me greater details on how you would troubleshoot this?

> But, do watch the voltage rating on the probes and think carefully about your grounding. The probe grounds are tied
> together and to the 110 V ground. The FFT function is great for tracking down harmonic problems.
 
Condern1... some Xfmr-related questions:

1) Is Primary-winding (480Vac, Controller-side) Delta-connected?

2) Is Secondary-winding (107Vac, Heater-side) Wye-Connected?

3) Is Heater 3-wire or 4-wire connected?

4) Is Xfmr showing any sign of distress, i.e, Temperature, or humming sound?

5) What are the three line-currents to the Heater?

6) If Heater is 4-wire, what is the neutral current?

Regards,
Phil Corso
 
C
No problem:

The general goal is to catch an intermittent problem in as few iterations as possible, and use the time machine to look at the cause. This is extremely handy when you get a fault that can have multiple causes. The DSO can sit for days recording data, and with a deep memory scope and the right choice of timebase and sample rates, can provide a detailed record of the events leading up to the trigger event. I've done this with both the Rigol and two Hanteks, one mine and one that belongs to the company.

I am most often dealing with 24V signals but it works with anything the scope can handle.

First you need to find a trigger. The fault indication is usually used, whether it be a light or a relay, or? Something that changes state when there is a fault. If you know the voltages, the DSO will let you set the trigger at midrange and the edge to whether it goes low or high or on or off. If possible you should test this so you know it will trigger. Then you set the two vertical channels to show what you want to compare, for example 2 phases of your incoming power. Set the time base, which also has implications for the sample rate, to show what you want, bearing in mind on a deep memory scope there is usually more record than you see on the screen at one time. A trip with overloads is usually a millisecond event so a second or two of event is usually plenty. Once you get everything set, Go to one shot mode and hit run. The scope screen will fill and keep recording. On a trigger event, it stops. The record now contains the events up to the trigger and the scope will let you scroll through the record.

It would be nice to have more channels, as in your case, you might have to run again if the event is on the phase you weren't looking at. But you have a 66% chance of catching what caused the fault if it was a line event. And 100% in two runs. If this shows nothing interesting, or maybe something really interesting, you can go to the load side with your probes and see what happens there.

Of course, you still have to know what to make of the data. And, yes getting familiar with the DSO will take time and effort, especially if you have never used an oscilloscope. Forty years playing with electronics does help. But the technique is so powerful for intermittent problems that I think it is well worthwhile to learn to use a time machine. And many very ordinary problems become trivial when you can see what is going on. If asked, and the moderator permits, I can provide real world examples.

Regards
cww
 
J

John Correa T.

The over current trip occurs always on starting or it is a random fault? Also can you give me the model of the SCR heating controller.
 
Top