awright

One resource for information on systems installed in isolated areas with basic system data, but not detailed schematics, is Home Power Magazine. Lots of discussion and examples of system design and capacity calculation. They offer CDs of multiple years of back issues on each. I certainly do not speak for the publisher, but perhaps you can appeal for a free set of CDs. awright

Have you looked at thedata sheet for the tlp741g? http://www.toshiba.com/taec/components2/Datasheet_Sync//209/4322.pdf It is a low-current (150ma), moderately high voltage (400v) SCR with a photo sensitive gate that can be triggered by light from the embedded LED with LED forward current up to 10 ma. A resistor of 27K to 33K should be installed from gate to cathode of the SCR to suppress self-triggering due to leakage from anode to gate. It can be applied like any SCR except that you trigger it with forward current though the LED rather than by a voltage applied to the gate. The LED terminals are insulated from the SCR terminals for up to 4000 volts. If you don't require isolation from LED to SCR then there's not much point in using an opto-isolated SCR. The Photo SCR will block current in both directions until triggered by light from the embedded LED. Once triggered it will conduct (conventional) current from anode to cathode until current drops below the holding current of 0.2 ma. If forward current drops below 0.2 ma (and there is no current through the LED) the SCR will "commutate," that is, turn off. Reverse voltage (i.e., SCR cathode more positive than SCR anode) will be blocked with or without LED current. The photo triac operates like a conventional triac but can be triggered by current through the embedded LED. It differs from the photo SCR in that it can conduct in the "reverse" direction if triggered by LED current. Applications for either device are similar as those for conventional thyristors except that the photo-thyristors can be triggered by current through the LED. I'm sure youcan find tons of applications for conventional SCRs and triacs on the 'net. awright

PWM is not the only way to soft start a DC motor but it is certainly the most efficient way in most cases. If you just want to soft start the motor and have no servo control requirements, I would think any basic PWM generating IC driven by a ramp and driving a single IGBT or MOSFET of sufficient current/voltage ratings would do the job. I haven't actually done this myself, so do not have design details to offer but I don't think a great deal of sophistication is required. If you want something that's been designed for you, look at this CANA-KIT <http://www.sparkfun.com/products/9668>, a 24 volt 50 amp basic PWM controller with built-in soft start function. A brief instruction manual (not including dircuit schematic) is offered at that site. $59. awright

Sorry to make such an obvious suggestion, but have you asked the manufacturer of the power supply how his unit will handle your load and what precautions to take? Looks OK to me but then, it's not my $12K. awright

While it is true that a neon sign transformer is current limited by design, do NOT assume that you can be casual about dealing with the high voltage. It only takes 10-20 ma through the heart to throw it into fibrulation with possibly lethal results and I believe that neon signs operate on higher currents than this. I presume you do not have a spec sheet on your transformer telling you what the current limit is. In any case, the nominal current for fibrulation is an average. You or your guests may be more sensitive. Be sure there is no possibility of human contact with the hot secondary lead and that the other end of the secondary is solidly grounded. I urge you to read this Wikipedia entry on NSTs before proceeding. http://en.wikipedia.org/wiki/Neon_sign_transformer awright

Why is this question coming up again? Did you post the same question on a different forum? The answer is the same as before. There is nothing at all critical about the SCRs used in this circuit. Any voltage rating will do because you can't find an SCR with too low a voltage rating for a full wave rectified 17 volt supply. Look up the current ratings of the two specified SCRs and find replacements with at least those current ratings. Specifications for both SCRs are available on line. You can find satisfactory replacements for less than a dollar at digiKey, Mouser, and many surplus outlets like All Electronics and Alltronics. awright

Using a commercial HV probe like the Fluke suggested by Ken is definitely the best way to go. However, such an accessory must be used with a DVM having the proper input resistance. This is automatically taken care of if you use the probe with most Fluke DVMs. It will definitely NOT work with any arbitrary 30 volt analog panel meter. The reason is that the probe contains a high impedance voltage divider consisting of an approximately 1000 Megohm series HIGH VOLTAGE resistor (one that resists flashover across the resistor) and a nominally 1 Megohm shunt resistor to ground. The actual values of the resistors form a 1000:1 voltage divider when the shunt resistor is loaded with the 10 Megohm input resistance of the DVM. Loading the probe with an analog d'Arsonval (moving coil) panel meter will seriously degrade the voltage division ratio to the point that you will have no idea of what you are measuring (in the absence of careful experimentation or calculations to determine the actual voltage division ratio). Look at the spec sheet for the probe here: <http://assets.fluke.com/manuals/80k40___iseng0900.pdf> Note that the probe meets specs only if used with a DVM having an input resistance of 10 Megohms +/- 1%. You can fabricate you own HV probe using an analog meter but you will be messing with potentially lethal voltages. In fact, you can buy relatively low cost HV probes with built-in analog meters. To make your own voltage divider for use with a meter you have on hand you must determine the exact resistance of the meter so that the required resistor values can be calculated. Do not try this with a VOM Ohmeter as you will probably burn out the meter under test. Rather than me trying to explain all the details and the cautions of working with high voltages, check out the very useful and detailed discussion of high voltage probe construction here: <http://www.repairfaq.org/sam/hvprobe.htm> Incidentally, Sam Goldwasser has collected and created an enormous amount of useful information on the web site linked to above. Note especially the safety considerations. awright

I wouldn't try to switch the secondary current. Switch the primary on and off, as is done in a soldering gun. You will pay dearly for a relay that will work in your secondary circuit. Also, you will be adding contact resistance to the secondary loop where it will affect performance. Remember that you have 12 volts supply in the car, but only a few tenths of a volt in your rig. The voltage drop across any contacts in the secondary loop would seriously affect performance. You are not going to see anything approaching 1 volt across the tip of a soldering gun. After writing a rambling speculative comment here I decided to just measure the voltage across the tip of my 140 watt soldering gun. It was 0.2 VAC across the tip conductors at the point of connection to the gun (after a few seconds warmup - voltage increases with temperature of the tip) and about 0.3 VAC across the conductor bars emerging from the gun body (indicating a low but finite resistance at the tip clamping nuts which I recently cleaned and tightened). Remember that the conductors inside the gun are a single loop of heavy copper alloy bar and the tip is a short, heavy copper conductor. You are not going to develop much voltage there. I wouldn't be surprised to find that the tip itself contains a section of some alloy with a higher resistance in order to concentrate the heat where it is useful, but I don'know if that's true. The required conductor size has a lot to do with your duty cycle. My gun is rated at 120 volts, 1.2 amps primary, or 144 watts and I measured 0.3 volts across the gun conductor bars. That implies about 500 amps in the secondary loop and this is only a medium size gun. It would be quite impractical to design your secondary conductors for that continuous current. Look at the size recommendations for 500 amp welding cable that carries current for much longer periods than your rig will. Welding cable conductors for 500 amps is about the size of your thumb. I'd take a clue from the soldering gun manufacturers, although part of their conductor selection may have to do with physical strength. Time for some experimentation on your part. Consider using copper tubing. Since you are going to be dirculating water anyway for cooling your mold, you could circulate water through your secondary conductors, allowing much higher current to be carried for the amount of copper in the conductors. Use a switch on the transformer primary with the rating of at least that of the switch on the microwave. That's all the transformer can handle. This looks like a good application for a SSR (Solid State Relay) with a current rating of, say, double your calculated primary current since you will be controlling it with a timer. You won't have as large an inrush current to deal with as you would with a motor or incandescent lamp load but you might have a brief surge and an inductive kick to deal with. Use a suppressor. The SSR manufacturers have app notes to guide your selection for various types of loads. awright

Google "Watchdog Timers." TI, Maxim, Analog Devices, and many other chip makers make devices with various levels of complexity and sophistication to restart a system if voltage levels drop below some threshold or if the watchdog input senses an absence of normal system activity or if it receives a reset pulse. They come with preset or adjustable time delays of milliseconds to seconds. They put out a delayed logic "0" or "1" that can be used for whatever you desire. If you already have and are happy with a pulse from your system to command a reset you can use a TDR (Time Delay Relay) They also come in all flavors of adjustability and functionality including delayed on, delayed off, one-shot, and many other programs. Sounds like you may be seeking a one-shot function that will switch the contacts for a selectable time period from a fraction of a second to many hours upon receipt of a pulse or contact closure. As I recall, Grainger's catalog offers a variety of TDRs and has a writeup explaining all the variations of programs available. awright

Your machine will run fine on the original fuses but you will lose protection in the event of a failure that would have blown the recommended fuses. I strongly recommend installing the recommended fuses but if it will take some time to obtain them from the states, I would go ahead and use the original fuses temporarily. Utility line voltages are rarely exactly the nominal voltages stated by the utilities. The long-term average there is probably a few volts different from 230 volts and the instantaneous voltage will fluctuate with instantaneous loading on the line. If you can obtain a meter, measure the actual line voltage at your receptacles several times over the day and night, determine the average, and use the setting closest to your average. It is normally not critical. For a reason that I cannot remember right now I measured the line voltage in my home recently and found 127 volts. Line voltage here is typically described as 120 volts. If you don'e want to measure your voltage, I concur with Hero999 that you should start on 240 volt setting and see how your equipment functions. awright

That's exactly what the new "Smart Meters" that the utilities are at this moment investing multi-millions installing do. They are queried periodically and respond with the electrical usage information. I do not know what technology they use for the link to the utility. I also do not know how often the meter "reports in," but I believe that it will be several times per hour. One of the guys installing "Smart Meters" here in the San Francisco Bay Area for PG&E told me he did not know anything about the technology, but that he thought a satellite link was involved. He did not know if the satellite communicated with each "Smart Meter" individually or if data from many meters was collected at a central point and the accumulated data was linked to the utility via satellite. Conceivably, the meters could use cell phone links locally. The "Smart Meters" provide the utilities much more detail in the quasi-instantaneous time-of-usage of energy by each customer which some (including me) regard as yet another intrusion into the privacy of the individual. The flip side is that it facilitates time-of-use fee structures, i.e., cost per KWH based upon when you use the energy which can be advantageous to the thoughtful energy consumer. awright

What makes you think that you must smooth out the ripple on the full-wave rectified waveform coming out of the rectifier bridge for welding? Why would a smooth DC voltage/current work better than the raw rectified wave? The heating value of the arc will not be significantly affected. Now, there IS good reason to provide an inductor on the output, but it is generally referred to as an arc stabilizer because it helps maintain the arc across the nulls of the rectified waveform. It also affects the handling characteristics of the arc, but not, I believe, because the DC is smoother (except that the nulls are rounded out). Since you are not necessarily trying to "smooth" the ripple out, conventional L-C filter design concepts are not particularly applicable. Wire size should be selected for the highest average (actually, true RMS of the AC+DC ) current you expect to use over any, say, ten minute period, taking into consideration the fact that you are dealing with a coil, not a straight wire which would have better heat dissipation. You may want to provide an air gap in the magnetic core to avoid magnetic saturation of the core but I'm not sure that is really necessary. An "air gap" is normally provided by inserting a fiberboard shim between the "E" stack and the "I" stack. I think your best bet is to look at the inductors in commercial welders of the same current capability that you want and try to roughly copy the inductor core size, wire size, and coil size that you see. I don't believe the inductor value is particularly critical to the quality of the arc or the resulting weld. awright

Are you still there and still interested, 89Panadol? Most likely your relay has a coil with too low a resistance. What type of relay is it and what is its coil resistance? How much current does the coil draw if connected directly from your 12 volt supply to ground? How does that compare with the maximum current rating of your darlington output stage transistors? What is the current capability of your 12 volt power supply? Is it capable of driving the relay coil directly? Measure the 12 volt supply while the device is trying to close the relay to be sure it stays at 12 volts. Your darlington output stage would have a current gain of at least 10,000, which is plenty for the drive current available from the 4093 gate via the 100K resistor. The BC547/8 has a collector current rating of only 100ma which is OK for many small relays but not necessarily for a large relay with a low resistance coil. But I would expect the transistor to fry if you tried to pull too much current. Are both transistors still OK? If TR2 burned out TR1 alone would probably lack sufficient gain to drive a low resistance relay coil. Try a more sensitive relay, perhaps a solid-state-relay, depending upon what your load is. And check your output transistors. awright

It would be more useful for you to tell us what you are trying to heat. It is straightforward to calculate the POWER dissipated in a given length of a given type of wire at a given voltage. Except for very standardized conditions and media, calculating the TEMPERATURE attained (In the wire? In the fluid medium? In the surrounding insulator?) is a very complex calculation involving surface areas and shapes, fluid flow, Reynold's numbers, thermal conductivities, etc. Consider a small space heater. The thermal POWER put into the room is fixed by the electrical characteristics of the heating element and the voltage but the TEMPERATURE of the element or of the air could vary over an extremely wide range depending upon the airflow. Or consider an immersion type tea water heater. The temperature will stay below 212 degrees F until the water boils away, at which time the temparature will shoot up by several hundred degrees and the heater will melt. So, what are you trying to do? Oh, yeah. Wire. Almost any wire material CAN be used as a heater but only a few are practical. In general, copper makes a lousy heater because it is too conductive. 10, or 100 amps through (approximately) zero ohms produces (approximately) zero power. Nichrome alloys are most often used for electrical heating elements because they have relatively low electrical conductivity so you can generate a useful amount of power using a practical length and size of wire. It is also resistant to oxidation. For special, short-term applications you can use iron wire but it will rust and has much higher conductivity than Nichrome. I have used a couple of inches of iron wire on a small very low voltage transformer as a quick and dirty nylon rope cutter for years.

stube40, what ever happened with your high power test load? awright