Rougher than Normal Finish?

Question: My Elox Astra model A25S power supply is giving us a rougher than normal surface finish. We have determined that the ON-Time switches are at fault. We can get what we want by lowering the ON-Time, but something is definitely wrong. How can we debug this?

Answer: There could be a very simple explanation. The Elox model A25S power supply includes an additional multiplier switch in the center of the thumbwheel assembly. The intention of this switch was to allow ON/OFF settings of greater than 999 µs when using metallic electrodes.

In the U.S., many companies use only graphite and it is not uncommon to ignore the multiplier switch. Some EDM operators may not even know what the multiplier switch does.

It is quite possible that the operator inadvertently changed the multiplier to X2. This would cause the ON and OFF time to essentially double. A programmed value of 100 would behave as 200. This would also explain why you are able to compensate by changing the ON/OFF settings.

If the multiplier switch is correctly set at X1, we need to look a little deeper into the cause. The thumbwheel switch connects to the Elox MVB board through a multi-conductor ribbon cable.


Attach the scope probe tip to TP305 and clip the scope ground clip to the “GND” test point on the MVB board. With the scope connected, turn the power on and look at the waveform. There should be a TTL (5V) logic-high. TP305 will show an inverse signal. For example, if the thumbwheel was set to 100-ON and 30-OFF, you would see the 100-ON as logic low and the 35-OFF as logic high.

Try the programmed value that is not behaving correctly. If the ON/OFF time is not correctly found at TP305, then a malfunction of the MVB assembly has been confirmed. The MVB board, thumbwheel switch assembly and ribbon cable should be serviced as a set.

If the scope reveals that the ON/OFF time is being correctly developed, then the problem likely is caused by peak current relays on the output module that are stuck on. Degraded surface finish can be caused by several factors. ON-Time and peak current are the two most prevalent. If a peak current relay is stuck on, it can cause a poor surface finish. Lowering the ON-Time will compensate and improve the finish. It is easy to misdiagnose an ON-Time problem when the fault lies with the peak current.



Elox Astra Resistor Failures?

Question: My Elox Astra power supply is experiencing repeated resistor failures on the output module. We replace the resistors, but they only last a few months. Why is this happening? What can we do?

Answer: There can be many reasons that module resistors fail. The most common is excess heat. The image above shows two Elox resistors. The upper resistor appears as typical in use. The lower resistor shows windings that have become so hot that they have damaged the vitreous enamel coating. The area highlighted in yellow has become so hot that the windings have failed.

The resistors shown have a factory rating of 175-Watts. However, in service, the resistors are actually subjected to slightly more than 750-Watts. This is more than 400% of its rated value. The 175-watt rating is based on convection cooling. With forced-air cooling, the rating is extended. The Astra power supply uses a large blower to draw air in through a filter on top of the power supply and out the rear. That air flow increases the wattage rating of the resistors. Figure 2 shows the blower used in the Astra.

The blower has a rating of 965 CFM in free air. However, as dust and debris becomes trapped in the air filter, the restriction reduces the air flow. At just 0.3 inHg of static pressure, the air flow drops off to 850 CFM. At 0.6 inHg of static pressure, the air flow falls to 710 CFM. As the air flow drops, so does the effective wattage rating of the resistors. The propensity for resistor failure increases proportionally.  

Perhaps the solution for premature resistor failure can be as simple as increasing the frequency of the filter change. However, it is also possible to measure the air flow to determine whether there is a problem with the blower motor itself. Knowing the discharge area on the back of the power supply, we can calculate the speed at which the air should be exhausted from the back. A typical Elox Astra has a discharge port that equates to 10.324” x 10.324” square. With no filter installed, the exhaust air speed is 397 meters/minute, or 14.76 MPH. A simple hand held anemometer can be used to check exhaust speed and determine whether the blower is functioning properly. Below is a spreadsheet to calculate the exhaust speed based on the exit port size and blower CFM in your power supply.

Spreadsheet: Power Supply Exhaust Speed

Another option to minimize resistor failure is to upgrade to a higher capacity resistor. Figure 4 shows a ribwound resistor. The extended ribbing visible in the photo greatly increases surface area. The result is a 171% increase in cooling capability. Or another way to look at it is an ability to survive in a 58% decreased air flow.

The resistor upgrade is a particularly good idea if the EDM is operated in an area subjected to elevated ambient temperatures. It is also a good idea if there is a need to achieve high levels of repeatability from cut to cut. As a resistor gets hot, its resistance changes. This causes the amperage going into the cut to change slightly. The amperage change affects surface finish and cavity size. The ribwound resistors run cooler, so there is less of a temperature related impact on surface finish and cavity size.