Willem PCB3B and Later Variant Series - Power Notes 2020/01/02 by Ami Sapphire Updated: 2020/03/15 2020 [0103] Added polarity choices to power suppply section [0315] Found and fixed typo in Large Resistor section Slight grammar correction in Power Supply Section This documentation is to attempt to clear up the power issues with the Willem programmers that use the MC34063 as a voltage regulator for the Vpp side on the DC power circuit. This mostly applies to the PCB3B and its later variants due to the fact that they have a Vpp section that has 12.5V, 15V/16V, 21V, and 25V settings. This is a stark contrast to the Willem programmers that have a 7805 regulator for Vcc and a 7812 regulator for Vpp (or 2x LM317 DC regulators that regulate Vcc and Vpp). To put this simply, the 7805+7812 version needs an unregulated 12V, 500mA-1A power supply for old ICs up to 16V and a 16V-24V 1A supply for even older ICs up to 25V, while the 2x LM317 version needs a 16V-25V 1A supply. -- 1. Power Supply The most common issue with the MC34063-regulated Vpp design nowadays is technically the power supply. (Before then, it was the under-specced 100uH inductor, but that will be up next.) First, the type of supply for this programmer should technically be an unregulated, linear power supply, 5.5mm x 2.1mm. (For properly-designed regulated switching supplies, the voltage will have to be upped to around 15V-19V and 1A minimum. However, that is not thoroughly tested since they were not common in a low profile form factor when these programmers were first made.) Second, for these programmers, the polarity is positive (center pin positive), and the minimum voltage is 9VDC and the maximum voltage is 12VDC. Strangely, they can actually tolerate 24VDC despite the design not calling for it, not to mention the 5V 7805 regulator not really liking that too much in the first place. Also, for the PCB5 series, the negative polarity plugs somehow power on these programmers, and that is not recommended. Third, depending on who you talked to, the minimum amperage is 500mA. That is fine for old ICs up to 16V, but the old NMOS units actually saturate on the original design so that it reaches around 2/3 of that supply! Some aren't too specced up to that point and the supply actually becomes even less efficient in the process while attempting to program ICS more than 16V, hence the voltage drop, and the original design itself does not help the situation. Hence, the recommended amperage for ICs up to 25V (according to me anyway) is 2A for 9V and 1.5A for 12V. 2. Inductor / Inductor Coil Back then, some people would assemble the single-sided PCB3B boards and manage to under-spec the inductor coil on the unit. (Not-so-great component choices included wrong transistor types and wrong logic ICs for the programmer itself, but these were hobbyists, so at least they learn in the process.) Another issue that could be combined with the under-specced inductor coil would be the 500mA power supplies, and even worse, they would usually be... regulated! Some were unregulated, but still. Oh, and the MC34063 in this design really takes a hit on the inductor as well, so it should be well-specced. First, the schematic for any Willem design that used the MC34063 power design specified a 100uH inductor. From what was archived over the years, no one seemed to have screwed this up, so that was good. Second, the inductor should be at least 1.5A. Too low, and the inductor coil overheats (oscillation) and may smoke and pop, or may actually explode. Since the amperage was never marked on the schematic, inductors actually varied on some non-Willem-built boards over time. (And yes, the late Willem himself built some and sold some on his website back then.) An example: one person actually installed an 84mA-specced inductor and then tried to program a pretty old EPROM. POP! went the inductor. I personally recommend an inductor coil specced around 2A-3A (assuming they fit at all, but some never give up), though I have seen some get away with one slightly lower-specced, and the lowest-specced unit was probably 1.2A. Third, any newer Willem clones use a magnetic ferrite bead wound inductor coil. Those are usually higher specced anyways and can tolerate oscillation a bit better that most other inductor types within the same amperage. This was found out over time, probably around the late 2000s. Fourth, the inductor's DC resistance should be at least 10 MOhms. Remember the 84mA inductor incident? It apparently had a DC resistance of 8 ohms! Oh, and that entire incident killed the EPROM IC as well, likely due to some heavy oscillation. Most larger 2A inductors are in the mOhm range anyway, and even then, it's usually around 30-40mOhms, so that should not be too much of an issue. Fifth, the type of coil should be a filter type rather than a large RF coil type, though they both work. Filter-type inductors are much smaller and can fit the programmer board, while the RF coil type inductors tend to be quite large in profile and may not even fit in the smallish area of the board. 3. Large Resistor The resistor that is located next to the 100uH inductor is marked 0.5 ohms (0R5 on the schematic). Some actually suggest you should use a 0.33 ohm resistor or even a jumper, which is effectively 0 ohms (rather use a 0 ohm resistor myself if I actually bothered,) but there may have been a reason why 0.5 ohms may have been used, and it may have probably been inrush current-related to protect the programmer's components, but in a quite cheap manner. First, I would leave the resistor at 0.5 ohms as the schematic dictates unless there is actually an issue. So, basically as a last resort thing. As mentioned earlier, 0 - 0.33 ohms can be used otherwise. Humor me, and install a variable resistor there instead. Second, it should be a 1W (1 watt) size resistor, like the more modern Willem clone units. Older ones use a 1/2W (0.5 watt) size resistor or even a 1/4W (0.25 watt) size resistor. Third, for tolerance it depends. Nowadays, 1% tolerance resistors are used in these units, but older ones used 5% tolerance resistors without much issue.