Skip navigation
1 2 3 Previous Next

John Wiltrout's Blog

102 posts

It has been a while since I did the last Shop Tip Blog. These are blogs about some of the simple things I have discovered that make working in my shop a little easier. Today I realized that I had not mentioned how useful a Lint Roller can be.


When I have finished a build or a salvage operation on a piece of electronic equipment there is always a mess of debris, metal chips, wire snips, and solder splatter left over. To me, it is very important that none of this debris gets to the floor. The floor is cheap laminate wood and if the small bits of metal get picked up by the casters on my chair they act like little grinders and will eventually wreck my floor.


I begin my clean up by picking up the larger pieces or whisking them into the garbage. The solder splatter is scraped off the glass plate that I usually work on and saved for recycling. This leaves the glass plate and bench relatively clean but there are always smaller pieces that are left over. This is where the Shop Tip comes in.


This is an inexpensive lint roller that is intended to clean lint from clothing. I get them for $1 at the dollar store.




It is a roll of very thin sheets that have a tacky surface. When the sheet becomes saturated with debris it can be pealed off to expose a new surface. I use the roller to clean the surface of the work bench as a final touch to pickup the debris that is missed by the whisking and scraping. This has kept my floor free of caster damage.


Here is what the roller looks like after a typical use.






It's Not Just the ESD (Static)

Posted by jw0752 Top Member May 18, 2017

I had a very challenging time last night. I have been working on producing a decent prototype of a machine I described in a previous blog.


The construction had been proceeding very well until last night as I neared completion and initial testing.




To power the LED Scanner I had decided to use the 2N7000 400mA  logic level MOSFETs. These, eleven in all,  had all been soldered into the board. and all the necessary connections had been made to the reverse side of the board. It was all neat and pretty except for one thing. It didn't work properly. It looked like most of the MOSFETs were showing a partial short from D to S.






A quick check with the meter showed that this was the case. 10 of the 11 MOSFET had from 10 to several K ohms of short between their drains and sources. My initial thought was that I had been careless with handling them and that static had gotten them. I hated what I had to do but I began to peal back the wiring from their contacts and with all the care and grounding straps I could find I replaced them one by one. At last reassembled I once again applied power and voila! Same problem, same appearance of partially shorted MOSFETS. So we dove back in for another go at it. This time I monitored the status of the MOSFETS as I went along. To my surprise they were being killed by me as they were being installed. At this point there was no way that static was to blame. Too many precautions and ground wires. It dawned on me at this point I was perhaps heat damaging them as I soldered them in. Because of space considerations the MOSFETs had been pulled down fairly tight to the board leaving about 3 mm of lead before entering the board. My solder iron does not have temperature control and runs at about 325 C. The data sheet for the 2N7000 says that it can tolerate 300 C on a 1/16 inch lead for 10 seconds. I think, from my experience, this is a bit optimistic. My usual technique is to minimize time on the contact but obviously something was damaging these parts. I got out the heat sink clips for leads and carefully began to replaced the remaining bad FETs. I continued to monitor them as I installed them. Since I was tacking wires and resistors to their solder pads It was necessary to make extra certain that time on the pad was minimized. I also sprayed coolant on the heat sink clips. Despite these precautions I still managed to kill a couple more. Finally after 6 hours and lots of frustration I had the machine up and running properly.


In all I killed 19 of the 2N7000 MOSFETS




Fortunately for the pocket book they are not expensive and while I did not have any left to spare after this debacle I did have exactly enough to complete the job.

Here is a picture of the completed project:






Here is a short video showing how the lights scan back and forth while hand held tactile devices buzz periodically in the patient's hands and a tone is emitted in the patient's ear as the light on the bar hits each end. I personally question the therapeutic value of this device but then what do I know, I am a technician who can't even seem to install eleven MOSFETs without Killing 19 of them.




Despite 60 years of playing with electronics this is the first time I have seen a DC motor wired in this fashion.  At first I did not expect it to work on DC current but to my surprise it took off and ran perfectly. Furthermore it did not matter if the polarity was forward or reversed. The motor continued to spin happily along. Just for the fun I also applied an AC current and once again the motor continued to whirl away. Here are some pictures of the motor wiring:










Just to clarify how the motor is wired I also produced this small schematic:


Novel DC Motor Wiring.bmp


On first appearances I assumed the motor would start to turn and then stop when the small tantalum capacitor became charged. I did not initially understand why this did not happen. I am hoping some of my friends on the Forum with more experience and knowledge will confirm my speculation that the counter EMF in the motor actually discharges the capacitor allowing it to accept a second charge and continue to maintain the magnetic fields that spin the motor. I suspect that the low ESR of the Tantalum caps makes their charge and discharge rapid enough to facilitate this process. I do not know why the designer of the piece of equipment where this small cooling fan was used chose to use this particular circuit. It was something new to me and it caught my attention. Now I just want to understand it.



If you are looking for an inexpensive kit build project that comes together nicely and works well to boot check out the DSO Shell 15001K Oscilloscope Kit.




At $22.00 including shipping from China this is a no brainer. With a 200K bandwidth (Perhaps a bit of an exaggeration) it is undersized for many applications but still is a good tool for Audio and lower frequency servicing as well as a good educational tool for the experimenter or enthusiast. This model of the unit comes with all the SMD components installed. The through the hole components, switches and connectors are left for the kit builder. Here is what comes in the package:




Note that while you see a probe in some of my pictures this is not included in the kit, nor is the 9 volt power supply. I found that the small wall warts that I use for the Arduinos worked well. I also put a battery pack on the back of one unit so that I have a portable model. The manufacturer cautions that 10 volts is the high limit of the power supply and damage may result if the unit is powered with a higher voltage.


Here are what the circuit boards look like out of the box:




The kit comes with 4 pages of assembly instructions, calibration, and operation manual. While it is over all adequate it takes a lot of looking close and a thorough reading to get it right the first time. I have build three so far and I still can't quite get it right. It is not that the correct procedures aren't explained or illustrated somewhere on the pages it is just that it is not a totally smooth sail and I have a tendency to use my intuition too much. Fortunately a couple redos are permitted. The solder points are very small and a little challenging so this may not be a kit for a first time builder. Better to have some soldering experience first though it is doable if someone with experience is supervising a beginner.




The resistors supplied are very small and 5 banded. My 68 year old eyes had a challenge with them so I just metered them and put them where they belonged. If you happen to have some of my special tungsten carbide meter probe tips you can check 1/8 watt resistors one handed like this:




When the main board is fully populated they have you plug it in to check for proper operation. The instructions are a little out of sequence so it is necessary to populate the board and then go back to the first instruction on the list telling you to test it. At this point all one can do it see the screen and toggle the function buttons.




I found that the vias did not pull the solder in quite as easily as I would have liked so a good inspection of both sides of the board after soldering allowed me to dress up a couple questionable looking joints. The Analog board is the board where the bulk of the through the hole components must be installed. Over all the instructions were good and the components easy to identify. On the first two units that I built I mounted the adjust control backwards on its board. This does not harm anything but the unit will not work properly. If the instructions are read carefully they have a small box with a red print warning not to do it the way I did. Once this error was corrected both of the first two units worked properly. Here is the small board showing the side not to mount the control on.




The analog board comes with two sets of header pins so that during calibration the main board can be off set to allow access to the adjustments. While in this position one can check voltages and tune the input to display a clean square wave.




One feature that I was really pleased with is the ability to toggle on and off a list of measurements. Usually on a meter this size and cost one is left to guess what the digital data is. Here is what the unit looks like with the measurements displayed:




I put a battery pack on the back of one unit to make it more portable and purchased a set of $10.00  Oscilloscope probes for the units. The probes like the oscilloscopes came with free shipping but this time the Chinese had them shipped to me from the Seychelles. The ways of the electronics world are hard to believe. I am enjoying this era of free trade and exchange of products. I wonder how long before the politicians see what is going on and screw everything up. Any how if you like to build things this was a fun little project with very little financial risk involved. Here is a picture of the back of my unit with the battery pack and a picture of a completed unit.






Have Fun.




Zener & LED Tester

Posted by jw0752 Top Member Apr 23, 2017

While sorting out and reorganizing my collection of mostly salvaged zeners the other day I began to imagine what I would have to do to make a simple tester. In the past I have set up the power supply, milliammeter, voltmeter, and a variety of current limiting resistors to test the questionable zeners. After considering the problem for a while I decided that a custom power supply with a few modifications would make a perfect zener tester. I also realized that it would also do double duty as an LED tester.


Here are the parameters of the modified power supply that I envisioned.


Variable voltage control from zero to thirty volts.


Variable current limit from zero to 1000 mA.


Digital display of output current in mA and voltage across the device under test.


A means to calibrate a target mA limit prior to connecting the part to be tested.


A way to connect the component to be tested easily.



A quick inventory of my parts graveyard revealed that all the necessary components for this build were already at hand. I would use the same Chinese power supply kit that I had been using in my recent bench power supply builds.




This board which is available from a Chinese Source for $10 accepts an AC input of 24 Volts and will control 0 to 30 volts output at a max 3 Amps. It also has a current limit control so that one can turn the voltage up all the way and it will self regulate the voltage to produce a constant current output. This constant current control was exactly what I needed to make the zener tester work. Since the applications that I will be testing will require currents in the low mA range I began by seeing if I could make the board work well in this range. The unit comes with a 10K potentiometer that with a 270 degree turn moves the current limit from 0 to 3 amps. For this project my preference would be a ten turn pot that controls the current between 0 and 1000 mA. After some experimentation I found that by substituting the following circuit in place of the original 10 K pot I got the result I wanted.


Current Pot Modification.jpg


The next problem that I would have to overcome was the display of the digital meter that I hoped to use. I have used these meters on my power supply builds and they will display 0 to 100 volts to a resolution of one tenth of a volt. This will be satisfactory for my needs but the current section with a built in shunt that displays 0 to 10.0 Amps will not have the resolution that I need. I began modification of the meter by removing the factory current shunt and the installation of some low ohm resistors. After several failures I finally got close with (2) 0.1 ohm resistors in parallel. I was then able to fine tune them by adding extra solder to the leads to slightly lower their resistance further. The end result was not off as much as the linearity of the meter over the 0 to 1000 mA range so I was satisfied. The beauty of making test gear for oneself is that you can tolerate the accompanying imperfections with more grace. Here are pictures of the meter front and back.




I decided to put a switch on the output that would allow me to switch the control board's output between a dummy load that I could use to precalibrate the target mA and the actual output to the test leads. By having a center off for the switch I have the added benefit of a parking position between tests. This idea of the switch was given to me a couple years ago by mcb1 in one of the many inspirations he has contributed to my devices over the years.


By now I had collected the case, transformer, control board, meter, and other parts I would need.


I tested the transformer that was a perfect fit for the enclosure and I was dismayed that it put out 30 VAC instead of the 24 VAC required by the board. I have had enough experience with the board to know that this voltage level would stress and probably damage the circuit. In a normal build I would have gone in search of a different transformer but two things made me reconsider. Number one, it was a perfect physical fit which is not always the easy part. Secondly, this application had a relatively low power requirement. I decided to try an old trick where loss of overall power isn't a problem and went in search of a small transformer or inductor that I could put in series with the main transformer. Just as we can use a resistor divider to adjust voltages I would make an inductive divider to lower the voltage to the primary of the transformer. With some luck I found the secondary of a small transformer that moved the primary voltage of my main transformer to 96 VAC and the secondary voltage to a very acceptable 24.4 VAC. In the pictures that follow you can look for the small blue transformer serving as an inductive voltage drop for the main transformer. I did not check it but I was probably inducing a hundred volts or more on its primary. I cut the leads off flush and sealed them with epoxy so they would not be a factor. I have made a note for an experiment some night to see if I can adjust the voltage output of one transformer by loading the secondary of a second transformer that is in series with it.


Here are some pictures of the build sequence. You may notice that I do not try to give instructions for building any of my projects. This is because my builds all come from my own pile of junk parts and most would not be able to be duplicated by a second builder. Even I am not usually able to duplicate a build. The idea of the build is however always offered to anyone interested and I will always answer questions below as to how you can integrate your own pile of parts to do something similar.



The case came from an old dental cure light. I have used these cases in the past for single channel bench power supplies.



Before wiring has begun



The little blue inverted transformer in the back left area is the inductor that is reducing the voltage of the main transformer primary.



The completed wiring prior to closure of the case.



The finished front panel.



Testing a Blue LED at 16 mA and the meter showing the junction drop of 3.2V.


The bright CC Mode light comes on whenever the board is controlling the voltage to maintain the target current.








Prototyping an EMDR Machine

Posted by jw0752 Top Member Apr 17, 2017

One of my Grandsons recently was telling me about a type of psychotherapy that is used to treat PTSD and other traumatic incidents called EMDR, Eye Movement Desensitization and Reprocessing. He described a machine that was being used by therapists that consisted of a light bar mounted on a tripod. The row of LEDs on the light bar would move back and forth from end to end. The machine also had hand held vibration devices that are placed in each of the patients hands and a set of headphones. During the therapy, between counseling, the patient is instructed to sit with their head fixed and follow the movement of the LEDs with their eyes. While the LEDs of the left side of the bar are lit the vibration device in the left hand vibrates and while the lights on the right side of the bar are lighting the right vibration device vibrates. As the terminal LED at each end bar lights a tone sounds in the corresponding ear of the headphones.


If you have further interest in this Therapy Topic here are a couple links to get you started:


My Grandson's question to me at this point was whether it would be possible for us to make one of these light bar machines. Since I like a little purpose to my explorations I decided it would be a good project to explore. I began with a simple scan circuit based on a 4017


4017 Chaser Circuit.JPG 


This very properly makes the lights scan from left to right but rather than returning it jumps back to a left right scan each time. I decided to extend the counter by one more stage and use the outputs of the second 4017 to reflect back on the LEDs so that the first 4017 would light the string in one direction and the second 4017 would light the lights in the opposite right to left sequence. I also had to be aware that ultimately the left side of the string would have to be distinguished from the right side. The Data Sheet for the 4017 provided the information necessary to add another 4017 to the string. An MC14081 "AND" gate was also necessary for the logic of this addition. The circuit was built and now I had a string of 17 LEDs flashing from left to right. Now I had to find a way to power the LEDs of the first 4017 using the second 4017 in a reverse order.  I began by putting 1N4148 diodes on each of the 4017's outputs. This would allow me to power the LEDs from each 4017 without any interference. I also removed the current limitation resistor(s) on the LEDs and isolated the return leads of the left from the right side LEDs. Instead of current limiting resistors I installed a separate simple one transistor current sink for each side of the nine LED string. My hope was that ultimately I would be able to use an OPAMP to sense when each current sink was being used so that I could turn on the correct vibrator to correspond to the right or left side of the bar. The tone in the headphones could be driven by a MOSFET linked to the end LEDs. A small capacitor and resistor on the drain of the MOSFET would allow me to control the amount of time the MOSFET remained on and hence the duration of the tone. For this prototype I just powered small piezo buzzers (sonalerts) whereas a production unit would have to have oscillators and a small audio amp. To simulate the hand vibration unit I salvaged vibration motors from some old cell phones. The Op Amps that I used to discriminate the left current sink from the right one were a bit sensitive but functioned properly. I believe that a production model with improved inter device connections would improve this problem. A nine volt linear regulator was installed to stabilize the voltage to the unit so that the op amps stayed in calibration.


Here is a picture of the finished prototype board:




I am sorry but I have not produced a schematic of the unit and I probably won't. While it is very functional it is unlikely that it has been anything but a learning experience and is not likely to ever be produced in this configuration. As it got more and more complicated it occurred to me that a Micro Processor like an Arduino would easily handle this function with many less components and complexity. I have made a video of this board in operation and I will attach it at the end of the blog.


At this point I decided to develop a second prototype based on an Arduino Duemilanove. The program was very simple and linear with very little logic involved. I wanted to control the speed of the scan of the lights so I put in a potentiometer that could be read by an analog input and then the value used as a value in the "delay" command. Once that was done it was simply a matter of telling the arduino to turn on and off LEDs, vibrators, and sonalerts in the proper sequence. The circuit itself was a lot simpler although the use of the jumper wires from the Arduino outputs to the LEDs makes the board look quite cluttered and obstructs the view of the LEDs. While the sonalert beepers could run directly off the outputs of the Arduino the 150 mA at 3 volt draw of the vibration motors was too much current for the outputs. To handle this I used the 9 volt power supply and MOSFETs driven by the outputs of the Arduino to power the vibration motors. To control the current to the motors to the proper level I once again constructed a simple one transistor current sink to ground and fed both motors to it.


Here is a picture of the finished prototype board.




Here are videos of the two different prototypes each doing basically the same thing.


EMDR Prototype built using discrete components.


EMDR Prototype using Arduino Duemilanove.


This has been a good experiment as I have had the opportunity to refresh my experience with the 4017, 555, and 14081 IC devices. I have also had some fun playing with simple current sinks that can be used in lieu of current limiting resistors when supply voltage isn't stable or other considerations are important. Further the utility of MOSFETs continues to impress me. I have had the basic skills to use MOSFETs for about three years now and I do not know how I survive the other 65 years without them. This project has also forced me to once again remember that despite my old fashion gravitation to the use of discrete ICs and other components the microprocessor is a better solution in many cases.



I was busy salvaging a bunch of Dental electronic equipment for interesting parts when I came across this interesting circuit board construction and I thought you might get a kick out of it.












This is a driver circuit from a handheld LED composite cure light. It allows the operator to select a timer setting and drives a high power LED that has it's light output towards the blue end of the spectrum. The unit also controls a cooling fan the size of a dime to help cool the LED. In order to fit the needed components into a body witha diameter on the order of a broom handle they had to stack a set of circuit board disks 2.2 cm in diameter.  Pretty cool!


Hacking a Toroidal Transformer

Posted by jw0752 Top Member Mar 15, 2017

I have this beautiful Toroidal Transformer but it is 19 Volts and I need 24 Volts for my application.




It has a nice big donut hole and I have plenty of 16 GA solid enameled wire so I decided to add a few more windings. The first step was to figure how many windings I would need. I used some lighter gauge wire and wound 10 windings, hooked up the transformer and measured the voltage. I had 1.7 volts for the ten windings so 29 windings should give me the boost I need. This can't be too hard, I thought. I made a loose coil of wire and kept winding it around and around until I had the thirty that I needed. Next I secured one end of the wire to one of the existing wires and started to work backwards. I pulled the wire as tight as I could and went around and around until I had worked my way back to the last wind. Here is what it looked like.




Now that looks simple enough but I am writing this blog to let my hands stop aching. That was really hard work. I had to wonder how they wind these transformers when they have to put hundreds of winds on them. I can't imagine what kind of machine would be able to perform what I just did with my hands. If anyone knows how they do it I would really like to know.


I next used some cloth tape that I had and began to wrap strips around the transformer to hold the windings in place and to protect the wire. Here is the transformer after this procedure.




Finally I figured out the correct phase so that the voltage in my windings would add to the existing secondary and made a good solder connection. A few more winds of cloth tape to polish things up and I was done. A test of the transformer revealed 25 volts output with 118 Volts input.




Now that my hands are rested I am going to go back and make another one.



There have been many times when I have wished that I had a hot air heat source that could be controlled. The rework station has a controlled heat source but it will not work below 100 C. For this project I have decided to build a modified hot air gun that will have controlled temperature output between room temperature and 350 C. This should be very useful for testing and calibrating devices using thermistors. I will begin by hacking my inexpensive heat gun from Spark Fun.




I began by opening the gun and removing the connection of the motor from the heater taps. The common practice in these devices is to put the mains across the heater element and then take a tap off the heater element around the 15 volt level, rectify it and use it to power the fan motor. For this project I wanted the fan to be tied to a fixed 12 volt source and independent from the variable AC voltage I will put on the heater element. Once the motor was cut loose from the heater element taps I removed the diodes from the back of the motor housing and installed a length of high strand flexible wire. Later I would pull a silicone sheath over the heater and motor wires to make the cord from the control to the gun functional. You will see this cord in later pictures.


The heart of this project is a small variac, about the size of a soft ball, that I salvaged from an old Ritter Dental X-ray.




Besides the full 0 to 110 volt sweep contact this unit also had hard taps at 8, 17, and 55 volts. My plan is to rectify the 17 volt tap and use it to drive a 7812 regulator which in turn will provide the power for the fan in the gun. My test of the fan motor showed that it drew 0.25 Amps at 12 Volts and ran with enough RPM to serve my purpose. The 7812 can handle 1 amp when heat sunk so with a small radiator it should be able to dissipate the watt or so it will be required to shed.


It is not my intention to build a precision device but I do want to have some indication of what my input settings are so I can estimate the output temperature. Therefore an AC voltmeter will be used to monitor the voltage that is powering the heater element. I will combine the voltage with an empirical listing of approximate output temperatures on the dial of the variac to get my output temperature as close as practical to my target temperature. There are many variables that will contribute to the output temperature of the air. I have attempted to limit and control some of the variables such as the fan speed and the voltage to the element but ambient temperature will also be able to influence the output. I will be using a small 0 to 55 volt 4 digit 3 wire digital voltmeter module to display the AC volts.




The challenges in using this module are that it is DC, doesn't read directly as high as I want to go, and requires an isolated power supply from the variac to do what I want it to do. I experimented with using the 8 volt and 17 volt taps on the variac to power it but I was not able to get it to zero. Finally I took an old linear 7 volt wall wart, broke it apart and used it to supply the power for the LEDs and logic. This allowed me to get accurate readings on the sense line. In the final design I used a 100K potentiometer as a voltage divider to to bring the voltage to the meter down one magnitude and then manually moved the decimal to properly reflect the real voltage. With a 4 digit display and no need for that much accuracy it left me with a good solution and the meter now displays 0 to 110 volts while actually sensing 0 to 11 volts.


Before I began the actual build I wanted to do some bench testing and to run some empirical measurements of current versus output temperature. For those who feel my bench is always too organized here is proof to the contrary.






While I am working with some mains voltages in these tests I have taken special precautions to use isolated sources and good test procedures. Unfortunately sometimes there is just no way to make it look nice and get results in a reasonable time frame. Here is a graph of the Temperature versus Current characteristics of the heat gun. I have a very inexpensive two channel thermocouple based thermometer. I tested the thermocouple for accuracy using boiling water and ice in water. It actually tested extremely accurate considering it only cost $13.00. I placed one of the thermocouples 2 cm in front of the outlet of the heat gun and taped it in place. I wanted my readings to be consistent. Finally I fired the test up and took readings at 10 volt increments from 10 to 100 volts. I also used a second also inexpensive thermometer and you can see the difference between the black and red graph lines.


Temp vs Voltage Graph.jpg


This will give me a rough estimate of my temperature output based on the voltage shown on the meter. Once I was convinced that things would work as envisioned I began the build. I had to make the small circuit for powering the fan motor and also one to adapt the AC voltage from the variac sweeper to what would be acceptable for the meter. The small isolated power supply for the meter's electronic was already built. Here is the motor board in progress.




and the salvaged wall wart:




Here is the schematic for the entire unit:


Temp Controlled Heat Gun.bmp

The construction preceded with the customary glitches. I had a bout of dyslexia and hooked the variac up backwards on the first test. I put way too much voltage on the fan motor for a split second and had to make a repair to the motor board. By the time I got things straightened out it was 3:30 AM. Once I knew it was working I went to bed and planned to put the finishing touches on it in the morning which begins about 10:00 AM for me. Here is the finished unit. The little thermocouple thermometer is still attached in the picture as I was running final tests and I wanted to place temperature labels on the face of the unit so I wouldn't have to use the graph to convert voltage to temperature. Here is the finished unit.






Finally I took a short video of how it works.




If you don't want to be disappointed by this blog don't read any further cause I lied.


The problem is that I have just signed up for a new Garbage Service. The new bin is to be left on the curb by my house in a very specific position so that the truck can come by and use its robotic arm to pick up the bin and tip it upside down to empty it. The arm then returns the bin to the curb in exactly the same position that I left it. I do not know when the truck comes so I find my self checking the bin several times a day to see if it has been dumped yet. The exercise involved in walking to the curb and looking into the bin is OK and probably actually good for me but the neighbors may notice and wonder why I am obsessed with the trash bin. I wanted a way that I could just look out my door and know immediately if the bin had been tipped or not. Here comes the Extremely Sophisticated Electronics part:


Tilt Alarm.jpg


OK its not that sophisticated but one has to at least admit that its simplicity is beautiful. It is powered by 4 AA batteries and uses a Tilt Switch to trigger a small SCR. The SCR turns on a RED / BLUE 1 Hz LED flasher that continues to flash until the reset button is pushed. The circuit has been built into a small plastic box and protected from rain by a sealed plastic freezer bag. Here are pictures of the unit untriggered and triggered.






Before attaching the unit to the Trash Bin I called the company to see if they minded if I drilled (2) 3.175 mm holes so that I could mount the alarm. They told me "sure you can drill the holes as long as you don't mind paying for the bin". No Holes! OK. Back to the drawing board yielded an the idea to mount a bracket to the Alarm and then use a hose clamp to attach to the handle of the bin without drilling any holes or making any marks. Here is a picture of the Alarm mounted to the bin.








Now all I have to do is count the minutes until Tuesday when I can test it out.



A while back I explored the Ideal Bridge available using the LT 4320 Controller and 4 N Channel MOSFETs.


I also built a two channel linear bench supply using some Chinese Kits.


In this blog I will explore the results when I built another module with the Chinese Power Supply Kit but this time instead of using the 1N5408 Diodes supplied I substituted the LT 4320 Controller with 4 MOSFETs. Here are some pictures of the completed Module.




The small board in the upper left is the LT 4320 with the MOSFETs. The Yellow wires supply the AC voltage and the RED and Black are the rectified output back to the board. The small auxiliary board in the lower left is the fan controller which supplies about 11 volts to the 24 volt fan at room temperature and increases the voltage to the full 24 volts as the heat sink approaches 100 C.




The Red lead with the black coupler is the 12 volt supply to the meter circuitry.




The nice Radio Shack CPU heat sink that I used on the first two modules was not available any more so I had to use this less impressive but still adequate sink. The fan for this sink came at 24 volts instead of the previous 12 volts so the fan control had to be modified to make the adjustment. Since 24 volts was so close to the max raw voltage from the unregulated supply I used it directly instead of placing a regulator in the circuit.




Here is the final side view of the module. The next step was to remove one of the Diode Modules from the Linear Supply and install the LT4320 version.




The meter probes are looking at the voltage to the fan as the heat sink rises in temperature under a 36 Watt dissipation.




This is just the view from the back.


The goal of this experiment was to see what the difference would be between a channel using the LT4320 as opposed to the 1N5408 Diodes. Keep in mind the supply has identical Toroid Transformers and support controls for each channel. It turn out that the LT4320 side had a Max 27.6 Volts and the Diode side had 25.7 Volts. Besides this unloaded voltage difference at the top the load capability also showed the same differential. For example I turned both channels to their max output and then load each channel with a 2 amp load. The LT 4320 side dropped to 22 volts before it could support 2 amps. The Diode side dropped to 20.4 volts before it could support the 2 amp load. Even under a 3 amp load which is the max for this supply the MOSFETs ran very close to room temperature. While I like the LT4320 from a standpoint of improved efficiency and sophistication the improvement is probably not enough to justify its use in most cases.




Picture showing the no load difference between the two channels.




Comparing Power Supplies

Posted by jw0752 Top Member Feb 12, 2017

I recently built a couple of Bench Power Supplies using salvaged parts and inexpensive kits and modules from the Chinese electronics suppliers. I Blogged about these builds here on element-14.




It was suggested by michaelkellett that I should do a comparison test to see how well the power supplies perform against each other. I have also added in a comparison to a relatively inexpensive commercial bench supply that is my primary supply.


Here are the three supplies that we will be comparing in this experiment. I will refer to them throughout this blog as the Commercial Supply ( Mastech HY 3005F-3), Linear Supply, and Switching Supply.




The Commercial Supply is a 0 to 30V 5 Amp power supply.




This is the Linear Supply as described in the Blog "Oh No! Not Another #@&* Power Supply" and has a range of 0 to 27 volts with a max 3 Amp current.




Here is the Switching Supply as described in the Blog "Using the Coarse + Fine Control Circuit in My New Bench Supply"  Which has a voltage output range of 1.2 Volts to 26 Volts with a 5 Amp max current.


The experiment will look at the ripple of each supply with no load and with full load. Each supply will also be tested and compared to see how it reacts to a load as well as how it reacts to the removal of a load. The test parameters will be 10 volts from channel one of each supply. The load will be an automotive brake light which has a beginning resistance of approximately 0.8 Ohm and an operating resistance of approximately 5.7 Ohms. The low beginning resistance of the bulb will serve to highlight the ability of the power supply to respond to a high current load.




The test instrument is a Rigol DS 1102E DSO.


Here is a layout of the test rig wiring.


Experimental Circuit.jpg





I have set up the rig so that both power supplies are switched to their loads simultaneously. One of the weak spots of this experiment is my assumption that the two load bulbs that I will be using are identical. This may not be true in the strictest sense but they will be close enough for the tolerance of this experiment.


Our first test will be to look at supply ripple without and with the load bulbs in the circuit. As a convention when displaying comparisons I will display the Commercial Supply first followed by the Linear Supply and then the Switching Supply. First however let's look at the base line noise on the scope. This is 32 mV PP with the scope probe shorted to its ground wire.




The No Load Trace of the three power supplies is as follows:


IMG_0856.JPG    IMG_0840.JPG

The Commercial Supply has 56 mV PP, the Linear Supply has 60 mV PP and the Switching Supply has 40 mV.


1.75 Amp load on each supply gave the following results:


IMG_0846.JPG    IMG_0841.JPG

The Commercial Supply has 80 mV PP, the Linear Supply has 700 mV, and the Switching Supply has 640 mV. The characteristic of the switching supply scan is however showing a much more regular pattern which is likely a vestige of the switching frequency.


Next I tested the supply for how they would react to being switched onto the test load.


IMG_0853.JPG    IMG_0844.JPG

The Commercial reacts with a 6 volt instantaneous drop that over corrects slightly and stabalizes after 350 ms, The Linear Supply also dips slightly less than 6 volts and also over corrects with a stabilization reached in about 450 ms, and the switching supply reacts with a 3 volt drop and recovers without an over correction to a stable level in about 100 mS.


The last test that I ran was to look at the situation when the load was removed from the supply.


IMG_0854.JPG     IMG_0852.JPG

The Commercial supply has the lease reaction with only a 400 mV PP disruption while the Linear had a 960 mV deflection and the Switching had an 800 mV deflection.


My conclusion is that the Commercial at 10 times the cost of the other supplies was the best performer with respect to ripple on the voltage under all test conditions. The surprise for me was that the Switching Supply actually out performed the Linear Supply in my opinion. This was not expected as I had always assumed that the switching supplies would be worse. The obvious switch noise on the Switching Supplies output may be more of a concern than the apparently random noise on the Linear supplies output but I do not know enough to make a proper evaluation of this. I have not gotten to the stage of precision in my experiments so far where I have had any problems . I want to thank Michael for suggesting this experiment as it has given me some potentially valuable insights into the performance of these three supplies.





Sorry I couldn't Resist.


It all started when I ordered a couple of these kits from Bangood in China:


Bangood Kit.JPG


At $6.87 including postage I could not resist. The kit came with all the components needed except the heat sink and transformer. There were even pin jacks and wire pigtails so that the Voltage and Current control potentiometers could be extended off the board. The instructions which were downloadable were better than average for a kit from China and it even included the semblance of a schematic.


PS-2 Schematic.jpg

After assembling the circuit board I ran tests on it to see if it was stable and worked properly. I found that it could be used either in a constant current mode from 2 mA to 3 Amps and also in a constant voltage mode from zero to 30 volts. The circuit even had an offset adjustment trimmer on the voltage control op amp so that the potentiometer could be zeroed. The circuit requested a 24 volt transformer to work this full range. After checking my salvage stock the closest I could find to the required transformer were a couple of 21 volt toroidals. This would ultimately provide a 0 to 24 volt output which was acceptable. Unlike my previous switching power supply build this would be a linear supply. Dissipating heat would be a serious consideration. A check of the salvage showed I had an old CPU heat sink with a mounted fan that I got from the Radio Shack 20 years ago. Unfortunately there was only one. A check of EBay turned up a second heat sink for the second channel. Here is a picture of the build when I was still waiting for the second heat sink.




While it is not on the schematic the board provides a separate 7824 voltage regulator to drive a fan. It is visible in the upper left corner of the circuit board. Since the CPU fan was 12 volt I initially replaced the 7824 with a 7812.


I am using an identical case for this linear dual bench power supply that I used for the previous switching supply. Even though the switching power supply had larger toroidal transformers space was more of a problem this time since the regulator boards with their heat sinks were so much bigger than the switching modules. Here are three pictures of the entire unit so that the layout is more apparent.








The display units are only 3 digit and are just inexpensive $4.00 modules from China but they had the advantage of having internal current shunts this time so I did not have to try to make my own. Each display also had its own current and voltage calibration trimmers. Don't get me wrong if you need mA precission this is probably not the meter for you but for most of my kinds of experiments it will be more than adequate and there are always the more precision bench meters if I need better accuracy.


I had to wait about a week for the second CPU heat sink fan assembly to arrive. After it was installed however I was not happy with the noise that the two cooling fans made as each were running at the full 12 volts even when no load was being driven. The answer was to build a simpleTLE 2142 op amp circuit with a thermistor as a sensor and a MOSFET to drive the fan. I still wanted to be able to run the fans at the full 12 volts when things got hot so I had to upgrade the 7812 regulators on the boards to 7815. This gave me a 15 volt supply and the drop out of my little sensor driver circuit was about 3 volts. Here is a picture of the unit with the second heat sink installed as well as the two op amp circuits mounted above the lower left side of each regulator board.



Now we were in business. The unit under no load has the fans driven with 6 volts which ramps up to 12 volts as the heat sinks warm up. The front panel of this unit is very similar to the front panel of the previous build. In fact I used the same template to mark and cut it. Unfortunately I slipped and scratched it which inspired me to flip it over and use the mirror side for this unit.  Here is a picture of the front panel.




Here is the completed unit in its case and also a picture of it with its brother.






Now I would like to tell you that this will be the last power supply for a while but  son Mike just dropped by and brought 5 more dental camera units for me to salvage out. Aye! my work is never finished.



A while back I heard that a diode full wave bridge could be replaced with MOSFETs that were switched in the proper sequence. The advantage of doing this would be the very low Rds(on) available from the MOSFETs which results in much lower waste heat generation and more power delivered to the load. For this experiment I am going to use the Linear Technologies LT4320 controller and (4) IRLZ34 N channel MOSFETs. I will also set up a conventional diode bridge and compare the readings that I get from each system. Here is a link to the data sheet for the LT4320:


Here is a simple schematic from the data sheet for how the LT4320 connects to the MOSFETs


LT4320 MOSFET Bridge.JPG


While the LT4320-1 is capable to 600 Hz the chip I am using is the LT4320 and it is rated for up to 60 Hz operation.


I will begin by constructing a simple full wave bridge using a common 4 diode package and take some readings. I have also installed a 680 uF capacitor across the output.




The transformer in the picture has been connected to a variac and the output has been dialed to 25.75 Volts AC under load. This may appear a bit arbitrary but when the variac you are using isn't precision you take the first stable value that is close to the desired value. I have loaded the circuit with (4) 12 volt auto tail light bulbs in series. With this load we measure the DC output of the system at 27.97 Volts DC and the current is 390 mA.




The AC ripple on the voltage is 2.68 Volts.




Here is what the wave form across one junction looks like:




The next step in the experiment was to wire up the LT4320 and the (4) MOSFETs. While the schematic looks very simple it turned out to be rather more complicated. In the end I used wire jumpers instead of the more solid staple connectors. It is likely that this will result in a higher connection resistance and will lower the performance difference that I hope to observe. Here are a couple pictures of the circuit:






The MOSFET circuit performed very nicely. Once again our input voltage under load was 25.75 VAC. The output voltage of the system into the load was 29.5 Volts DC and despite the lighter wiring we had a current of 400 mA. The ripple of the MOSFET system was also 2.68 VAC. Here are shots of the meters and the oscilloscope screens for the ripple and across the drain - source on one MOSFET.








In both experimental setups I monitored the temperature of the devices with the finger tip thermometer. The diode bridge got almost hot to the touch while the MOSFETS remained cool. This was not very scientific and may also be explained but the much larger surface area of the MOSFETs compared to the Diode Bridge. Here are the data from the two experiments side by side.


                              Diode Bridge                              LT4320 & MOSFETs


Input Voltage          25.75 VAC  60 Hz                      25.75 VAC 60 Hz


Output Voltage       27.95 V DC                                29.4 V DC


Ripple                    2.68 V PP                                   2.68 V PP


Load Current          390 mA                                        400 mA


The MOSFET Ideal Diode Bridge using the LT4320 controller seems to be a much better system for rectification as it delivers more power to the load and has less heat generated. Keep in mind that the LT4320 and MOSFETs still showed an improvement despite being wired with the higher resistance jumper wires. On the other hand the complexity of the circuit and the magnitude higher cost over the diode bridge will not make this a practical replacement in most applications.







A couple weeks ago I asked the forum for some ideas for a Coarse/Fine adjustment circuit for a basic bench power supply that I was going to build. While the power supply is nothing special and the build was rather routine I thought it might be fun to show you how the final product turned out and how I incorporated the Coarse and Fine circuit into the unit. Here is a picture of the inside of the completed power supply.




The unit has a nice project box that was originally a intra oral dental camera. Since I had two of these units I was able to salvage the nice toroid transformers from each unit so that I would have two isolated DC power channels on the outputs. The small green board in the back left corner was part of the original unit's circuit board that has been modified and retained as a line power entrance for the unit. Here is a close up of that board.




The board and the toroids are capable of switching from 120 VAC to 240 VAC input power though in my area, at least, only the 120 VAC option will be used. Along with the toroids from the original dental cameras I have also retained the bridge rectifiers and the filter caps. I will need 12 volts to power the LED meters that I have chosen for the front control panel and I will also need a voltage source for the power on LED and a small cooling fan. To do this I built a 7812 based linear regulator circuit for each channel. Besides the channel's meter this circuit will power the LED on one side and the fan on the other. Here is a better picture of the voltage regulator circuit:




Now is the point where I take a short cut and use a couple of Chinese DC to DC converters that I purchased for $4.00 each. Amazingly these came for that price including shipping. Every time I order some of these cheap items and get the shipping included I suspect that it may be the last time. It doesn't make sense how anyone can make a profit doing it this way. Here is a close up of the converter:




This little DC converter is based on an XL4016 chip and is capable of input up to 36 volts and output of 1.2 volts to 35 volts at 8 amps. Since my transformers will only produce unregulated 27 volts my upper range for this power supply will be limited by the voltage drop of the transformers under load. Experiments have shown me that I can expect 24 volts at 1 Amp and 20 volts at 5 amps final output from the unit. I was a little disappointed that I can not lower the output below 1.2 volts but this is more of an aesthetic complaint as opposed to a practical one. Most of my experiments are at 3.3 volts and above and should I need a lower voltage for some reason I could always use another means to get it. The DC/DC converters came with small multi-turn 10K trimmers which I have removed and installed wiring harnesses to attach to the control panel potentiometers. Here is a schematic of the circuit I finally settled on for my coarse / fine control. Since the converter only looks at a variable resistance between 0 and 10K to control the output voltage I have tried to stay close to that parameter in the control design.


Coarse Fine Power Supply Circuit.bmp


Here is a picture of the inside of the control panel but unfortunately it doesn't do a very good job of showing the wiring to the controls. The Coarse Potentiometer will change the output voltage from 1.2 volts to 27 volts over it's 270 degree sweep while the Fine Potentiometer will change the voltage output by at the most plus or minus 1 volt over its 270 degree sweep. The fine control is more effective for voltage levels in the middle of the range and its effectiveness drops as the coarse control approaches its limits.




The meters come with very large shunt bars to be used to measure up to 100 amps but this would not be practical for this unit so it was necessary to make and calibrate some shunts more practical to this application. I used some 16 GA brass wire and soldered the leads from the ammeter to it. With a lab ammeter in series I was able to move and resolder the meter leads back and forth on the brass wire until I got the meter reading to match the Lab Meter. If you follow the large black and red wires from each meter you should be able to spot the brass shunt wires for each channel. Here is a picture of the unit completed and assembled:




I have also made a short video of me using the fine controls to zero in on a specific voltage on each channel. Besides my time which nowadays is practically worthless I have about $35 dollars invested in this unit thanks to my sons for the donation of the two scrap dental camera units and cheap Chinese electronic modules and meters.



I also want to thank my friends on E-14 for their excellent and inspiring suggestions for adding the fine voltage control.