DIY – Page 6 – VoltLog

Designing A T10 LED Bulb With Constant Current Driver | Voltlog #319

Welcome to a new Voltlog, in this video I’m gonna walk you through the design process of this small but important piece of circuit which is a T10 LED based lamp. We’ll go through why I need a T10 LED lamp, why I made the various design choices and in the end we will test the resulting PCB to see if it functions according to our specs.

So I started thinking about designing my own T10 LED bulb. First I picked some high quality Cree LEDs that would fit the purpose. These have a forward voltage of 3V, they have a high CRI of over 95 and I can pick from a variety of output colors, after a bit of experimenting, I ended up going with 4000K which is this nice neutral white.

Next I had to figure out how to drive these at constant current and really here I needed to decide between a switch mode drive and a linear driver. A switch mode driver would be more efficient, generating less heat but it will need more parts, which would increase the cost and complexity so in the end I went with a linear driver, which is the AL5809, this is a nice two terminal device, it comes in a variety of current presets and you just place it in series with your LEDs.

There is one constraint though, it needs 2.5V over the LED string voltage to operate correctly. That means that with our 3V rated LEDs, we can place two of those in series, plus the driver, which raises the min voltage to 8.5V. It might be possible to add a third led on board to bring it up to 11.5V and it would still work in the car but installing that 3rd led would be difficult so I decided to stick to just 2 LEDs.

CH341A Programmer 3.3V Fix | Voltlog #318

Welcome to a new Voltlog, if you watched the previous mailbag video I showed getting this CH341 programmer tool for EEPROMS and FLASH chips and my viewers were quick to point out this actually has a design problem regarding the voltage levels it puts out. So it appears there is a 3.3V regulator on board for providing power to the eeprom or flash chip via the zif socket but the CH341 itself is running at 5V which means it outputs 5V signals on the SPI bus.

Someone has reverse engineered this little board and put together a schematic which clearly shows pin 28 VCC connected to 5V but I want to make sure nothing has changed and this still present. And checking with a multimeter, pin 28 is still connected to 5V USB.

Connecting a 3.3V rated part for programming might result in damaging that part. You might get lucky, you might have a chip that will tolerate that but in general that’s bad practice and so we’re going to attempt to fix it by simply cutting the 5V line going into the CH341 pin 28 and supplying it with 3.3V from the onboard regulator instead. And the datasheet also specifies that V3 pin 9 should be connected to VCC and 3.3V as well.

The Curious Case Of A Very Sensitive PIR Light | Voltlog #315

This is a portable PIR activated LED light, it ships with some double sided adhesive backing and it’s pretty convenient to use around the house or when going camping I quite like it. It works with 4 AAA batteries and personally I use this inside a wardrobe to help me find stuff. If positioned correctly it turns on as soon as I open the wardrobe door. Because I like it so much I decided to order another one to place in a different closet for the same purpose. I don’t know if I ordered from the same seller, probably not, I just went on Aliexpress, did a quick search and ordered another one.

And here it is, after a long wait it finally arrived, at first sight they look identical, apart from a minor difference in the LED output color but that’s to be expected because whatever they can find at the market that particular day right? But in practice I also noticed something different, the new one is much more sensitive in picking up movement, to the point where it turns on even if it’s behind a closet wall and I walk past by, or even behind a wall. I know it doesn’t make much sense, because these are supposed to be infrared sensors so they can’t pass through walls. But it must be bouncing the IR rays off the walls and it’s so sensitive that it is picking up those reflections, I would have no other explanation for this but if you do, let me know in the comments.

Identifying & Replacing A Fake FTDI FT232R Chip | Voltlog #314

Lately I’ve been getting a few of these serial interfaces from aliexpress, this is just a RS232 to TTL level converter while this one is a USB to RS485 interface and it’s great that we can buy these for cheap for we are getting exactly that, something cheap. This converter chip is likely not of the best quality and not to mention this FT232 chip is certainly not a genuine one.

There is nothing worse than having to deal with communication issues and debugging your tools instead of the actual project you are working on so I’ve decided to replace these chips with some genuine, good parts because the pcb and the rest of the circuit should be fine as long as we have a good conversion chip in there.

You might be wondering if there is any way to tell for sure that you’ve got an FTDI FT232R clone. There are at least 2 methods that I know you can use. First one is to install an older version of the driver which FTDI released in 2014, after plugging your fake FTDI depending on the driver version it will either disable the FT232 chip by writing something to it’s internal EEPROM or it will modify it so that it shows a custom alert message over serial instead of your actual data. I don’t like this method, because there is a better, simpler way, you can simply check for the serial number of the FT232 which you can get from the device manager under windows or by using the FT_Prog utility. If your serial number is A50285BI or 00000000 then you most certainly have a fake chip because that is a popular serial number which is written to fake chips. They don’t bother changing the serial number on the fakes, they mostly write the same number or set it to zero.

Choosing, Placing and Routing A Shunt Resistor | Voltlog #313

If you are looking for the correct way to select, place and route your current measuring shunt resistor on a PCB then stick around as I will be sharing a bunch of tips & tricks that I learned while designing and building my own boards over the past years.

A shunt resistor can be used in multiple ways but for the purpose of this video we are interested in using it for measuring current, this means it needs to be inserted in series with the device under test. It can either be inserted in a high side configuration as shown in this example but it can also be inserted in a low side configuration. These two typologies have their advantages and disadvantages.

Here is a side by side comparison, when used in a low side configuration there is a ground offset which depending on your application may or may not pose problems. When used in a high side configuration, you must use a differential input amplifier for sensing the voltage drop while in a low side configuration you can get away with single ended.

Another thing that might be important, in a high side configuration you have the ability to detect a short on the load, while on the low side configuration you can’t do that. Also in a low side sensing configuration, you are not able to sense and account for additional leakage currents through stray secondary paths from the load to ground. This list might not be complete, there might be other differences to consider, these are just the ones that I had to consider in my designs.

But let’s say you’ve figured out the topology you want to use and you are now faced with choosing your current shunt resistor. Can you just pick your typical metal film 1ohm resistor from your favorite distributor? You can, but you won’t get the best results. Probably the most important factor you want to consider when choosing your shunt resistor is the temperature coefficient or the temp co as engineers like to refer to it. This will tell you how much your resistor value is going to change with variations in temperature.

Every system you build will probably see a variation in temperature at board level and so if you calibrate your system to calculate the current for a given resistor value and that resistor value changes with temperature, you are going to introduce significant errors in your measurement.

ScopeShunt Visualising The Current Waveform With Your Oscilloscope | Voltlog #310

We usually use an oscilloscope for visualizing a voltage over time but sometimes it’s also useful to visualize the current waveform over time. The right way to do it is to get a current probe which can sense the current and convert that to a voltage that the oscilloscope can display however such devices are pretty expensive, they can be around $1000 even for an entry level one like the Rigol RP1001C which is only rated up to 300KHz bandwidth.

But we can improvise something for a much lower cost and it should allow us to visualize the current waveform on the oscilloscope. You’ve probably seen me use a shunt resistor when testing power supply to take a look at the current waveform. Because as you know passing a current through a resistor will generate a voltage drop.

That voltage drop is directly proportional with the passing current and with a round value resistor we can have an easy to use transformation ratio between voltage and current. All we have to do is o introduce this resistor inline between our power supply and the device under test

For example if I have a 1ohm resistor, we have a 1:1 ration, for each mA passed through that resistor we will have 1mV of voltage drop that our oscilloscope can display. Such a circuit will of course have it’s limitations, for example it won’t work very well when testing low voltage low power devices because our resistor will introduce a burden voltage, which will drop our supply voltage to the device under test. This is also not an isolated measurement so it might not be safe when connected with higher voltage circuits.

But there are still a lot of scenarios where you could use this successfully on the electronics workbench so it might be worth building something like this. I want to make this nicer by building it inside an enclosure with the required bnc connector for connecting to the oscilloscope and 4mm banana plugs for passing the current through. I picked this small aluminium enclosure which would be enough to house the resistor, actually the resistors, because there are several advantages to using multiple resistors in parallel.

Alternative to this simple shunt resistor measuring method include the Joulescope which is a fully featured dc energy measurement test instrument with incredibly wide dynamic range that allows you to capture the smallest currents next to a jump to a higher current. I reviewed the Joulescope in Voltlog #211.

Reverse Engineering Gopher NPS-1601 Front Panel Circuit | Voltlog #305

In this video I’m gonna show you how I reverse engineered the front panel of the Gopher NPS1601. First I’ll show the method I used for reverse engineering and then I’ll show the results I got This is in my opinion the best power supply you can get in the $50 range and if you would like to know more about this, I did a review in Voltlog #255.

Why did I decide to reverse engineer the front panel of this power supply you might ask? Well it was a topic I had on my todo list ever since I got this unit because I feel like the user interface on this is not as nice as it could be if it had a color TFT display to show more info than what you get on those small 7 segment displays. And I’m not just talking about improving visibility cause you could do that just by using brighter 7-segment displays, I’m talking about showing more data like: calculated wattage or capacity, showing set current and voltage at the same time with actual current and voltage you know the kind of info you get on the more expensive Riden power supply but hopefully in a better color scheme.

Voltlog #297 – How To Parfocal A Trinocular Microscope

Welcome to a new Voltlog, in this video I’m gonna show you how to parfocal your trinocular microscope, this means getting both the eyepieces and the third camera port in focus at the same time and maintaining them in focus while adjusting through the zoom range of microscope from low to high.

This is extremely useful if for example you are using the microscope through the eyepieces, soldering on some board and at the same time you are using the camera port for recording or live streaming. Once you do this procedure, the camera and the eyepieces will be synced in focus even as you adjust a different zoom level but with some limitations, as I will show in a moment, these are the result of the camera optics which will limit the effective zoom range in focus.

Now if you are using a 0.5x barlow lens like I do and like most people doing electronics will be using, you need to consider the fact that the working distance is about 16.5cm between the barlow lens and the work surface. So keep that in mind, because the adjustment we will do later will be happening around that working distance.

Voltlog #296 – Microscope Camera Focus Issue Fixed!

Welcome to a new Voltlog, as you may remember in Voltlog #282 I got a new trinocular microscope and then in Voltlog #292 I got a new camera because the one shipped with the microscope had issues. Banggood till this day did not replace the camera which I believe is defective, they kept asking me to try these various settings in the camera menu to improve the brightness but none worked.

Now to get a good image and have both the camera in focus and the eyepieces you will need some kind of reduction, an optical adapter between the camera and the microscope port. So first I got this SZMCTV 0.3X adapter, after reading some reviews who were claiming this will give the best field of view.

After installing this adapter I noticed the field of view was almost the same as through the eyepiece which is great but I couldn’t get it in perfect focus with the limited adjustment available on the adapter and I was getting this tunnel vision which I couldn’t quite explain but it was obviously not what you want to see.

And so I got this newer wide body 0.5 adapter and installed it on my microscope and the problem was solved I could now get my camera into focus at the same time with the eye pieces but with a loss in the field of view due to the 0.5x adapter. Getting the SZMCTV 0.5X adapter would probably also work, that’s a bit cheaper, but I went with the recommendation of getting the newer type of adapter with better optics.

Voltlog #295 – Weekend Update

Welcome to a new Voltlog, this will be just a short weekend update video, to let you know what I’ve been doing in the past couple of weeks. So I’m staying at home and only going out for important stuff, which gave me more time to tinker and work on projects. I hope you are doing the same because social distancing is a good way to slow the spread of this disease.

One of the things I’ve been thinking of doing for a long time is a bench tool holder, the kind that will hold my tweezers, soldering iron tips, flux syringes, stuff like that. So far I’ve been keeping them in one of these pen holding cups but they fill up and it’s hard to grab the thing you need quickly. So I’ve started designing my own in Fusion 360, I included 6 deep pockets, 2 shallow ones in the front and 5 tubes for holding syringes.

The next thing I did with my 3d printer is related to the current pandemic, as you may know, medical staff in many countries which have been hit hard are lacking protective equipment, they just don’t have enough mask or suits or face shields, not to mention enough ventilators which are desperately needed for those in critical condition.

In parallel I’ve been designing the battery monitor circuit for the power drill I reviewed in Voltlog #285, this is the first revision of the PCB, the circuit works as expected but I’m still waiting on some parts to be delivered and I’ll do a separate video on this subject. It probably needs a second revision of the PCB because this one doesn’t quite fit in, the way I thought it will.

I also started designing a T10 automotive LED bulb. It’s using a linear constant current driver and some decent LEDs, because what I’m currently using in my car are the aliexpress bulbs which kinda start dying after a few months of usage. I’ll probably place an order for this panel today but who knows when I’ll receive it.