DIY LB7 FICM Repair – Part 2: Keeping your working FICM alive

I didn’t think I’d be following up so soon after my last post exploring how to disassemble an LB7 FICM and examining the reparability since it was a failed unit. That FICM ended up being damaged beyond repair due to multiple layers on the PCB getting burned by the leaking electrolyte. This observation lead me to recapping my working FICM shortly after realizing that failures should be avoided at all costs. There are numerous people on the internet offering reman FICM’s at a substantial cost, but my searches revealed no information about reconditioning FICM’s on your own. The purpose of this post is to document how I overhauled my FICM to help others that may want to take on this task on their own. Hopefully this helps.

Tools and parts needed to recap the LB7 FICM:

Disassembly

LB7 FICM Just removed from my truck

The FICM I removed from my truck was much cleaner than the damaged one I had disassembled for the previous post. As a result, the four T15 screws that need to be removed are much easier to see in this picture:

Bottom of LB7 FICM, the four T15 screws that need to be removed circled in red

Once the screws are removed it’s time to remove the cover. The cover is sealed to the cooling plate that will require some effort to separate. In order to get the cover off you will need use a flat-blade screwdriver to lightly pry the cover off of the cooling plate. On the connector side of the FICM, at each corner there is a gap between the two halves that allow you to get leverage to pry the FICM open:

Lightly pry in this area to start separating the cover from the cooling plate

Once the sealant starts to let go the cover will pop right off. Once the cover is removed you can examine the board. Thankfully, mine was in good shape with no signs of leaking electrolyte. Since there’s no record of failures of the 100V film capacitors or the 40v polymer capacitor at the bottom right, the only capacitors I will be replacing are the two pairs of electrolytic capacitors nearest to the connector on the board.

Now the fun part </sarcasm> lifting the board off of the cooling plate. The board is adhered to the cooling plate with a thermal transfer silicone. Much like removing the upper rocker cover on the LB7, the key here is to get a small spot in the silicone to lift and the rest will come right up. In my last post I was able to use a small 90 degree pick to start lifting the board from the corner. That did not work well for lifting this board. I ended up going against my own advice in my last post and using the connector to lift my board off of the cooling plate. By grabbing the cooling plate and using my thumbs to lightly push up on the connector I was able to get my board to start lifting. Once I got the board to lift it peeled right off of the cooling plate. Once the board has been removed from the cooling plate you can bring it to your work bench to start the desoldering process.

LB7 FICM board ready for desoldering.

Desoldering

By this point the solder on these board is over two decades old. It will have some layer of oxides on it and will require some amount of work to get the solder to flow properly to remove. One interesting issue I ran into on both this board and the board from the previous post is that there are multiple layers of large traces connected to these capacitors. As a result, my 40 watt Weller iron had difficulty melting the original solder without the heat getting wicked by the traces. This situation is less than ideal because now you’re heating components and not the solder. After running into this issue, I highly recommend using at least a 60 watt soldering iron to reflow the old solder and solder the new capacitors to the board. Thankfully the desoldering iron I used was able to fully melt the solder quickly and remove it. Before going any further, lets identify the points we’re going to rework and desolder. Also now is a good time to use a hobby knife to slice the silicone between each pair of electrolytic capacitors we will be replacing.

Bottom of the board, red circles showing reflow and desoldering points.
Carefully slice the silicone between each pair of capacitors. Take care not to place a cut anywhere on the board. This is needed to remove each capacitor individually.

In order to get the solder to flow well enough to get a clean desolder I started by melting fresh solder into the existing solder joints. Once I got a little bit of solder to flow in I got onto the joint with the desoldering iron to suck out the solder. Since my soldering iron was inadequate for this task the desoldering iron only removed about half of the solder from the joint. I then had to flow more fresh solder into the joint to get to the solder that didn’t get reflowed from my first attempt. It took about three reflow / desolder attempts per joint which is less than ideal as it applies too much heat to the board, traces, and solder mask. Using the right tools here will prevent damage from overheating the board and / or surrounding components. Be absolute sure you have a clean desolder job before attempting to remove the capacitors. Any attempt to pull on the capacitors with even the slightest amount of solder still affixing the legs to the traces my result in pulling up a trace and turning your one hour job into a five hour job. Once the old capacitors are desoldered and removed, thoroughly clean the pad and surrounding area with isopropyl alcohol. If your board has some leaked electrolyte on it, you may also want to neutralize that with a basic substance before cleaning with alcohol.

Top of board with capacitors removed and pads cleaned up.
Bottom of board with capacitors removed and pads cleaned up.

Install New Capacitors!

The hard parts are over. With the board all cleaned up you can install the new capacitors. These capacitors won’t fail like the old ones and should outlast the vehicle. The new capacitors have a black dash on the top to indicate the negative leg. These all face away from the connector on the board. Even when soldering in the new capacitors I found my 40 watt Weller wasn’t up to the task. Again, using a good 60+ watt iron here will properly flow the solder in the joint the fist time. I ended up having to use my desoldering iron to flow the solder into the joint once I got it on the board. Once all four caps are soldered in, it would not be a bad idea to apply some non-conductive silicon to each pair, as they were before. You can use the thermal glue here we’ll be using to reapply the board to the cooling plate later on.

Board with new capacitors installed, prior to applying silicone between each capacitor

Reassembly

Prior to reassembling the FICM, we need to clean the remaining silicone off of the bottom of the board and the cooling plate. For cleaning the board and cooling plate I used a tungsten carbide gasket scraper. When cleaning the board itself, be sure not to put any cuts into the board itself. Unfortunately I forgot to take a picture of the cleaned up cooling plate, but make sure the surface that bonds to the board is clean and flat. Isopropyl alcohol was used after all of the silicone was removed to do a final clean on the bottom of the board and cooling plate were new thermal glue will be applied.

Cleaned up thermal transfer areas on the bottom of the board. Sadly, no picture of the cooling plate was taken, but you get the point here.

Reassembling the FICM will happen the reverse of disassembly. There’s no way to properly align the board directly to the cooling plate. Instead, the cover has the alignment bumps that align the board. Note alignment bumps in the cover circled in red below. Now would also be a good time to inspect the vent cover. This should flow a little bit of air, allowing the internal components to breathe.

Alignment pins in the top cover circled in red. These will align the board during reassembly. Also inspect the vent plug while here.

I do not know what was originally used to seal the case to the cooling plate, but I also know it’s not going to come off and trying to get it off will likely cause more damage than what the effort is worth. When reassembling my FICM I used a very small amount of Permatex Gray RTV to go over the grooves where the old sealant was. I started by laying a small bead in the connector area of the cover. Once the connector area had a small bead of RTV I placed the FICM board into the top cover, being sure the connector sat down in the sealing groove and the board was aligned with the pins in the cover.

Small bead of RTV in the connector groove to reseal the connector area of the cover.
Board set in the cover with the connector properly aligned in the cover groove and board aligned to the pins on the cover.

Once the board was placed into the cover it was time to apply the thermal transfer glue and a small bead of RTV around the rest of the sealing surface on the cover. After doing some testing on the bench, I found the GENNEL thermal glue to be quite effective at heat transfer and maintaining a good bond. This is essential as the board needs to be well bonded to the cooling plate with good thermal transfer characteristics to keep the driver FETs cool. The video below shows a test I did prior to using this glue to verify the heat transfer. A thermal camera was used to observe a heat sink glued with the GENNEL glue to the hot side of a Peltier device. When energized, the heat sink thermal image was observed to make sure it was rising proportionally across the surface of the heat sink without the surface of the Peltier getting considerably hotter than the heat sink:

Since the glue is non electrically conductive I was pretty liberal in applying a large bead to the thermal transfer surface of the board. I did not care to spread it across the surface. When cleaning the board with the carbide scraper I found the board is not perfectly flat. By applying the bead and allowing the glue to spread when the board is pressed to the cooling plate it will fill in the uneven voids between the board and the cooling plate.

Bead of thermal glue applied to both thermal transfer surfaces and small bead of RTV around the cover seal area. The thermal glue is not electrically conductive so if some spreads to components it will not affect them.

Once the thermal glue and RTV is applied it’s time to put the cooling plate on. Leaving the cover and board assembly as it is pictured above and line up the cooling plate with the screw holes and carefully lower the cooling plate to the board and cover, ensuring proper alignment the whole time. Once the cooling plate is set on top of the cover and board the screws can be installed. Slowly work the screws in going cross-corner each time you move to a new screw. Using this method will allow the thermal glue to properly spread without applying too much stress to the board or components as the glue spreads. It should take about two minutes of this progressive tightening to get the screws to the point where they are fully tightened. Once the FICM cases are tightened together clean up any RTV that squished out of the case halves or connector area. You are now done with reassembly. You may immediately reinstall the FICM, but allow 24 hours for the thermal glue to cure before driving the vehicle. Trying to drive the vehicle before the glue is cured my result in improper thermal transfer and overheating of the driver FET’s.

Cooling plate dropped onto the board and top cover assembly, all screws fully tightened.
LB7 FICM, recapped and reassembled.

Hope this helps anyone who may want to attempt to recap their LB7 FICM on their own.

DIY LB7 FICM Repair – Part 1: Dissecting a deceased specimen

A Facebook post in a Duramax group recently alerted me to common failures of the Fuel Injection Control Module (FICM) on my LB7 Duramax. Some of the posters in the group suggested that all LB7 FICM’s are a ticking timebomb whereas this post in the DuramaxDiesels group suggest that the use of superior components in the LB7 FICM should not be cause for concern. I wanted to do some more research to see if this is a cause for concern, because OEM FICM’s are not readily available anymore and the fly-by-night remanufactured units are questionable at best at a high cost ($1000+ without core exchange at the time of this writing. From what I have read, LB7 FICM failures are well documented, but they are not as widespread as FICM failures on the LLY engine that was short lived after the LB7. The common mode of failure on both controllers is the electrolytic capacitors begin to leak, slowly damaging the board until either a short occurs or the traces erode away, losing continuity across a trace. It’s a well known fact that the electrolytic capacitors used in automotive applications have a finite lifespan, roughly 25 years. My 2002 Lb7 is quickly approaching that age, so it’s time to get a gameplan together to keep my truck running as well as it does now.

This issue with electrolytic capacitors is nothing new. A friend recently informed me about the Capacitor Plague that affected many electrolytic capacitors between 1999 and 2007, well within the window of when the LB7 FICM’s were made. A little over twenty years ago I performed many services repairing DSM ECU’s that were prone to the same type of damage. Those were only failing at 12 years of age. Since the issue was so widespread it was well documented how to perform the repair on your own and what parts were needed. Oddly enough, the same knowledge base doesn’t exist for the Duramax FICM’s. Aside from the thread I linked in the first paragraph of this post, really the only thing you find when researching LB7 FICM’s are a flood of individuals offering repair services. The purpose of this thread is to explore the LB7 FICM to assist the DIY’er in either repairing or refreshing their LB7 FICM to keep their vehicles running.

Before tearing into my working FICM I decided to buy the cheapest core on eBay, to see if I can repair it (assuming it was failed) and to figure out how to disassemble and reassemble it. Even if the FICM was beyond repair the information I collect along the way would hopefully help those with an adequate aptitude in electronics overhaul their FICM, or at least evaluate if their failed FICM is repairable.

I ended up getting a non-working core FICM on ebay for $120 shipped. It was a greasy mess, clearly the engine it came off of was not well taken care of:

2003 model year LB7 FICM obtained on eBay.

First was to get the case open. This was pretty simple. There are four T15 torx screws on the bottom of FICM that hold the case together (circled in red):

Four T15 screws circled in red. Remove these.

Once the case screws are removed you will need to pry the two case halves apart. Avoid prying on the connector itself. There are some places around the case that you can get a flat blade screwdriver between to start to break the seal. Once the seal is broken, the case halves are easy to separate. On this particular FICM the seal was a black goo that smelled like Permatex RTV, but wasn’t cured like RTV. I do not know if this was someone’s attempt to reseal this FICM or if it is like this from Bosch, but avoid touching this stuff. It will stick to anything it touches and it is a BPMF to get off:

Black seal goo found on my FICM. Don’t let anything touch this stuff. It will stick to it and never come off!

Once you get the cover off you will notice one or two prominent things on the board: If it’s failed you will see a black crater under the failed capacitors. The other thing is the board is glued to the lower part of the case:

LB7 FICM with the cover removed. Note the failed area under the blue capacitors on the upper left side of the board. The board is retained with a heat transfer silicone. Its consistency is similar to cured RTV. You can see pieces of excess I started pulling up on the right side of the board.

If you are needing to repair or overhaul the board it will need to be lifted off of the case. It’s glued to the case with a thermal transfer silicone that has a set identical to cured RTV. I found that removing the board from the lower case is much like removing the upper valve cover on the LB7, or really any other large area surface that’s held together with RTV. Do not pry on the connector, it is only retained by the soldered pins. I found that trying to pry with a flat blade screwdriver started to chip the edges of the board, which may or may not contain traces. After trying a few things I found the best way to lift the board off of the lower case was to use a small pick and start probing around the board for a weak spot in the silicone. Much like valve covers and oil pans, once you get one area lifted the rest comes off very easily.

Hard to see, but using a small pick and very very light pressure, I was able to get a section of board to lift. Once this corner came up the rest of the board popped right off the silicone.

Once the board came off I evaluated the damage from the capacitors. A solid repair did not look promising:

Extensive damage from leaking capacitors. The electrolyte has eroded through the capacitor labeling, the plastic pin retainer for the connector, and has eaten clear through the solder mask on the circuit board.
Damage to the bottom of the board. The solder ball on one of the connector pins and lack of solder on one of the capacitor legs suggest someone has already been in here, but it’s also entirely possible that was melted away from the short that occurred from this damage.

With the board off I moved on to removing all of the electrolytic capacitors on the board. I started with the non-failed capacitors to evaluate where the high current traces on the board connect, to provide a frame of reference for the damaged side. One thing I will note, this is a multilayer board and most of the high current traces are connected to the capacitor legs. You will need a HOT and powerful iron to get through the solder on these capacitors. My 35 watt Weller station barely got the task done to reflow the joints prior to desoldering. There are a lot of small components and traces around the capacitor legs. You have to be absolutely careful when reflowing and desoldering the capacitor legs. If you are not absolutely confident in your reflowing and desoldering abilities, it would be wise to practice on a junk board, any junk board prior to operating on your FICM.

Once I got the failed capacitors off of this FICM, the level of damage started to become apparent. If you look closely at the crater, you can see two small traces going under the large trace that was eroded away. This is a multi layer board. The damage on this FICM was so substantial, the electrolyte had torn through three of the layers. The small traces visible was the fourth layer. I started to clean the area around the damage to see if this board was salvageable. Once I started clearing away the fried board, it was apparent that this board was not salvageable.

Extent of damage form the leaking capacitors.

After discovering the extent of damage on this FICM was beyond any reasonable repair, I stopped moving forward with working on this one. Hopefully up to this point I’ve provided some useful information for those looking to potentially DIY their own FICM repair or overhaul. Those with working FICM’s it would be wise to change at least the four electrolytic capacitors prone to leakage. This will at least guarantee that you won’t experience the common mode of failure on these modules.

With that said, I will be at least opening my FICM in the near future to inspect its condition to determine a course of action. More to come in another post on that topic.

Until then, Here’s some close-ups of the Electrolytic and power Film capacitors on the LB7 FICM:

More to come later……

Golf Cart STL’s Now Available

Over the past year I have been designing small parts in CAD for friends and my own golf cart. Since there seems to be some demand for these I have decided to make them available here, in case they are of use to anyone else.

These parts are mostly for Club Car DS, but there is one charge port adapter for EZGO TXT as well. The link to each part has a brief description of what the part is for, a 3D preview of the part, and some pictures of the part in use.

Click here for Golf Cart STL’s

Enjoy!

Home Energy Redefined: The Beginnings of the NSFab Home Microgrid

This post introduces my custom designed and built home microgrid. In the near future I’ll be creating posts that detail some of the inner workings of this amalgamation of off-the-shelf parts. None of it will make much sense if the bigger picture isn’t introduced to those unfamiliar with it. First I should probably address a part of the title you may be wondering about. What is NSFab? I made my first internet presence in 2004 with www.nsfabrication.com selling hand fabricated parts for DSM’s and Turbofords. While it’s been well over a decade since I’ve owned that business I still use the NSFab moniker for anything that I design for personal use or public domain. I designed and built the topology, it’s an NSFab piece.

The current NSFab badge worn by parts I design for personal use or in the public domain. A clear copyright infringement of NEMA’s Mr.Ouch. It best represents what NSFab has evolved to in its 23 year existence.

Everything starts back in May 2022. I’m riding to back to San Diego from a day of go kart racing at the now extinct California Speedway for a friends bachelor party. Flipping through FB I find someone selling a bunch of 230 watt solar panels they obtained from a building demo job. I shot an offer of $800 for twenty of these 230 watt panels. The seller accepted the offer and the next day I brought home twenty 230 watt panels. At this time I had no idea what I was going to do with this 4.6kW of solar, but I got it for cheap so I can worry about that later. These particular panels were right a ten years old. Solar panels live a hard life so there is some amount of degradation that occurs with them over time. The first thing I needed to do is check if these things are even worth using.

Testing used PV panels

When looking at the data tag on a solar panel (we’ll call it PV panel moving forward) you will find two ratings that allow us to quickly evaluate the condition of the panel. The first rating is the irradiance for STC (Standard Test Conditions). The second is the Short Circuit Current or Isc.

Data tag off one of my PV panels, for reference.

A quick primer on irradiance. Solar Irradiance is measured in Watts per square meter (W/m²). Lets say we have a solar panel that had a photovoltaic area of one square meter. This theoretical solar panel is 100% efficient. We have some instrumentation that shows solar irradiance at the panel is 800W/m². If that panel was facing directly at the sun it would produce 800 Watts of electricity. Solar irradiance is the potential of solar energy across a given surface. We measure this energy potential in Watts per square meter. Solar panels aren’t very efficient, so for a given surface area they generally only turn about 22-28% of that potential into useable electricity.

A vast majority of solar panels are tested at 1000W/m² irradiance. This means all of the STC ratings for that solar panel are measured at 1000W/m² energy input. Even on the sunniest days here in SoCal seeing 1000W/m² on a clear day is unheard of. Anything higher than high 800’s at solar meridian near the summer solstice on a clear day is not very common here. What this means is on a normal clear day, the panel will likely produce less than the STC rating due to the input energy being less than the nameplate test conditions. To test the panel I need to know what the solar irradiance is at the time of testing. I have a full suite weather station here at the house that tells me the solar irradiance in real time. A quick calculation to determine the percentage difference between the STC rating and the actual solar irradiance will give me an idea of what the rest of the ratings should look like. If the solar irradiance is 80% of STC conditions, then I would expect the output of the panel to be 20% less than nameplate, plus or minus 10%. Now that we know what to expect, an easy test of a solar panel is to short the outputs together and measure the short current facing directly at the sun. Most PV panels have an Isc rating, or Short Circuit Current rating. If the measured short current is within 10% of the corrected Isc from the STC rating the assumption is the panels are good. Thankfully all of my panels tested within the corrected Isc rating. Now I know I have some PV panels I can do something with and that was the motivation I needed to move forward.

Building the system

I had zero interest in the traditional grid-tied residential PV scheme. In net-billing areas feeding PV energy back into the grid has no benefit to the equipment owner. When I started to put this system together I was still a field engineer for a company that serviced generators and switchgear. My knowledge was incredibly strong in traditional power generation but not alternative energy solutions. Before investing a lot of money into equipment I wanted to feel out the potential for storing PV generated energy and using it at a later time. The test bench for this was simple: a single 230 watt panel sitting up on my patio pergola feeding into a Victron 20 amp MPPT charge controller. The charge controller maintained a regular old group 24 automotive battery that fed a 300w modified sine wave inverter. Stuff I had laying around. I used this arrangement to power a 55″ LCD TV. I just let the TV keep running on a public broadcasting channel to evaluate how long the TV could run on such a modest system. After 72 hours the TV kept running. Multiply this potential by twenty and it seemed as if I could run my entire 744 square foot house (yeah, I know, big pimpin’) on the amount of PV potential I had. And thus, the first iteration of my microgrid was born:

The birth of the NSFab home microgrid. June 2022. A simple energy storage system.


The first Microgrid was simple. It was really less of a microgrid and just a larger version of the test bench. The small blue box in the upper right corner of the module is a Victron MPPT 150/70 charge controller. This took the nominal 135 volts DC from the full PV array (4s/5p configuration) and turned it into the 48 volts needed to charge and maintain the battery array. The larger blue module is a Victron Multiplus II 48/3000 inverter. The inverter had a grid connection that allowed the inverter to switch to grid when the inverter became overloaded or battery SOC was low. The output of the inverter was wired to a single circuit in my house: the living room. This was really just lighting and my network gear. The inverter runs a 48VDC input which at the time was 4 heavily used group 24 batteries wired in series. This arrangement only stored about 1kWh of energy which was fully charged by 11am and fully discharged by 6pm. PV production was curtailed most of the day due to the batteries being fully charged at such an early time and such little draw on the inverter. This system clearly had room to grow. The goal here was to maximize what was there before expanding. With exception of the garage and kitchen, due to having single loads greater than the inverter capacity, the rest of the house was transferred to this single inverter. 99% of my daily usage was now being covered by solar power during the day and still curtailing PV power before midday. I was at a point where I was at inverter capacity and still had plenty of headroom in PV capacity. Time to store more energy. That’s when the next iteration of the system was assembled:

Second iteration of the NSFab microgrid. July 2022.

The next iteration of the system took place about a month later in July 2022. This saw the replacement of the group 24 batteries with brand new 8D deep cycle batteries, a Victron Smart Shunt for battery SOC monitoring, manual transfer switch to allow switching between generator and utility power, and a Victron Cerbo GX that added cloud monitoring and microgrid functionality to the entire system. The new 8D batteries increased storage capacity to 4.2kWh of storage. This allowed the house to have enough energy storage to make it through the peak times of our utility supplier between the times of 4-9pm. This iteration allowed the house to run completely independent of utility power from ~9am to ~9pm every day, the time frame where the house saw the most energy usage. Having actual data produced by the Cerbo GX and Victron’s VRM interface allowed for detailed data collection that I could analyze to start really refining the system.

The home microgrid gets kicked into high gear

A large change happened in my life shortly after building the second iteration. A change that I didn’t realize would rapidly accelerate the development of this system. I had landed my dream job at ComAp as an Application Engineer in August 2022. Five weeks into this job I attended the RE+ trade show in Anaheim, CA. This show was a massive eye opener for the possibilities of building the systems that I was already developing prior to coming into this industry. I felt like I was just given a cheat code to design and build something that would actually sustain my house 24/7. Something that was self-healing, zero maintenance. The goal post was moved…. Significantly. It was time to put some investment into this system and make it shine.

RE+ September 2022. 5 weeks into the job at ComAp.

Shortly after RE+ came the most money I’ve put into the system yet. Detailed analysis showed I was still putting a lot on the table in terms of PV that was being curtailed after the batteries were charged. That was free energy being thrown away. Two short months of using FLA batteries cycling at 50% SOC also showed significant degradation in battery capacity. After attending RE+ I knew lead acid batteries had no future in the energy storage arena. It was time to change to a different battery chemistry. Lithium Iron Phosphate, LiFePo4, or LFP was the future of safe and cost effective home energy storage. This saw the last iteration of what I considered “Version 1” of the NSFab home microgrid:

The final iteration of the NSFab home microgrid in the Version 1 form.

This final iteration of “Version 1” involved removing the 8D FLA batteries and installing two EG4 LL server rack batteries. This brought the usable storage capacity up to 8.1kWh. This was a game changer in the system. When I started practicing energy conservation techniques this iteration of the system I was able to run the house entirely grid independent (off-grid for those that want to be pedantic about it). This addition was an eye opener. There is zero reason, ZERO REASON to use any sort of lead-acid battery chemistry in energy storage systems. Being able to use 80% of the battery capacity in a solution that took up half the space of the previous FLA solution, providing nearly twice the capacity for little increase in cost is a no-brainer. Looking at the timeline in this blog, I had gone from acquiring solar panels in May 2022 to an efficient, fully functioning microgrid in October 2022. Five short months of evolution. This system remained in this configuration until it all had to come down in the summer of 2023 due to construction on the house where the microgrid was mounted. The last large investment came in “Version 2” of the system in the summer of 2023. That’s where I’m at today:

The current iteration, Version 2 of the NSFab home microgrid. October 2023.

Pictured above is Version 2 of the NSFab home microgrid. This iteration was built in October of 2023 and has remained largely untouched since then. The detail of this iteration will be covered in posts of their own. The last large investments into this system is the addition of a second inverter to create a 240V split-phase arrangement, and a control cabinet that houses the Cerbo GX and higher level controls. This was all the result of telling myself “If I have to take this system down, I’m going to put it back up right.” But we’ll get to that at a later time. Before concluding this post, lets take a look at the control cabinet:

Computing power within the control cabinet. Pretty, isn’t it?

There’s a lot of higher level control and computing power in that cabinet. Again, that’s all for another post. I got to this point simply by looking at a FB post while riding home from a friends bachelor party. More to details to come soon….

16-bit Nick goes beyond 16 bits with SSL

One thing that really annoyed me about getting the 16-bit Nick blog online is that I started out using only http. Since I only post from the server that the site is hosted on I didn’t pay much mind to the secure aspect of browsing the blog. After all, how unsecure can reading my blog be? Anyone can read it. However, being that it’s been nearly a decade since I’ve done any font end based work I glossed over the fact that most web browsers inherently try to establish SSL connections with every website it tries to connect to. This lead to a slew of issues connecting to the site if anyone tried to come here directly from one of my domain names. I want my blog to be easily readable and lack of access means it was anything but. For twenty some-odd years I had all of my domains registered with Powweb and until I ended hosting services in 2018, they were my hosting provider as well. In the early days I found Powweb to be easy to use and for the most part pretty reliable. As time went on and the hosting company passed ownership around I found them to be very unreliable and support was non existent. If it weren’t for the downfall of Powweb’s customer service, I likely would not have ever stopped web hosting services. I have more details about all of that in my blog post Coming Full Circle, if you need to catch up. At a time where webhosting didn’t make much sense anymore I just couldn’t justify the time to search for a new host, backup the entire server, and migrate it elsewhere.

Now that this new blog is alive and well, it was time to make some changes. Powweb at some point in time became iPage and let me tell you, the reliability and customer service isn’t any better. All of my domains were still registered with Powweb/iPage so I still have to use them to change the DNS records to point to the server that 16-bit Nick is hosted on. Last year when I had the idea of making training videos I wanted to grab the 16bitnick.com/.net/.org domain names as well. Since Powweb/iPage is currently the registrar of the rest of my domain names I went ahead and registered the 16bitnick names through them, just to have for future use. Getting those domains pointed to the 16-bit Nick server turned out to be a pain, and it was the point of realization that if I wanted this to work well and reliably, I had to get away from Powweb/iPage. Without getting into too much detail, the 16bitnick domains were placed on ClientHold for no known reason. It took Powweb/iPage about 72 hours to come back with a response that basically read “oopsie” and all of the sudden my domains were working again.

Basically. If I was going to reliably run an SSL connection to my server, it was time to find a provider that can provide reliable service.

After searching around reddit (I find reddit far more reliable than Google for peer-review feedback) I decided to give Cloudflare a shot. I started by simply registering another 16bitnick domain with them, 16bitnick(dot)me to try them out. I was immediately impressed. Within minutes I started the process of transferring nicksalyer.com over as that’s the primary domain currently for this blog, everything revolves around that domain. That was yesterday. Today, I have the SSL certs installed and 16-bit Nick is now an officially certificated, secured website. Thanks, Cloudflare!

Ok, well, it wasn’t THAT easy. Recall above where I mentioned that it’s been many years since I’ve done this kind of stuff. Back then a hosting service dealt with all of this for me at a cost. Now I am managing my own physical server on it’s own 300/300 fiber connection. I’m just using Cloudflare as the DNS provider. I’ve never installed SSL certificates before, let alone from just a Linux command line. The learning curve was steep. A few times I about gave up. But this was a necessary step in making the 16-bit Nick blog easily accessible and readable. This morning I got it all figured out, thanks to the error logs in Apache. This is a significant step for not only my continued learning in life, but to make this place easy to access, read, and understand.

Coming Full Circle

Hi internet, it’s been a while. Not that anyone reading this right now read my blog years ago, but I look back on where I came from to how I got here and it illustrates how the internet has come full circle. From 2009 to 2018 I hosted a blog, among other websites, on one of the domain names that brought you here. When I started the Nixworld blog back in 2009 I had a focus on sharing all of my stories from traveling around the west racing cars to writing technical blogs of all the things I learned along the way. Even a blurb about my personal life here or there. Facebook was already three years into the public domain and Myspace was just a memory. Social media back then didn’t allow for much more than a small paragraph of text in a single post and you were limited on how many pictures you were allowed to store. People were given a small stage to a limited audience back then so it didn’t make much sense to try to use it as a knowledge sharing platform. Internet forums were a big thing at the time and while they were great as a knowledge sharing platform, the signal-to-noise ratio in the automotive groups was quite low. It was difficult to hold meaningful discussions without contributions from those that hang out at the top left of the Dunning-Kruger chart. I wanted to do something else that would reach the proper audience.

In 2006 I bought into a web hosting service and started nsfabrication.com to create a web site and web store for my business. Since I already owned server space it made sense to host my own information sharing platform of some sort. I tried hosting my own forums using phpBB. I quickly found that was less than ideal for what I wanted to do. The forum had all of three or four members if I recall correctly and it was mostly discussion about how I assembled the products I sold. I took the forum down and explored other options. During this time I also hosted a photo gallery on that server. While social media was gaining traction in the late 2000’s, as I pointed out above, you were quite limited on the amount of pictures you were allowed to host on your page. By hosting my own gallery I was able to overcome this limitation and store as many pictures as I wanted, only limited by the amount of server space I owned at the time. Blogging was quite popular in the 2000’s as it was a way of writing more than what was allowed on social media as well as being able to format the presentation any way you pleased. This seemed like what I was looking for in my quest to share knowledge. Even though I was capable of managing my own web server back then I was still pretty inept at developing my own PHP code. I needed to find a package or product that allowed me to host my own blog and build it how I wanted it to look. That’s when I found WordPress and Nixworld was born in the spring of 2009.

My first post on Nixworld was basically a “hello world!” and an introduction and disclaimer into my lack of confidence in sharing knowledge. Spring of 2009 is also when I started touring the west with Fair Enough Competition and racing Andrew Brilliant’s time attack Eclipse. Most of my posts were focused on sharing that journey, a behind the scenes on what it’s like to race cars out of your own pocket. Sleeping in cars and tents was the lifestyle back then, if I wasn’t rebuilding a DSM transmission on the side of the road. There was a lot of other automotive tech posts back then. There were two very popular posts that I found being shared across a plethora of platforms. One of the posts I made early on in the blog that took a deep dive into using HID retrofit lamps in standard refractor headlights, why certain color temperatures should be used, and why projector housings were superior if you wanted to use that type of lighting. The final and most popular post of Nixworld was published in the summer of 2016. This post explained the black magic that was measuring fuel injector deadtimes across an array of supply voltages in order to properly populate deadtime/latency tables in engine ECU’s for aftermarket and unknown injectors. After the summer of 2016 there were no more posts made to the Nixworld blog.

I left the full-time automotive industry in late 2013 to enter the industrial controls arena, working on gas turbine power plants and their control systems. By 2016 I was done with the automotive industry as a whole. I was burnt out, done with cars. I had zero motivation to work on my own vehicles. I removed my website at nsfabrication.com and pointed that domain (and all of the other domains I own) to the Nixworld blog. I had stopped working on my own project car at that time, learning some years later that was the last time I’d ever work on it after selling it in the summer of 2022. While my career started blooming in the industrial controls arena in 2016-2017 it was a niche area that was relatively uninteresting to those who weren’t directly involved in it. I didn’t have much to share about it. What I did have to share was a quick blurb here and there with maybe a picture. This is where social media reigned supreme. Just a quick post about some weird stuff I was working on that supplied electricity to thousands of peoples homes. Social media took over for my postings in late 2016. Everyone was there. My friends, my audience, people in the industry, and people I knew form the automotive industry that saw my new flourishing adventure as interesting. When 2018 rolled around I was evaluating my budget and decided that paying for webhosting was no longer necessary. I wasn’t using it, all of the traffic had died off in favor of social media, and the hosting service I was using was starting to face a lot of reliability issues. I pulled the plug, thinking I’d never look back. Social media was the future, or so I thought.

Fast forward to June 2025. I’m about tired of social media as I was the automotive industry a decade prior. Things have changed a lot on the internet in that decade. In the last few years advancements in machine learning, now coined AI, has given humans the ability to instantly create life-like content based on anything they imagine. This blur between fantasy and reality has taken over social media. You no longer engage with your friends, you engage with a series of ads and creators you’ve never heard of, being groomed by algorithmic curation. Knowledge sharing has no place in todays social media landscape. Interacting with friends and people with common interests is quickly fading away. It’s time to take my ball and go play somewhere else.

My current career in controls engineering over the past three years has armed with tremendous knowledge in many different programming languages and hardware platforms. I’ve found myself back in the instructors position, wanting to share my knowledge with anyone who wants to get into this field. Here’s the thing though. I have a solid career with a company that takes really good care of me. My quality of life is the best it’s ever been. I don’t need a side hustle nor do I have an desire to try to compete with the content produced by the peanut gallery on social media (refer back to those who like to hang out on the top left of the Dunning-Kruger chart). So how do I share my knowledge where it can be received without having to filter through the rest of the bullshit forced in front of you by a simple math equation?

Walp, here I am. Back on WordPress. Instead of Nixworld I’m calling this blog 16-bit Nick. The naming convention explained in a previous blog HERE. Shortly I’ll be posting about things relevant to my hobbies, industrial controls engineering, and maybe even a personal thing here or there. I’ve come full circle on the internet and found myself back in the same place where I started this idea nearly two decades ago.

What’s in a Name?

In the past year I’ve decided that I should start producing a series of training videos aimed towards teaching the Codesys IDE. While there are many great resources out there already for Codesys, the videos aimed towards teaching them can be difficult to understand as there are not many people skilled in this area that speak English natively. I want to fill that gap, but I also wanted to create a user name that represented the old school digital era that I come from. 8-bit Nick was the first choice. Declaring anything 8-bit represents the period of time where the video game crash had ended and the 8-bit era of gaming was born. That’s how old I am. But, since declaring anything as 8-bit represents something from a classical timeframe it’s use has become rather stale, a race to the bottom in classical naming convention. I chose the next era of iconic data bus size: the 16-bit era. While still classical in convention, 16-bit Nick encompasses not only the dinosaur that I am in today’s tech market but it also represents the most commonly used data register length within industrial control systems. And since Codesys is an IDE for industrial control systems, 16-bit Nick just seemed to fit. Plus, I’ve always had a soft spot for Modbus which uses 16-bit registers to carry data. So there you go, the reason behind the name 16-bit Nick.