Welcome to the Zerksus web site
(If you cannot see the full framed site, click here)
Last added to: 2008/10/08 (Previous Tips, Rotator (Asymmetric Capacitor - lifter on bearings) , Soldering Station, Fun1, Fun 2)
2008/10/14 SO8/14/16 Carrier modules
Prototype SMD ICs on breadboard or Veroboard
2008/10/12 F310 module prototype design
Check out www.re-package.com
We are splitting the design section of Zerksus and the new module/repacked component section to its own business. Things will go slow at first (as many of the PCB companies do not take orders after the middle of November, for Xmas shut down), but we are planning to do at least the F310, F350, Q26 and Cinterion MC55i modules.
2008/10/05 Ultra Low Power Tracking Device
We are designing an ultra low power tracking system for a company in Cape Town. Why ultra low power? Well, many times you don't have any access to power and/or the next time you see your asset is a year from now. Now there is nothing you can do about GSM power requirements as the unit will draw its 2A peak when required, but you can do a lot about the standby current.
Initial testing show that this device uses a miniscule 22uA when inactive. This gives the unit enough power to run a wake-up clock (programmable from seconds to 24 hours), maintain I/O levels (crucial for the low power design) and retains the GPS data (making 1 to 2 second 'hot start' times possible). At 22uA you have189 days (more than 1/2 year) of standby capacity per 100mAh battery !!!
The unit is designed to wake up periodically and obtain and transmit its location. There is also an alarm input option to wake up the unit.
This is the first of a few tracking devices that will run on our dirt cheap DIY server system. The client can use Pay-As-You-Go cards and will get access to the Zerksus Tracking Server. This will allow the client to obtain the latest location anywhere over the internet. We are also making provision to make the data compatible with just about any mapping package on the market. We are also negotiating for mapping rights that the client will pay for per month per unit. All of this way below the cheapest package on the market.
With this technology you will be able to drop a unit in your car and track it yourself or start up your own small tracking company with very little infrastructure. At the moment I cannot list this unit's price as I have already negotiated price and supply with my major client. If anyone is interested then you are welcome to contact me (firstname.lastname@example.org) and then we can make a plan.
(The pictures below just shows the circuit board. Currently there is an enclosure that is custom designed for a particular application. The waterproof enclosure houses the circuit, GSM and GPS antennas as well as the battery array. Batteries are a major part of the system as you need specialized non-rechargeable lithium batteries, because of their long shelf life. Unfortunately lithium cells cannot handle large currents and you have to parallel (expensive) several batteries. You can use NiMH (3x 1.2V) batteries but, although cheaper, they discharge quicker and you might end up replacing batteries or charging every month or two. For normal vehicle tracking that should not be a problem as you would have 12V/24V close by.
If anybody is wondering, yes, the circuit board is populated on both sides (the bottom side is where the magic lives). The top is basically housing just the GSM and GPS modules.
Things have been going hectic as Zerksus in the last few month. We have spent a lot of time ranging from contract work to concept designs for new products.
HHO gas generation for fuel saving on vehicles.
As the oil prices shot up the petrol and diesel prices followed. As the oil prices come down the "powers" became accustomed to their new found cash flow and you probably won't see the fuel prices drop in the same way. So HHO gas generation started becoming a hot topic again, as a means of saving fuel.
Although Zerksus does currently have its own HHO reactor design we decided to focus on the electronics part of the problem. Please note that this product is not the Stan Meyer free energy design and is basically a Faraday limited electrolysis design. Tests do show savings from 30% to as much as 71% (but I know that can be wildly debated).
The HHO gas does not really provide you with extra energy. The PSU runs at 120W with between 85% and 90% efficiency. In the end the total gas energy is only about 30W. So thinking of running an engine just on HHO produced gas is unrealistic. The HHO gas, however, does provide faster spark burning and lowers gas exit temperature (water vapor). This causes the engine to run more effectively. The car and oil producers do not like this and are building features into the vehicles (like new onboard computers and lambda sensors) to force higher use of pertrol (if you do use a HHO system). BTW, we are in the process of manufacturing a device to modify lambda sensor data to bypass this problem.
The reactor unit produces gas in proportion to current supplied. My clients initially used linear regulators to achieve this but due to the hot engine compartment and the 15A+ currents their circuits soon turned to smoke.
My design is a very efficient switch mode power supply that provides up to 90% efficient current conversion. This is very important as any excess heat adds to the already hot engine compartment.
There are many designs on internet but knowing what I know now I can just smile as I think what is happening under the hood. There is a reason why there is an "automotive specification". Many people just don't get it. The system we are working on, for Africa wide distribution, has full support all all facets of the product. Watch this space.
Continual advancing technology leaving the DIY-er in the cold.
I remember in the 80's I could plug a 8051, EEPROM and RAM into a breadboard and build something useful very quickly. BUT in the last few years, as switching speeds have gone up, electronic packages have shrunk. These days most components are mostly available in SMD format with pins so small it is nearly impossible to solder them by hand.
If you are lucky you can get some of the newer CPUs in a QFP (quad flat pack) but more and more it is moving to absurdly small BGA packages. If you really want to build something you need to lay out the board in CAD, have it made (R1000+), plus a stencil (R2000+) and have it mounted (+ setup costs R1000+). And then you have the chance of a serious malfunction when you finally switch it on.
Sure, you can buy just about any component in development board format, but that is expensive, the board is over populated, difficult to get hold of components and large size. I know of very few people that can take a dev board into some final product.
So what is the plan? Well, remember the days when you could safely plug something into a breadboard and apply 5V? That is what I want to bring back. A few days a spent the day at a large components supplier. The RF guy brought out his entire collection. Each new module was better than the previous. More sensitive, lower cost, more power, less space required....
The "less space" was my problem. It is nearly impossible, for a normal DIY human, to solder those things! So I looked at which components were "taken away" from the DIY market. It gets large quickly. First of all I noticed all the new GSM and GPS modules were out of scope for most people. Pin pitches of 0.5mm, weird voltages, complex software APIs etc... Then I looked at most of the newer CPUs. They are all fun packed but many of the newer ones require them to be reflowed. You cannot even solder them under a microscope. Another supplier showed me an awesome accelerometer. Low cost, high specs.... only in BGA. ...And as a scraped though more modern datasheets the number of new RAM and Flash chips, special function ICs etc. all are moving toward unmanageable packages.
So who gets excluded from this game? Well, firstly young people. I remember when I started electronics I would beg for money from my parents. Today they would have never sponsored something so expensive. What about students? What about primarily non-technical people (e.g. someone that wants to control his gate from his cell phone, but cannot program to save his life). Well, even people like me. Sometimes I just want to quickly hook a small CPU up on a piece of Veroboard. I recently bought two new Cinterion MC55i modems. Do you have any idea how difficult it is to 'quickly' get it going?
The plan of action:
Currently on my shelf I have technology that includes the following:
High speed 8051 CPUs
USB2.0 interface ICs
Apart from the PICs(only some) the rest requires a surgeon's dexterity. The plan it to mount these (because I know the guys who can) on a PCB with 100mil / 2.54mm spacing connectors. These boards will fit easily on Veroboard or plastic breadboard. All the high frequency items will be on the module. The fastest speeds will probably things like RS232 (@9600 bits/s). Also all special power supplies will included on the module. You just need to supply 5V.
Scenario: You want to go on holiday and want to connect your home alarm system to a GSM modem to SMS your own cell phone if something happens. You want to combine 8 zones into a simple byte of data and send a text message when the event happens. When it happens you can call your friends/police/security company and they will sort out the problem. Also, if this works then you can sell this to your friends as well and make some pocket money (maybe sell a lot).
(Option 1. South African prices). You download free CAD software and start to design. You buy an unfamiliar modem. You spend an additional (up to ) R4000 making double layer board minimum and placing the components. You battle with software. In fact I reckon you will spend several weeks. The chances that the system will work is very slim.
(Option 2). Seeing that you are in a hurry you buy a Zerksus GSM module. Sure it will be more expensive than just a normal modem but it is pre-built with a simple power supply, CPU (that excepts very simple RS232 commands), EEPROM (for e.g. PIN storage, destination cell number) and easy to solder pins. You have the option to either add something simple as a PIC (sending RS232) or trigger a pin to send a packaged SMS. You rather spend some time working on your alarm interface and then solder the entire thing on cheap Veroboard. At this stage I am looking at 20% of the cost of option 1. At a later stage you CAD a simple single layer board and use the module on top of that as well and go into production with module.
Ending this section: I want to simplify some of the components so much that a 14 year could easily use it. At the same time I want to provide quick and effective parts to engineering companies to increase time to market. Currently the options are so wide that I will first need to get a few enquiries before I start launching such a line. Keep in mind that it will cost me $1000 (R8000) plus for every prototype module (only 1 or 2 ) I repackage like this and several thousand more to make a decent preproduction run per component. Thus it will be impossible for me to have a range on the shelf in the short term.
Well, things have been busy, very busy. So before I left this site un-updated for almost a year I thought it might be a good idea to start adding some content.
New products and capabilities that I will add more information on later:
10 Amp, 55V DC power supply, with 5V (1A) and 24V(1A) PSUs, for NC machine.
Silabs F310 8051 board with 11x MOSFET (8A per channel at a time limited), 8 inputs and RS485.
The second and improved version of the vehicle tracker.
A long period tracking device. Basically a close-and-forget device for long term maintenance free use.
4 motor stepper controller, with integral CPU. Basic target systems are NC tables. (Brushed DC motor controller also planned). The idea behind this design is to make a 4 channel driver and controller that is a lot cheaper than buying a controller board and 4 discrete drivers.
I am also not involved with the Magnum 2005 anymore, so please direct any enquiries to Magnum.
I haven't updated the site for a while. My dad passed away last month and things have remained hectic ever since.
New vehicle tracker
But life goes on and I am currently waiting for the final paperwork for a sizeable order for my new vehicle tracking system that I have named "Tiny". As you can see below it quite small and only consists of 3 basic parts. As with the Model-T Ford you can also get it in any colour you want as long as it is black :-)
Now, Tiny boasts megabytes of Flash, ARM9 CPU, sleep mode CPU, controllable power supplies, easy to hide away antennas and a GPS with a -159dBm sensitivity.
This is actually only one of 3 parts of the system of which the other 2 are the internet server and client software. We are currently designing a server that will provide all the routing you need to start up your own vehicle tracking system (at home if you really wanted to). The client software retrieves the data and stores it for reporting. You can also purchase just about any NMEA compatible GPS software to display a single vehicle on. We are even making it compatible with the commercial Google Earth program (Roughly $20 per year license).
Why yet another vehicle tracking system? Well because a client asked us to develop one, and now that it is operational there is no reason to just leave it sitting on the shelf.
The installation is a no-brainer. We organise a GSM contract for you, and supply you with the unit(s), programmed SIM(s), client software and an account on the server. If you want to install it yourself you just need to find permanent 12V power and connect the unit, along with the antennas. The high sensitivity GPS makes it possible to install it in places normally unsuitable for GPS, thus simplifying the installation. You purchase mapping software of your choice, install and configure the software and you are done.
The basic version of Tiny has no extra I/O ports but the unit can be expanded to, in the future, include sensor inputs, panic button input (auto SMS), relay outputs and even voice.
We are busy with monthly pricing for the use of the server, GSM contract, data and additional transactions. So far the cost is cheaper than the current competition. I must just mention that the bigger companies do provide vehicle recovery that we don't because the tracking unit is under your control.
Why buy something like this? Well, apart from "where is my wife or my child" you can run your business more efficiently. With fuel costs so high you cannot afford a driver taking personal detours with your company vehicle. If a driver takes a 2x 5km detour everyday, for whatever reason (including personal or too long route), then after 22 average working days this will end up to 220km. My one client commented that this will cost him R200 just in petrol. Taking other factors in consideration like tyres, services and depreciation of value then this cost can easily reach R400 to R500 per month. So if you are OK with a total loss of R6000 per year or R30 000 over 5 years then don't invest in something like this. My one friend always jokes by saying: "The best off-road car is a company car". Do you really know where your driver goes and how fast he gets there?
I mentioned more features in the future. So does that mean you are stuck with old technology when you buy now? No. You are not stuck. As small as this device is I must point out that it contains a small motherboard. For the cost of a future change of motherboard and upgrade in software you can migrate to a more fancy system with minimal cost relative to a new system.
We have partnered with a distribution company, that will also handle the access controllers as well, so I won't be selling this product directly from Zerksus. Between the two partners we will be able to organise the physical hardware, software, internet connectivity, GSM contracts, installation as well as application for credit via a reputable financial service supplier.
Here is something I literally did in 30 minutes. It operates in the same way as the antigravity lifters you will find elsewhere on the web. Follow the link to see more.
Below is a photo of the new dual channel (2 x 50 Watt) Magnum Soldering station PCB. Soldering will never be the same.
Unfortunately the reader-controller project is hanging in the air due to supplier problems relating to the enclosure. Currently I need to disappoint clients asking for quotes. These include larger companies like Eagle Tech in Cape Town. A real damn shame. Our plan B is now in full swing but at a tremendous cost.
Since the middle of last year Zerksus has been involved with C&S Centre with the design of a large format printer (4 meter to 6 meter bed lengths). We are getting the X-axis mechanical system in on the 21st of March. Will post images where possible. Printers, like these, will cost a lot less than imported systems and can launch a new business with advantages like high speed (1 meter per second printing, 57 mm stroke), 5 different ink types, local support etc... Imagine printing full wall paper artwork without edges. C&S is working hard to make an August date for an international printer show.
The Magnum soldering station prototype PCBs should go out for manufacturing in January 2007. We have received a lot of interest from businesses as well as the public. When the project is completed Magnum plans to launch a large promotion of the product. Watch out for that.
On the reader-controller front we are still experiencing heavy delays with the enclosures. In January we will proceed with plan B (you should always have a plan B:-). As to the possible jobs we have decided to lean towards a major established electronics supplier in South Africa. If anyone still wants to make a R1 000 000 per year (after tax) then please contact me before I sign the agency over to them.
On the horizon is another major FPGA project. This project involves video capture on a massive scale. One(minor) problem is that JPEG is not free anymore. Some company has won the court case as being the owner of the JPEG algorithm. So each camera you buy contains a portion of JPEG royalties. If you want any information about this JPEG problem then email me and I will send the URL.
Merry Xmas to everyone.
Before I start below please checkout the link in the upper right corner of this page. The world is slowly (inaccurate wording) running out of energy. The link will take you to Bill Anderson's energy site where you will also find some free downloads. In his books he discusses both practical energy generation and saving as well as fringe technologies. He is a practical hands-on guy and describes in detail (part numbers and prices in some instances) where to buy equipment to produce your own electricity. Currently there is a 3x eBook special which covers the above, bio-diesel (in depth do it yourself- And I mean detail) and a scary perspective of the current and future oil industry. So before you find out all about the latest PS3 game console rather find out how to power it yourself and keep you future life out of the dark age.
Credit cards: Zerksus will be accepting credit card sales from early 2007. We have automated the Setcom service into our server. As said before we will only focus on South Africa sales until further notice. The delivery charge is R75 (ZAR) for door-to-door delivery ( up to 30kg). For large orders we will need to telephonically contact the client to confirm details including a fixed line telephone and physical address. For oversees clients please contact me so that we can arrange a normal bank transfer approach.
Pressure, Pressure, Pressure. Many things to get out by January. We are still experiencing problems on the delivery of the reader-controller enclosures, but I have been promised that it will be sorted out soon. Only when the full package is ready will I publish the hidden and unique qualities of the product. Samples are still planned for end of January 2007.
The Magnum soldering station is also progressing well. I gave Angelo Angeli (the owner of Magnum) a demo on Saturday. After stabilizing the unit (in still air) even had long periods where it stayed within +-1 degree Celsius. The product target is still only +-4 °C. The first ADM version will go into production within the next week, dependent only on the PCB manufacturer. The uses will include lead-free soldering which required stable temperatures. The unit will also contain features to give owners, of multiple units, more control over what their operators are doing. The Magnum Soldering Stations are planned in a dual version, of which a single (more cost effective solution) version as well as a soldering iron/hot air pen with pump for SMD rework. If you are based in South Africa then you will not find a better (and even cheaper) option than this new range of Magnum soldering stations.
About lead-free. Many people think it is the next best thing. In fact it is crap. RoHS requirements include reducing the amount of lead pollution. The fact is that lead-free solder is an inferior product which we have to get used to and deploy in the most effective way. Tin compounds are brittle, does not have the long life of leaded solder, crystallizes over time, causes microscopic cracks, does not bond well to gold, grows crystals on RF circuits and is overall inferior to leaded solder. Add to that the problems that it does not solder as easily as leaded solder and you have real problems if you are in electronic production, no matter what level you are on. That old trusty soldering iron is going to get you nowhere. If you attempt to solder products, using lead-free solder, with your old soldering station then you will have so many comebacks that you will lose more money than several new soldering stations combined. Add to that the slower production rate and related problems and you will definitely feel the productivity and financial loss.
The FPGA based CPU boards are doing well. A new batch will be ready for a client by January 2007. The nice thing. I did not have to do anything. Pure passive income based on 1 to 2 year old designs. The client does want a few changes but that will be done in VHDL software and not hardware, one of the main attractive features of FPGAs.
FPGAs to the masses??? Every few days I get somebody asking me for FPGA development boards, information on FPGAs, how to program them and eventually how to make their own products with FPGAs. The last part usually discourages them. FPGAs are extremely cool. You effectively design your own microchip with it. So why do people become discouraged? Have you, as a novice, ever seen the implementation of a FPGA?
There are quite a few small things you have to keep in mind when designing the circuit board. How the configuration EPROM is hooked up, control lines and general layout.
Power supplies. These chips no longer work on a single 5V supply but on at least 2 PSUs (e.g. 1.5V core and 3.3V I/O).
Mounting and layout problems. Most of the people giving up did so because they did not know how to layout a board (or have never done so or have limited single sided or Vero prototyping board experience). These boards cannot be done at home and requires a PCB fab to manufacture. If you have the skill to make/manufacture the PCB then you still have to mount it. Apart from one manufacturer that still has a 40 pin DIP version and another that has very old FPGA types in a PLCC package you will be stuck with, at best, QFP (Quad Flat Packages). These have a 0.5mm pin spacing. Don't even look at BGA (Ball Grid Array) and FBGA/UBGA (Fine line/Ultra Fine Line) packages as they have to be professionally mounted with precision tools and ovens.
Why not just buy a kit to develop and play with? Well, no problem with that. A development kit allows you get started quickly and design pretty wonderful circuits. But getting from dev board to a small production run is almost impossible unless you have mastered the 3 topics above. So most of the experimenters stop after their first flashing LED attempt as their super-duper quad CPU design will never hit the market. You could take your expensive dev kit and build it into a box and sell that (and some people do that because in the end you IP is worth much more than the hardware and the dev kit usually looks kind-of professional already) ... but that might be a bit too expensive for lower cost products.
Solution: Go the ZEN way and find the middle path? When I develop something I still use Vero board and bread board. There is no way that I am going to CAD and manufacture without making sure the final board is going to work. So how do you plonk a 144 pin QFP FPGA on a piece of Vero board? Simple... use a FPGA breakout board that has standard 100 mil pins. Here is how the basic product will look like in various versions:
PCB with FPGA and configuration EPROM.
Optional oscillator with different frequency ordering options.
On board JTAG connector
Specialized PSUs on selected versions (e.g. powered from 5V only)
One or two test LEDs
Connectors connecting to bottom via 100 mil standard pins to provide I/O to your circuit.
Optional connectors to the top for higher speed future expansion
You don't have to worry about the FPGA chip's specifics.
You don't have to worry about the power supply (certain versions only).
You don't have to layout the fine pitched PCB.
You don't have to solder the fine pitched pins risking component destruction.
You can develop you product using the board and
you can go into production with the same board.
The board won't cost near that of a dev kit's price.
Repairs will be a breeze. If the FPGA is blown then install and re-program a new one. No de-soldering and re-soldering.
You don't have to order and/or keep large volume stock.
You develop on either a development mother board (suggested for first time effort), Vero board, dual strip bread board (single is too small) or your own custom development or production board.
The bottom line is that you will use the FPGA board very much as a single chip solution that you can solder yourself.
We are currently supplying a similar solution to a client. The client only has to interface to the 100 mil spaced pins. The board also contains RAM and Flash, which makes it more a CPU card, but the entire board is slightly larger than a credit card and designed with cheap double sided board. For the public we are initially planning a board with just the FPGA, EPROM and connectors (with possibly an external PSU module in a similar pin configuration). The idea is to use it as a component and to add your own LEDs and other interfaces.
If anybody is interested in something like this before middle of February 2007 then email email@example.com
I haven't been this busy for some time now. For about a month now I have asked people if they wanted to become involved in the reader-controller project as either in the sales or installation of it. As the potential income is huge I thought people would jump at the chance. I was shocked to see how unmotivated most people are. I spoke to a few people that rather wanted to work in my garden or drive me around than to learn something new and free themselves financially. Eventually I gave up and approached the bigger guys.
The PC software is progressing well even featuring 3D graphics and voice feedback. We are also putting a lot of emphasis on help. The software is full of tool-tips and text files. We are also busy with training videos showing you step-by-step how to operate the system. Lots of work now but a real time-saver in the future.
I am now in the process of talking to a few larger companies that might help take this project to the next level. Currently it looks quite promising. I cannot name any names yet but will do so when the time is right.
Almost have the credit card facilities in place. I still have to do a test this week. When operational I will place products on the site with "Add to Cart" and "Buy Now" buttons. For simplicity sake we will only process South African orders for now, as delivery is quite simple (+- R35 to R75 to closest post office depending on weight).
So, what is happening? For the next 2 to 3 months I am only focussing on 3 areas namely the reader-controller, FPGA based CPU cards (video overlay and small footprint controller) and a soldering station project.
The reader-controller project is basically finished but we are having a few problems with the manufacturing of the enclosures (to be sorted out soon). If you are technically minded, have a feeling for security and want to make lots of money then let me know if you want to resell this product. I already have a few agents in the Gauteng area but nothing is set in stone.
I am waiting for an order for another batch of the video overlay cards, and then some more after that. This quickie project turned out to have a lot more potential than I originally thought. The one nice about it is that you save on having to include an additional CPU card as with normal PC104 system cards. The FPGA boards provide a regular stream of income for Zerksus but I still want to focus on the educational part (dev boards and programming cables).
The soldering station project was something that started slowly and but is proving to be a potentially viable product. I am planning to be finished with the project in November 2006 with tests running during December. If you are moving over to lead free solder then watch this space. The old way of soldering is going bye-bye...
As you can see there is nothing really happening at the online store as everything happens via email or direct client contact. People hear about Zerksus and then place a contract or organize a joint venture deal. So far I have received many CVs from people looking for work. Please note that I don't employ people. Everything gets contracted out (except my wife's work :-). I do not mind if you do email your CVs. I read most of them and even sometimes give feedback (spelling mistakes etc...). Most of the time I will put the person in contact with my contacts and hope for the best. I am interested in people doing contract work at home and also empowering these driven people. If you do development at home and think you have what takes then drop me an email.
I can already see that next year is going to be hectic. There are a few large projects, using mainly my FPGA boards. Zerksus will be involved in the hardware, software and system design.
Stack of 4 video overlay cards. It takes BW video in and overlays 512x Y-lines pixels (transparent, colour1,2 and 3) over the picture.
All these products are developed in South Africa, by South Africans. We aim to supply both the local and international markets.
Info: Please note that we currently can only receive money via bank deposits and bank transfers. We are currently putting the facilities in place to receive secure credit card payments as well.