Today we received the first samples of final SimuCUBE board designs. Testing of it shows no flaws so we’re giving the green light for the first 250 pcs batch assembly! All parts are already in stock, so assembly should not take too many days.
SimuCUBE with the needle bed SimuCUBE production tester device
The photo above shows also the work-in-progress test bed for SimuCUBE that will be used to test & verify the operation of every shipped SimuCUBE. The device under test will be placed on the top of the bed and the spring loaded pins will contact certain pads on the bottom of the SimuCUBE board to allow tester to contact all necessary signals at once.
The decision to build a racing simulator with servo motor based force feedback has yielded first tangible results. The picture below shows the current setup of a force feedback wheel with a 36 Nm (peak) motor equipped with a sin/cos encoder. The system is being driven by IONI Pro HC and a STM32 Discovery based interface.
Simulator direct drive FFB wheel is operational but currently it lacks other controls. Mechanical parts design and tooling was implemented by Aki.
The first test drive on iRacing put a wide smile on our faces immediately and gave more motivation to make faster progress on it :)
The plan is to experiment with wheel effect calculations, such as friction, inertia, damping and spring, inside the drive firmware to take full advantage of high resolution feedback devices. Also it’s a good platform to test how far IONI can be pushed in output current ratings in this kind of system.
We run number of tests to each manufactured drive, one of them being a load endurance test. The test is done by driving constant 9 Amp sinusoidal current from all four power lanes and clocking the time of how long it takes drive to heat up by 25 Celsius. This test is used to verify proper characteristics of the power stage. Each tested drive must score above a certain level of endurance to be passed.
Load endurance test with IONI Pro and IONI Pro HC
The statistics above shows endurance results from about 200 tested drives. From here we can see that Pro HC model has average endurance times of 47.7 seconds while Pro has average of 27.5 seconds. This means that on average HC model takes 1.73 times longer to get equally warm during the test. On the other hand, it could be understood that on average HC model dissipates 42% less heat.
The test also reveals that the variance is higher on HC model. There are few extraordinary good units probably due to varying characteristics of the MOSFETs even though both model MOSFETs are made by same manufacturer. Pro HC model uses the newest and lowest loss type that is available today in this size while Pro uses couple of steps lower rated devices. I hope some day the MOSFET manufacturer manages to avoid the variance and start offering transistors that perform like the best ones seen here.
Despite of the variance, the best thing is that there is clear bottom level on both types, so any unit will not have troubles meeting the specs.
We have started tweaking the IONI drive in order to push it beyond it’s present output current limits. This includes testing of lower loss power MOSFETs and current sensing resistors to achieve reduced heat dissipation. So far we have achieved 35-40% reduction of heat dissipation at high currents.
MOSFET replacement work in progress. The new ones are already installed on right side. These particular transistors have exposed top side thermal pad for external heat sinking.
During Easter weekend I was testing IONI on my own DIY CNC router. Steppers ran above 2000 rpm making the machine run faster than it ever has.
IONI has native resolution of 25600 steps/rev (or 128 fold microstepping), which may lead to insufficient step rate from the CNC controller. In my case the LinuxCNC was able to generate only 33 kHz step rate, so step mutliplier in drive has been cranked up to 25x. This normally would reduce microstepping and cause noisy operation, but not with IONI when setpoint smoothing feature is turned on. At 0:32 you will see a comparison between setpoint smoothing off and on.
We have been quietly experimenting on a special IONI motherboard that has discrete power stage to extend IONI power capability. We already have a proof-of-concept board that is equipped with leftover VSD-E mosfets (61A 200V rated). After couple of weeks of use, we could not tell the difference between integrated and external power stage by observing motor control performance and smoothness.
IONIZER proof-of-concept board
Bottom side showing the discrete power MOSFETs
IONIZER has been considered to be available as DIY board (bare PCB) and a reference design where users could modify it to match virtually any power need. However, it’s not ready and many changes are needed to have our approval (such as galvanic isolation of all I/O, short circuit protection and switch to PCI-E 8x socket for newer IONI). Finishing this might be a subject for another crowd funding project.
Some pros are:
Can be made to virtually any power
Not highly expensive
If built with assembly errors, can be hazardous
Some safety & protection features add complexity on circuity (STO, over temperature & short circuit protections)
Industrial manufacturing may not be cost effective (more complex solution than Argon)
Tell me in the comments how useful do you see IONIZER and which voltage & current range would be your preference :)
The prototype IONICUBE 1X has been now assembled and tested. Functionally everything is Ok, which is the most important thing, but few adjustments in layout and component selection will be made. No big deal – no delay!
IONICUBE 1X prototype (with temporary type pin header etc)
In the other news, we’re over $30k in the Indiegogo campaign! Pure awesomeness! This means that the very next IONI firmware release will buff current rating to 18 Amps on devices that are sent for the campaign contributors :)