MOD5213 projects

[Vernon]

Picture of the circuit board to my machine vision system. This device tracks the position of my internet model railroad and enables automatic trip navigation. The train usually stops within a couple of vertical lines of the intended point. The locomotive has an 850nm IR LED and the camera has an IR pass filter that passes anything longer than 780.

Video is processed by an LM1881 sync separator chip and the horizontal and vertical signals are fed though a 74LVC245A that cuts the level to 3.3V. They are then connected via short jumpers to MOD5213 IRQ inputs. Jumpers are used to allow the module to be used for other purposes. The bright white spot produced by the IR LED is detected by the LM 319 video comparator - the output of that chip triggers a one shot that produces a (approx.) 2 microsecond pulse that is also reduced to 3.3V and applied to a MOD5213 IRQ. Works great and is better than having to run 20 wires to sensors in order to know where the train is.

The MOD5213 sends position data to the MOD5270 server via RS232.

[Vernon]

This next project has not been started - I am waiting to see if Donald Trump is competitive in the general election polls. If he is I will pull the trigger. Because you know he will nuke somebody and we will have to measure radioactive fallout.

The gamma spectrometry kit is made in Australia and designed to be used with the sound card in a PC. A scintillator crystal produces minute flashes of light when a gamma photon is absorbed and this light is amplified by a PMT (photo multiplier tube). This produces pulses where the amplitude can be used to identify the isotope. CS137 is one isotope of interest and the attached picture shows it's distinctive 660 KEV pulse. Normally the A/D in the sound card measures the amplitude but I am going to use the 12 Bit A/D in the MOD 5213.

On the circuit board the same LM319 as the previous project generates an IRQ when a pulse reaches a threshold. A one shot extends this pulse by about 15 microseconds to allow for rise time and the IRQ will trigger on the high to low transition of this pulse (the end of it). The IRQ will then read the A/D and that will permit the main loop to identify the isotope. The pulse is amplified by the AD8055 to get about a 2V signal for the desired isotope - that does not require much gain, about 2.5, so bandwidth will be excellent. The output of the amp connects to the A/D. Once again jumpers are used to allow application of the MOD5213 to other uses.

[Vernon]

This picture was obtained from the people who make the gamma instrument - it shows the result of an ionizing event. As you can see it takes about 20 microseconds to peak - so the IRQ is delayed by about that before the A/D is read.

Between the Donald and that guy in North Korea - you need one of these.

[Vernon]

Got an Ebay AD7606 module and temporarily hooked it to the MOD5213 SPI with jumpers - works great and I am going to migrate any and all high accuracy A/D projects to this. It has a 200Ksps sample rate and gets great accuracy in any application - DC or AC. I made a 5V scale DC meter just for test and pictures show a spreadsheet of the accuracy - one can remove the 0.14% with the scale factor and the bending at the low end is due to a half a millivolt shift in the wiring clusterfuck. The other pic shows the Keithly reading, the Keithly is parts per million, and the data coming out the serial port of the MOD5213 to Tera Term on the PC. The MOD5213 setup will also be parts per million once I get it on a circuit board with fat ground traces. I also have a lot of stuff floating that should be grounded.

The modules are $17.76 - you do have to move a resistor to put it in SPI mode but I sent a message asking if they would do that for me because I have a hard time soldering microscopic stuff. In the picture of the PC screen the "SPI VALUE" reading is the DC volts from the big multi-turn potentiometer. The other values and phase angle is a simulation of sampled data from two 60HZ AC signals. That will be tested once I get the permanent circuit board done. The 8 simultaneous sampled channels will allow measurement of the angle, amplitude, watts, vars and so on of an AC power system.

[Vernon]

I connected AC signals to the AD7606 module and essentially made a voltage, current, and phase angle meter. I have it slowed down so it can be read – but these are one cycle values with the extra noise of the A/D not being on the circuit board, certain pins floating that should not float, and running off the digital supply. With an exponential filter a meter reading would be glass smooth. This is raw data.

https://www.youtube.com/watch?v=rN5NNHVpAxc

Check out the video. Note that it has 8 channels so you could do three phase voltage and current if you wanted. I just used some transformers and a shunt for the current to do one V and one I.

[Vernon]

Work continues and the permanent circuit board is in work in China as I type! Performance is great even with it hanging on wires and I am sure the noise will be a lot less once it is on a board. Right now, measuring AC it is better than what might be expected for DC. This data is with an 8 cycle circular buffer and my AC measurement algorithm. Noise is +/- 0.01% of range (range is 50V) or .005 and all values recorded are actually within .005 of the Keithley. So I can claim +/- .02% of reading + 0.01% of range as compared to the standard (the Keithley) – and, because the results are all actually within 0.01% of range, I still have a huge margin. I don’t really need the 0.02% of reading. And it ain’t on the circuit board with the dedicated linear regulator and so on - yet.

Parts per million soon!

[Vernon]

https://www.youtube.com/watch?v=pcfs2EZnDUQ

Got the circuit boards from China for the AC amplitude and phase angle meter and put it together. Works great - video above.

I need to get all the isolation magnetics, get the clamp on (en-route), put the instrument amplifiers in and so forth. And pick a case. Using the spare channels (no RC filter) right now. I have some 20VA transformers stepping down the voltage from the Epoch test source and the same isolating the current. Obviously I need smaller ones. The angle is on the money with little noise.

Then I need a menu to select E to E, I(binding post) to I(clamp), E to I(clamp) and so on. On the E to I there will also be an alternate screen to read Watts, Vars, and PF. Makin' progress!

[Vernon]

A little out of focus – but showing Volts, Amps, Angle, Power Factor, Watts and Vars on one screen. Makes it easier. The only thing selected in the menu is what two inputs you want to measure. A2 would be the clamp on for example. Then, of course, if you are measuring two voltages or two currents the PF, watts, vars and so on do not display.

[Vernon]

[youtube]_OkoAiddujQ[/youtube]

https://www.youtube.com/watch?v=_OkoAiddujQ

Accurate !

[Vernon]

A MOD5213 project complete today! I am calling it the TSM 2017 with TSM standing for Test Switch Meter – because you use it to check CT and PT secondary values at a test switch. One of the pictures shows the power on or reset menu the other shows the V1 – I1 (hardwire current) mode. The voltage is 120.24 (actual voltage out of the Epoch relay test set), the current is 1.998, the angle (Epoch is set to 330) is 330.09. It calculates and displays the watts, vars, and power factor. I moved the battery % down to the lower right corner and left a space for SY. SY will represent GPS sync and will be the angular difference between the GPS PPS positive edge and V1. Parameters, except the clamp on, are measured to an accuracy of 0.1%. Note that you can measure the amplitudes and angles of two voltages and two currents and use a clamp on with V1.

Then, since some people say I am a space cadet, I also invented what I call AstroSync. This option will add a GPS to the meter and allow remote phase angle measurement. Did you pull a new cable to a substation at an oil refinery? Want to make sure A phase is still A, B is still B, C is still C? Getting ready to commission a pilot wire scheme? Need to monitor power angle? Want to know the angle before closing a breaker? Want to validate line impedance? This will do all that and operate by synching the V1 phase angle measurement to the GPS PPS pulse that is within nanoseconds of a Navy standard. The 60 HZ is not synched to GPS, the observed angle could be anywhere from 0 to 360, but the frequency error is very small and the angle changes extremely slowly. So – if I am reading 134 degrees and you are reading 134 degrees at a remote location we are connected to the same phase. The difference is what matters. Two hours from now it might read 140 degrees – but, of course, the same phase will read the same. I call it AstroSync because both astronauts and GPS satellites are out in space.