MOD54415 Projects.

[Vernon]

Got the board from China (www.pcbway.com - free plug for them) and put some of the parts in. Works great. AD7606 works, display and touch work, port expander works, RS232 works and the internet works. Displaying channel 8 A/D DC volt value on both the display and a web page. I still have to install the DC to DC converters for the + and - 15V and the instrument amp circuits. It will have one AMP02EP, 2 x AD620 and 1 x AD626.

The AD7606 is available for about $17 in an Ebay module. It is a great chip - 16 bit, 8 channel, simultaneous sampling with 1 bit INL. A Hi-Z input means no special and/or fast OP-AMP is needed to drive it. Voltage dividers and IA outputs work fine.

[Vernon]

Bad day! I lost the UPS for the PC about a month ago. I found out last night that if lightning strikes and the power goes off just as you are saving a file - you lose the whole file and it turns up empty. A day's work just gone. I Now have a UPS on there.

Recovered from that disaster and continued testing of the MOD54415 board. Made a voltage, current, and phase angle meter just to prove out the AD7606 - and it works great. Using channel 7 and 8 with transformer isolation and the on board voltage dividers. You can see the potentiometers in one of the pictures and you can see that it reads the same as the relay test set. Will do more work on functions and then install the + and - 15V converters for the IAs. Once all that is proven I will install the servo module - and test the 16 channels of PWM! This does it all! It even measures phase angles to within a few hundredths of a degree with just the sampling A/D.

[Vernon]

And - of course you need the phase angle meter web page. Watts, Vars and other parameters tomorrow.

[Vernon]

PCA9685 I2C PWM modules are available on Ebay for cheap and the board will mount one of those. They are great for servos, control of DC motors and so on - but I just came up with a new application; PWM control of 16 channels of ordinary 120V AC lighting!

This is especially useful considering the availability of dimmable LED lamps with a rated power around 10 watts. The circuit attached will (I hope) allow control of up to 60 watts per circuit. It uses a 400V 0.8A sensitive gate SCR (cheaper and has a more sensitive gate than a triac) in a bridge circuit to maintain SCR polarity. A channel of a ULN2803 driver (actually a Darlington but drawn as an NPN for simplicity) regulates the current through an LED that illuminates a photoresistor (range ~ 1K to ~ 1 MEG). This controls the conduction angle of the SCR and the brightness of the lamp. It operates in a manner similar to a dimmer with a potentiometer (not rocket science) and is simple and very cheap. The AD7606 is fast enough to digitize audio and the MOD54415 is fast enough to run an FFT to determine the spectral content. You can then have 16 different colored lights that set a mood with the music (and marijuana may now be legal where you are!!). The actual build will have a 150 ohm resistor in series with the LED - I forgot that. Let me know if you want a circuit board !

[Vernon]

Added watts and vars. 0.1% accuracy across the ranges. Angle reads +0.07 degree anywhere you test it.

[Vernon]

All the parts are in, the IAs are tested, and everything works so far. Have not tested or installed the AD626, it is connected to a 12bit A/D input, but will do that at some point. I also need to install the servo module - also for later.

[Vernon]

http://www.cttestset.com/PhaseTEST.htm

250PPM AC accuracy! As you can see from the inset in the picture the calibration will support +/- 0.025% +/- 1 LSD.

While I couldn’t beat 0.1% with the MOD5213 board I changed the algorithm and sample rate and this one is awesome. My best AC measurement circuit board yet - and this will go in the next generation of CTMP test sets (see www.cttestset.com). This will offer 4x better accuracy. The much faster measurement board processor supports a higher sample rate and that makes for a YUUGE improvement. It is Internet of Things ready and you can see the live results and a higher res picture at the link above (I will leave in on for a couple of hours). Freescale MOD54415 processor at 250 MHZ. AD7606, 8 channel, simultaneous sampling A/D. You just can't measure two signals and the angle between them with any greater precision and I am going to have to start rating in ppm!

Some circuit board layouts are just better than others and this one is perfect. Same basic schematic as the MOD5213 phase angle meter but less noise. Go figure.

[Vernon]

Massive power and fat wires!

Getting ready to run servos and you can't do that with your old cell phone charger. So I attached this YUUUGE, 6 dollars from Electronic Goldmine, Condor 65 watt power supply. That will give me 4+ Amps spare +5V power for servos. Rewired with #18. You can see the mounting detail of the servo module - and you can't put the chip and parts on your board for the price of the module. And besides who wants to solder that small stuff.

The 4-40 screws are a self tap into the module so that works out. Using nylon nuts to prevent shorts.

[Vernon]

Running a servo back and forth and measuring AC parameters to 250 PPM at the same time!

I didn't make a movie but if you all don't believe me I will.

[Vernon]

Have you all ever heard of Cooley and Tukey?

Well they invented the FFT time to frequency domain function!

As a first – the phase angle meter in the next CTMP will offer power quality (% THD). Since I have a good sample set and now I have some really serious number crunching power I decided to have power quality where I never did before - because I can’t really crunch numbers on the old 16MHZ 68HS12 on the current ct test set measurement board -- and even the 66MHZ MOD5213 TSM processor is not ideal. But now, with the 250 MHZ MOD54415, it can be done without having it dog out. It can do the FFT at least as fast as if can draw the vector graph so I simply have a key to swap out the graph for the FFT! As you can see in the pictures I changed the Epoch protective relay calibration source from fundamental in one picture, second harmonic in the second picture, and then third harmonic. The FFT results correctly display the amplitudes of the voltage at the respective harmonic and that is how you can tell if it is working. Cooley and Tukey were clever indeed.