I use the analog input no1 with a 10 kilohm potentiometer, but cannot get a linear value, even if I don’t touch the potentiometer. The value changes in a range of +/- 4, which is too much for my project. Is there anything I can change about it?
Miguel
The analog inputs of a K8055 are supposed to get an input voltage. With jumper SK2 open and ATT1 turned fully clockwise, that input voltage should be in the range of 0 to +5V.
You can use a 10K potentiometer for that if you properly wire it as a variable voltage divider. In that case the slider (wiper), which usually is the middle contact of the pot, goes to A1. One of the side contacts goes to ground and the other to a +5V power source (you can use the upper pin of SK2 as the power source).
This is exactly what I did, and it’s working fine.
When I start the K8055-Demo and connect the card I can read the digital value of “data1” below the “AD1”-Slider. ANd this is where my problem starts, because the value is not constant but keeps changing in the above mentioned +/- 5 - range, which is too much for my project because I have to divide the 0 … 255 range possible in 32 parts of 8 each.
The question was if I can avoid this flickering.
Thanks
Miguel
If you have long wires between the potentiometer and the K8055 input, then there may be some extra “hum” captured be the wires. You may try to filter the input by adding an electrolytic capacitor (e.g. 10uF) between analog input terminal and GND terminal.
Thks for your effort. Meantime I tried a 10uF capacitor which gives better result but the “hum” remains at +/- 1/2, which is better but still too much …
The K8055 card uses the USB +5V supply voltage as a reference the analog-to-digital converter.
You’ll get unstable conversion results if the power supply is not stable enough.
You may try self powered USB HUB with a stable power supply.
For some more info please see also the last post of this thread: viewtopic.php?f=3&t=7822
I assume that the unpredictability and instability of USB supply power is what caused the design changes on the K8055N compared to the K8055.
If you compare those two schematics, the N-board uses an LM317 voltage regulator to get 3.3V from the 5V supply. Only the op-amp and the digital input pullup resistors in front of the darlington array are powered by the 5V supply. Everything else is running on the stabilized 3.3V. The AD input is scaled down behind the op-amp with a 1K/2K2 voltage divider to match that.
I may be wrong, but I don’t think the voltage regulator used is actually a stock LM317. Probably an LM1117 or another low dropout version.
One has to keep in mind that the title of these kits is “USB Experiment Interface Board”. And they IMHO do a fantastic job when used as such. But I would never expect something with that title to be usable as a production grade precision measuring device.
I will have to study the PCB layout in more detail to see how difficult it would be to use a combination of LM7805 and LM7833 together with a 9-15V power supply and a few other components, to make the board completely self powered. At a first glance it looks like 2 to 5 cut traces and a couple of wires and resistors plus the RF filter capacitors. I should have all those components laying around. Sounds like a fun Labor day project to me.
Regards,
Jan
I just wanted to follow up on this with my results.
There are many changes from the old K8055 to the new K8055N:
[ul][li]Pull up resistors between the ULN2803 and the digital input ports changed from 47K to 10K.[/li]
[li]There is now a voltage divider between the analog input OP-Amps and the PIC’s ADC inputs to adjust from 5V to 3.3V.[/li]
[li]All resistors in the PWM output circuit have changed due to the lower operating voltage.[/li]
[li]The resistors around the DAC OP-Amps have changed. The - input is now connected to the output via a voltage divider so that the OP-Amp amplifies from 3.3V to 5V.[/li]
[li]All LED current limiter resistors changed from 1K to 680R.[/li][/ul]
I don’t think it is worth the effort to try this rather difficult modification. There is nothing a stock K8055 board would gain from this under “normal” circumstances.
Regards,
Jan