Digital Oscilloscope for Testing & Debugging Astro Remote Controls

[Gina]

This starts off with a relatively cheap kit of parts to build a low frequency digital oscilloscope with TFT LCD display.  This should prove very useful for testing such things as remote focusers or dew heaters for telescopes and similar microprocessor controlled devices using stepper or DC motors or PWM control.

My project will expand on the kit to include a rechargeable battery so that this becomes a fully fledged test instrument which can be used in the field with battery powered equipment as well as indoors on the work bench.  A self contained unit free of an external power supply has many more uses than a mains powered instrument.  I shall also be desinging and producing my own case to contain the oscilloscope and its battery power source.

The idea for this project was provided by Dave_D in another thread viz. Toys for arduino based add-ons  but since this project is not a "toy" in the conventional sense and also since I'm extending it into a fully fledged DIY workshop and field instrument, I thought it warranted a new thread of its own.

[Gina]

I bought the kit from Amazon UK with Prime delivery - LeaningTech DSO138 TFT Digital Oscilloscope Kit With DIY Parts And Probe Open Source, 1 Msps, 2.4"  for a pound less than they are selling them now.

The kit comes with PCB with some component fitted plus all the parts, probe and full instructions, circuit diagram etc. - very comprehensive and easy to follow IMO.

This is the spec :-

• Max sampling rate :- 1M samples/sec
• Anologue B/W :- 0-200KHz
• Y range :- 10mv-5v / div
• Max I/P voltage :- 50v with 1x probe
• I/P impedance :- 1MΩ/20pF
• AC or DC input coupling
• Resolution :- 12 bits
• X display :- 1024 points
• Timebase range :- 10μs - 500s / div
• Trigger modes :- Auto, Mormal, Single Shot
• Power :- 9v DC (8-12v)
• Current :- 120mA
• Dimensions :- 117 x 76 x 15mm
• Weight :- 70g (without probe)

Also, not mentioned in the spec but can be seen from the photo :- Screen size 12 div (timebase) x 8 div signal. 

I'll repost the video of the build instructions - this shows a case which is available separately but I shall use my own design as mentioned above.

 

[Gina]

I have started assembling the kit.  The resistors are very small and the colour coding is not the familiar E24 range but seems to be a 3 figure value plus multiplier etc.  Even with reading glasses and a 3x illuminated magnifier I find the coding hard to see so I took the advice and measured each and every resistor before fitting.  I have also now fitted some of the other components too.

[johnfosteruk]

I missed Dave_D's original thread, I like this a lot. I have limited use for it at the moment but can see it getting plenty of use in time - I'll be following this one Gina.

[NEM]

Are the surface mount chips already installed on the board when you receive it?  I can't imagine soldering U1 in place by hand.

[Gina]

Yes, in the version I purchased.  But for anyone whos does fancy soldering SMDs there is a version without most of the SMD soldered onto the board at a slightly lower price.  Personally, I now find soldering ordinary wired components quite enough, thank you very much!!!  In fact if there had been a version with all components ready mounted at a few pounds more, I would have bought that.

[Gina]

Capacitors all in place and some soldered but my present soldering iron has rather a large bit and I have a nice small Antex iron arriving shortly so I might leave any more soldering until then.  Meanwhile I can work on designing the case and power supply.

I have 4 4000mAh Li-Ion cells that I bought for a clock project but not using ATM.  All these Li-Ion cells seem to be the same size called 18650 and 18mm diameter x 68mm long.  I have some battery holders on order due to arrive in a day or two.  These cells will last a long time at something like 200mA load, viz. 20 hours.  I thought of using something smaller - I have a number of Canon 7.4v 860mAh batteries and spare battery boxes from stripped down 1100D cameras but the discharge time would be only 4 hours, maybe less.

[Stub Mandrel]

This will be interesting. I id something similar but different many years ago with an add-on box I made for a BBC micro... I still have a PCB for an avr based version waiting to be populated.

http://stubmandrel.co.uk/16-electronics/92-beeb-scope

[ChrisLX200]

Mine came as a pre-soldered, pre-built kit - is that cheating?

ChrisH

[Alien 13]

I do miss not having a scope I used to borrow them from work when needed looking forward to this topic.

Alan

[Gina]

That looks like a nice scope but not very portable, Chris  

[ChrisLX200]

For years before that I had an old analog Tektronix 465 which I bought for £80 (s/h) and it was great, sadly it decided to explode one day... I still miss it, but that was a heavy beast - this one is very light indeed.

ChrisH

[Stub Mandrel]

I used to use my Dad's Heathkit scope, now I have a lovely old twin-trace Hameg which has the same analogue appeal as vinyl records.

[Alien 13]

The benchmark for analogue scopes was the HP 1740A 100MHz scope many an engineer cut his teeth on one of these the last of the greats was the  Tek 2465 range I had so much fun with those the later Digital TEKs and LeCroys just didnt have the same feel and you always had the feeling you had missed something and had to recheck any measurements.

Alan

[Gina]

Soldered some more parts onto the PCB.

[Gina]

The up voltage converters arrived today so I shall be checking them out for converting the nominal 7.4v from a pair of Li-Ion cells to 9v to power the oscilloscope.  The battery/cell boxes should arrive tomorrow but I can test the voltage converters from my bench PSU.

[Gina]

That'll be tomorrow - up early and out all day today so not tonight (Josephine ).

[Gina]

Testing voltage up converter...  Set bench supply to 7.4v and adjusted output from up converter to 9.34v - the test voltage used for the oscilloscope.  Varied the bench PSU voltage and up converter output stayed as set with input from 4.5v to 9v.  Above the set output voltage, the output rose with the input voltage with a small volt drop.  Since these are designated as up converters this is to be expected and perfectly acceptable.

[Gina]

Been looking at the discharge curve of Li-Ion cells and this seems to vary slightly with manufacturer.  Li-Ion cells should not be completely discharged or they will be damaged so an indication to recharge well before this point is reached is essential.  A voltage in the range of 3.3 v to 3.5v would seem to represent almost discharged so to be safe I think I shall indicate need to recharge at 3.5v representing around 90% to 95% discharged.

Some battery powered equipment has a meter to indicate charge level or a LED bargraph.  I think I should have some bargraphs but not seen them recently in my sorting out.  A simpler device would be a three colour LED with green for OK, orange or yellow for warning and red for "change batteries now".  A bi-colour red/green LED can give red/yellow/green with the yellow being both red and green LED on.

A bi-colour LED will need some circuitry to measure the cell voltage and drive the LED adding complication, alternatively there are battery voltage displays which do it all in one go which I'm looking at.  I don't really want anything as complicated as a digital voltage display - just a 3 state indication would be quite adequate IMO.

So far I've found the following :-
Ultra Bright LEDs 3mm/5mm/10mm Bi-Colour Tri-Colour
Li-ion Battery Voltage Test Monitor Low Voltage Indicator Buzzer Alarm

I'll continue Googling and looking at various devices but another possibility is to use an Arduino to read cell voltages and drive the LEDs - well, not quite but an ATMega328P chip with sketch uploaded to it in an Arduino Uno.  I have a number of ATMega chips plus crystal and capacitor kits that I was going to use in my now shelved CCTV control & display project.

[Gina]

Li - Ion battery level indicator

[Stub Mandrel]

12 minutes ago, Gina said:

Li - Ion battery level indicator

Wow, that's designed to guzzle your milliamphours!

The use of an LED as voltage refernce is plausible, but may not be accurate enough for this. You can use MUCH bigger resistors for the LEDs and use high efficiency ones. And  multiply all the resistors in the potential dividers by 10.

 

[iapa]

On ‎10‎/‎07‎/‎2016 at 23:09, Gina said:

I have started assembling the kit.  The resistors are very small and the colour coding is not the familiar E24 range but seems to be a 3 figure value plus multiplier etc.  Even with reading glasses and a 3x illuminated magnifier I find the coding hard to see so I took the advice and measured each and every resistor before fitting.  I have also now fitted some of the other components too.

 

Colour coding looks like the 4 band scheme for wire wound resistors, the 1st two band are a simple base 10, the third is the multiplier as you say (10^n).

• n colour
• 0 black
• 1 brown
• 2 red
• 3 orange
• 4 yellow
• 5 green
• 6 blue
• 7 violet
• 8 grey
• 9 white

brown black brown would be a 10 x 10^0 = 10ohm; green blue yellow is 56 x 10^4 =560K.

A fourth band is a % tolerance. The real trick is which end to read from :), the 4th band should be spaced further apart from the other. If there are 5 bands, then 1st,2nd,3rd are the value. 4th the multiplier and 5th %tolerance.

We used that when I were a lad building things - about 40 or so years ago, transistors were around, as were the very first CPUs (4004/8008 etc. )

 

Good luck, look forward to seeing completed item.

 

[Gina]

They are 5 band resistors.

1 hour ago, Stub Mandrel said:

Wow, that's designed to guzzle your milliamphours!

The use of an LED as voltage refernce is plausible, but may not be accurate enough for this. You can use MUCH bigger resistors for the LEDs and use high efficiency ones. And  multiply all the resistors in the potential dividers by 10.

I quite agree on both points.  I certainly won't be using an LED as voltage reference.  Not sure I like this approach anyway.  Using an ATMega328P would use fewer components and I have them in stock.  It may seem like overkill to use a microprocessor just to check battery voltage but this seems the way things are going.  If it means fewer components then it's a "no brainer".

[Stub Mandrel]

It usually only adds one pin to with AVR/Arduino as they mostly have a built in A2D.

[Gina]

Yes indeed   That's what I thought - use the built-in ADC.  And they'll drive LEDs with just a current limiting resistor.