Hi Bill

I've been looking at 'PC scope's' on and off for around a year. Mainly because I can't afford a real oscilloscope as yet, even with ebay. How has your design worked out? I have a second PC which just needs a drive and ram which I would fancy using as some sort of test tool for 'the' hobby. I also have an eeepc which I'd use as well.

regards

Fenris

 

It's one of those projects I've done about 70%, and stopped. You're right, I need to finish it. It is a good idea.

I bought several 4 way telephone and RS232 switches for a pitance. The rotory switch is reusable, as is the box. Don't ask my why I haven't finished it, no good excuse.

 

Hi Bill

I think it's called life! I'm sure you would have picked it up again at some point. I had the software on my eeepc and plugged a mic in to watch the display when I spoke. I was intrigued. I know there is a limitation to the range that sound cards can handle but it still seems to be a useful enough tool to have handy.

regards

Fenris

 

A sound card is not suited for most electronic measurements. As a thumb rule sound cards is AC coupled. Even if they have sample rate up 96Khz the bandwidth is limited to about 20Khz by digital filtering. Since most sound cards use delta sigma converters. If a sound card is to be used for measurements it must be calibrated before use each time as they tend to drift somewhat.

 

Bill has included a calibration circuit with the main circuit so that end is covered.

regards

Fenris

 

A problem we can not overcome is that most of (if not all) sound card is AC coupled so they can not be used for signals below 20 Hz. But if we are willing to accept the sound card limitations I still think the hobbyist can have much fun with both sound card input and output. I have also posted a sound card project. But not on this site. My project was an AC skin conductance meter. I made it by using Labview software and some circuits. And the Labview is very important. That is why I did not post it here

 

Truth, PC Oscopes have been out for a while, I just tried to formalize it. It has it's uses, but a true oscope is definately better.

I'd like to find a relatively simple circuit to interface either with the USB or printer port that would offer a true oscope. I know the commercial models are out there, I'm talking something home brewed and hopefully, cheap.

 

Ya true, if someone could come up with a USB based scope, it would have been cool.

 

Awrite, it's time for a major bump in this thread, with Bill's permission and all.

I had a soundcard o'scope software (http://www.zeitnitz.de/Christian/scope_en private use only) in my hands for quite some time now and I wanted to make an input module. I had seen the Witham schematic but never really got to build it.
I started examining it the other day and I came to the same conclusions that Bill did, about the large output voltage, so I decided to give it a go myself. I haven't found any other schematic, other than the one that Bill did, pointed by Google (oh, the irony). I didn't feel quite satisfied by it and besides I did a few remarks about the soundcard and software performance that I would like to incorporate, so I decided to build my own. Here it is:
If you want me to make the image smaller, say so.

Let's start the analysis.
First I 'd like to discuss about some of the goals and of the system:

• I measured that the soundcard of my backup computer clips the input if it exceeds 155mVp. That means that any input going to the mic. in plug broader than that will skew the result. That is also an output limitation for the circuit, as I don't want to push the input further than that.
• I also discovered that the software waveform differed in amplitude from the actual input, with the input mixer on max. Maybe this is a compatibility issue of the soundcard or the software.
• I 'd like to use an oscilloscope probe to take my measurements, just for convenience and a bit of show-off.
• I 'd like to make the system portable, so Bill's idea of two 9V batteries for power supply seems ideal.

I 'll discuss the units of my module and comment on it separately.

1. Input AC coupling capacitor.
I just took Witham's value for granted. If anyone thinks a different value would be more suitable, please say so

2. LP and HP filters
I know that the system limits high and low frequencies as it is, but I though that a HP filter could do good on rejecting noise and reducing the consumption on the OpAmps. I have serious doubts about the use of the LP filter.

3. Input buffer
As I plan to use the scope to possibly debug amplifier/radio applications, I think it would be nice if I could measure faint signals without skewing them. Thus the existence of the buffer. Great input resistance.

OpAmps
For now I plan to keep the TL082. However I don't know if the OpAmp can operate adequately with a +-9V supply. I 'd really appreciate some input on that.

4. Signal Amplifier/Reducer
Here we have the usual inverting amplifier. Its job is to scale the signal down. For the maximum scale of 1:1000, the expected input signal is 155mV*1000=15V, well above our range, but still in the same order of magnitude.
The purpose of the parallel resistor with the switch is to be compatible with x10 probes. Have I got this right? Is a x10 probe reducing the signal 10 times, raising the need of an x10 amplification?

5. Amplitude trimmer
As I said before, the ratio of input amplitude to displayed amplitude isn't 1:1. For example, with a 150mV 1kHz sine input, the software displayed 800mV (if memory serves right). Thus the need to trim the amplitude, if we want to actually read the width of the signal, as well as its shape.

6. Input voltage divider
If I am to operate the module on +-9V, that means that my input can't be larger than that, right? However, I feel that this is a little restraining. 90V on the other hand is more than I need, so I have incorporated the selection of a voltage divider right after the AC coupling.
If the signal is large (and probably strong) we can scale it down. If it is small and weak we can send it straight to the buffer.

Now we come to the parts that aren't displayed in the schematic.

7. Calibration circuit
I have used the software with my microphone for a while and it doesn't seem to need calibration input. Actually I don't think there is one. I 'll leave that out as a result.

8. Output level indicator
An extra measure of precaution and a nice visual one too, would be if we could see the amplitude of the waveform before actually plugging the module to the PC. That way we could be sure that the input would be safe.
I though of the LM3915 logarithmic LED bar driver. It works nice with 9V supply, can detect peak amplitude and the log output is ideal for zeroing-in the output amplitude.
One issue is that I need to do a small statistic on the clipping voltage of various soundcards, so that I can set the top of my scale correctly.

9. Output overvoltage protection diodes
No one wants to fry his motherboard and I feel I can't be too careful. I thought of adding a pair of opposite-facing diodes between the module output and the Ground. That way the output voltage will never reach more than 0.7V of amplitude. If I find that all the soundcards that I test clip at about 200mV, I 'll think of adding germanium diodes. Can they handle possible failure situations?
I 'm also a bit concerned about any skewing the diodes might add to the input signal. Any thoughts?

10. Calibration Circuit
No, this isn't a contradiction. I 'm talking about the software/hardware amplitude scaling again. If I happen to find that each system scales the input in a different ratio, it would be nice to have a way to calibrate it without the need of a signal generator. I 'm thinking of a circuit that will produce a sine wave of fixed frequency and amplitude (say 100mV) internally. When I plug the module in the soundcard I can then see what impact that has and trim the amplitude trimmer accordingly.
However I don't have the faintest idea of what that circuit could be. Any input is wellcome.


I 'd also like to know your thoughts about the consumption of the system. Will the 9V batteries handle it?

I think that would be all for now. Thanks for having the patience to read so far. Looking forward to reading your feedback.

George

 

Something to think about, you can get the 4 way phone or data switches cheap nowdays. Instant case and rotary switch for a song.

 

(post #29) Bumpity bump...

 

What is the question? One of the things I meant to do is compare the finished design with a real Oscope. I suspect a sound card just isn't going to be that good. Frequency response is not likely to be flat, their is no DC associated with it, in short it is barely adequate for some one that doesn't have a scope. It is a way of seeing what the wave for looks like.

The board I built also has sockets for the op amps, so I can check out what difference various op amps make. This is so low quality I suspect 1458's will do well.

 

There are a few questions throughout the project description. I posted it for comments and possible improvements.

If you don't have to add anything Bill, that's fine. You don't have to do all of the work around here. I 'm already thankful for taking the time to post your two cents.