TGSL Experiments

[nakky]

Do you mean this way ?

Grt Nakky

[Dennis the Mennis]

Do you mean this way ?

Grt Nakky

Hi Nakky ,

Yes , I was suggesting the bridge as a possiblility for the -5 volt connection .
Have you tried it allready ?

regards

Dennis the Mennis

[simonbaker]

Euros

Hey Stefano - thanks again for Euros - but you didn't have to go that far!

Euro Crash: Motorists Steal Spilled Coins

Italian Passersby Dig In after Truck Carrying 2 Million Euros in Coins Overturns and Dumps its Contents onto Highway

http://www.cbsnews.com/stories/2010/...n6651493.shtml

[Stefano]

Hey Stefano - thanks again for Euros - but you didn't have to go that far!

Euro Crash: Motorists Steal Spilled Coins

Italian Passersby Dig In after Truck Carrying 2 Million Euros in Coins Overturns and Dumps its Contents onto Highway

http://www.cbsnews.com/stories/2010/...n6651493.shtml

"where the accident occurred Monday, said Tuesday it was impossible to establish how much money had been stolen..."

If we think how much we pay to our government someone could have considered it a sort of tax refund.

Stefano

[simonbaker]

"where the accident occurred Monday, said Tuesday it was impossible to establish how much money had been stolen..."

If we think how much we pay to our government someone could have considered it a sort of tax refund.

Stefano

It all goes back into the economy!

Better get your TGSL is working, now's your chance to find some Euros that rolled away!

[dfbowers]

While working on TGSL Final circuit, I wondered how the resistance of the TX coil affects the current going through it, and thus the magnetic field strength, in the oscillator circuit.

I ran some LTSpice sims trying different resistances.

Here is graph I made in Excel showing the current in the coil vs. the resistance of the coil. It may be helpful for choosing wire guage. My test coil, about 30 gauge wire, is about 30 ohms.

Interesting to me is that lower resistance does not dramatically increase current, but it does give you some boost.

-SB

Simon,

Can I look at the circuit that you did the simulation on? Did you do the entire TGSL circuit vs. coil resistance?

Don

[simonbaker]

Simon,

Can I look at the circuit that you did the simulation on? Did you do the entire TGSL circuit vs. coil resistance?

Don

I'm pretty sure I used a stripped down oscillator, and was graphing the current in the TX coil vs the TX coil resistance. The current is available by right-clicking on the coil after a simulation. I don't know what I plotted! Probably peak current.

Attached are a few circuits.

1. One is just the bare oscillator like in the TGSL, no JFet feedback (TGSoscCorrected.zip).

2. One is oscillator like in original TGS with JFet feedback, but parts
are not exact (JFet and zener) (TGSoscJfet2.zip).

3. One has the bare oscillator plus discriminator front end (if LF353 doesn't function replace with something else). The GB front end is not there, but easy enough to add (testTGSosc.zip).

Probably the one I used is (1) for the graph, varying the internal resistance of the coil (by right-clicking on coil and changing param).

-SB

[simonbaker]

packaging

PCB packaged for next field test, wires less free to move, container to keep light out.

Tea anyone?

[simonbaker]

low freq response

I'll post here in my own thread so not clutter main TGSL troubleshooting thread.

Personally I'm interested in trying to see difference in PCBs because... I feel like it. No, I'm intentionally reinventing the wheel because often the rules become progressively restrictive and while excellent practice (which I wouldn't contest), I want to see what really are the critical factors in this particular case. It is sometimes important to not over-design, especially if you want to save money. Commercial manufacturers often do research (I would guess) to see how far they can cut corners to save cost. So the most beautiful robust designs are not always the end point.

To say a handwired pcb can't succeed may be too dismissive -- much of the TGSL is an extremely low frequency circuit where one might expect board layout effects to be much less important. Also, there is evidence that the dominent noise is from the coil acting as an antenna, not stray pickup because you didn't trim a solder blob low enough.

For me, it's fun to troubleshoot and investigate.

Granted, boring to others.

I write because it helps me think. So I'll post progress here even if not earth-shattering news, and people can ignore this thread. If anything useful pops up, I'll post it in TGSL Troubleshooting thread.

-----------------------------
More gain tests
--------------

I decided to check one more thing regarding the main filter stage of the TGSL. I previously injected a signal at 10 Hz and noted that dfbowers and my PCB both had virtually identical output.

However, noise has a spectrum, and if our filter stages have different spectral shape, that could also affect the resulting noise signal.

So I did the same test at 2 Hz and at 20 Hz.

At 20 Hz, no difference. I'm going to check 80 Hz to be complete.

But at 2 Hz, waveforms looked different. dfbowers output was a little higher and had a "peaked" look. Mine slightly less, more sinusoidal.

Not sure if it means anything. But I'll check some other things. I always wonder about the back-to-back capacitors. Doesn't make sense to me that at 2 Hz there should be a noticeable difference.

Video:

http://www.youtube.com/watch?v=xzaqLA99tZM

-SB

[Qiaozhi]

Simon - if you switch your scope inputs to AC, what is the peak-to-peak AC noise level at pin 5 U106b on both your and Don's PCBs?
I'm not sure if that's what you measured last time, but I want to be certain that you measured AC rather than DC.

[simonbaker]

Simon - if you switch your scope inputs to AC, what is the peak-to-peak AC noise level at pin 5 U106b on both your and Don's PCBs?
I'm not sure if that's what you measured last time, but I want to be certain that you measured AC rather than DC.

I'm actually typically measuring Pin 3(U106a) DISC channel, but theoretically should be identical.

My recent tests were for injected signal, not noise. I really can't measure the noise accurately it is so non-stationary. So it is an impression that whenever I switch the probes to Don's PCB, the noise seems lower. When I switch to mine, it seems higher. I just can't get a solid noise figure. I'll try to figure out some strategy to quantify it better.

For the gain tests with injected signal, I'm using DC coupling with my scope probes. At 2 Hz I wouldn't trust AC coupling measurements.

I generally use the vertical adjustment to find a convenient baseline to measure peak-to-peak from.

I'm going to repeat and also check the output of the previous stage so see I see the same distortion with dfbowers PCB.

BTW: I'm getting a weird anomaly with dfbowers PCB when I do same gain test at 80 Hz. The output is a "double" waveform at LM308 output -- meaning you see it twice but one slightly lower than the other.

I don't think I saw that with my PCB. But my scope could be going bad too. All part of troubleshooting, it's like trying to figure out a magic trick.

Here's a video of the anomaly. Probably of no significance whatsoever, or a mistake, but I'll check because didn't see it with my PCB.

http://www.youtube.com/watch?v=9BY29EKK7Kk

-SB

[simonbaker]

one mystery solved -- ?

I believe I solved the mystery of dfbowers "peaked" wave during my signal injection test to test the filter sections of our PCBs.

And it is the opposite of what I would expect given the excellent performance of the the MD.

(To recap -- injecting a tiny signal at the input of the LM356 gives an amplified signal at the output of the LM308 (using DISC channel).

I noticed that the output signal on dfbowers PCB looks less sinusoidal than mine; in fact quite "peaked" with occasional wiggles and jumps.)

I randomly noticed that when I plugged in some headphones, the output signal looked really distorted, jumping sharply with each beep. (why is it beeping when the GB channel has no input... geez I'll check...)

Anyway, I unplugged the headphones and pondered it some more. The obvious suspect is one or more of the "rails" (ground, V+, probably not V- because audio not connected there).

So I ran a jumper from the "speaker" ground to the battery ground.

Voila - the signal became much more sinusoidal. I reconnected the headphones, and instead of a short, choked beep, it was now a longer, smooth beep.

Conclusion -- the audio circuit (even with headphones unplugged) -- causes distortion of signals running through the filter/amplifier section, and it seems to be because of loading of the ground buss.

The irony in all this is that dfbowers TGSL is an excellent machine! Apparently my PCB has less loading from the audio, but his PCB seems to be superior -- maybe because of better handling of coil noise somehow (which I have yet to identify the reason).

However, it reinforces my desire to adopt a PCB layout where the audio ground connects directly to the battery ground. (Unless it turns out this loading effect actually improves performance somehow!!!)

I have attached video showing the difference in audio beep and output waveform with and without the "ground jumper". Note the output waveform is the smaller one at the top -- sorry, it gets confused with the larger input waveform at the bottom a little.

http://www.youtube.com/watch?v=JldwJ9HeqXQ

BTW: using scope probe, it seems that the audio affects the output of the LM358 also, but more difficult to evaluate because of smaller signal.

-SB

[dfbowers]

I believe I solved the mystery of dfbowers "peaked" wave during my signal injection test to test the filter sections of our PCBs.

And it is the opposite of what I would expect given the excellent performance of the the MD.

(To recap -- injecting a tiny signal at the input of the LM356 gives an amplified signal at the output of the LM308 (using DISC channel).

I noticed that the output signal on dfbowers PCB looks less sinusoidal than mine; in fact quite "peaked" with occasional wiggles and jumps.)

I randomly noticed that when I plugged in some headphones, the output signal looked really distorted, jumping sharply with each beep. (why is it beeping when the GB channel has no input... geez I'll check...)

Anyway, I unplugged the headphones and pondered it some more. The obvious suspect is one or more of the "rails" (ground, V+, probably not V- because audio not connected there).

So I ran a jumper from the "speaker" ground to the battery ground.

Voila - the signal became much more sinusoidal. I reconnected the headphones, and instead of a short, choked beep, it was now a longer, smooth beep.

Conclusion -- the audio circuit (even with headphones unplugged) -- causes distortion of signals running through the filter/amplifier section, and it seems to be because of loading of the ground buss.

The irony in all this is that dfbowers TGSL is an excellent machine! Apparently my PCB has less loading from the audio, but his PCB seems to be superior -- maybe because of better handling of coil noise somehow (which I have yet to identify the reason).

However, it reinforces my desire to adopt a PCB layout where the audio ground connects directly to the battery ground. (Unless it turns out this loading effect actually improves performance somehow!!!)

I have attached video showing the difference in audio beep and output waveform with and without the "ground jumper". Note the output waveform is the smaller one at the top -- sorry, it gets confused with the larger input waveform at the bottom a little.

http://www.youtube.com/watch?v=JldwJ9HeqXQ

BTW: using scope probe, it seems that the audio affects the output of the LM358 also, but more difficult to evaluate because of smaller signal.

-SB

Simon,

That's quite an observation that you have made. Now are you isolating the speaker ground from the rest of the grounding circuit? Can you do quick sketch on the schematic so I understand how you wired it? I guess I'm just making sure that I don't somehow short the entire grounding circuit to something I'm not supposed to. I would like to duplicate your results if possible. If we can't add any depth, maybe a way of improving audio? Also would be interested is seeing what SPICE models..

Thanks!
Don

[lunamay]

what do you think about speakers..i mean which is bether: piezoelectric or standard 4 or 8 ohm

another thing about jack conector to headphones..should i conected it directly to speaker(some jacks has something i can called switch. when you put jack in jack conector speaker is"cut of circuit "and all signal is going directly to phones)any thoughts??

best regards

martin

[simonbaker]

Simon,

That's quite an observation that you have made. Now are you isolating the speaker ground from the rest of the grounding circuit? Can you do quick sketch on the schematic so I understand how you wired it? I guess I'm just making sure that I don't somehow short the entire grounding circuit to something I'm not supposed to. I would like to duplicate your results if possible. If we can't add any depth, maybe a way of improving audio? Also would be interested is seeing what SPICE models..

Thanks!
Don

Hi Don:

First, I don't trust any test I do unless I do it at least three to thirty-three times! So will continue to check it out.

Quick answer -- I didn't isolate the audio ground, just jumpered it in parallel, sort of load relief on the main ground buss -- arterial bypass so to speak. Yes, I created a "ground loop".

My test was somewhat artificial as I injected the signal I was studying. Still, adds to my feeling that a better audio ground may be a good idea.

See attached photo. The green lines are jumpers I clipped on to perform the test. One is just to short the RX coil input to keep it from adding another signal. The other is the ground-to-ground shunt that I theorized would relieve some of the ground buss lifting by the audio pulse. In a revised circuit, you would also cut the ground connection from the speaker to it's current ground buss connection.

You can also see where I injected the signal and measured the output.

It was dang tricky to get clips to stay put, so always some room for error.

Now -- why did I get a beep with only one channel? That troubles me, so I need to check to make sure this isn't all BS.

-SB (BS backwards???)