EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: Porch on March 11, 2016, 05:33:22 am
-
I just picked up a GM328A http://www.banggood.com/GM328A-Transistor-Tester-Graphic-Wave-Signal-LCRRLCPWMESR-Meter-Inductance-p-997582.html (http://www.banggood.com/GM328A-Transistor-Tester-Graphic-Wave-Signal-LCRRLCPWMESR-Meter-Inductance-p-997582.html)Transistor Tester From BangGood. It's a sweet little device with a little frequency generator up to 2Mz.
So I connected to by Tek 465 scope and see what It will do. I don't have a better frequency generator and my home circuits don't go above 50Kz, so I have never tried this before.
My probes are PP-150 https://www.circuitspecialists.com/pp-150.html (https://www.circuitspecialists.com/pp-150.html)
Bandwidth:
X1: DC-6MHz
X10: DC-100MHz
See photos
With the probe set to X10, the waveform looks something like a square wave. I am sure the GM328A is not perfect, but this is just for fun.
With the probe set to X1, it's a sawtooth.
I watched Dave's video on 1X probes, and I would expect some difference between the X1 and X10 settings, but this is surprising. At less the half the probes rated bandwidth at X1, and the waveform is a total mess.
Is this the result of cheap probes not getting near their specification, or something else? Thanks for the education.
-
Test details matter... What's that frequency? Did you compensate the *10 probe? What does the trace look like on a standard logic output?
Failing those, a high-impedance output driving the probe tip capacitance will produce the textbook exponential transition.
-
You need to consider that a "2 MHz" square wave isn't simply a "2 MHz" signal.
A 2 MHz sine wave is a 2 MHz signal. Any other repeating waveform has harmonics (multiples) of the fundamental frequency (2 MHz in this case). A square wave is a sort of worst-case scenario as far as harmonics; it has more of them at higher levels than other other wave shape.
If you pass a really good clean 2 MHz square wave through a 3 MHz "brick wall" low pass filter, you will see only a 2 MHz sine wave at the output, because you have removed all the high frequency components (the harmonics) that make a square wave "square".
In your specific case, there are two effects at work. First, the limited rated frequency response of the probe in 1x mode, only 6 MHz, which is quite insufficient to display a 2 MHz square wave. More important though is the high capacitance of the probe in 1x mode. When in 1x mode, the input at the probe tip is simply connected directly into the coax cable that leads to the scope. The problem is that this cable has a fairly significant capacitance. This is directly responsible for the triangle wave that you see: the capacitance is being slowly charged then discharged each cycle, leading to a ramp up/ ramp down effect. This also depends on the output impedance (drive strength) of the signal generator. A lower impedance means a faster ramp up/down, so something closer to a square wave result. In this case, the transistor tester signal generator has 680 ohm resistors in series with all outputs, so you have a relatively very high impedance source, hence slow ramp up/down and a very triangular looking result. (actually it clearly looks like an exponential capacitor charging curve (https://www.google.ca/search?q=capacitor+exponential+curve&source=lnms&tbm=isch&sa=X&ved=0ahUKEwj7wtrQ-rjLAhUnnoMKHcZ5Cb0Q_AUIBygB&biw=1283&bih=818), which it is).
When the probe is set to 10x mode, there is a 9 MOhm resistor (along with other things) in series with the signal, and this decouples the signal from the capacitive load of the cable. 10x probes have their own limitations, but they are worlds better than 1x probes for anything other than DC or very low frequency, very low level signals.
-
In 1X mode, the input capacitance is going to be on the order of 100pF or so - that will place a pretty heavy load on your signal.
You might want to review my video on 1X and 10X probes:
https://www.youtube.com/watch?v=SX4HGNWBe5M (https://www.youtube.com/watch?v=SX4HGNWBe5M)
-
Looks like the main problem was the transistor tester could not overcome the capacitance of the probes set to 1X.
I used a Mosfet driver and the waveform was nearly the same. No more sawtooth.
See attached images. One at 1X, other at 10X.
Note that at 10X, it shows the noise in the signal, and it 1X, it's clean.
I had fun changing the noise by trying different bypass caps in the circuit. Strangely, putting a bypass cap on the power supply rail created more noise on the waveform.
Didn't know the probes are rated with a sine wave, but I can see how that works now.
Thanks for all the advice, I learned a lot reading and experimenting.