Waveforms + Measurements

[moodz]

Representative waveforms at the output of the diff amp.

The pulse repetition frequency is quite high at 5.3 kHz however in this application the DSP code needs the high rate for adequate sensitivity.

[WM6]

Representative waveforms at the output of the diff amp.

The pulse repetition frequency is quite high at 5.3 kHz however in this application the DSP code needs the high rate for adequate sensitivity.

Is this lasted under target or free run?

[moodz]

Is this lasted under target or free run?

The little bump you see is a result of slight imbalance. There is no target in this shot. The bump comes from slight underdamping on input ... it can be tuned by playing with damping resistor value and balancing capacitance on differential coil A / B leads however it will still work fine for sample and hold purposes.

[moodz]

sample target info

I am using a coil with 32 turns / 22 cm dia. It appears to have a 20 usec decay.
The amp gain in this case is set to 100 not 1000 as there is too much gain at this setting.

The trace below shows a plot for target / no target. The purple trace is the no-target reference trace. The target is an australian $1 coin ( .... this is not a gold coin BTW ) and was placed in the centre of the coil.
There is about 200mV of deflection caused by placing the target coin. IMHO this is quite good.

[WM6]

Not bad really.
Interesting to know if ideal balanced (without bump) can give better or worst deflection rate.

[moodz]

Not bad really.
Interesting to know if ideal balanced (without bump) can give better or worst deflection rate.

I had a look at the input to the diff amp post the clamping diodes and I cannot see the "bump" however there is a fast transitiion "spike" at this point where the clamping diodes stop conducting. One of the reasons the diff amp is DC connected to the coil is because it needs reference bias on its inputs ( see data sheet ) in order to operate. Because the coil centre tap is connected to analogue ground ( VPP ) this provides the bias to the amp. It is possible that the amp only starts working properly after the diodes stop conducting ... not sure. I think I will leave it till I get to a more advance input scheme like active clamping as Carl had suggested.

moodz.

[Carl-NC]

This could be an overdrive glitch from the amp; current-feedback amps often do this. As long as you're inside the range of the ADC at first sample, I wouldn't worry about it right now.

[moodz]

variable damping

Substitute this circuit for the main 1K damping resistor. ( ends of 330 ohm resistors )
I used 3 x 1k resitors in parallel.
VPH is the +VE bias voltage input to 78L05 reg on the diff amp supply.
Note this does add some parasitic cap which will add a few usec to flyback.
The pot is 10K however not that critical.



the installation .... neat huh ?

various settings of the pot are obviously changing the damping value.

[WM6]

Suppose that different fly-back shape signals on last 4 scope picture you get from different damping R value but need some explanation?

[KingJL]

Decay

One thing that the waveforms show is that the under-damped RX wants to resonate at ~110kHz. That somewhat explains the slow decay of ~25usec. The 22 cm diameter coil with 34 turns of twisted pair connected as described results in a total RX inductance of ~2.2mH (68 turns). This combined with a combined circuit capacitance of ~650pf-750pf (MOSFET 330pf, C1 220p, 5 41418 diodes total 20pf, coil 80pf (?), cable 50pf(?) calculates to ~120kHz which corresponds with what the waveforms tell us. I think moodz has done some great work and has advanced the PI technology discussion. I do like the concept of the differential front-end and I am excited to explore the possibilities, but I think we need to get a faster decay to be a viable PI design. I have been toying with a negative supply variant that exhibits ~175pf total capacitance and 800uH RX (effective 212uH TX) inductance that results in 12.5us-13us total decay (10mv @ 12us after amplification G=100). As I get the opportunity and time, I am planning to rig this and test it.
Great work, moodz!

[WM6]

One thing that the waveforms show is that the under-damped RX wants to resonate at ~110kHz. That somewhat explains the slow decay of ~25usec. The 22 cm diameter coil with 34 turns of twisted pair connected as described results in a total RX inductance of ~2.2mH (68 turns). This combined with a combined circuit capacitance of ~650pf-750pf (MOSFET 330pf, C1 220p, 5 41418 diodes total 20pf, coil 80pf (?), cable 50pf(?) calculates to ~120kHz which corresponds with what the waveforms tell us.

I am excited to explore the possibilities, but I think we need to get a faster decay to be a viable PI design.

Thanks for explanation KingJL.
It would be great if you and others from design group join this great moodz development.
With faster decay you mean something like this?:

[KingJL]

Thanks for explanation KingJL.
It would be great if you and others from design group join this great moodz development.
With faster decay you mean something like this?:

Not really. The base that you are indicating (the negative section of moodz display) is actually the value during TX. The decay goes to the proper base, but takes 25-30usec to get there (the 25-30usec positive section of moodz display). Remember you are seeing the decay amplified by G=100. For the 1st 20-25usec it is limited at the positive rail by the amplifier being in saturation (all this is normal), thus appears as a positive pulse. In a faster decay the apparent pulse would be shorter as in the attached waveform (the attached waveform does not show the TX pulse as it is sync'ed to start at TX turn-off and is not of long enough duration to show the next TX pulse).

[WM6]

Not really. The base that you are indicating (the negative section of moodz display) is actually the value during TX. The decay goes to the proper base, but takes 25-30usec to get there (the 25-30usec positive section of moodz display). Remember you are seeing the decay amplified by G=100. For the 1st 20-25usec it is limited at the positive rail by the amplifier being in saturation (all this is normal), thus appears as a positive pulse. In a faster decay the apparent pulse would be shorter as in the attached waveform (the attached waveform does not show the TX pulse as it is sync'ed to start at TX turn-off and is not of long enough duration to show the next TX pulse).

Thank you KingJL.

Evidently I need some term of part signal explanation for my "inglish" English.

Is this correct (regarding by letters marked parts of signal on attached picture)?:

A => rising pulse
B => pulse delay (pulse duration)
C => point of pulse decay (point of decay)
D => pulse decay
F => fly-back pulse duration
G => fly-back nulling
H => waiting time to next pulse (waiting time), waiting Period

I would be very thankful for help in this term clarification.

[KingJL]

Thank you KingJL.

Evidently I need some term of part signal explanation for my "inglish" English.

Is this correct (regarding by letters marked parts of signal on attached picture)?:

A => rising pulse
B => pulse delay (pulse duration)
C => point of pulse decay (point of decay)
D => pulse decay
F => fly-back pulse duration
G => fly-back nulling
H => waiting time to next pulse (waiting time), waiting Period

I would be very thankful for help in this term clarification.

The way I see them :
A=> start of flyback (end of TX)
B=> saturation during flyback
C=> coming out of saturation during decay
D=> ending of decay
F=> quiesent level (fully decayed)
G=> start of TX pulse
H=> TX pulse duration

Kind regards,
J. L. King

[moodz]

The way I see them :
A=> start of flyback (end of TX)
B=> saturation during flyback
C=> coming out of saturation during decay
D=> ending of decay
F=> quiesent level (fully decayed)
G=> start of TX pulse
H=> TX pulse duration

Kind regards,
J. L. King

Thanks for this analysis and your comments JL & WM .... this last sequence is correct as I understand it. The ADC starts a sample run from C to G ideally at 1 sample each usec.

regards, moodz.