IF YOU HAVE MAKE , TRY THIS .

[sisco]

Phase-angle shift between TX and RX is used to disc the object.

this mean :in golden sabre four FETs after preamp use for switch in difference

phase-angle (infact like difference delay for PI).

if this is true can tell about timing (phase-angle)?

[Max]

this mean :in golden sabre four FETs after preamp use for switch in difference

phase-angle (infact like difference delay for PI).

if this is true can tell about timing (phase-angle)?

Hi,
it's not the same thing.

So there isn't any direct relationship between phase angle you can see in IBs and "timings" you use in PIs.

IB/VLF use continuos sine wave/signal at TX (at least very often) and RX is sine wave too.

The phase angle results from many factors related to the object composition but also e.g. on ground minerals influence that is nulled by GEB circuitry.

In IB/VLF we talk about "frequency domain" analisys, where in PI we talk about "time domain" analisys.

Problem in time domain analisys is that you haven't a parameter like phase-angle to directly get disc informations. That's why some PI with disc use sampling and DSP to guess e.g. conductivity of target and then giving an estimate of what metal could be.

So, as far as I know, just timing is not enough to build a reliable disc system for PIs.

Best regards,
Max

[sisco]

I could't find AD797 and OPA37 so change may PCB layout for use NE5534P

(OR AD844) .I think can use NE5534P (OR AD844) for preamp and first amp

(before sampler).for after sampler must test between TL0X1 series .

this is my PCB for test my idea . first I want make this with four CHANNEL ,

but after try do this saw must use large PCB with near 22 ICs , so for first

step design this PCB with two CHANNEL for test.

forget told , for power part use GOLDSCAN 4 power unit . I think better than

ICL7660 , because must supply for near 11 ICs.(9 OP-AMP + 2 4051)

this power unit enough or not ?

[Max]

I could't find AD797 and OPA37 so change may PCB layout for use NE5534P

(OR AD844) .I think can use NE5534P (OR AD844) for preamp and first amp

(before sampler).for after sampler must test between TL0X1 series .

this is my PCB for test my idea . first I want make this with four CHANNEL ,

but after try do this saw must use large PCB with near 22 ICs , so for first

step design this PCB with two CHANNEL for test.

forget told , for power part use GOLDSCAN 4 power unit . I think better than

ICL7660 , because must supply for near 11 ICs.(9 OP-AMP + 2 4051)

this power unit enough or not ?

Hi,
depends which op. amps. but if you look at GS4 schematic there are TL074 etc powered that way, so I think it could be ok for 9 single op. amp. and 2 cmos.

NE5534 need more current than others TL0x1 but is just one, so it's ok.

Best regards,
Max

[sisco]

you think difference between single op-amp and dual op-amp and what is

better ? single or dual or not difference .

if offsetvoltage not important I think use dual op-amp better .

(or use single op-amp with adj offsetvoltage for sensivity PI is better ?)

[Max]

you think difference between single op-amp and dual op-amp and what is

better ? single or dual or not difference .

if offsetvoltage not important I think use dual op-amp better .

(or use single op-amp with adj offsetvoltage for sensivity PI is better ?)

Hi,
offset is a problem in preamp. if too big... so there I suggest always use a single op. amp. with eventually a nulling trimpot, if needed.

For others I think you just need low noise op. amps. with limited offset, no need of manual nulling there.

If you use e.g. TLC2262 you have low noise and low offset... but also using TL0xx it's about the same... you don't need offset null in most of the cases, even with hi-aplification.

About single-double-quad... well not big differences if you use all op. amp. inside the package... or connect inputs to gnd if don't use (don't leave them floating)... then you can use e.g. guard rings or shielding planes to block self-oscillations danger when use more op. amps that e.g. share same supply (e.g. a TL074), most of the times you don't need that cause of hi-separation between "channels" (op. amps) in a single container product.

But are good pratices.

Best regards,
Max

[sisco]

1 hour ago told me find OPA37 and I ordred 3 OPA37 , because this is very

heavy for me see price :

NE5534 : 0.28 Euro

LF356 : 0.33 Euro

LF357 (metall pack) : 3.75 Euro

AD844 : 5 Euro

OPA37 : 11 Euro

so I use for PREAMPs (before sampler) OPA37 and for use amp after sampler

what is better use TL serie or NE5534 ?

TLC2262 for preamp or amp ?

[Max]

1 hour ago told me find OPA37 and I ordred 3 OPA37 , because this is very

heavy for me see price :

NE5534 : 0.28 Euro

LF356 : 0.33 Euro

LF357 (metall pack) : 3.75 Euro

AD844 : 5 Euro

OPA37 : 11 Euro

so I use for PREAMPs (before sampler) OPA37 and for use amp after sampler

what is better use TL serie or NE5534 ?

TLC2262 for preamp or amp ?

Hi,
yes AD and OPA are generally speaking more expensive... than older and easy stuff like TL0xx.

I suggest TLC2262 (that is very cheap) for amplifier/integrators section.
You could use also older stuff for this like LS404 (quad low noise, very cheap now but obsolete) MC33079 (very good, cheap quad) or also TL074 (quad) or TL072 (dual); also TL084 and TL082 are fine for test like TL074/72.

I don't suggest using NE5532, dual version of 5534 cause one ic need 6mA of current and seems very much respect to the others e.g. TLC2262 (muche less current required and similar performances).

For preamp. I'll use good one OPA37 is very good one, you made good choice with it.

You can use also 5534 or lf356/357 for preamp... but OPA is better if you can find/buy it at a reasonable price... 11eur each is very expensive !

Kind regards,
Max

[sisco]

tomorrow I will be start build my new PI with this parameters .

about preamp GAIN , as yet not decide .

i think for total gain , 1500 is good , but how must Apportion between two part

of preamp , i don't know. (50 - 30 or 60 - 25 or 30 - 50 .....?)

[Max]

tomorrow I will be start build my new PI with this parameters .

about preamp GAIN , as yet not decide .

i think for total gain , 1500 is good , but how must Apportion between two part

of preamp , i don't know. (50 - 30 or 60 - 25 or 30 - 50 .....?)

Hi,
in your circuit change few ! Cause you use always OPA37 there.
Same noise level, same offset etc...

In older designs the preamp. must give a small amplification cause of noise and offset higher than you have with OPA37.

But is always a good idea for me. I'll use 30-50 -> 1500.
Using small 1st preamp. amplification is good, always, cause that way you minimize impact of e.g. offset due to gain before further amplification.

Best regards,
Max

[sisco]

Hi Max

my PCB is ready but parts is not complete Assembly.

see schematic and tell , my part value is good or must change ?

please see OPA37 datasheet and tell me C1 value is good or must use 5pf

capacitor or not or better replace with capacitor + resistor(serial) ?

[sisco]

see this compare between three PI:

PI.................preamp gain...................amp1 gain.................amp2 gain

minipuls.........1000............................. 220..........................1 - 23

hammerhead...1000............................100.. .........................100

this PI...........1500............................370.. .........................68

total gain:

minipuls = 1000 x 220 x 23 = 5,060,000
hh = 1000 x 100 x 100 = 10,000,000
this PI = 1500 x 370 x 68 = 37,740,000

you don't think my total gain is very high or cause i use ultralow noise

op-amp this is not problem?

[Max]

see this compare between three PI:

PI.................preamp gain...................amp1 gain.................amp2 gain

minipuls.........1000............................. 220..........................1 - 23

hammerhead...1000............................100.. .........................100

this PI...........1500............................370.. .........................68

total gain:

minipuls = 1000 x 220 x 23 = 5,060,000
hh = 1000 x 100 x 100 = 10,000,000
this PI = 1500 x 370 x 68 = 37,740,000

you don't think my total gain is very high or cause i use ultralow noise

op-amp this is not problem?

Hi,
yes I think it's high.
Anything from 10^6 to 10^7 is ok... but more could give instability of circuit.
Yor preamp gain seems ok to me, so you maybe have to consider reducing the gain in the central section, something like this:

1500x100x100 = 1.5*10^7 or 15,000,000

Of course, using very low noise components there is a smaller risk of having erratic behaviour at higher gain, but you must consider also that signal due to target decay very fast (exponentially) with target distance, so a huge amplification can't make you gain more than few cms with small targets.

I think that 15,000,000 could be a good compromise between stability and sensitivity of the whole chain of amplifiers.

Best regards,
Max

[sisco]

about C1 how ? this is determinable with frequency operation (i do't think)

or OP-AMP parameters ?

about total gain , OK I reduce amp1 and amp2 gain to 10,000.

but how must Apportion between amp1 and amp2 ? you told 100 x 100

can like preamp1 and preamp2 for amp1 gain more than amp2 for example

50 x 200 or 200 x 50 or really not difference ?

amp2 after couple C (C5) so must use gain more than amp1 or not?

[Max]

about C1 how ? this is determinable with frequency operation (i do't think)

or OP-AMP parameters ?

about total gain , OK I reduce amp1 and amp2 gain to 10,000.

but how must Apportion between amp1 and amp2 ? you told 100 x 100

can like preamp1 and preamp2 for amp1 gain more than amp2 for example

50 x 200 or 200 x 50 or really not difference ?

amp2 after couple C (C5) so must use gain more than amp1 or not?

Hi,
C1 is needed to eliminate the pole due to Rf and Cin... feedback resistor and input capacitance.

This Rf+C1 effect cause your preamp will act like 1st order low pass filter with frequency -3db depending on C1 and Rf values.

A small vaue is needed to prevent the pole action give you self-oscillations and to not introduce a cutting-edge at few Khz from filter effect.

I've not done all calculations but can suggest you use 22pF with your setup.

Best regards,
Max