Delta Pulse

[kt315]

Vadim18 I buy all chips and components from china. The detection depth is same for 5kg+ weight pure iron pieces of same size. The 1mx1m coil can detect these 5kg iron pieces at the same distance. This detection depth is trash. Absolute scam. It is very disappointing after spending so much effort and money. Such false claims and fake videos only to sell trash pcb. Now it is very clear why the seller of this design is always playing angry when you ask him questions about the poor performance. He will always attribute the wrong to the customer and if nothing works then will start abusive remarks and language. No wonder he would always advise to get parts from trash. Do not invest much time and money in to this. Its once in life time maybe that you would got search for train burried 15 feet deep.

i get orders again and again. despite on your negative claims.


hi

i am from (deleted for privacy) , i want to buy 2 pc's delta pulse pcb
Previusly i bought 2 pcs from you and worked great.how i must pay?

[dbanner]

Big coil requires proper gauge wire, connector and cable AND correct damping.
How in the world you're gonna test 1x1 meter big coil in the lab. NO, NO! Take it to the bush!

[kt315]

https://www.geotech1.com/forums/showthread.php?11831-Cache-Test

really testing on cache. 24 inch is 60.96cm.

[6666]

I begin my tests with LF355, TL071, LF356. With LF355 my unit work super stable, can detect aluminium cola can to 60 cm distance with 27cm coil.
Will run with LF357 better,

Why are you not happy with 60cm ?, those 357 are hard to get.

[Vadim18]

They are good chips, but they both wil not run in this circuit, by "common-mode input voltage range" limits.
I begin my tests with LF355, TL071, LF356. With LF355 my unit work super stable, can detect aluminium cola can to 60 cm distance with 27cm coil.
Will run with LF357 better, but not so much. Will detect more noise than signal. Main key of sensivity are integrator stage.

[chuubaba]

Vadim18 the opamp required by BOM is 50v/us. I have CA3130N and LM318n with slew rates of 30v/uS and 150v/us. Is this not sufficient? The chips are direct from the manufacturer.

[Vadim18]

I think problem not in PCB, but in components in ic-s.
Today parts from trash may be not so atractive but they are 99% original.
By the way! I got 2 fake LF357( remarked TL071) and 1 original not from China, but from №1 reputable store in Ukraine. Last week i got 5 pieces of TL072, all with oxidized legs also from reputable store. Thats a worldwide tendency
You can build simple square wave generator to check and compare how fast you op amps. Tl062 may be remarked lm358, LF357 may be TL071 (like in my case)

Chek you op amps first.

[kt315]

if my pcb is trash you can take files i began from and to do your new. begin from zero point.
thats not so hard. then you may blame just yourself. for curve hands.

[chuubaba]

Vadim18 I buy all chips and components from china. The detection depth is same for 5kg+ weight pure iron pieces of same size. The 1mx1m coil can detect these 5kg iron pieces at the same distance. This detection depth is trash. Absolute scam. It is very disappointing after spending so much effort and money. Such false claims and fake videos only to sell trash pcb. Now it is very clear why the seller of this design is always playing angry when you ask him questions about the poor performance. He will always attribute the wrong to the customer and if nothing works then will start abusive remarks and language. No wonder he would always advise to get parts from trash. Do not invest much time and money in to this. Its once in life time maybe that you would got search for train burried 15 feet deep.

[Vadim18]

Great news chuubaba!
You doing a good progress.

There is dificult for me to point wich component make a problem. But at this time you get a working unit, you must be able to detect a cola can at this distance (60-90 cm is most realistic). Stainless steel are bad target for detection, even in my case.
Here a common ways to make a bad working detector:

1 bimetal wire for coil, from utp5 twisted pair;
2 metal wire type resistor, R5 and R6.
3 fake op amps or op amp with damaged inputs.You can also broke inputs by static then soldering it. LF357 are 95% fake at nowedays, use instead LF355, LF356 or TL071.;
4 unshielded and long wires at signal path, in case of DP thats are Sensitivity pot wires;
5 bad "used condition" electrolityc capacitors, especially near IRF740 and power input.

Also ceramic capacitors in place of c16 c18 c22 c25 are not good.
In case you not have an Oscilloscope or logic analyser - all resistors and caps must be werified with DMM with +-1 % tolerance.

Dont try to understand all at once, i start to make my first MD in 2016 (SurfPI) and only after 2 years i make a good working unit) I made a lot of mistakes, also learned more.


By the way big thanks to Anatoliy, i choosed his schematics as most adequate for bulding DP.

[kt315]

all holes on pcb are metallized. if you desolder a component you must do that accuratelly. VERY ACCURATELY.
you must remember that pcb was done by chinese, NOT BY UNION OF SOVIET SOCILALIST REPUBLICS/USSR.
USSR made board had FAT METALLIC LAYER EVERYWHERE, chinese put thin layer. if i will order pcb batch with 2x fatter metal layer
the price also increases in 2x time.

NOT USSR MADE BOARDS -! CHINESE -! JUST YOUR ACCURACY AND HIGH ATTENTION -!
LOWEST ADHESION OF THE METAL TO TEXTOLITE -! THIN METAL LAYER -!
you pull out metal layer from holes and do not want to see this fact.

all your problems appear just you begin to desolder something and change.
----
The entire PCB and POTs are inside a special metal box to block EMF
----
again just your crazy fantazy. YOU CAN NOT PROTECT THE PCB FROM EMF!!!!! your coil is OPENED FROM ALL SIDES FOR EMF!!!!!

[chuubaba]

Hi chuubaba.
..........Often this problem may be coused by broken diodes 1n4148 ....................

Thank you Vadim18. The 1N4148 diode located at VD3 was faulty.


I am sure you have very good understanding of this DP PCB and you can suggest a fix for "low detection depth" problem with big coil. I am using all components as written in BOM for rev 5.0
The frequency is 100Hzs. All resistors are 1%. MOSFET is 740 and VD1 is DSEI2010. Damping resistor is 680R 1W. Coil diameter is 400mm inductance is 396uH. Coil and connecting cable resistance is 6.7R. The entire PCB and POTs are inside a special metal box to block EMF . I can detect the iron in my concrete roof from 3 meters. A stainless steel pot size 95mmx105mm at distance of 76CM.

[Vadim18]

Delta pulse is giving nonstop tone and not detecting anything at all. I remove the MOSFET and the diode but it is still beeping.

Hi chuubaba.
If your IC (LF357, TL062) seats in sockets, try to take out each of them, only one ic step-by step and see what happens. At first take out D4 (LF357), or desolder Sensitivity resistor, or one leg of R20 if you not have sockets on the board.
Often this problem may be coused by broken diodes 1n4148 (low quality) and D4, or simply by fake D4 op-amp.

[chuubaba]

Delta pulse is giving nonstop tone and not detecting anything at all. I remove the MOSFET and the diode but it is still beeping.

[kt315]

i give pots' presets because a newbie for the primary testing has not to fear big risk.
if newbie got first result he can encrease PULSE WIDTH/DEPTH and SENS to more hight value seeking the edge he gets the maximal sensitivity/depth with the acceptable stability.
so follow step for newbie is to set DEPTH/PULSE WIDTH to 250-300us, SENS to 7-8 point (if pot max graduation/scale is 10).
TRESH pot is working specifically and has to be always in middle/neitral position.

Anatloy