Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Topics - woody

Pages: 1 2 [3] 4 5 6
General Discussion / Inside the GPX-4500 andGPX-5000
« on: December 20, 2016, 09:18:39 pm »
Has anyone ever wondered just how they work?  These detectors have 2 main hearts, a CLPD, a complex logic device that is located right next to the Hitachi processor ic, these 2 devices control the timing for various tx and rx functions and general housekeeping as like controlling the electronics when switches are changed. The input stake is much the same as the GP series except they are using a reverse biased n channel mosfet as the second device in the tx on voltage and back emf blocking, a dual gate mosfet is used to hold the input ad797 inputs to ground when not in receive mode. The only difference is that in previous detectors the mosfets were connected to ground but now there are 1k resistors in series to ground. Next in line are the usual bank of 4053 switches and as a slight departure from old designs 10 sets of dual opamps. power supply and feedback is much the same as the old GP design.

General Discussion / DHL and Customs clearance.
« on: December 16, 2016, 06:28:45 pm »
Here, Express service from Kazakstan to Australia......using DHL,   FEDEX......NO PROBLEM    DHL....Always a bloody problem!

If you send me a detector make sure you add "For repair and return" Exempt under section 136.2 of the Australian Customs act.

IT technically is a "Temporary import"  DHL do not do Temporary imports and hand ball it off.   The message is DO NOT USE DHL!

2016   Location   Time   Piece
36   Customs status updated   SYDNEY - AUSTRALIA   14:17   
35   Clearance Delay   SYDNEY - AUSTRALIA   05:33   
Thursday, December 15, 2016   Location   Time   Piece
34   Clearance Delay   SYDNEY - AUSTRALIA   07:11   
Wednesday, December 14, 2016   Location   Time   Piece
33   Clearance Delay   SYDNEY - AUSTRALIA   05:42   
Tuesday, December 13, 2016   Location   Time   Piece
32   Clearance Delay   SYDNEY - AUSTRALIA   05:28   
Monday, December 12, 2016   Location   Time   Piece
31   Clearance Delay   SYDNEY - AUSTRALIA   07:42   
Sunday, December 11, 2016   Location   Time   Piece
30   Clearance Delay   SYDNEY - AUSTRALIA   05:24   
Saturday, December 10, 2016   Location   Time   Piece
29   Clearance Delay   SYDNEY - AUSTRALIA   06:25   
Friday, December 09, 2016   Location   Time   Piece
28   Clearance Delay   SYDNEY - AUSTRALIA   04:59   
Thursday, December 08, 2016   Location   Time   Piece
27   Clearance Delay   SYDNEY - AUSTRALIA   07:01   
Wednesday, December 07, 2016   Location   Time   Piece
26   Clearance Delay   SYDNEY - AUSTRALIA   08:37   
Tuesday, December 06, 2016   Location   Time   Piece
25   Clearance Delay   SYDNEY - AUSTRALIA   05:24   
Monday, December 05, 2016   Location   Time   Piece
24   Clearance Delay   SYDNEY - AUSTRALIA   02:54   
Sunday, December 04, 2016   Location   Time   Piece
23   Clearance Delay   SYDNEY - AUSTRALIA   06:49   
Friday, December 02, 2016   Location   Time   Piece
22   Customs status updated   SYDNEY - AUSTRALIA   09:17   
21   Clearance Delay   SYDNEY - AUSTRALIA   06:38   
Thursday, December 01, 2016   Location   Time   Piece
20   Processed for clearance at SYDNEY - AUSTRALIA   SYDNEY - AUSTRALIA   11:17   
1 Piece
19   Clearance Delay   SYDNEY - AUSTRALIA   11:17   
1 Piece
18   Arrived at Sort Facility SYDNEY - AUSTRALIA   SYDNEY - AUSTRALIA   11:14   
1 Piece
17   Customs status updated   SYDNEY - AUSTRALIA   06:52   
Wednesday, November 30, 2016   Location   Time   Piece
1 Piece
1 Piece
1 Piece
1 Piece
12   Customs status updated   INCHEON OUTSKIRT OF SEOUL - KOREA, REPUBLIC OF (SOUTH K.)   01:43   
Tuesday, November 29, 2016   Location   Time   Piece
11   Departed Facility in ALMATY - KAZAKHSTAN   ALMATY - KAZAKHSTAN   18:44   
1 Piece
10   Processed at ALMATY - KAZAKHSTAN   ALMATY - KAZAKHSTAN   16:21   
1 Piece
9   Clearance processing complete at ALMATY - KAZAKHSTAN   ALMATY - KAZAKHSTAN   16:11   
1 Piece
8   Uncontrollable Clearance Delay   ALMATY - KAZAKHSTAN   08:42   
1 Piece
Monday, November 28, 2016   Location   Time   Piece
7   Uncontrollable Clearance Delay   ALMATY - KAZAKHSTAN   18:41   
1 Piece
Sunday, November 27, 2016   Location   Time   Piece
6   Uncontrollable Clearance Delay   ALMATY - KAZAKHSTAN   23:23   
1 Piece
5   Processed for clearance at ALMATY - KAZAKHSTAN   ALMATY - KAZAKHSTAN   22:57   
1 Piece
4   Arrived at Sort Facility ALMATY - KAZAKHSTAN   ALMATY - KAZAKHSTAN   22:16   
1 Piece
Saturday, November 26, 2016   Location   Time   Piece
1 Piece
1 Piece
1   Shipment picked up   UST-KAMENOGORSK - KAZAKHSTAN   11:57   
1 Piece

General Discussion / Modification to get rid of power line noise and EMI
« on: December 16, 2016, 06:17:43 pm »
Who would be interested in a mod that totally removes power line noise and 90% of all EMI over 500Khz? 

I have carried out some testing using ADC and DAC with digital filters, seems to work really well but have to sample at very high rates like 5-10 MSPS or it causes signal loss, it may be better to try notching in the time and frequency domain after the 4051

Had a crazy fault today, the detector works without the coil connected but shuts down when a coil is connected, LCD screen stays on and the negative 15 volt rail that supplies the output Fets hits 25 volts and shuts down the negative supply. Hmmm thought it might be in the voltage comparator circuit around the TLE2022 , voltage reference or divider resistors gone high?  No, I made the detector force itself on without the coil and removed the negative tx rail capacitor as it was getting over voltages. then hmmm a burning smell, the second blocking Fet to the AD797 went up in smoke.  Great so I removed it and the input gets as these possibly had gone short and allowed tx voltage to take out the second Fet.  Then hmmm more burning smell , the ADG333 was growing a brown spot in the middle so I put in a new ADG333 and fired the detector up.  Well it worked, there are probably some parts that were stressed out as the ADG333 supplies switching voltages to the damping resistor control Fet, the input protection Fets, the GAas Fet clamps and TX drive.  Why a failure in the ADG333 caused the shutdown was strange as all waveforms were correct before it smoked itself.  So if you get this fault and the detector shuts down it would be wise to think the ADG333 is the problem as trying to find a shutdown problem like this could be a real bastard fault to track out. I did have another customers detector that did a similar fault and came up with the dreaded coil over current problem, it was a sneaky fault as it was a loading problem that only showed when the coil was connected, it turned out to be the FMMT718 pop bipolar that is in part of the 5 volts power supply. These transistors are a low loss switching transistor and if the VCE rises due to thermal issues the detector drops out the +5 volt rail. If you test the transistor it will test fine but the detector will shut down, change the 718 and all is good again.   Anyway it is interesting working on the Minelab detectors, sometimes frustrating but mainly straight forward.

Matt was out using the latest mods for small gold, Variable Frequency set to 50, front gain set to 4 and back gain set to 5. 
This was detected in NORMAL Timings, not enhance and it was on very hot ground not far from Moliagul in Victoria.
The Gold ended up weighing in at 0.2 grams.  Not bad considering the size of the coil and using such a low frequency of 50 and low gain.
Notice how quiet the detector is running, this is exceptional for normal timings, I doubt the SDC2300 would do any better.
All testing and claims are now made with a corresponding video to show it how it is.

Technical / Inside the GPZ7000
« on: October 02, 2016, 12:17:41 am »

The soft start or current inrush protection mosfet has burned a hole straight throng the circuit board.

This is starting to happen on older detectors, the reason could be just thermal cycling or metal migration causing the failure.

The small part that burns out is the Fairchild 455 mosfet, it is pictured on top of the white transformer core, the replacement is a 40 amp mosfet as shown with its 3 legs soldered to where the small
device was located.  The lower the on resistance as in the more amps it handles the better.

The FET that i used is a 100 volt rated 40 amp N channel mosfet but any high current mosfet will do the job, I left the legs on the set long as to make the connections easier to see.

This circuit board is not repairable as the power was left connected to the detector after it failed. The message here is to remove power from the detector if it stops working or more damage can be caused
making the detector unrepairable.

Technical / Isolating the input gate switching mosfets from signal ground.
« on: September 19, 2016, 01:03:48 am »
In the GPX series detectors the gate drive to the input gets is directly driven from the ADG333 via a BC857 directly to the gates, my take on this is that the designer tried to get as much gate saturation as possible lowering the conduction resistance of the fets. To get lower input noise and less gate capacitance to ground it is advisable to cut the track from the emitter of the 857 and add a smd 10 ohm resistor in series. This value can be played with but take it too high and the Fets may get hot due to resistive losses. On small gold tests I gained near 10% extra depth performance on tiny gold of 0.01 gram. Also a reduction in 1/f noise possibly by reducing saturated carrier electron storage but one would think that the long time duration between transitional states would negate this effect.  Less capacitance = more initial signal input. As i do not design mosfets it is a guess why the noise performance is better, I just observe the effects and take advantage of the circuit improvements.  Maybe optical or galvanic isolation is the next thing to test. So far the modded GPX detectors work much better than the SDC2300 and the GPZ7000 on most gold sizes subject to ground conditions.

General Discussion / GPZ7000 Video.
« on: July 29, 2016, 07:17:18 pm »
I did a video on initial impressions.  More to come.


How to repair faulty detectors / Faulty gpx lcd displays
« on: July 23, 2016, 01:28:21 am »
No display?  Most people blame the atmel micro on the board, it is not that chip.  If the display stops workingit is the the lcd itself. The very thin glass on the leg to board area can get fine hair line cracks and the display dies.

Displays are available through the manufacturer in France. Je-An  or the complete module is available in China for $100 usd. I tested some Chinese modules and they are 100% perfectly ok.

How to repair faulty detectors / Crazy fault in SD2000 detectors
« on: July 23, 2016, 01:22:04 am »
The detector is deaf as a post, it will pick up the side of a car and not much else.

First check the n channel fet near the coil connections, these fail due to metal migration in the semiconductor layers. 

Near the lm 394 ic on the input there is a 68 pf capacitor that is leaky, that is that it still measures like a capacitor but it starts becoming resistive, i have found this capacitor to fail on countless sd2000 detectors. It s located near he 4093 ic that controls the gate timing of the receive fets, it is used as a timing delay to give a small delay before turning on the fets after the end of the transmit pulse.

For sale and swap / Gpx4500 fully modded for sale
« on: July 22, 2016, 01:33:03 am »
I have a gpx4500 for sale, has the variable gain and variable frequency upgrades.
Message me for details.

For sale and swap / SD2000 for sale SOLD
« on: July 22, 2016, 01:29:04 am »
I have an sd2000 that is for sale.. Message me if you are interested.

After testing so many boards and trying out new settings to specific customers orders and can now say that obe board does not fit all detectors.  Boards that were created for the GPX4500 do not work on the GPX5000 due to the different input gain profiles of each detector. I was baffled as to why the input stage gain was only a multiplier of 33 and the GPX4500 has a gain multipier of 47.  It is easy to figure out, if you measure the resistors on both detectors you will find that both detectors have an inverting input connected to a 10 ohm resistor while the feedback resistor in the 5000 is set at 330 ohm and the 4500 is set at 470 ohms. When compared side by side in normal mode the 5000 shows significant depth loss on most solid gold targets. The point is that the 4500 is a better deep gold detector in both enhance and normal modes. So why is the 5000 taking a hit in the gain of the first input stage?  The answer is the fine gold mode. Fine gold uses an output pulse that is much the same as enhance but what is differerent is when the receiver comes on in relation to the transmitter turn off and ground decay. I would be thinking that having the receiver turn on in fine gold with similar front end gain of a 4500 would cause too much residual ground decay to get into the detector and make it noisy. So the cure was to use less gain but this causes the other modes to lose performance. It is very simple to test this and prove it by using a 4500 and 5000 and trying depth tests by using a test bed with buried objects. Apart from this difference between the detectors one could increase the gain in a 5000 to match the deep gpld performance but fine gold would become noisy. The only other way is to use a variable gain input stage so you can run as much gain as the ground minerals will allow. We invented the low noise variable gain input stage some 5 years ago and have fitted well over 1200 detectors with this sub board. The next phase is nower noise input stages that can make use of the low impedances of the coil and low ohmic resistance of the input back emf blocking and conductance Fets, All the later model detector use the standard low noise AD797 as the input stage but after some observation i discovered that the very low noise of these components could be made even better. The 1Hz to 10Hz noise specification is a lot higher than the higher frequencies that the ic is designed for, it has very low noise of 1nv/hz but not at the very low frequencies that the pulse induction detector looks at, the AD797 jumps up to 7nv/hz at these low frequencies and as this low frequency component drive the integration stages it seems that some sensitivity can be lost in the noise. The mission was to keep the noise floor under this level and keep the low frequency components under 1nv/Hz at the critical 1Hz to 10hz range.  Well it has been done with a choice of 2 sub boards 1 of 0.82Nv/Hz and one of 0.52Nv/Hz.  The greatest trick was to keep the slew rate very high,dampen any overshoot and keeping common mode noise under control by using a differential 2 stage amplifier and converting back to single ended output.  The whole design was built from scratch and it is a serious upgrade to install in any pulse induction detector, be it a SD2000 or later detector.

Pages: 1 2 [3] 4 5 6