Full m/lab SD2000 schematics

[vovgut]

op!

[sdowman]

there are errors on the schematic

[sdowman]

there are errors on the schematic

there are errors on the schematic

[eclipse]

Yes I've mentioned it in my post above on this page. Refer only to the original schematic posted by ZED.

[eclipse]

Got some PRO questions for GB & Modifications:
If I change TX (240us) and the 4x60uS to 4x100uS for example for ease of calculations, and reduce Freq 400Hz (enough to fit in the period) do I need to change GB samples:

- Do I need to change Transmit OFF time (240uS to 400us)? If so perhaps I will also need to change:
- Channel 2 Sample 2 (84us Low + 16uS High)
- Channel 1 Sample 2 (340uS Low + 60uS High)
- Ground Balance Sample 1 - (250uS Low + 150uS High)
- Ground Balance Sample 2 as close to the next TX cycle or I should keep the 300uS delay after the last TX off time?!


If I add adjustable Primary Delay pot to adjust (15us & 45us) Delays do I need to adjust anything else besides CH2 S1 & CH1 S1 positions

[IBGold]

Eclipse, When we used to change the timings on the 2000 all we used to do is change the crystal frequency that way the other functions all changed in unison the standard crystal frequency was 2.000 KHz approx and mine is processor controlled with ( Mario's Mod board ) the frequency ranging from 1.250 KHz to 4.000 KHz the lower frequency's ( Longer periods ) for larger deeper targets the higher frequency's for smaller targets ( Shorter periods ) I hope that helps.

Regards, Ian.

[eclipse]

Thanks Ian, based on the information I can try to increase/decrease ALL signals with same ratio, and see if it still works

[eclipse]

Been working on modified / simplified version of this thing. V1.3 of the boards arrived. What I did is:

- Removed the entire pulse generation, transmitter & power sections
- Removed the complex snubber I know nothing about, would probably try without one, or try and use the optional 2x zeners (or one fast rectifier HV) diode/resitor/cap
I can also try and disconnect the snubber section with the blocking signal for the front-end mosfets.
- Mosfet should be able to reach 400-500V with or without the snubber. 600V is the maximum due to the limit of the blocking FET

- power supply is simple 7660/LT1054 boost converter (+5V), 79L05 for (-5V) and one more for 3.3V. Can be connected to battery of any sort because I've added a fixed
7808/7810/7812 regulator (not decided on which voltage it would work best so I need to try that).
- Parts removed allowed me to fit the whole thing on 2 boards 90x100, sliding in Hammond case, one will go on top of the other connected with 8 PIN molex connector
- Timing will be controlled from a single 20 PIN SSOP PIC32 CPU (thing is smaller than 8 PIN op amp!!!), - will generate TX, SAMPLE1-SAMPLE6 and the blocking signal for the protection mosfets (fed to dual inverting/non-inverting mosfet driver), I can probably get it to work down to 2.5uS IF I can get it to work at all
- No inductors, SSM2210 for the preamp (I should check and compare with AD797 setup), integrator gain increased to x2.2 from the original
- level shifting and stuff needed to be done, 4066 won't going to work so I tried something else instead to drive directly from the MCU, things may go horribly wrong there
- P1-P3 are connections for some software upgrade (FREQ/DELAY so on adjustments).
- 1206 decoupling caps on all damn ICs on the bottom, 2 each

PROBABLY won't work at all but hey I will try anyways.
Almost same front end (PCB on the left) can be used for the GoldScan 5.

No share / no sale at this time guys.

[pustareka]

Been working on modified / simplified version of this thing. V1.3 of the boards arrived. What I did is:

- Removed the entire pulse generation, transmitter & power sections
- Removed the complex snubber I know nothing about, would probably try without one, or try and use the optional 2x zeners (or one fast rectifier HV) diode/resitor/cap
I can also try and disconnect the snubber section with the blocking signal for the front-end mosfets.
- Mosfet should be able to reach 400-500V with or without the snubber. 600V is the maximum due to the limit of the blocking FET

- power supply is simple 7660/LT1054 boost converter (+5V), 79L05 for (-5V) and one more for 3.3V. Can be connected to battery of any sort because I've added a fixed
7808/7810/7812 regulator (not decided on which voltage it would work best so I need to try that).
- Parts removed allowed me to fit the whole thing on 2 boards 90x100, sliding in Hammond case, one will go on top of the other connected with 8 PIN molex connector
- Timing will be controlled from a single 20 PIN SSOP PIC32 CPU (thing is smaller than 8 PIN op amp!!!), - will generate TX, SAMPLE1-SAMPLE6 and the blocking signal for the protection mosfets (fed to dual inverting/non-inverting mosfet driver), I can probably get it to work down to 2.5uS IF I can get it to work at all
- No inductors, SSM2210 for the preamp (I should check and compare with AD797 setup), integrator gain increased to x2.2 from the original
- level shifting and stuff needed to be done, 4066 won't going to work so I tried something else instead to drive directly from the MCU, things may go horribly wrong there
- P1-P3 are connections for some software upgrade (FREQ/DELAY so on adjustments).
- 1206 decoupling caps on all damn ICs on the bottom, 2 each

PROBABLY won't work at all but hey I will try anyways.
Almost same front end (PCB on the left) can be used for the GoldScan 5.

No share / no sale at this time guys.

Zdravo,hi eclipse

News good, infa ok ?

remove snubber burnt all transistore. bad idea. very bad.

zeners solder to bad, much noise, bad voltge drift.

Mosfet 400-500V bad, max 200V, best 100-200a curent. very high fast need. ocean front wave crash big boat, not wave. front crash nuget gold not pulse. gold sing when crash. ok

7660/LT1054 must synce global osc./atenuate koefficeint. cros-inference blok.

7808/7810/7812 too bad no ok. noise grat rise no choise.

inductors must did. avoid inference.

integral gain rise to bad - noise rise quadratico. need optimale integral koefficeinte.

SSM2210 put 4 paralel all, one at two at tree at four, noise got NF/4. not many impedanze need match valid.

need excelent design new way, not discover hot water.

good day sir.
luck with you

[pustareka]

m-lab SD2000 need high skil enginers not fantasers, new bipolare impulse times. need bipolare sampling chanal by chanal. positive charge no equale negative charge. high velocity mag. field spedup charges at gold nuget, very high stop mag.field crash gold, made inertia negative charge. disperisone negative charge make gold signal. gold sing when high hit. agains clok windin make charge impulse run + → -, clok winding run charges - → +. tranzistor swich coil cold alvays run + → -. very hot burnt finger run - → +. yes big secreto coil windin. need charge science no electorne science. my help - think bipolare no monopolare. think charges are you. good day.

[usefi5961]

dear all friends:
i want build-in sd2000 metal detector,i think this detector is good for making but i want know this detector realy work(schematic is ok and not problem in schematic)and so if any one have .lay (pcb layout file)please send me([email protected])please detail search coil and pcb file.
good luck.

[ne555]

Why not be working to the Relic striker is stronger than the CD 2000 relic striker is 8.4khz good depth

[ricalbertvig]

http://depositfiles.com/files/e2b9wzfj3
the file for the PCB is no longer available

[ricalbertvig]

pcb layout sd2000


hello does anyone have a pcb layout that I can share? - I have read that those that are available have errors .. and in others it is no longer available to download ... attentive to your help ..... .. mail: [email protected]

[Altra]

Found on EBay, no tech info in the listing. Probably need to pm the seller?

https://www.ebay.com/itm/SD-2000-Pul...UAAOSwFyhaKFGV