Full m/lab SD2000 schematics

[unregistered]

Sorry for the silly question, but the power part I don't understand.
How come the battery (+VB) is equal to GND? What is PCB's shield then.. ( normally is - of battery and here GND is + of battery). Im confused ..

[ZED]

The 2000 uses a dual supply as one of its supplys,hence the +6v or +vb is boosted to +5v above the batterys +vb supply,the -6v is regulated to -5v using the +vb as a referance,so using the +vb as a referance you have +5v and -5v and +vb becomes grnd and also it becomes the grnd plane for most of the ccts.The supply you mentioned is refered to as a single supply.

If yer still a little puzzled do a goggle on dual supply.

Zed

[eclipse]

What is the high power rating resistor on the board?
Also are there any other high power components on board (except this resistor, which appears to be in the supply part) and the 200V capacitor connected to the IRF710?

[Unregistered]

That isn't a resistor it is a 1.2mH inductor....

Regards Ismael

[eclipse]

Thank you Ismael, I suppose you mean it's the 1.2mH connected between +- on the battery. Should be 1000-1500mA probably.

[ZED]

Err no !

Some ccts use the -v supply of the battery as ground,the inductor is used as a filter for the -v battery supply on certain ccts.

Zed

[eclipse]

Thank you guys for the answers, the schematic is really interesting lots of stuff to learn on PI MD

[porkluvr]

Off topic?

This may be a little off topic - but here it is anyway. I never would of thought of using a complex PI waveform except for finding this thread.
I tried to run a SD2000 timing simulation in LTspice but was unable to obtain a complex waveform; only a steady pulse stream at 8kHz. I probably wired the test circuit incorrectly but I can't find my mistake, and I gave up trying. I have -no- initiative to build a complex circuit without being confident of its accuracy and so devised another method for a complex pulse train:

A CD4017 decade counter, logic gates, and analog switches - varying 555 timing. Target sample pulses come from one-shots (similar to the Goldscan IV or Hammerhead). Two different blanking pulse intervals are obtained by using a CD4007 (SPDT connected) to switch the RC network on 1/2 CD14538 (or use CD453. Some manufacturers produce "4538" devices with timing component restrictions - they are not all the same. Beware.

The two 1/4_DG411 are normally closed switches. Two "0" sections of a CD4053 could be used instead. Could use 1/3 CD4053 in the receiver blanking. Ahhh, that adds up to 3/3 CD4053 - but I have CD4007 and DG411's in my spare parts box so that's how I designed.

This circuit -may- be more complicated than the SD2000, but transmit and sample pulses are variable. As shown, there is one long pulse followed by four short pulses. In real life this would require two Quad NAND packages to implement. By rewiring the CD4017 reset to Q4 instead of Q5, a 1+3 pulse train would be obtained and only one quad NAND package would be necessary (delete U16). Change the RC time constants if so desired.
As shown, this circuit draws about 200ma.

There is only one long listening interval, so ground cancelling needs done before the short pulses. The timing is different from the SD2000 and has not been tested. Transmit power is relatively low, so be sure and use a low-noise preamp. (Transmit timing can be changed to increase power output, but my physical body is out of shape and do not want to lug around a great big battery.)

LTspice is a free-be, but some simulation models used (CD4007 and 1/4_DG411) are my own creation and not included with LTspice. Also, the CMOS logic are not included. And, IRF640 is not automatically invoked; the NMOS symbol needs IRF640 entered into "Spice Model" and "Value" fields. There may be other "'gotchas", (and if this was EASY, then everybody would be doing it) so....

In "ComplexWave.Zip" I've tried to package my LTspice simulation file together with all necessary models and symbols, but some of the details may have escaped me... I can't think of everything. If anybody needs something more to get this going please let me know and I will try to accomodate. Of course, you'll do your part - use the LTspice "Help" system. Also, there are more than one excellent LTspice tutorials available.
Here's a good one: http://highered.mcgraw-hill.com/site...ort_files.html

I show a -10.5V supply, and drive two IRF640 MOSFETS. That's because I have a bunch of IRF640 on hand and no money to purchase BUK455. (And, I do not have a BUK455 simulation model.) IRF640 will not work with a -6V power supply, by the way.

That's all for now. Good luck. The SD2000 is a monster.

[Goodrat]

As some may know, I am always interested in the audio output circuitry.
Does anyone know if the audio output of this SD2000 shcematic has any similarities to the Explorer models? Because it looks like it is a bridged output. Plus the way Minelab wires their headphone jack (wish shematics here would include that minor detai!) the tip and ring are wired so that stereo phones will have the speakers in series. There is no sleeve connection. This could be why mono detector phones don't get enough volume. Mono phones short out half the bridge.

[eclipse]

BTW there is an error in the version done by vladimir,
the LM394 pins aren't correct so watchout for that if
you are using it.

[Aziz]

Hi porkluvr,

I tried your model with LTspice and it won't start the simulation. There are missing parts (CD40xx).
I had a look to LTspice and it seams a little bit unlogical in integrating other parts.
Do you know how I can integrate parts other than supplied?
I have downloaded some parts (NE5534, AD797, ..) and would like make some simulations on the pre-amp.

Aziz

[eclipse]

I'm running multisim 10, it's available 30 days trial from
http://www.ni.com/multisim/
It has over 16000 models integrated already, so if you want to go step up (or two) start using it .

[Goodrat]

As some may know, I am always interested in the audio output circuitry.
Does anyone know if the audio output of this SD2000 shcematic has any similarities to the Explorer models? Because it looks like it is a bridged output. Plus the way Minelab wires their headphone jack (wish shematics here would include that minor detai!) the tip and ring are wired so that stereo phones will have the speakers in series. There is no sleeve connection. This could be why mono detector phones don't get enough volume. Mono phones short out half the bridge.

Let me correct that thaught.
Yes Minelab connects stereo phones in series with their jack wiring, but there is no sleeve connection. The push-pull circuit of the audio output does not get defeated with mono phones.
Interesting though how to figure the best load impedance. They have two 220 ohm and two 180 ohm. So what do we call this? 800 ohm output impedance, or 440? Their manual on the SE suggests 100 ohm stereo. Since wired series that would suggest 200 ohm headphones. Which would be the load if you connect the Koss UR-30's they supply.

[porkluvr]

update

Hi porkluvr,
I tried your model with LTspice and it won't start the simulation. There are missing parts (CD40xx). I had a look to LTspice and it seams a little bit unlogical in integrating other parts. Do you know how I can integrate parts other than supplied? I have downloaded some parts (NE5534, AD797, ..) and would like make some simulations on the pre-amp.
Aziz

Aziz, I downloaded my Complex/zip folder and unzipped it, just as you would have... let's see what happens... . OK, please ignore (and DELETE) the file "ver2TX+TIMING+BLNK.asc" I should have deleted that because "TX+TIMING+BLNK.asc" is more up-to date.

OK, here's what I forgot to mention:In the LTspice installation folder, go to "C:\Program Files\LTC\SwCADIII\lib\sym" (or whatever is appropriate for your system) and create an "add-ons" folder so that you have
"C:\Program Files\LTC\SwCADIII\lib\sym\Add-ons. You can rename Add-ons, but that is what I use and so that is what the simulation file looks for as it stands.

Place the CD4000 folder in \Add-ons. Also, put the NE555.asy and potentiometer2.asy (also potentiometer.asy). Add the "switches" folder.
You can put the symbols anywhere you like,but unless you set things up exactly as I had them you will need to place (and then edit attribute fields of) the symbols manually. Place the parts according to the simulation screenshot from the other day (I did "CTL+PRINTSCREEN" to get that image), then observe the jpgs I have posted today and make sure the various parts' fields are properly filled out. (Right click the various components' symbols to open their "Component Attribute Editor", then click inside the various fields to enter data.) Place the model files according to the .inc command - or better still: put the various model files in "C:\Program Files\LTC\SwCADIII\lib\sub", and won't need the .inc statements, if you edit the "Spice Model" field (you can drag and drop an asy file, or click for editing) - but that is a more advanced concept... try it and see. (Place a "*" in front of the various .inc commands to ignore them and see if it works).
Replace "standard.mos" (in the "C:\ProgramFiles\LTC\SwCADIII\lib\cmp" folder) so that VN2410L and TP0606 will will be active. You will have to restart LTspice before the new standard.mos will be recognized. I also added Standard.jft and Standard.diode. Replace the original files with these if you want the extra models, but they are not used in this simulation. I also have included "opamps.sub" and cmps.sub files, but they are not used either. All of those models conform to the most often used virtual pin-out (and use the "X-OPA.asy"). NE5534 does not - but I have included it's model and symbol as well. Third party models must be kludged in, that is something that I have learned to deal with, but it takes some effort to master. Sorry about that. I have found LTspice easier to use than P-Spice, but that' just my opinion.

I see that the VN2410L may be corrupted and need be massaged, so right click on it to change the nmos.asy attributes for that device. No, no, do this instead: use the "Add Symbol" icon, then select NMOS. Rotate (CTL+R) and mirror (CTL+E) the symbol to place it in its proper orientation. Right click and scroll to the bottom for "VN2410L". I don't know exactly why the corruption happened but there are some different formats possible for various components models, it's hard to explain. There may be other issues that pop up.

I appologize for my slacking, but like I said - I can't think of everything. There are some inconsistencies in my methodology that I hope you can see through, but let me know if you continue to have problems. I have included another "TX+TIMING+BLNK.asc" file that hopefully is repaired, but sometimes wierd things happen when simulations are transferred between computers, it depends on where the various files are located and some.
I have NOT re-posted all the necessary model and symbol files, anybody needing these should download "complex.zip" from my previous post. (you'll need the CD4000 folder but I am re-posting the switch models today).
Look at the CD4007_SPDT.sub file for a pin connection description for that device. Look at TI's CD4007 datasheet http://focus.ti.com/lit/ds/symlink/cd4007ub.pdf
(fig.h) to see one possible SPDT configuration. See the CD4007.sub model file for another possible connection. "Third Party" models and symbols pin-outs must agree before they will work - see "netlist_order.jpg" to get an idea.

If you make a copy of the simulation file and then delete the "Target Sample Timing", that will run much faster. The multitude of CD14538s cause the program to run very slowly. Once you have it up and running, then run the whole schmiel. There is also a CD14538 in the receiver blanking section - that section can also be temporarily removed.

I would have answered sooner but for some reason I don't always receive new message flags in my email inbox, and new posts to this forum seem to inexplicably come and go at random as I browse.
I hope I have covered everything.

[Aziz]

Hi porkluvr,

many thanks for your help. I appreciate that. I had a more look into LTspice in the meantime and I like it due to easy use. On the following site, there are more models and hints for download (you must register for membership):
http://tech.groups.yahoo.com/group/LTspice

I also tried the National Instruments Multisim 10.1. This contains lots of SPICE models, which can be exported and saved.
So I got many parts out of the Multisim and will adapt to LTspice for further use.

It is quite difficult to get all the models for used parts in a schematic. I have to read more in the LTspice tutorial and I am sure, this will be a good investment.

Regards,
Aziz