Announcement

Collapse
No announcement yet.

Super TGSL With H Bridge TX Diff RX Quad Sampling

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • The last picture is bad. Are you sure feedback connections pins are 1-3 and 5-7?
    The one before the last - did it stand still or was there some wander? I could try to push amplitude to the limit.

    Comment


    • Success! I took the board to work today to look at it under the scope. I found some flux
      that my thorough cleaning missed but also realized one of the pots was wired wrong. I knew
      on the board the center is not in the center for one side as I didn't have room to do that,
      but I wired it wrong so had oscillation at both ends of the pot. Here are the pics today with
      the 22N caps and pots wired correctly;

      DISC Min
      Click image for larger version

Name:	Disc Min.JPG
Views:	1
Size:	45.1 KB
ID:	339614
      DISC Max
      Click image for larger version

Name:	Disc Max.JPG
Views:	1
Size:	52.7 KB
ID:	339615
      GEB Min
      Click image for larger version

Name:	GEB Min.JPG
Views:	1
Size:	66.2 KB
ID:	339616
      GEB Max
      Click image for larger version

Name:	GEB Max.JPG
Views:	1
Size:	59.9 KB
ID:	339617

      So finally we can move on and get this puppy finished!

      Comment


      • OK, this is much better, but still there are some oscillations. This seem workable. I'll have in mind these when I design a next generation of this phase shift to:
        - reduce impedance (potentiometers could be lower impedance but I did not check how "low" I can go)
        - increase amplitude to the maximum that does not cause phase transfer problems.

        It is a worthy goal to wipe the oscillation, because the CMOS switchers tend to inject charge at each transition, and those oscillations seem like potentially several charge injections at once. I'm even considering a true double balanced configuration to deal with charge injections the only feasible way. It would complicate the build, but may prove to be essential for going deeper. I'll see.

        Comment


        • Originally posted by Silver Dollar View Post
          Success! I took the board to work today to look at it under the scope. I found some flux
          that my thorough cleaning missed but also realized one of the pots was wired wrong. I knew
          on the board the center is not in the center for one side as I didn't have room to do that,
          but I wired it wrong so had oscillation at both ends of the pot. Here are the pics today with
          the 22N caps and pots wired correctly;

          DISC Min
          [ATTACH]28781[/ATTACH]
          DISC Max
          [ATTACH]28782[/ATTACH]
          GEB Min
          [ATTACH]28783[/ATTACH]
          GEB Max
          [ATTACH]28784[/ATTACH]

          So finally we can move on and get this puppy finished!
          BRAVO,,, Silver-dollar & DAVOR ,,I have been watching this project with great interest ,as now that I finished my 1st M/D DIY IDX-pro ,with I might add great help from the site members I am keen to build another ..

          DAVOR,,, you often speak of a Centre-tapped coil's What advantage other than does not need shielding ????? ..
          I am afraid your technical work is above my head ,[I am 74Y/old & although I can build nice circuits The theory is beyond me] ,,but I appreciate your expertese as you suggest mods to other projects on the site such as the ''S-M-W'' which seems a good unit also ..
          I live in Australia & we have a few gold-fields & I cannot afford a high cost detector & would like to build a VFL [so as to discriminate mineralised ground ] to detect maybe gold ..[I guess higher frequency is needed]???????

          GREAT work by the 2 of you fella's ,All the best from Australia,,,,BARRY..

          Comment


          • Hi Barry,

            it is said that Australian gold fields were re-examined after the PI machines were introduced there. I had no chance so far to check how bad the ground there actually is, but everything indicates that if you use a VLF you'd need a machine with extended ground balancing range to find something.
            This project is about building a machine with extended GB, while keeping the ferrite reference intact. This should keep iron and coloured metals apart even at very bad lands. I have no idea if that will suffice in OZ, but will surely work on the beaches.

            As for centre tapped coils, there is a bit more to it than just the absent shielding. The unbalanced (and shielded) coil is internally ... unbalanced. It means that there will always be some coupling between the unbalanced Tx and unbalanced Rx, but not in a case Rx is balanced.
            The other reason is only practical - it is much easier to build an unshielded coil than a shielded one, and the real good shield is a royal pain in the nether regions because it involves graphite - very messy.

            Comment


            • Well I got rid of the 5pf caps for hysteresis and no more oscillation!

              Click image for larger version

Name:	GEB Min.jpg
Views:	2
Size:	71.4 KB
ID:	339642

              Click image for larger version

Name:	GEB Max.jpg
Views:	1
Size:	80.4 KB
ID:	339643

              Click image for larger version

Name:	DISC Min.jpg
Views:	1
Size:	75.2 KB
ID:	339644

              Click image for larger version

Name:	DISC Max.jpg
Views:	1
Size:	71.5 KB
ID:	339645

              Click image for larger version

Name:	3RD Pulse Min.jpg
Views:	1
Size:	73.9 KB
ID:	339646

              Click image for larger version

Name:	3RD Pulse Max.jpg
Views:	1
Size:	72.2 KB
ID:	339647

              Click image for larger version

Name:	Third Pulse.jpg
Views:	1
Size:	81.7 KB
ID:	339648

              Comment


              • Here's a picture of the project so far;

                Click image for larger version

Name:	SuperSoFar.JPG
Views:	1
Size:	143.3 KB
ID:	339652

                I ran out of 10K resistors or I might have had it working today. I have an order in to Mouser
                but chose the inexpensive shipping so maybe next week it will be here!

                Comment


                • Amazing!
                  I may have a little detail that I noticed - it is always a good practice to weave the wires that go to a potentiometer. Especially if it is at a high impedance. You never know what the lose wires may pick inductively.
                  The next step will be to confirm if you may null the ferrite exclusion channel (all metal). Once it is done we'll see to the rest of the device, one function at a time.
                  The all metal channel is very important because - the other detectors do not have one.

                  Comment


                  • The wire I'm using comes as twisted pair so is easy to weave but I thought we didn't want the wires
                    together. On the TGSL thread it was pointed out that the original Tesoro units had rats nest of wires
                    in them. I hope it fits in a case when I'm done!

                    Comment


                    • Originally posted by Davor View Post
                      Hi Barry,

                      it is said that Australian gold fields were re-examined after the PI machines were introduced there. I had no chance so far to check how bad the ground there actually is, but everything indicates that if you use a VLF you'd need a machine with extended ground balancing range to find something.
                      This project is about building a machine with extended GB, while keeping the ferrite reference intact. This should keep iron and coloured metals apart even at very bad lands. I have no idea if that will suffice in OZ, but will surely work on the beaches.

                      As for centre tapped coils, there is a bit more to it than just the absent shielding. The unbalanced (and shielded) coil is internally ... unbalanced. It means that there will always be some coupling between the unbalanced Tx and unbalanced Rx, but not in a case Rx is balanced.
                      The other reason is only practical - it is much easier to build an unshielded coil than a shielded one, and the real good shield is a royal pain in the nether regions because it involves graphite - very messy.
                      HI DAVOR ,, THANKYOU for the reply & I must say your English is very very Good ,as I see you live in Croatia..?????

                      AS said earlia,I am 74 y/old & have just taken-up this hobby DIY-DETECTOR,, I have built many electronic projects for my other hobby Electric-Guitar ,AMPS [VALVE] Effects units,[MAINLY MULTI-TAP ECHO UNITS] solid-state but mainly from Schematics as my theory on electronics is not-really good & I do not do SMD work ..

                      I can obtain any componants here & also have a good collection on hand ..

                      I have only joined this site 8 -9 weeks ago & have learned a great deal from the Members here including ''silver-Dollar'',I successfully have built the IDX-PRO & made 3x coils for that & it works well [But not been-out on the ground detecting as yet only my Back-yard..

                      my computer skills are only average so cannot do spice ETC ..I have good soldering skills ,several good multimeters,an old single scope.good frequency-counter & a signal generator , ..
                      I have ''never'' Etched my own boards !!!But used perf-board etc etc..

                      OK I have not been to our gold fields to know what the mineralization is ,,But from reading articals from prospectors They mainly use The australian MINE-LAB [Made here in my State 200 ks away] but ''whites'' have made a detector VLF -called a goldmaster GMT which is reguarded highly here it has 48kz although much cheaper than M/L it is out of my budget !!!!

                      IT seems that somewhat higher frequency is needed for small objects as sa Gold etc..

                      I very much like your idea's on the c/tapped coils such as this project ,So if I was to try this circuit-out [if Silver-Dollar does not mind me copying his project???] WOULD I need to build the TGSL ,or could it be adapted to an IDX ??????..

                      As for Graphite shielding I finally succeded in a good mixture [after many many failures] I gring my own carbon-graphite from huge brushes used in wind turbines I get 1/2 worn ones for free ..I found that nitro-Cellulose clear furnture finnish works the best stcks to housings, & micropore-tape & is very conductive with one coat ,But as you say it is very messy to use ,but effective !!!& better than alluminium tape [less resistance] ..

                      GREAT PROJECT you guys are doing & I watch with great interest ,,,,..

                      CHEERS FROM ''Down-Under'',,,,,,,,,,,BARRY..

                      Comment


                      • Copy away! I took many of Davor's designs and incorporated them into a Golden Saber here.
                        I bought a Tracker IV and it only got 3" of depth so I was looking for MORE! You should be
                        able to add on to the IDX but have to make sure it still functions as it works a little differently
                        from the TGSL.

                        I found some old chips in my junk box the other day, one was a 1024 reticon delay chip. I saw
                        one on EBAY sold for $85! So here might be gold in my old collection of IC's!

                        I pushed this project up to 12.5 khz as it was suggested that is a good all around freq. You might
                        like to go higher for your conditions/targets. I think Davor was contemplating a LF vs VLF machine
                        at one point, there;s lesss noise up there too!

                        Comment


                        • Nice job guys. I have been following this thread from the beginning with interest.
                          Don.

                          Comment


                          • Originally posted by Silver Dollar View Post
                            Copy away! I took many of Davor's designs and incorporated them into a Golden Saber here.
                            I bought a Tracker IV and it only got 3" of depth so I was looking for MORE! You should be
                            able to add on to the IDX but have to make sure it still functions as it works a little differently
                            from the TGSL.

                            I found some old chips in my junk box the other day, one was a 1024 reticon delay chip. I saw
                            one on EBAY sold for $85! So here might be gold in my old collection of IC's!

                            I pushed this project up to 12.5 khz as it was suggested that is a good all around freq. You might
                            like to go higher for your conditions/targets. I think Davor was contemplating a LF vs VLF machine
                            at one point, there;s lesss noise up there too!
                            HI,, S/Dollar thanks for the OK to copy your hard work I appreciate that -Mate..

                            OK i wish I could help with the technical issues But I am afraid i am not compitant in the theory side of things ,but IF i can help somehow I will ..

                            As reguards to the frequency is this ''governed by the coil design'' ??? & caps that tune the coils??? or the actual circuit used [ie]TGSL,IDX-- ETC.. to what I have learned so far it seems the Coil inductance & caps .. BUT I see some detectors have 3 odd frequency's that can be switched ,,Maybe to complex for the Diy etc,,,, chips-relays-etc- in the coil-head- etc etc ..I guess as you say a happy medium is better as say your 12.5kz !!!??

                            ON the delay chip I once years ago bought a delay chip ''bucket-Brigade'' from Tandy's,Iwill check I think its a 1024??] & never used it & still have it in its original pack..
                            The multi-tap Echo units I built are for the ''shadows sound mainly'' they use a programed FV1 chip on a little daughter board SMD already built [made in USA& it gets attatched to a PIET designed mother board with a couple fets & 3x 8 pin OPamps ,a 9v reg & 4x pots exturnal its all analog audio & you have a great Echo unit [Better than you can Buy any where]!!!!..with 8 programs to choose !!!total board size is aprox 4x4'' need's a 12v plug pack only..
                            ,I cannot program chips [The language etc is beyond me] But I was involved in a program with a Dutch Guru & ''PIET'' If you are into guitar etc I can give the details to get hold of a Kit from the Netherlands, the unit is cheap & fantastic ,The unit was 3-4 years in the setting-up which I was a first of 6 guys that were sent the hex files to load into the chip,using a programmer. [PIC-IT] & give reports to the developer [Great Guy PIET] he did all this work for parts only cost for the guitar DIY fellows ..
                            The Kit comes pre-programed now..& you can buy the programed daughter board from the US & could build the rest from schematic I have all that here ..if interested ..I can email pics-etc etc..daughter board is attatched [piggy-back] by a 14 pin socket ..
                            I also have a great Tape echo unit Multi-Tap which I converted to All valve operation & head timings copied from the Famous Italian MEAZZI MULTI-TAP Analog Machine..bUT this above unit gives the same results ,,Etap2.. Thats its Title..

                            BUT I have now changed hobbies for this DIY M/D machines to have a go on the ground !!!.FIND those tin cans of gold Coins HA.Haaar.. [not likely but I will have some Fun]..

                            GOODONYA-MATE Take it Easy,,,,,,,,,,BARRY..[PS I will send a PM & give my Email address if thats OK] if not NO-WORRIES..]
                            Last edited by hoadlies; 03-17-2014, 01:52 PM. Reason: missed something out

                            Comment


                            • Yeah, I'm Croatian. Many people here also have a vibe for languages.

                              IDX is a nice device on its own, and other than replacing chips with some better choices there is little you can do to improve it. Say, you could improve Tx by introducing a trimmer to set operation point. And perhaps some tweaks with response time bases. Other than that it would be a complete overhaul.

                              There are details in VLF that are not tackled properly:
                              - charge injection from switchers - fixing it would require some serious redo in a mixing stage
                              - 1/f noise - it may be fixed by a chopper, however, the closest thing to a chopper is an AD convertor
                              - ferrite phase reference, and a wide range GB as separate channels - with regular detectors difficult soils ring as a machine gun, not because of the bad soil, but because of the iron response phase reversal
                              - only digitally processing detectors have a 4-quadrant discrimination
                              - sound quality is generally bad, and also quenched hard to avoid chatters

                              So there is plenty of room for improvements. Only not easily done on the existing rigs.

                              Comment


                              • Originally posted by Davor View Post
                                Yeah, I'm Croatian. Many people here also have a vibe for languages.

                                IDX is a nice device on its own, and other than replacing chips with some better choices there is little you can do to improve it. Say, you could improve Tx by introducing a trimmer to set operation point. And perhaps some tweaks with response time bases. Other than that it would be a complete overhaul.

                                There are details in VLF that are not tackled properly:
                                - charge injection from switchers - fixing it would require some serious redo in a mixing stage
                                - 1/f noise - it may be fixed by a chopper, however, the closest thing to a chopper is an AD convertor
                                - ferrite phase reference, and a wide range GB as separate channels - with regular detectors difficult soils ring as a machine gun, not because of the bad soil, but because of the iron response phase reversal
                                - only digitally processing detectors have a 4-quadrant discrimination
                                - sound quality is generally bad, and also quenched hard to avoid chatters

                                So there is plenty of room for improvements. Only not easily done on the existing rigs.
                                HI DAVOR,, Thanks for reply You mention trimmer on TX????to set operation point ??? could you explain please [IDX],..

                                CHEERS BARRY..

                                Comment

                                Working...
                                X