TESORO GOLDEN SABRE

[Fred]

I noticed in Compadre thread that a mod is to drive TX oscillation directly from battery voltage instead of regulated supply to get higher voltage, more depth.

Anyone try that for TGS?

The Relic Hawk also has a unregulated power supply at oscillator...

[Unregistered]

Hello Eclipse

Transistors are standard onces: BF245, PN2222/PN2907 (plastic ones).

Look for the e.b.c. pins !

[ivconic]

For the things Max says ( about old types of the components in TGS)
I ordered 3 weeks ago components worth $180 (and damn im still waiting for them) just to push up as much as I can the electronics. The components include all (except pf caps) Wima - series FKP2 and MKP2 and electrolytics caps: Nichicon muse fx. All resistors are ultra low noise 10uV CMF55 series. Transistors are standard onces: BF245, PN2222/PN2907 (plastic ones). The IC: LT1057 in the front stage, two LT1018 comparators, LT1008 insted the 308, and 2xOPA2277.
The board, I decided to use Ivconic's, but i had to change almost all capacitors with other bigger footprints for the Wima's. Also the box I ordered is large, 160x100 i can add some stuff.
The layer with the traces I decided to add ground plane, some of the jumpers will be gone cos of that and also will order for manifacturing, its about 30$ so not that high.
As soon as I am done with all that will post the results, hope I am not throwing money on the wind ...

Ground plane is ok idea; will improve S/N ratio noticeable. I tried once and it was successfull. Reason i am not doing it anymore are new enclosures - lets me to add foil "cage" inside...
Yet...adding ground plane is good idea.
One more thing;
i suggest you not to connect ground plane on every through hole. Keep gnd traces connected at bottom layer as those are already and use just one through hole connection to gnd plane. Just one...not many.
On component side try to void gnd connections on every component that goes to gnd.
If oposite...gnd plane may turns to produce more noise!!! Ha,ha,ha!
Strange but true!
Already checked and experimented...
Cheers!

[simonbaker]

>>On component side try to void gnd connections on every component that goes to gnd.

What does that mean?

Regards,

SB

[Max]

>>On component side try to void gnd connections on every component that goes to gnd.

What does that mean?

Regards,

SB

Avoid ground loops.

[eclipse]

Ok good good news all the parts are here! Nice nice a can now start building
the TGSL! Will post the more interesting data here
Almost all parts (except Bourns pots which don't look so healthy for this price) are damn good.
Yep Ivconic Ive seen that trick on ML machines - foil cage attached to cable with paper clip (hah!)

[simonbaker]

For the things Max says ( about old types of the components in TGS)
I ordered 3 weeks ago components worth $180 (and damn im still waiting for them) just to push up as much as I can the electronics. The components include all (except pf caps) Wima - series FKP2 and MKP2 and electrolytics caps: Nichicon muse fx. All resistors are ultra low noise 10uV CMF55 series. Transistors are standard onces: BF245, PN2222/PN2907 (plastic ones). The IC: LT1057 in the front stage, two LT1018 comparators, LT1008 insted the 308, and 2xOPA2277.
The board, I decided to use Ivconic's, but i had to change almost all capacitors with other bigger footprints for the Wima's. Also the box I ordered is large, 160x100 i can add some stuff.
The layer with the traces I decided to add ground plane, some of the jumpers will be gone cos of that and also will order for manifacturing, its about 30$ so not that high.
As soon as I am done with all that will post the results, hope I am not throwing money on the wind ...

If you can, it would be nice to have a "minimal" PCB layout available for TGSL based on Ivconic's design with small changes:

- Use a "dual" IC instead of two LM308

- Remove unused resistors R48, R49, R50, R51 (including the back-to-back 100K substituting for the 200K R48, R51)

- Replace back-to-back 10 uF capacitors with single 4.7 uF Bipolar for C14 and C17

- Add the missing R27 resistor if it is still missing

- R47, R55 -- where are they? They seem to be on schematic, not on PCB. Not needed?

A goal is to make the PCB small enough to fit in a CD Tray of an inkjet printer, it would open possibilities for printing directly on board. It almost fits now.

Other PCB layouts would also be useful with different part substitutions so we have alternatives depending on available parts.

Are you using the Sprint layout program, or something else?

Regards,

SB

[eclipse]

Hey,
1.The LM308 as I understand reading the whole topic probably twice is that there isn't any other good alternative (except the LT1008 which I now have)
so I don't think it's a good idea;
2. The BiPolar - definately yes I am removing those since I got bipolars by Nichicon, also the the 2 pairs of resistors could be replaced with 220K
3. Yep the 1K is missing, Ill add that, also the new audio mod
4. Adding ground plane - thanks i will watch for not doing loops in the GND plane

Honestly said I am making the board bigger, my box is 160x100 and have lots of room, I had smaller board working on but after ordering those parts (and mainly the Wimas - huge footprints) - nope not smaller but bigger board I need.

New idea:
I already asked Max about it, but still.... look at this picture, see the similarity? Could it be used for this project (it's the audio from relic hawk)??

[simonbaker]

Hey,
1.The LM308 as I understand reading the whole topic probably twice is that there isn't any other good alternative (except the LT1008 which I now have)
so I don't think it's a good idea;
2. The BiPolar - definately yes I am removing those since I got bipolars by Nichicon, also the the 2 pairs of resistors could be replaced with 220K
3. Yep the 1K is missing, Ill add that, also the new audio mod
4. Adding ground plane - thanks i will watch for not doing loops in the GND plane

Honestly said I am making the board bigger, my box is 160x100 and have lots of room, I had smaller board working on but after ordering those parts (and mainly the Wimas - huge footprints) - nope not smaller but bigger board I need.

New idea:
I already asked Max about it, but still.... look at this picture, see the similarity? Could it be used for this project (it's the audio from relic hawk)??

Are you using the battery check? If not, get rid of R48, R49, R50, R51 (includes back-to-back resistors) -- you won't need them, i.e. the 220K at all! That part of the circuit left over by mistake.

Can somebody comment on R47, R55?

I think there may be many ways to experiment with the audio, try it!

Cheers,

-SB

P.S. Thanks again to Ivconic for original excellent layouts -- very appreciated.

[eclipse]

Actually simonbaker you are absolutely right for the 9V batteries!
There are NiMH GP batteries 200mA and i could find also 480mA ALCAXLINE (hard to get) LI ion. The difference in $ is big between those two but 200mA would not give more than 3 hours of work. Do you think i should remove 7808 in this case? Thus the device will drain the maximum power (~10V fully charged, and 7V discharged)?
And please if someone can, give more info on the resistors we mentionted here, can they be removed?
Cheers

[simonbaker]

Would greatly appreciate if someone would comment on why each op amp chosen for its stage.

Some guesses:

LF353 - low noise high bandwidth, basically a preamp for the J107 lock-in amplifier to boost everything from the RX coil with as little noise and distortion as possible.

LM358 - just a cost effective, good old workhorse efficient op amp, high gain, bandwith not so important because part of low-pass filtering? Why not ultra-low noise for this also?

LM308 - precision op amp to provide very accurate gain relative to ground (because of very low offsets) of phase signals prior to comparing by the LM393. Is precision more important than low noise in this stage?

People tried substitutes that didn't work as well -- can they explain the reason?

Understanding reasons for each choice would help in finding substitutes.

Regards,

-SB

[DIRCEU]

Hi! ALL

Recently I bought 20 pieces of uA 308N (metalic).
Have anyone some info if compared with LM 308 AH (metalic)???
Thanks in advance

DIRCEU

[eclipse]

I draw what I mean, do you think this is going to work?

[Max]

Actually simonbaker you are absolutely right for the 9V batteries!
There are NiMH GP batteries 200mA and i could find also 480mA ALCAXLINE (hard to get) LI ion. The difference in $ is big between those two but 200mA would not give more than 3 hours of work. Do you think i should remove 7808 in this case? Thus the device will drain the maximum power (~10V fully charged, and 7V discharged)?
And please if someone can, give more info on the resistors we mentionted here, can they be removed?
Cheers

Hi,
9V battery on TGS ? No good. Will last few, expecially on the speaker.

IF you mean on TGSL (smaller circuit, smaller consumption) it could be good... but you'll get lowered audio cause audio stage runs on battery voltage... so the more you have, the more it sounds strong.

In TGSL with headphones you can use 9V battery... but just using a very low dropout 8V regulator instead of 7808 (that requires at least 11V on input to work good).

A 9V alkaline (e.g. duracell's one) could last also 15hours if device is used on headphones only. This is similar choice to bandidoII strategy... use headphones and battery will last also 20hours in bandidoII, for real (this is also due to lowered consumption of bandidoII components).

NiMh of small capacity must be avoided on TGS/TGSL... 200mha are not enough for decent performances and duty. Better something around (at least) 500mah.

Kind regards,
Max

[Max]

Would greatly appreciate if someone would comment on why each op amp chosen for its stage.

Some guesses:

LF353 - low noise high bandwidth, basically a preamp for the J107 lock-in amplifier to boost everything from the RX coil with as little noise and distortion as possible.

LM358 - just a cost effective, good old workhorse efficient op amp, high gain, bandwith not so important because part of low-pass filtering? Why not ultra-low noise for this also?

LM308 - precision op amp to provide very accurate gain relative to ground (because of very low offsets) of phase signals prior to comparing by the LM393. Is precision more important than low noise in this stage?

People tried substitutes that didn't work as well -- can they explain the reason?

Understanding reasons for each choice would help in finding substitutes.

Regards,

-SB

Hi,
1. it's not lock-in amplifier, you haven't critical integration there, but it's similar cause you mix reference and input signal of same frequency, then filter with a low pass: this filters out many noises but not very good... that's why you can get interferences in TGS from power lines, switching regulators etc. LF353 is a sensitive one... hi bandwidth. Can use different but seems it's good choice having one there instead of other stuff.

2. LM358 : uhm... is old good... but not so good, good replacements are possible and suggested... lower noise like TS27m7, they work better than LM358

3. LM308 : replacement possible... but worth ? I think not at now. Seems metal LM308 are the best there, for now.

4. LM393 : can replace but change few cause source of noise is not there, but early in the chain

5. TL062, avoid them at all. Use lower noise, even TL072 is fairly better.

Kind regards,
Max