Did the TDI SL second mod yesterday...

[Jim in Idaho]

This mod has been done at the factory on all SL's made since 2015. So, only applies to earlier models (mine is 2013). It consists of removing the diodes D20, and D21, and capacitor C56. You also have to install a jumper across two of the three diode leads on D21. I'll post a pic showing what I did.
The effects of this mod are much more subtle than the sweep speed mod. Maybe a 5-8% gain in sensitivity in the air test on the 1 grain gold bar. But, still some gain, though I had to be paying attention to hear it.
Getting the jumper in place is tough...took me 15 minutes of fooling around to get it soldered in. I'm an old Air Force radar tech, and don't have the skills for this tiny work. Not to mention the fingers of a 40 year concrete construction guy...LOL The mod pushed my air test distance on the 1 grainer to 3 1/4", without headphones. Might have done a bit better using the hp's. there is also a slight increase in ground noise to contend with, but nothing that really bothered me. In really ugly ground, it may be worse. We'll see. In the pic, the red triangles are where the diodes were. On D21, you can see the jumper under the lowest leg of the triangle. That jumper is 3/16" long. The straight red line, on the right, is where C56 was located.

[Orbit]

Yes I know about this mod , thank you !

[Orbit]

Hi Carl-NC !
Would the SAT/SW the same worked if instead of cd 40..at the orginal, used the fet j113 ??
I'd like as little as possible IC on pcb !

[Carl-NC]

Yes, but I replaced the JFETs with 4066 switches to eliminate hand-matching the JFETs. It's a much better solution.

[Orbit]

Thanks Carl !
Then I'll leave it anyway cd4066 ....

[Mark-VA]

Great information, thanks all

[Orbit]

Hi Karl !
Is it better to use LT7660 or lt1054 or transistors due to synchronization as TDI , of course I mean positive powerof course I mean positive power TDI ,is there any reason why it was used on TDI a pair of transistors etc, is it possible that if it is not synchronized the power supply creates so much noise ?

[Carl-NC]

The '7660 doesn't have enough input voltage range, although there are ways of making it work. The LT1054 would work but the discrete transistor solution is much cheaper and does the same thing. Yes, you want to synchronize the charge pump.

[Orbit]

The '7660 doesn't have enough input voltage range, although there are ways of making it work. The LT1054 would work but the discrete transistor solution is much cheaper and does the same thing. Yes, you want to synchronize the charge pump.

Thanks Carl!

[Orbit]

Hi Carl ! This is a TDI timeline diagram not drawn pulse synchronization due to discretion !
My question is : Should TX be inverted at the exit of the MCU for example, at Goldscan 4 The TX pulse is inverted would the synchronization pulse completes its work in case TX is inverted *? Or all pulses must be positive from the processor as with TDI ?

By this I mean if synchronization is possible now ?

[Carl-NC]

[ATTACH]47375[/ATTACH] Hi Carl ! This is a TDI timeline diagram not drawn pulse synchronization due to discretion !
My question is : Should TX be inverted at the exit of the MCU for example, at Goldscan 4 The TX pulse is inverted would the synchronization pulse completes its work in case TX is inverted *? Or all pulses must be positive from the processor as with TDI ?

[ATTACH]47376[/ATTACH] By this I mean if synchronization is possible now ?

I'm not quite sure of the question. Do you mean the sync pulse for the charge pump? As far as pulse polarities, they all need to be whatever they need to be, and that depends on the design of the drive circuits. For example, if you directly drive the NMOS from the uC (generally not a good idea) then you need a positive pulse. If you have a bipolar inverting driver for the NMOS, then you need a negative pulse.

[Orbit]

I'm not quite sure of the question. Do you mean the sync pulse for the charge pump? As far as pulse polarities, they all need to be whatever they need to be, and that depends on the design of the drive circuits. For example, if you directly drive the NMOS from the uC (generally not a good idea) then you need a positive pulse. If you have a bipolar inverting driver for the NMOS, then you need a negative pulse.

Do you mean the sync pulse for the charge pump? ( That's what I mean ! ) It is of negative polarity ! That's why I ask ,since sync is starting from uC ....And all impulses are of positive polarity tx .. s1 , s2 ,s3 , s4 ..... s5 sync of course synchronization is negative ! So the question is whether you can be inverted tx and that the sync pump is doing the job in terms of noise elimination..

I don't know if you understand the translator does not translate properly?

--------------------------------------------------------------------------------------------------
Concretely : tx Negative ! s1 and s2 snd s3 and s4 Positives ! Sync negative !!!! Is the noise lower then?

[Carl-NC]

Do you mean the sync pulse for the charge pump? ( That's what I mean ! ) It is of negative polarity ! That's why I ask ,since sync is starting from PIC-UC ....And all impulses are of positive polarity tx .. s1 , s2 ,s3 , s4 ..... s5 sync of course synchronization is negative ! So the question is whether you can be inverted tx and that the sync pump is doing the job in terms of noise elimination..I don't know if you understand the translator does not translate properly?

--------------------------------------------------------------------------------------------------
concretely : tx negative s1 and s2 snd s3 and s4 Positives ! Sync negative !!!! Is the noise lower then?

Yes, you can simply invert the TX pulse and feed it to the charge pump. This assumes the charge pump does not require a 50% duty cycle. The 7660-type chips usually do not, they divide the frequency in half to get the 50% duty cycle. If you are using a 7660, ideally you would want to feed it a 2x frequency clock. That is, if your TX pulse is 1kHz, create a 2kHz clock for the charge pump.

[Orbit]

Thanks Carl ! I understand you things are clearer to me !

[eclipse]

You have to be aware that when changing sync polarity is where the point of oscillations occur. You want to keep the switching point far away from sampling times, in the TDI it happens when TX turns on. Oscillations last only few microseconds, so there is no noise.