WELCOME!!

[Altra]

As for the GS5, it is synchronized and only changes states during the tx on and off. So during sampling it is quiet. The only problem
with a voltage doubler design is that you need a enough voltage on the drive side to keep the regulator from dropping out. For a
+5 output you want at least -8v feeding the doubler.

[Old cart]

Altra, one more question. Using a charge pump as described how much P-P noise should we expect on the supply lines? Have you ever measured you GS 5 for example?

[Monolith]

One of the thoughts that I have had for some time is to use higher voltage supplies for op amps to increase the dynamic range. For example if you used +/-10 volts then the 1st preamp would not saturate if it had a gain of 10 and used the traditional diode protected inputs. Without saturation the amplifier would recover quickly. We could do this with a charge pump like the LTC3260. This device has integrated low drop out linear regulars for low noise and operates at high frequencies for efficiency. The frequency is adjusted by a single resistor. I think we could then tap into the device clock, divide the frequency by 100 and use it to drive the transmitter and audio for a fully synchronized system. The adjustable frequency could be used to reduce the effects of external noise and still remain synchronized.

Any thoughts on whether this will work?

The attached Sepic converter generates a variable positive and negative output. I think it runs at 100kHz, need to check.

Schematic and PCB layout attached

[Old cart]

You can look at the original surf pulse or GS4 / 5 schematics. Or without a DC to Dc converter see Dooley's TD3X project. The main thing
is your coil ground needs to be connected about halfway between your analog supply.

Without the dc to dc converter doesn't this then constrain you max coil voltage of about 1/2 battery voltage? Also the rail splitter has to absorb the big current pulses from the coil.

[Old cart]

One of the thoughts that I have had for some time is to use higher voltage supplies for op amps to increase the dynamic range. For example if you used +/-10 volts then the 1st preamp would not saturate if it had a gain of 10 and used the traditional diode protected inputs. Without saturation the amplifier would recover quickly. We could do this with a charge pump like the LTC3260. This device has integrated low drop out linear regulars for low noise and operates at high frequencies for efficiency. The frequency is adjusted by a single resistor. I think we could then tap into the device clock, divide the frequency by 100 and use it to drive the transmitter and audio for a fully synchronized system. The adjustable frequency could be used to reduce the effects of external noise and still remain synchronized.

Any thoughts on whether this will work?

[Altra]

You can look at the original surf pulse or GS4 / 5 schematics. Or without a DC to Dc converter see Dooley's TD3X project. The main thing
is your coil ground needs to be connected about halfway between your analog supply.

[Old cart]

Altra, we could solve the offset problem by AC coupling the input after the blocking diodes (don't know what impacts that might have) or use ground ( really +4 relative to the - side of the battery). I think I have this right. For some reason these supplies for PI really confuse me. I really like real plus and minus supplies as are traditionally done in electronics.
Any input on how to do this correctly are greatly appreciated.

[Old cart]

Altra, that is yet another possibility. Do you have suggestions for such a design? I have seen many designs used in metal detectors that are synchronized but none that I know of are particularly low noise. The only such designs that I found that are low noise are high frequency non sync'd.

[Altra]

Don't you guy's want to design a syncronized low noise positive DC to DC converter? Not a rail splitter? Based on the schematic in post 7. The preamp going to see a large input offset. Your signal is going to be to be centered at about 10v which is above your positive 4v rail. Most PI's use a DC to DC converter to solve this problem.

[Sean_Goddard]

I have that in my workshop ;-)

OK I've asked to have su sections for the different modules created then we can move the convos for each to the relevant section.

I like the debate on the Power supply, I was going to use a TLE with X1 1A Op Amp as the splitter. As per Old Carts advice, I was going to drop that 10R in the line.

When we ge the threads set up can all those who have proposed a solution post in this main part with a brief of what they have done? I'll need to keep this top level tidy.

Thanks ALL.

[green]

No, it was just a quick and dirty simulation for the fun of it.

In any case it's so simple it can be built on a breadboard in a few minutes with off-the-shelf parts.

The following is a low voltage version for splitting a 5V source:





[ATTACH]34520[/ATTACH]
[ATTACH]34521[/ATTACH]

Have you tried making V2 a 100 Hz square wave?

[Teleno]

One of the things I like about the Teleno splitter is the tweakability, and no likely future unobtainium involved. But a 2426 could be thrown in, in place of the OPamp.

Replacing the op amp by rail splitter TLE2426 removes the gain in the feedback loop which is necessary for regulation. As a result the ripple will be orders of magnitude worse.

[Monolith]

Can any of the members here comment on the availability of parts in their region? Is the problem that the parts are too expensive or that you can just not get them?

the time when one could go to RADIOSHACK and pick up some parts, are over. Parts have to be ordered. Digikey, Mouser Farnell, and many more have almost everything, however the shipping cost a lot, unless one makes an order large enough like 80$ or so, to get free shipping.

Then there is Ebay. Works fine most of the time, if one watches out for shipping charges. AliExpress works good too. Most of the time the shipping is FREE, that means it is included in the price. Amazing how they manage to get 20 diodes delivered to your door by the post man for a total of less than 2$.

It takes a bit of work but one can find almost everything as long as one can send the money.

So, why use obsolete parts? They are usually cheaper.

[Thomas]

I do like the idea of a rail splitter though because it eliminates the need to have to think about non standard supply voltages. With the rail splitter there may be issues with the big discharge taking out the splitter itself. Does anyone have experience with this ?

One of the things I like about the Teleno splitter is the tweakability, and no likely future unobtainium involved. But a 2426 could be thrown in, in place of the OPamp.

[Old cart]

Another alternative is to just do a supply like was in the TDX3 project. This apparently works quite well in that detector. Note the battery in this case is about 12 volts. The battery alarm can be eliminated. You could also use 4 and 8 volt fixed regulators. There may have to be some modifications to the coil drive voltage as it is way above the highest amp rail voltage. this may be mitigated by the input clamping diodes.


I do like the idea of a rail splitter though because it eliminates the need to have to think about non standard supply voltages. With the rail splitter there may be issues with the big discharge taking out the splitter itself. Does anyone have experience with this ?