Let's start from square one...

[KingJL]

My preference is to ditch the voltage doubler charge pump, which means no mono coils.

I tend to agree, but I was hesitant to eliminate an option that may be beneficial. All of my limited experience has been with dual-coil. I tend to prefer them from a performance point of view. The question is, does the mono-coil provide any befit other that simplicity of design or ease of interface?

While we consider the area of coils (funny how we jump from PS to coil), is there any benefit (S/N) of a centertapped RX coil feeding a differential input stage?
Rergards,
J. L. King

[GT Blocker]

coil vs PWR

J.L. King,

"While we consider the area of coils (funny how we jump from PS to coil),"

Actually, since powering the coil and the board are so inter-related, I think that's a great topic to go with the pwr supply Q.

I do have another probably dumb question, however...
Would there be any benefit to using separate batteries for the coil and the rest? Just a thought.........

[Aziz]

Hi all,

you can still use a charge-pump DC/DC converter for the analog part. This should be controllable, which means, it should be deactivated by the micro during critical sampling processes. This can be done via PWM output disabling to avoid clocking of the DC/DC converter. Never use a DC/DC converter you can not give the clock timings. The DC/DC converter must be synchronized to the micros clock.

See the SPICE circuit simulation for some ideas (the second rar-file):
http://ibcoils.tripod.com/files/index.html

You will also see the regulation of the transmit coil power!

Aziz

[Tinkerer]

I tend to agree, but I was hesitant to eliminate an option that may be beneficial. All of my limited experience has been with dual-coil. I tend to prefer them from a performance point of view. The question is, does the mono-coil provide any befit other that simplicity of design or ease of interface?

While we consider the area of coils (funny how we jump from PS to coil), is there any benefit (S/N) of a centertapped RX coil feeding a differential input stage?
Rergards,
J. L. King

I don't see any advantage in the mono coil. My reasons for wanting a high TX voltage are for experimenting with entirely new and different methods involving several coils.
The charge pump also helps with the LI-Ion polymer batteries. These batteries pack a lot of power for little weight, but have the problem of a wide range of voltage.
Example: A nominal 14.8V LI-Ion polymer battery has a peak voltage of 16.8V and a cutoff at 11 Volt. This is just too much range to use without regulating.
Power consumption and battery weight have always been
the major problems with PI detectors. To solve these problems would be a considerable improvement.

Tinkerer

[Carl-NC]

The question is, does the mono-coil provide any befit other that simplicity of design or ease of interface?

Coil weight?

While we consider the area of coils (funny how we jump from PS to coil), is there any benefit (S/N) of a centertapped RX coil feeding a differential input stage?

Maybe, but I don't see it off-hand.

J.L. King,

"While we consider the area of coils (funny how we jump from PS to coil),"

Actually, since powering the coil and the board are so inter-related, I think that's a great topic to go with the pwr supply Q.

I agree, don't put off the power supply until the end, but consider it up front, and make changes as necessary.

I do have another probably dumb question, however...
Would there be any benefit to using separate batteries for the coil and the rest? Just a thought.........

While it might be more efficient, I don't care for this approach. I prefer a single battery (pack).

you can still use a charge-pump DC/DC converter for the analog part. This should be controllable, which means, it should be deactivated by the micro during critical sampling processes. This can be done via PWM output disabling to avoid clocking of the DC/DC converter. Never use a DC/DC converter you can not give the clock timings. The DC/DC converter must be synchronized to the micros clock.

Certainly if you use a switching reg, it should be synchronous. But it can still cause problems because, as the load current varies, the impact of the switching reg will vary.

The charge pump also helps with the LI-Ion polymer batteries. These batteries pack a lot of power for little weight, but have the problem of a wide range of voltage.
Example: A nominal 14.8V LI-Ion polymer battery has a peak voltage of 16.8V and a cutoff at 11 Volt. This is just too much range to use without regulating.

Do you mean for the coil TX circuitry, or all the circuitry? The TX circuitry can tolerate a lot of variation. Everything else needs to be regulated, but not necessarily using switching regs.

- Carl

[Tinkerer]

Coil weight?

Maybe, but I don't see it off-hand.



I agree, don't put off the power supply until the end, but consider it up front, and make changes as necessary.

While it might be more efficient, I don't care for this approach. I prefer a single battery (pack).



Certainly if you use a switching reg, it should be synchronous. But it can still cause problems because, as the load current varies, the impact of the switching reg will vary.



Do you mean for the coil TX circuitry, or all the circuitry? The TX circuitry can tolerate a lot of variation. Everything else needs to be regulated, but not necessarily using switching regs.

- Carl

With a single coil configuration the voltage difference makes a difference in the TX coil current. This in turn demands a change in the damping resistor. It is not critical, but will reduce sensitivity to small fast targets.
With IB coils the change in coil current is VERY critical.

One way out could be a coil current control or constant current control.

Using Nicads or NImh batteries, the voltage range is less, but the future seems to be in LI-Ion.
I agree that everything should be regulated.
To get high performance the noise factor is probably the most important one.

Tinkerer

[gwzd]

Do you mean for the coil TX circuitry, or all the circuitry? The TX circuitry can tolerate a lot of variation. Everything else needs to be regulated, but not necessarily using switching regs.

- Carl

Although not necessarily, but not excluding neither.
Why are you people so afraid of switchers? Just use good ones not 20 years old stuff, put some thought into the layout, filtering, synchronization, and screening and you'll have something much more efficient, lighter, nicer than the lame linear regulators alone. But, yes, it takes a little bit more refinement and ingenuity to do those.
Regards,

[Carl-NC]

Although not necessarily, but not excluding neither.
Why are you people so afraid of switchers? Just use good ones not 20 years old stuff, put some thought into the layout, filtering, synchronization, and screening and you'll have something much more efficient, lighter, nicer than the lame linear regulators alone. But, yes, it takes a little bit more refinement and ingenuity to do those.
Regards,

Not afraid, just cautious. There are many situations where you want to avoid switchers. In this case, the trade-off is between a little more efficient power supply, versus potential performance degradation. In a first cut, I would go with linear, then if switchers can offer a decent improvement in power consumption try them out, and do a direct comparison.

- Carl

[GT Blocker]

Q, Carl,

Hey there Carl,
I have some thoughts. You said above,"While it might be more efficient, I don't care for this approach. I prefer a single battery (pack)."

What is the difference between the coil's needs and the circuit's needs?
Without the coil's needs, how long could the circuit run on a smaller Li?
And would there really be an advantage to a smaller Li for the circuit and a larger one for the coil?


I wasn't thinking all that much about pwr efficiency as much as cross talk, et al.

Thoughts..........

[Carl-NC]

There are advantages to a 2-battery system. For sure less crosstalk. For sure better tailored batteries for each function. But I've never liked multi-battery systems because of inconvenience and weight. And I think it's not difficult to design around a single battery. However, if there is a clear case for a second (i.e. a usable performance improvement) then I will certainly reconsider.

- Carl

[gwzd]

Not afraid, just cautious. There are many situations where you want to avoid switchers.

Yes, there are a few, I don't see this particular one being one of them

In this case, the trade-off is between a little more efficient power supply,

Are you serious? a "little"?

versus potential performance degradation.

In which way? noise, radiated EMI, current spikes making miniature magnetic fields? These are not retorical questions, but real ones. You know better the specifics of this kind of NDT instruments and can bring light into the subject.

In a first cut, I would go with linear,

Yes, a possible, but absolutely unnecessary step. Well, I'd spare the time and use it for something else, writing in these forums for instance.

then if switchers can offer a decent improvement in power consumption try them out, and do a direct comparison.

- Carl

Yes, well, as you said that would be the cautious way of doing things, and probably you'd be right. However, putting down the specs and sticking to them is another way and you slip the making of two boards.

I saw you used in your HH a 7660 with a 10kHz switching frequency, I don't see how that is better than using switchers with 1MHz frequencies that your slow opamps aren't even able to see. The only thing that comes to mind is price, but in no way performance. Are there any subtle reasons for that?

I most agree with you that doing a reasonably good compact design on a two layer board using switcher is a challenge, but I thought that is the beauty of these forums. You find here people with very unorthodox way of thinking and I was expecting something like that. By any chase have you look inside a Keithley 2502 picoamperemeter? I have and there are a few switchers there, that's to say the least. Well, they go down to 15fA without any problems.

Regards,

[Gimp]

Switcer Plus LDO

If you are concerned about the switcher noise, follow it with an LDO (low dropout linear regulator). Set your switcher voltage at the voltage you need plus the LDO dropout voltage plus some margine (say 0.2V). This will keep your efficency high, and still offer you the low noise of the linear regulator.

Steven

[Tinkerer]

Power supply

If you are concerned about the switcher noise, follow it with an LDO (low dropout linear regulator). Set your switcher voltage at the voltage you need plus the LDO dropout voltage plus some margine (say 0.2V). This will keep your efficency high, and still offer you the low noise of the linear regulator.

Steven

Hi Steven,

do you want to get involved? Here is what we need:

Li-Ion battery has 11.1 V nominal voltage, but peak charge of 12,75V
and a low limit of 8.55V
So we want to use a SEPIC to get regulated:
+12 V for TX
+5V for digital
+5V for analog
-5V for analog
+3.3 V for analog

Analog supply needs to be low noise. Best would be isolated.
All this with low losses and low cost.

Quite a challenge, isn't it?

Tinkerer

[Gimp]

I'm not sure how much help I can be. My only experience with switchers is using National Semi Simple Switcher (design on WebBench).

One thing I do know is ground layout is critical. Single point ground under the switcher with the feedback resistors kept as far from the inductor flux path as possible.

I've spent a couple of hours this evening on WebBench without any luck. The biggest problem I'm seeing is the low steady state current demand. Most of the switchers don't have good efficiency till you are running over 1A continuous output current. It looks to me like the average current it going to be around 100mA. (My Pulse II is right at 100mA depending on settings.)

I'll keep looking.

[Tinkerer]

Power supply

I'm not sure how much help I can be. My only experience with switchers is using National Semi Simple Switcher (design on WebBench).

One thing I do know is ground layout is critical. Single point ground under the switcher with the feedback resistors kept as far from the inductor flux path as possible.

I've spent a couple of hours this evening on WebBench without any luck. The biggest problem I'm seeing is the low steady state current demand. Most of the switchers don't have good efficiency till you are running over 1A continuous output current. It looks to me like the average current it going to be around 100mA. (My Pulse II is right at 100mA depending on settings.)

I'll keep looking.

Gimp,

thanks for the effort. Maybe we should look at this challenge from a different angle.
We want to design an improved PI detector.
Lets look at ways to improve the HH design.
There are a few decisions to take:

• How much power do we want to use, or, if we have the possibility to chose, what is the minimum power use and what is the maximum power use.
  • How do we want to divide the power use. 50% on the TX and 50% on the signal processing?
    Or 80% on the TX and 20% on the processing? Other ideas?
    Cost considerations:
    Using LI-ION Polymer batteries that are very light, we can use more power, but at higher cost. Higher battery voltage also means higher cost. What is more expensive, a SEPIC power regulator to produce 12V from 8.5V to 12.75V or a 14.5V battery and a BUCK configuration?
    
    Any input is welcome
    
    Tinkerer