TF-900_TM800

[ivconic]

Yes, 1020nH is no doubt the value. This means 1020nH/turns-squared.

The number of windings required to achieve an inductance of 13mH would then be:



which I assume would be too many to fit within the space.

Yes, there probably won't be much space left.
Another question bothers me, the color is blue.
I still don't understand if it's the type that stores the energy or the type that suppresses and "consumes" the energy (filter type)?

[Qiaozhi]

I stumbled on this and it might be usefull:
https://coil32.net/rf-toroid.html
Downloaded it and now trying to see some details.

Honestly, it doesn't need to be that complicated except in high-frequency circuits, and power applications where the core can become saturated. At the frequencies you're talking about the value is the simplest solution.

As an example, I needed to use an iron powder core to avoid saturation, and I was limited to only about 45 turns by the wire and core size.
The core I chose had an value of , which I calculated as requiring 45 turns to achieve an inductance of 259.2uH.
In practice I was only able to comfortably get 44 turns on the core, and the inductance measured as 260uH, so the calculations are very close to reality. As you might expect, in practice you may have to add or subtract a turn to get closer to the desired result.

[ivconic]

Honestly, it doesn't need to be that complicated except in high-frequency circuits, and power applications where the core can become saturated. At the frequencies you're talking about the value is the simplest solution.

As an example, I needed to use an iron powder core to avoid saturation, and I was limited to only about 45 turns by the wire and core size.
The core I chose had an value of , which I calculated as requiring 45 turns to achieve an inductance of 259.2uH.
In practice I was only able to comfortably get 44 turns on the core, and the inductance measured as 260uH, so the calculations are very close to reality. As you might expect, in practice you may have to add or subtract a turn to get closer to the desired result.

Yes, few turns more or less will not make a problem.
Original core with coils allows approx. 1mm extra space when pipe is put through.

[Qiaozhi]

I still don't understand if it's the type that stores the energy or the type that suppresses and "consumes" the energy (filter type)?

I think what you're talking about is just the way the core is used in the circuit. This is not a function of the core material.

Also, the colours do not mean anything. There is no colour coding of cores in the same way as resistors. It appears that the core colour means different things to different manufacturers.

[ivconic]

I think what you're talking about is just the way the core is used in the circuit. This is not a function of the core material.

Also, the colours do not mean anything. There is no colour coding of cores in the same way as resistors. It appears that the core colour means different things to different manufacturers.

Uh ... that's a real relief to hear!
One problem less to worry about!

[Skippy]

They all store energy .... but they all have some maximum useful frequency, above which they get lossy. The higher the u , the worse they perform at high frequencies. If you want to make a balun for a 50 MHz RF power amplifier, you're going to use materials with u less than 20, because they work at high freqs. If you choose some super-high u material, like 10000, the max useable freq drops to a few kHz before they get lossy. Example: common mode filters need to pass 50 / 60 Hz AC, but block crap from the SMPSU stage, which could run at 25kHz - 500 kHz typically, with harmonics creating interference into the MHz range. So they need lots of inductance, and absorbing energy at 50 kHz is beneficial. If you're making a SMPSU running at 50 kHz ( NatSemi Simple Switcher for example ) you need medium u material; you want it to store up energy, that's what SMPSU's do, so material with u =10 is no use. But u = 5000 cores are also useless, because they will absorb energy, and give an inefficient PSU. So such a SMPSU would use materials with u = 100, likely powdered iron type, ( this is probably what those white/yellow cores you have are).

It's mostly powdered iron cores that are colour-coded, and there is some sort of 'standard colour code', but you know that not many will use it. Here is one table, see the white/yellow is u = 75.

https://www.nutsvolts.com/questions-...ility-measured

[ivconic]

They all store energy .... but they all have some maximum useful frequency, above which they get lossy. The higher the u , the worse they perform at high frequencies. If you want to make a balun for a 50 MHz RF power amplifier, you're going to use materials with u less than 20, because they work at high freqs. If you choose some super-high u material, like 10000, the max useable freq drops to a few kHz before they get lossy. Example: common mode filters need to pass 50 / 60 Hz AC, but block crap from the SMPSU stage, which could run at 25kHz - 500 kHz typically, with harmonics creating interference into the MHz range. So they need lots of inductance, and absorbing energy at 50 kHz is beneficial. If you're making a SMPSU running at 50 kHz ( NatSemi Simple Switcher for example ) you need medium u material; you want it to store up energy, that's what SMPSU's do, so material with u =10 is no use. But u = 5000 cores are also useless, because they will absorb energy, and give an inefficient PSU. So such a SMPSU would use materials with u = 100, likely powdered iron type, ( this is probably what those white/yellow cores you have are).

It's mostly powdered iron cores that are colour-coded, and there is some sort of 'standard colour code', but you know that not many will use it. Here is one table, see the white/yellow is u = 75.

https://www.nutsvolts.com/questions-...ility-measured

Alright!
Now is all clear to me!
Guys, Skippy and Qiaozhi; combining your posts today I learned indeed much!
Big THANKS!

[ivconic]

... Green cores are from common-mode suppression filters, ferrite with u = 5000 -> 20000, but designed to soak up AC energy , not store it efficiently, so they have limited uses....

Since I don't have other cores available, these two greens are a bit closer to what I need. The real question now is; whether it will behave well on:

A) 12.5 Khz when used as cache locator
B) 73.5 Khz when used for cave mode

The fastest way to find out is to wind them up and try with a detector.
I have no other choice for now anyway. Until I locate where I could get more appropriate cores.

[Skippy]

If you are wanting to make some progress:
Use the RX toroid to construct a working TX stage. Then use your 'junk box' stuff to see if it's any use on the RX board.

Also: The polarity of the TX relative to the RX is almost certain to be important, so be prepared to swap the RX core wires over if the thing doesn't work OK.

[Jose]

One question I have with this detector is whether the RX stage should be tuned to the same frequency as the TX, or should be different as in the case of VLF detectors, where the RX tuning is slightly different from the TX.

[Skippy]

Jose: it looks like they will end up tuned close to each other. The only real difference is the RX PCB has additional pads so more tuning capacitance can be added in parallel with the 33nF. This would tune the RX lower than the TX, of course. And this assumes the previous reverse-engineering, stating N = 75 Turns on both cores is correct. If in fact the TX had 78T, for example, it would make a signficant difference ( about 4% in this example ) to the frequency.

[Jose]

Jose: it looks like they will end up tuned close to each other. The only real difference is the RX PCB has additional pads so more tuning capacitance can be added in parallel with the 33nF. This would tune the RX lower than the TX, of course. And this assumes the previous reverse-engineering, stating N = 75 Turns on both cores is correct. If in fact the TX had 78T, for example, it would make a signficant difference ( about 4% in this example ) to the frequency.

Thanks, I think the same as you, I tuned it the same as the TX frequency. I had observed phase differences with respect to the TX signal, if it is tuned differently. Hence the question and the doubt that he could have taken a different path.

[Qiaozhi]

LTSpice simulation attached.

[Skippy]

It would be interesting if George could model the addition of the shorted TX loop. Specifying the resistance of the loop would be tricky ... skin effect would play a part.
I'm thinking the loop would behave like adding a parallel resistor across the 13mH/33nF combination,
example if the loop resistance = 100 mR, this would equate to 75 x 75 x 0.1 = 560R. on LC .
X_C = X_L = 690R, for comparison.

[ivconic]