Toroid Rings.

Question

Choosing the correct toroid ring.....

I’d like guidance on choosing the best component for a try at making a 60KHz RF amplification strip for a (kind of) TRF radio receiver. Having tried to study the topic, I think I now know less than I did to begin with!

As a starting point consider the cost as no object, and a selection can only be made from the stock list of Farnell, with whom I have a trade A/C. The overriding need - as I understand it - will be for the best possible Q.

I have a small selection of ‘magnet wire’ + some 68 x 0.04mm Litz wire.

Wisdom, guidance, etc, most gratefully received.

Thanks.

John

michaelkellett · Answer

A toroid won't work like a ferrite rod aerial - put simply there is no way for the magnetic component of the RF to get into the core and generate emf in the coil;. I was interested in this too - so I've ordered some ferrite rods from Amazon https://www.amazon.co.uk/gp/product/B074W4RNCV/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1 I found this article which is fun: The Creative Science Centre - by Dr Jonathan P. Hare I'm thinking of a minimal RF amplifier and using an STM32Hxxxx processor with its 16 bit ADC to do a SDR (Software Defined Radio) and demodulator. How far I get will be time limited (as usual). @John 60kHz is quite a low frequency for LC filters - the Ls need to be big and can be expensive and/or quite lossy. But active RC filters work very well in this frequency range. The MSF signal uses on/off modulation with 100ms time slots, so a bandwidth of about 20 to 30Hz at the input to the de-modulator seems about right. That's quite hard to do with analogue filters but easy in maths - which is why I fancy the ADC and processor for the filtering. MK

johnsmalley · Answer

Guys .... Thanks for the responses. Just to clear it up, I’m *not* trying to use a toroid core as an antenna. I’ll mess with a ferrite rod for that. The toroids are for a kind of IF strip but, of course, it’s not IF it’s at the antenna frequency to make a TRF receiver (but at one, fixed frequency = 60KHz). These coils will be tuned with a C to make a tank cct amplified by a BJT or an FET or a chip. My reading tells me that the best Q is always a compromise, but it is to be had with the largest practical core. Then the fewest possible turns to reduce capacitive losses between windings and to the core. Further, the core material should be designed for the frequency of interest. In addition the skin effect @ frequency of the wire needs to be chosen correctly: I already have Litz wire that should work. (Did I miss anything to reduce losses?)! I have some 40mm 3E27 cores laying around. They should be ok. Experimentally, 10 turns of .75mm enamelled Cu wire makes about 1.5 mH. I resonated that with the desired C. Signal if fed in with a loop close by and is extracted to a scope with a pickup turn on the core. Or, the inductor (no capacitor) is connected to my B&K 880 LCR meter where Q, D, ESR, Z, DCR, R, theta can be measured. I’ve tried different gauge wires wound in different ways to get an idea of how these things affect the inductor properties for the same value of L - just for learning. As mentioned, all this is pretty new knowledge to me so I’m likely to have got stuff wrong and missed important points. John

14rhb · Answer

I'm old too . I think your question has been answered in my absence by several knowledgeable Element14 members. The blogs are a nice way of sharing our fun and experiences in electronics (and related disciplines) with like-minded folk. The modern day equivalent of a show-and-tell I guess. Once someone makes a blog (plenty of photos and pictures/diagrams) they often get a flurry on interesting comments - sometimes the discussions go on for weeks. If someone writes a blog on something they may highlight issues they had or are experiencing - again fellow E14 members often add helpful comments. It really is a great community - I encourage you to at the very least log in often and watch the content, or better still just jump straight in . If you are ever in doubt, just ask in the same way you started this discussion.

johnsmalley · Answer

Hello, Michael, and thanks for the reply. The request is serious in that I'm trying to learn, but hobby in that I now do electronics as a retired electronics man (with little knowledge of coils and most other things radio!!) You're right. It's about MSF. To that end I bought one of those Canadian receivers (yet to arrive) but I also use the opportunity to wind coils, try this, try that, try to understand the maths, etc. So I decided to see if I could make a simple TRF radio with a couple of tuned RF / audio stages. If all else fails I still have the bought receiver, but that will teach me nothing about LCR circuits. So you see my quest is a bit of both hobby and serious. I will probably try a crystal stage to see how I get on with that, too, but I kind-a want to make a tuned stage just for the hell of it ... to see if I can make it work. Learn something about ferrites and so on. I looked at toroids because my reading-up seemed to show they might work, and I get to learn a bit about those parts, their data sheets, etc; I get to learn even if toroids don't work like that!! Self teaching is always full of difficulties, and I'm at that stage with the coils where I need a hand as the Q of my built coils is always too low (B&K 880 LCR meter) with 90 being about the best no matter what I wind them on, or how, or with. There we are. That's the story. John

shabaz · Answer

Hi John, As Michael says, normally a ferrite stick is used, because it can provide some directionality without needing much of an antenna. With a toroid, you may need to experiment more with the antenna. Another reason (practical) is that since this is such a low frequency, it will need winding a lot of turns, and that's not fun with a toroid. Anyway, long story short, it is a fairly experimental process to build up the toroid. There is no optimum method for selecting toroids, some of it will need to be by experiment, and some of it by what is available out there. You can search online for the LC formula for your resonant circuit (if this is what you're making), and pick a value for C and look for reasonable values for L. So, for example, if C is about 700pF (this value can be achieved with a few hundred pF fixed capacitor, in parallel with a variable capacitor for trimming), then L is 10mH. Next step is how to construct such an inductor, and you can seek out the different types of toroid material, and manufacturer recommendations for frequency. For 60kHz, type 'J', type '61' and type '67' would all work (as would some others). The manufacturer will have formulas to work out how many turns, but it is just a ballpark (it varies with batch, and with temperature too). Since you need a lot of turns, the toroid needs to be large too. There is an online calculator here, and it is as accurate as possible (I've confirmed this in the past), but as mentioned expect variation, and you'll need to experimentally check : https://www.changpuak.ch/electronics/amidon_toroid_calculator_2.php Using that, something like FT-87-J would work, with about 58 turns, will equal 10mH approx. There's other things that will affect Q, but with such a high inductance value (relatively), you don't have much control over that once you've picked the toroid, because it is a lot of turns. Anyway, it is a niche thing and Farnell won't have a great selection for the toroids for radio applications, very few distributors sell them. Most of the toroid cores from distributors are intended for use as chokes, and they don't specify the nH/turns^2 value (which is what is needed for calculating inductance). For suitable cores, there is a US source (Amidon) but shipping and minimum order costs are extremely high last time I checked (it was not worthwhile for an order less than a couple of hundred $). So, it may be good to use the online calculator for a few different toroids, and hunt on ebay etc., to see if anyone is selling any of those particular toroids. If you don't want to wind a toroid, you could use an off-the-shelf 10mH unshielded inductor. It won't have very high Q, but you may not need very high Q anyway. You could always swap it out later if it wasn't performing as well. Anyway, it sounds like a fun project.

michaelkellett · Answer

It's dead easy on E14. When you are logged in, click the create button at the top right hand of the screen, next to your name, and then select Blog Post. From there on it's pretty straightforward. MK

michaelkellett · Answer

Here's a link to one I prepared earlier. Capacitor Leakage Current over Temperature It describes some experiments I did to measure some aspects of the performance of capacitors. When I get my ferrite rods I'll make an aerial and blog about it - just like writing the experiment up in a notebook like we did in the olden days. MK

michaelkellett · Answer

At 60kHz your best bet is a pot core, if you want to use an inductor. You need a gapped core (ones without gaps are good for power but less good for filters), made from a ferrite suitable for filters in the frequency range you are interested in. The EPCOS (TDK) https://www.mouser.co.uk/ProductDetail/EPCOS-TDK/B65811F315A48?qs=sGAEpiMZZMs2JV%252BnT%2FvX8JpxZdCKKXlRTnIVKnqpr08%3D will do. The Al value is 630nH per turn so 100 turns will get you 6.3mH. It's hard to find anyone in the UK with a good range of pot cores now so you'll need to go to Mouser for these. (They have a UK based web site). You will need a bobbin and clips to hold the cores together. I'll leave you to do some reading - download the data sheet from Mouser. This core can accept an adjuster. MK.

michaelkellett · Answer

Hello John, first, is it really 60kHz - ie 60000 Hz, nearer audio than what most would call RF If so - what is you are trying to build - I can't see what you would want a toroid - if you want a ferrite rod aerial that isn't a good start. Is the application for fun or for serious stuff, and I'm guessing you want to pick up the MSF signal from Anthorn ? MK

dougw · Answer

blog is short for web log . element14 provides tools and resources to create, store and publish notes and logs on projects that members create.

shabaz · Answer

Also, it's a compromise. If you choose a material and/or dimensions such that you need fewer turns, then Q can be higher. So you'd think you should go for a material that need fewer turns, but such material varies more with temperature, which then means your high Q resonant circuit might shift in frequency. Plus, the larger it is, the more resistance. So then you could choose thicker wire to reduce resistance but then you may not have space to get as many turns to reach the required inductance. You could space the turns but then that affects inductance and you may not be able to fit the amount of turns without thinner wire. In summary the parameters are interdependent, and I don't know of any software that would help. You could design such software to iterate, varying each parameter bit-by-bit, and give you an answer, but it's a big task maybe.

dougw · Answer

If (decades old) memory serves, high Q coils are characterized by large diameter air core windings with gaps between each turn of low impedance wire.

shabaz · Answer

Hi Michael, Oh that's a shame. They were of a decent length too.

I tried a whip antenna just now into a HF receiver to see if it would pick it up in CW mode to record a WAV file in case it helps for your code. I've put the zipped-up MSF wav file here it is recorded from about 12:11 BST (11:11 UTC) today (Fri 4th 0ct 2019) and the recording is just over four minutes long. The receiver was in CW mode, set to 1kHz tone. The tone is audible, but would need some signal processing (the antenna wasn't very good, it was just a short length antenna).

EDIT: After some filtering in an audio app (I don't have matlab on this PC) this is the spectrum, for about 5 seconds. The horizontal scale is seconds, i.e. running from 30.0 sec to 35.2 sec. The carrier off period is visible, seems to be off-periods of 200, 200, 300, 300, 300 msec in this portion.

Toroid Rings.

Top Replies