Amidon and Micrometals toroids.. other brands ?

[jpdesroc]

Hi everybody,
I'm designing with toroids.. mostly for MF RF input and output transformers
(500 Khz to 5 Mhz range) for RF power amplifiers.
I am facing a problem though..
Every single toroid transformer's design examples
I can find on the web are according to
2 toroids suppliers: Amidon and Micrometals.
These toroids seem to be available only from
these own companies.. (minimum qty buy with shipping delays..)
But I'd like to order from Digikey or Mouser
because of their low minimum parts count / 1 day delivery time..
The problem is when I try to figure out which
type of toroids would be the best for these frequencies
in their available toroids list.. ex Digikey here:
https://www.digikey.ca/products/en/i...=1&pageSize=25
I'm very confused on what type I should buy..
They all seem to be made for frequencies higher than 30Mhz and more..
From Amidon it looks like the T50-1 BLUE 500khz to 5Mhz would fit my needs.
From Digikey or Mouser I cannot figure another
brand / part number that would be equivalent..

Does anybody have an easy way for me to figure out
what parameters should be checked Al / Q / Freq self resonance.. etc..
Then I could do some comparisons between Amidon/Micrometals
and other easy to find/buy brands from again Digikey or Mouser because
of their fast shipping time. ??

Thanks.

[Argus25]

Hi,

If you are making RF input and output transformers in this frequency range the type of ferrite that is ideal is type 77. Also, it is actually better to use a different configuration than a toroid.

I would recommend the RF-400, RF800, and RF 2000 transformer cores from CCI.
They are easier to wind, easier to mount as the ends are solderable pcb material, they have an internal riveted tube construction that acts a a primary or secondary. After using these, I didn't go back to toroids for any of my broadband RF power amplifiers because these are superior and you can buy them in any amount you like:

http://www.communication-concepts.co...0-transformer/

Hugo.

[Radio Wrangler]

Bill Amidon of Hollywood is a small distributor who has been very helpful to radio amateurs in the US. He published a catalogue of the cores he sells which helped them become popular. The cores he sells are made by Fair-Rite inc for the ferrite ones and Micrometals for the dust-iron ones.

For MF and HF transmitter transformers, ferrite materials are needed rather than dust-iron.

As Hugo says, other core shapes (binocular cores) can be better in these applications.

You can get Fair-rite and Micrometals cores from a variety of different distributors, but they may not be interested in small quantities though you should try asking you might get them as free samples.

David

[jpdesroc]

Thanks for your nice suggestions.
But still I'm stuck with 3 brands..
Amidon, Micrometals and Fair-Rites.

The only new thing I get now is that 77 is the right type
of material to be used at my needed frequencies (500 Khz to 5 Mhz range).
As Hugo suggested the CCI RF-400 kit would work for output.
I will need to ask for samples to get at least one..
Browsing at other distributors I still don't know
what important --> parameters <-- I should pick to filter out
what would fit my objectives..
Again thanks guys for your nice replies.
JP

[Terry_VK5TM]

Amidon is not a manufacturer, it is a distributor of the other two, Micrometals and Fair-rite.

These are the two main manufacturers that are relatively easily accessible by people in our hobby through various outlets.

There are some made in China, but that is a mine field as far as knowing exactly what you will get.

Not 1 day shipping, but this is another place to buy toroids (and other things)
http://www.kitsandparts.com/toroids.php

Also, the toroid calculator may be of interest:
http://toroids.info/ (click the toroid part number in the blue panel to get the different calculators).

[Radio Wrangler]

Fair-Rite inc. make only ferrite cores.

Micrometals inc. make only metal dust cores.

Amidon is simply one of many distributors of these.

"Mix" numbers are not absolute. Different makers sometimes use the same number to name quite different materials, so you have to name the manufacturer along with the material designation. Otherwise you could get the wrong stuff.

Just to add to the confusion, different manufacturers apply unrelated colour codes to their materials. If you find a yellow core you don't know what it is unless you know who made it.

EG. I use a lot of cores in Fair-Rite number 43 material. It makes good transformers at MF and HF frequencies, and above 100MHz it acts as an EMC filter. It is a ferrite. The exact same cores, made by Fair-rite, can be bought from Amidon, who calls the material FT-43. It's the same stuff made in the same factory in New Jersey.

To choose a material, you need the manufacturer's data sheet. You need to look at the graphs showing saturation levels and loss factors versus frequency for the material. Then you need to look at the data for the different sizes and shapes of cores available in the material you've chosen.

Designing the transformers for an HF transmitter power amplifier is not easy and a lot of care has to be taken over the minimisation of strays and leakage inductance. Unless you get this right, you will keep destroying sets of power transistors.

If you want to get into the heavy end of serious RF design, I'd like to encourage you. There aren't many of us! (My day-job is designing transmitters for aircraft radar and radio systems).

If you want to make a low power transmitter similar to the many designs for QRP equipment, then the output devices are cheap and things at this power level are tolerant. It is a good place to start.

If you just want a high power transmitter for less money than buying one ready made, then developing your own usually costs more! Maybe the best approach would be one of the kits from Communications Concepts Incorporated, usually referred to as "CCI"

There were some superb applications notes published 30+ years ago by Motorola. They explain a lot. CCI did kits for the designs in them. Helge O Granberg was the author. Helge also published a book on power amp design. His work is still the foundation of most current PA designs.

David

[Andrew B]

For what it's worth, I needed some cores for a project, Micrometals kept on coming back with minimum quantity, packaging, DHL etc etc. I kept on pestering them saying that I would send a $20 note along with a self addressed envelope and that I would "take my chance" with USPS, in the end they relented and sent me half a dozen different "slug" samples FOC. Result

[kellys_eye]

Spectrum Communication have a reasonable range of cores with 'identification' and sell in one-off quantities. No connection with them, just a satisfied customer!

http://www.spectrumcomms.co.uk/Components.htm

[jpdesroc]

Originally Posted by Terry_VK5TM

Amidon is not a manufacturer, it is a distributor of the other two, Micrometals and Fair-rite.

These are the two main manufacturers that are relatively easily accessible by people in our hobby through various outlets.

There are some made in China, but that is a mine field as far as knowing exactly what you will get.

Not 1 day shipping, but this is another place to buy toroids (and other things)
http://www.kitsandparts.com/toroids.php

Also, the toroid calculator may be of interest:
http://toroids.info/ (click the toroid part number in the blue panel to get the different calculators).

very useful info !!
Thanks.

[jpdesroc]

Originally Posted by Radio Wrangler

Fair-Rite inc. make only ferrite cores.

Micrometals inc. make only metal dust cores.

Amidon is simply one of many distributors of these.

"Mix" numbers are not absolute. Different makers sometimes use the same number to name quite different materials, so you have to name the manufacturer along with the material designation. Otherwise you could get the wrong stuff.

Just to add to the confusion, different manufacturers apply unrelated colour codes to their materials. If you find a yellow core you don't know what it is unless you know who made it.

EG. I use a lot of cores in Fair-Rite number 43 material. It makes good transformers at MF and HF frequencies, and above 100MHz it acts as an EMC filter. It is a ferrite. The exact same cores, made by Fair-rite, can be bought from Amidon, who calls the material FT-43. It's the same stuff made in the same factory in New Jersey.

To choose a material, you need the manufacturer's data sheet. You need to look at the graphs showing saturation levels and loss factors versus frequency for the material. Then you need to look at the data for the different sizes and shapes of cores available in the material you've chosen.

Designing the transformers for an HF transmitter power amplifier is not easy and a lot of care has to be taken over the minimisation of strays and leakage inductance. Unless you get this right, you will keep destroying sets of power transistors.

If you want to get into the heavy end of serious RF design, I'd like to encourage you. There aren't many of us! (My day-job is designing transmitters for aircraft radar and radio systems).

If you want to make a low power transmitter similar to the many designs for QRP equipment, then the output devices are cheap and things at this power level are tolerant. It is a good place to start.

If you just want a high power transmitter for less money than buying one ready made, then developing your own usually costs more! Maybe the best approach would be one of the kits from Communications Concepts Incorporated, usually referred to as "CCI"

There were some superb applications notes published 30+ years ago by Motorola. They explain a lot. CCI did kits for the designs in them. Helge O Granberg was the author. Helge also published a book on power amp design. His work is still the foundation of most current PA designs.

David

David,
My aim is at building a 5-10w RF amp working below 2Mhz.
The T50-1 toroids and RF400 xformers from CCI seem to be a good choice here..

[Radio Wrangler]

Ah, There's a nice design for a push-pull 10 watter by Chris Honey which uses a pair of common CB output transistors.

For work below 2MHz you may find ferrite material better than the iron dust T50-1.

At this power level you can play and avoid all the expensive smoke that accompanies many higher power exploits.

10W used to be the British power input limit on topband.

David

[Argus25]

I would have a close look around on CCI's website, many of the amplifiers there there were designed by Mr Granberg.

One amplifier there, I have built many times is the AN779L, which uses inexpensive transistors that were intended for CB work. I'm pretty sure you can just get the circuits/manuals from CCI too.

http://www.communication-concepts.co...-for-leadtime/.

Be cautious in interpreting the 20W power output rating of these, it depends whether you intend to use them for SSB or continuous carrier output with amplitude modulation. In the latter case they get pretty hot even with a substantial heatsink over 7 to 10W and at the lower frequency end of their range and at high output powers the cores do result in some non linearity, so it depends how distortion free you want your RF carrier.

There is also the absolutely wonderful AN779H 12 which is my favorite. It was designed to be fed at its input by the MHW591 Motorola module, they are still available, most of the remaining ones are in Malaysia. Though all sorts of other drivers can be used, I have used wide band OP amps like the AD8056. (You really haven't lived until you have looked inside a module like the MHW591, and seen how Motorola executed the design, it is really quite something, I documented the circuit once some years ago)

Hugo.

[Argus25]

I've attached something that might be of interest, a 4 watt push pull RF amplifier using RF400 type 77 cores and three ceramic screw down 2N5945 RF transistors, but many types would work, I have a box of these so I thought I'd put them to good use.

The drive to get it to full output power (20v peak across a 50R load) is 200mV peak and the current consumption 0.61A. (With a 10V pk output level into 50R the current consumption is about 0.38A) The Z in is about 150 Ohms.

It was designed to operate on 1400KHz, but is broadband due to the RF400 transformer cores and all that requires adjusting for optimum operation and minimal distortion on other frequencies is the 1100pF cap on the output transformer. It gives a good result with a low distortion sine wave.

At these frequencies the schematic looks a lot like its audio amp counterpart, but of course it is important that the pcb is made with a good ground plane and an adequate heatsink for the transistors. The 1N4004 can be thermally coupled to the heat sink. Sorry about the messy hand drawn diagram.

[jpdesroc]

Thanks David,

Originally Posted by Radio Wrangler

Ah, There's a nice design for a push-pull 10 watter by Chris Honey which uses a pair of common CB output transistors.

Nice design.. but it works above 2Mhz..

For work below 2MHz you may find ferrite material better than the iron dust T50-1.

Ok then which ferrite material would you recommend for input and output transformers below 2Mhz for 5-10watters ?

[Radio Wrangler]

Fair-rite number 43 aka FT43 is what I normally go for with enough turns to get enough inductance at MF-HF. I also have some Siemens N28 and a similar material from SEI kicking around.

David

[jpdesroc]

Originally Posted by Argus25

I've attached something that might be of interest, a 4 watt push pull RF amplifier using RF400 type 77 cores and three ceramic screw down 2N5945 RF transistors, but many types would work, I have a box of these so I thought I'd put them to good use.

The drive to get it to full output power (20v peak across a 50R load) is 200mV peak and the current consumption 0.61A. (With a 10V pk output level into 50R the current consumption is about 0.38A) The Z in is about 150 Ohms.

It was designed to operate on 1400KHz, but is broadband due to the RF400 transformer cores and all that requires adjusting for optimum operation and minimal distortion on other frequencies is the 1100pF cap on the output transformer. It gives a good result with a low distortion sine wave.

At these frequencies the schematic looks a lot like its audio amp counterpart, but of course it is important that the pcb is made with a good ground plane and an adequate heatsink for the transistors. The 1N4004 can be thermally coupled to the heat sink. Sorry about the messy hand drawn diagram.

This is almost exactly what I was looking for..
A simple MF amplifier made of available parts with all
the needed windings well explained.
MANY THANKS !!!

[Argus25]

Originally Posted by jpdesroc

This is almost exactly what I was looking for..

One trick I could have mentioned mounting the 1N4004, it can be mounted on the PCB near one of the output transistors with leads long enough so it can be bent over so one edge of its long axis touches the side of transistor body and a blob of heat conducting compound added to help thermally couple it.

Another method is to use an earth lug, wrap the long flat end around the diode body and and put the lug under one of the transistor's mounting screws, say if its a TO-220 type.

Since its only a 4W amp and there is emitter stabilization too, it is actually ok just to have the 1N4004 on the board nearby, with an adequate heat sink on the transistors.

Many of the broadband RF power amps you will see have less or no emitter degeneration, to get more gain & power, and they have much more elaborate, often IC controlled and temperature compensated bias stabilization instead. The advantage of the emitter degeneration it is that it helps with thermal stability, simplifies the bias circuit and reduces distortion at the cost of some gain and output power, but for a low power amp, its worth that trade.

[jpdesroc]

Originally Posted by Argus25

I've attached something that might be of interest, a 4 watt push pull RF amplifier using RF400 type 77 cores and three ceramic screw down 2N5945 RF transistors, but many types would work, I have a box of these so I thought I'd put them to good use.

The drive to get it to full output power (20v peak across a 50R load) is 200mV peak and the current consumption 0.61A. (With a 10V pk output level into 50R the current consumption is about 0.38A) The Z in is about 150 Ohms.

It was designed to operate on 1400KHz, but is broadband due to the RF400 transformer cores and all that requires adjusting for optimum operation and minimal distortion on other frequencies is the 1100pF cap on the output transformer. It gives a good result with a low distortion sine wave.

At these frequencies the schematic looks a lot like its audio amp counterpart, but of course it is important that the pcb is made with a good ground plane and an adequate heatsink for the transistors. The 1N4004 can be thermally coupled to the heat sink. Sorry about the messy hand drawn diagram.

Newbie question..
Looking at the push-pull amp schematic
made of RF400 transformers you just supplied
can you tell me how to wind the primary
of the second transformer (the output xformer) ??
I've been on the web for a while looking at many
different ways to wind this type of RF transformers
at the separate connections side (not the tap side)
Please have a look at 2 possible ways I found to wind the primary:
http://www.arcenson.com/public/RF400_windings.pdf
According to your last push-pull class AB schematic,
which way is the good way to wind the output transformer primary
and why ?

Thanks
J-Pierre

[Argus25]

Originally Posted by jpdesroc

which way is the good way to wind the output transformer primary
and why ?

Thanks
J-Pierre

The thing to understand about transformers with the metal tubes, they were conductors created to give the opportunity for a very low Z primary, meaning the voltages are low and the currents high. So When you look at the transformer, what you are seeing is a "one turn center tapped primary" which is convenient for many applications. Also, it makes for convenient PCB mounting with the pcb strips they are soldered to.

For example with a two turn secondary on the RF400 core (making a 1:2 turns ratio transformer), the impedance ratio is 1:4 meaning in this case that if the primary center tap is used to supply power and the collectors of the push pull transistors are connected to either end, and say the secondary loaded with 50 Ohms, the impedance ratio (collector to collector) is 50/4 Ohms.

However, the 1 turn primary, in conjunction with the size of the RF 400 core, results in primary currents that are too high using the core at the low frequency end, less that 2MHz even for a 4W amp and core saturation degrades the linearity. And it needs more primary inductance and turns.

Two turns a side (4T center tapped) is ideal in this case, for the primary of the output transformer and the RF400 core. So the metal tubes in the core don't have to be used at all and just thread the teflon coated wire around. I still use the PCB ends for mounting it, but not for an electrical connection. However, as shown in the attached diagram, giving some examples, if you want, you could make use of the metal tubes and connections and get the same result for the primary. The reality is, if you have more than 1 turn required on any of the transformer windings there is no requirement to use the metal tubes as part of any winding.

I used teflon coated wire which is thinner OD than that supplied with the RF400 or it can get a bit crowded in the tubes.

PS : I think the arrangement in your diagram B is defective because the currents in the wires from the collectors and the metal tubes, end up running in opposite directions and the magnetic fields from those conductor pairs would cancel out.

[jpdesroc]

Very informative stuff you gave me here..
Nowhere on the net I could find a nicely explained example
of how to 'understand' these transformers behaviour,
specially at low frequencies..

By the way I want to use an RF choke between the +12vdc
and the output transformer primary tap.
What do you suggest ?
I have some 330uH@3.3A (100mOhms) powdered iron toroids on hand..
The 330uH would give around 2.9k ohms@1.4Mhz.
I would put 0.1uf on each side of it to the ground..