circuit design till 100MHz

100MHz is most certainly "high frequency", and signal integrity in PCB design will matter a lot.

Dave.

these days with such small components, you can make traces so short, as long as you have a good ground plane and with such good low capacitance devices upto 100 mhz its almost like playing with audio used to be.

depends what you are trying to do of course and trying to breadboard at 100mhz is a different matter, ofc theres still plenty of room to get into trouble, and pcbs do start to exhibit frequency dependant charecteristics at less than 100mhz if you have long enough traces. theres lots of old info on that with regard to making wideband 100mhz+ oscilloscope amplifiers etc.

2ghz can be a challenge, and my laser didnt like being modulated at 2ghz, ive not gone over 2ghz so far, unless you include harmonics, although my simple srd sampling head seemed to show harmonics quite a bit higher, I have no idea how acurate it was.

Colin =^.^=

hi, i am designing some power amplifiers. i can see most of the capacitor after few MHz is dominated by its parasitic inductance.Similarly inductors become capacitors. Transformer is no more an ideal one. the magentic core also behaves erratic way. Do check the permeability vs. freq characteristics. Damn,the permeability is flat till only few MHz. I am looking for some references where these issues are discussed. i am happy enough if i can design working circuit till 100Mhz. I taked to these RF people. They just say do not worry yours is not high frequency, they say for 100Mhz, the wavelength is 3m, so you do not need to worry much about the hish frquency effects. I am not convinced. Is there any information regarding this where i can read more on it. kristo

I have not ventured into high frequency design, even at this moderate level, but I found a lot of good information in application note AN47, for high speed amplifiers, on the

website. It is a well written, comprehensive, and informatively illustrated PDF document. There are also references to other documents that might help.

Paul

Thanks Paul. i am desparately looking for more references. I just need references to read. I want to learn the topic and master it.

snipped-for-privacy@hotmail.com a écrit :

Listen to them again. I don't think they told you you won't suffer from non ideal components effects, but rather that you won't suffer from transmission line effects. IOW, you can consider your PCB tracks as lumped elements rather than distributed ones. It's not the same thing at all.

As an example, an easily reached 1pF PCB track parasitics is only 1.5k at 100MHz.

--
Thanks,
Fred.

Thanks Fred. I got your point.Is there any literature on it? Thanks.

If you look at an RF book it will probably go into more detail than you want, if you say why you are wanting to use cored transformers we can see if theres an alternative. the parasitic inductance of capacitors at 100mhz can be minimised by using smal value capacitors, that are physicaly smal too, and very short trace lengths, ie zero length. aim to make your 100mhz circuit the size of a small postage stamp.

ive resorted to putting SMD capacitors through the hole of a 2 sided pcb before to reduce inductance, but thats not realy needed at 100mhz.

it very much depends on what it is your trying to do, ie if its wideband or narowband.

Colin =^.^=

if thats the one I think it is its excelent, its a bit humerous in places too :) gives good advice about grounding, supply decoupling, using short lead lengths, ground plane, sheilding etc, conecting scope leads too.

Colin =^.^=

Thanks colin. Yup i really do not want to goto RF design. Having said that i also do not want to just design the circuit for low frequency. Is there an alternative for cored transformer?

Its hard enough getting audio to stay stable never mind megahertz !

that depends entirly on what you are trying to do, but the answer is - probably.

Colin =^.^=

You should also ask your self why you don't want the transformer. You are likely to be better off using one if you are working with a wide band signal and need to change impedance levels or split or combine drives.

i cannot do wothout transformer. in my application i am using the transformer to couple the high frequency line into the power lines. i cannot do without it. But is there is some way to design transformer without core then it would be wonderful.

I gues you cant, however you can get cores wich go upto 100mhz or more, micrometals do them. the higher frequency the smaller the particles its made of but the lower the inductance too. guesing your bandwidth isnt that high, you can tune out the undesirable parasitics.

Colin =^.^=

Just a nitpick: Technically, 100 MHz is "Very high frequency"; HF is only

3-30 MHz, then VHF from 30-300 MHz, so I'd think either VHF construction techniques would apply, or broadband, which is even harder. =:-O

Cheers! Rich

When I read this: "We intend to supply useful high speed products and the level of support necessary for their successful application (such high minded community spirit is, of course, capitalism's deputy)."

I almost wept. :-)

Thanks! Rich

Only to radio operators. :-)

Check out Randy Rhea's book, "HF Filter Design," which details filters up to something pushing a gigahertz! :-)

What is the voltage of the power line? How much current is flowing? These will determine the size of wire that must be used for the power line.

At 100MHz, you can use an air core transformer if you don't need a wide bandwidth. You want the primary and secondary to be tightly coupled to each other. Making a twisted pair or using a coax will do this.

For air core, you need to calculate the inductance you need and then wind the coax or twisted to yeld that much. You won't want to resonate with anything less than about 22pF.