MC3361 Mixer IC

What can I say, I like to trawl through parts catalogues and find somewhat obscure parts or ultra cheep parts to play with. So, when i found the MC3361, which is an FM Cordless phone IC, I thought why not. It costs 25 cents in one offs, and has some conversion gain and I assume its a gilbert cell in there, so its way cheaper than an SA612 and others.

LO in is on pin one, you can also add an xtal here for the LO. RF in is on pin 16 and the IF out is on pin 3. Now i did not try and match the impedance’s which are around 1800 ohms, i just fed in 50 ohms from my function generator into the LO and RF ports and then attached the spectrum analyzer to the IF port.

Other than some DC blocking caps and a 50 ohm termination resistor on the input there is not much to add to this. And as you can see from the plot below from the spectrum analyzer its working doubley balanced. RF in is suppressed and the sum and difference are the dominate signals. RF in was 7Mhz and the LO was 1Mhz for the sum and difference of 6 and 8mhz IF frequencies.

Is it as good as other mixers? I don’t know, but it works and its cheap, so i think its worth a go. Oh and it also has an FM de-modulator, AF amp and a op-amp for filtering in there also. I am yet to play with the other bits, i just setup the board for mixing only and well, mixing works.

Facebooktwitterredditpinterestlinkedin

Filter Capacitors and Voltage Rating

As i explained in a previous post, I am building an amp that is going to be capable of doing more than 100w. No biggy for most people but when everything you have built previously has been qrp there becomes a whole new set of challenges. With qrp you can just she will be right mate and do what ever you like “Almost” and it will work to some degree. So with this in mind, i set forth on working out what voltage rating i would need for the capacitors in the low pass filter.

What i really wanted to know, was the WHY behind the decisions others had made in their design choices and I started to ask others WHY as well. Often the reply was just use Mica Caps. Did not matter what power level, the answer some gave was just use 500v Mica and you will be right. No explanation, just this is what I do.

So that just never sat well with me, the answer cannot always be use $60 worth of 500v caps in your low pass filter. It took a while but i finally came across some information that seems credible and more importantly reliable and my gut feeling was right, Mica while good, might often be over kill, but still there are some caveats that are important to understand.

Firstly the voltage that a capacitor in a LPF sees is the peak to peak voltage of the AC wave form. Might be obvious to some, but its not information that is easy to find on the internet. So knowing that we can then use the power formula to find out what voltage might be present in the filter. Power(watts) = Voltage^2 / Load(Ohms) where in this case, the voltage of the signal is RMS Volts and once you have solved for it, you need to convert RMS to Peak to Peak by multiplying by 1.4.

So in my case 100w = 70.7vrms/50ohms :: 70.7*1.4 = 100v peak to peak. Now obviously there can always be an impedance mismatch between the low pass filter and the antenna and with that a change in the signal voltage. In my case i also looked at 100w at 25ohms and 100ohms load to get a better picture of what might be expected of the capacitors in the low pass filter. Being that I am actually designing the amp deck to be capable of 400w even though i am only running it at 100w, it means the capacitors need to be able to withstand a peak voltage of about 350V.

So onto the next point, I wont be using Mica Caps, for the very simple reason that $50 worth of caps is a lot of money to throw at something that might be a dismal failure. I will be using SMD ceramics and here is something i learned just recently, ceramic caps derate in capacitance the closer you get to the maximum voltage rating. According to a TI white paper i was reading, you should probbaly derate ceramics in filters by at least 1/3. So I need at least 360V rating, that means using 1Kv or 2Kv ceramic caps should be perfectly fine. The good thing is, Kv rated SMD caps are cheap, in the sub 50 cent range in 1 of qualities on mouser and they are NP0 and 5% tollerance. So i have solved this problem now and can move onto other things like output transformer and core sizes i need to use for the power level.

More fun.

Facebooktwitterredditpinterestlinkedin

Volt and Ammeter

Volt and Ammeter

So i discovered something last week or so ago while building power amplifiers. Its a total pain the the butt to be pulling everything apart to wire in the DMM to measure current when setting the bias for fet devices. Also that the meters in my Chinese power supplies are just about useless for anything less than an amp.

So with that in mind, i grabbed out a couple of analog panel meters that I have had sitting here for just about for forever, drew up some artwork in my favorite 2d design package and then went out into the garage to cut and bend some sheet aluminum into something that resembles a case.

Yeah the case is not my best work, but its perfectly serviceable and does the job that its meant to do. IE it houses and protects the meters. So now i can plug this into the power supply, plug the device under test into this and monitor both current and voltage while keeping my DMM free to be used for other jobs.

Yeah I know that you can use 2DMM and achieve the same result, but why spend a lot on an expensive DMM that is going to sit doing nothing most of the time, when for about $10 you can make a bit of test gear that will be good enough.

73, Rob.

Facebooktwitterredditpinterestlinkedin