IRF510 Final Amplifier Stage

 

The IRF510 is almost like ladies "panty hose" ~ worn everywhere, easily accessed/removed and durable as hell. With but a few components and around 200 to 300 MW of drive it is easy to see 5 watts at the output of an IRF510 Amp. My most recent buys of "name quality devices" were around the $0.80 mark. So truly a lot of bang for your buck. You also will need a suitable heat sink -- just idling at 100 ma of bias and using a 13.8 VDC source is equal to 1.38 Watts of heat being dissipated--read hot! The "easy access part" becomes critical if you buy those unmarked Chinese specials sold for a quarter a piece! You will be changing a lot of IRF510's!!!!

Many questions have arisen regarding proper biasing -- if you are up to your neck in anal retentive things then shoot for around 100 ma with no signal applied. My process: I set the bias by listening to the signal (connected to a dummy load) on an outboard receiver . There is a sweet spot where the signal sounds very clear and there is no output on a scope or RF power meter with no modulation -- quite often that is awful close to 100 Ma. The difference is that 100 ma is like our generic panty hose and there typically is a bit of difference in the manufacturers. My approach is blind to the differences and sets a point where you get maxi um output consistent with a good quality signal. [At this point the uBitx and EMRFD illuminati may withdraw their slide rules from the K&E green cases and prove me wrong. Unless you are my age you probably have no clue about a K&E slide rule!]

 

There are several parts to the IRF510 Linear Amplifier Stage and include: the input circuit, the bias circuit the DC power filtering and the output matching circuit.

I will cover the bias circuit first as there are two elements here. Let us cover the first in that I supply power to the IRF510 at all times, since on transmit I do not want to hot switch the drain supply as the current draw would be large. I can see the looks of worry. Without any bias applied; but power connected to the Drain the current draw is miniscule. It is only when the bias is applied that we see significant current with an idle value near 100 ma. (Notice I said near!) So to actually turn on the linear we apply the bias.

Thus part one is that the bias application is like turning on the power switch. Part two is that the bias is adjustable and a few components take 12 VDC and make it look like 5 VDC. You can look at the bias specifications and you do not want to apply more than 5 VDC --or you will smoke the IRF510! Most circuits use a simple Zener regulator. The SPRAT Part II shows a Zener but I have removed the word Zener from my vocabulary and instead I used a 5 VDC three terminal regulator. The three terminal device is a bit more bullet proof and does a bit of power clean up and isolation from the 12 VDC rail.

Now a bit of tribal knowledge here. If you lift the center leg of the 78L05 regulator and insert a LED between that leg (which is normally ground) and connect the LED cathode end to ground --you have now raised the bias level above 6 VDC. So why would you do that as you just said don't apply more than 5 VDC. But you can take my IRF510 circuit and simply swap in one of those cool Mitsubishi RDHF series RF Power FET's and run it at the higher bias level. You will only get 6 watts out with no other circuit changes than the bias level BUT -- the RF out unlike the IRF510 does not fall off the cliff above 20M. The penalty --the RDHF runs hotter on idle so bigger heat sink needed and the higher bias level

Keep in mind I think modules and this IRF510 is used in many of my rigs. So I am always thinking general applicability versus a specific application.

There are but a few connection to the device itself but I do a lot of DC power filtering and that includes a type 43 sleeved ferrite core on the power lead going to the board. An extra added measure; but I have seen this as standard in many commercial rigs --so why not here.