Wednesday, January 28, 2015

MOSFET AMP IRF510, a little discussions-2

 A friend was discussing the IRF510 with the homebrew group.

"Any friend with Bitx linear original version success build are most welcome... I completed linear of bitx20 last day and found that the Transformer T1 at the drain of IRF510 getting very Hot just after switching the linear ON... core i used is FT37-43 with 2 turns bifiler my current bias is 50mA @ 19V DC.. any suggestions are most welcome..."

             Well I had made the first version linear by the VU2ESE. He sent his linears foto. (thanks friend)  I feel it is not  is not the first, version but the version 3.

             Important difference between the version1 and version3 are the linear amplifier. In the first version 2ESE  has tried the plastic former and balun core. Now a days foreign hams / Indian hams use the TOROIDs which comes along with KIT.

            During my quest with taming the linear ,VU2ESE had hinted me of using the tap washer reason it was not possible to find the suitable toroids/balun cores.... I gave it (tap washer) a try and it worked.... thanks Farhan.. The tapwashers many turns absorbs the capacitance of the FET. Which makes it more efficient... 2ESE has always multiple use for a part.

              At present My friend just want to using only two turns on the power amplifiers transformer. IMHO this is not right. There has to be more turns to increase the inductance.  May be the innocent looking balun core is a big culprit. The turns depends on the material, how much RF flux it can handle, operating frequency.   This RF amp fault  has burnt power supply IC and there is lots of frustration.... He will be happy if the linear gives him more than 2watt, enough to drive a big linear....

bitx ver.1--- linear
the top view of the linear.

          Well what I can say... Here is the actual linear ...Have a look. IMHO the culprit is two fold. big leads causing the self oscillations/vhf parasites. Core is not handling the power level........ he should try putting the negative feed back network (100k+0.1mfd) betweed drain and gate. At gate he should also put a 10 ohm resistor coming from the bias to the gate. Cut short the length of the gate. Try different heat sink. If ckt is on a plain copper plate, then try to make it sleep behind the pcb..... In the nut shell there is positive feed back.....

      Most probably his FET oscillated and gone QRT. There is no way to test it. Just replace the FET. make the Leads short and It should work. First try with 12v DC unregulated. with 1 amp fuse. That way can save the power supply and LM317 voltage regulator.

Here is my version of the above linear on veroboard. I have made two prototype with success.  Just keep the lead short whatever method/pcb you use....
VU3INJ, linear

 (Lets wait for his response......  This my own musing / hypothetical solutions I have omitted others comments-views and nothing to do with him or his skills personally.)

No comments:

Post a Comment