• You can now help support WorldwideDX when you shop on Amazon at no additional cost to you! Simply follow this Shop on Amazon link first and a portion of any purchase is sent to WorldwideDX to help with site costs.

DIY Mosfet amp questions

roggom

Member
Jul 14, 2011
6
0
11
Hi there, I picked up an RM Italy KL-40 and was amazed at its simplicity.

Quick into: I have been in the electronics field for over 20 years and work in medical instrumentation/ xray. I build SS and tube audio amplifiers from scratch, as I find it fun and intriguing. I understand that building things sometimes cost more than new, but I like building. I am new to RF so just looking for some insight.

Anyway, I took this amp design http://www.rmitaly.com/download/manuals/KL40-manual_rel_400.pdf and hand etched a pcb and built it.
Instead of the propitiatory FET, I used a TO-3 IRF-350. I figured the input impedance was similar to the IRF 520; which has been used in some Linear amps. Well it did not pass the carrier signal. So I scrapped it and rebuilt it on a proto board. Same thing, idle current at about 8mA, but no amplification.

Observations:

Inductors:
I crossed the metric magnet wire to an available 22 guage. I was only getting about 6volts peak (clean signal) at the gate Well, when I added 4 turns of 16 awg in parallel with the input coil (L1) I got 9 volts peak (distorted) and 200mA of conduction, but only about 700mV of gain. I have some 20 awg magnet wire on order.

Fet: If I increase the bias voltage I can make the fet conduct more. I added a voltage divider and increased idle to 12ma and have 700mA of draw when I key up. Still not transferring any gain to the signal. I tried a IRF520 in parallel but it only conducts (generates heat) but no signal gain. I have some MS1307's on order.

Also after adding the addidtional coil to the input of the fet I have a distorted signal.

So, have any of you made a 40w linear amp?

Did you use a BJT or FET?

Can someone explain what causes the signal to distort?

Is there an Ft for fets? What is the typical bandwidth of a power fet?

Lastly, I work with 27 mHz, I am using a 50 ohm dummy, and not transmitting. I am only doing this as an experiment so when I pass my tech test I can build other amps for use.


The first pic is the amp, the larger signal scope pic is at the gate of the fet, the smaller is the output.
 

Attachments

  • DSC03982 (640x427).jpg
    DSC03982 (640x427).jpg
    183.8 KB · Views: 1,166
  • DSC03990 (640x427).jpg
    DSC03990 (640x427).jpg
    203.9 KB · Views: 309
  • DSC03991 (640x427).jpg
    DSC03991 (640x427).jpg
    202.3 KB · Views: 335

im no amp builder but capacitance and lead inductances have a large impact on power bandwidth/ suitability of cheap switching fets for rf amps,
looking at the datasheet your irf350 seems to have too much input capacitance for 27mhz, even the irf520 switching fet is a little high but they do work in the right circuit,
i know from experience the irf520's don't work so well in the rm amps even when bias is increased from factory settings neither do the fairchild fets claimed on forums to be the same as ms1307's on a curve tracer, i think i saw about 20w from 4x irf520's in an rm amp, i took precautions against destroying the fets with static,

when i put the ms1307's back in it worked fine.
 
Thanks for the feedback, I am sure once I get the 1307's in there it will start working. Great to hear practical application info, all of the threads I have found out there did not share any results.
 
cool

very cool good luck with it.
and good luck with your ham test when you take it.
 
Waiting for the MS1307's, I decided to make a ball pen rf probe, the design is available all over the net. Armed with my new probe and an old school VTVM I started probing my amp. I found that the output coil was choking my signal. I also found a cold solder joint. The cool thing is that I am now conducting, the only thing is that I am only getting about 1 Vp of gain, using the rf probe the input is still a bit less than the input.

I will do some more tweaking when the other parts come in.

Some more observations:

I replaced my T3 coil ( 6T of 22 awg, with 4T of 16 awg) and now pass the signal

The FET is conducting and drawing 600mA

Unlike the original amp I can slam the input with 6w DK and 18 swing with no issue, with that I get about 1200mA of current draw and about 3Vp of gain over the input.


So the good thing is that it does conduct, just have to figure a way to increase the gain.
 
Making Progress

Well the MS1307 came in and were a bust. I am usually skeptical about eBay fakes, but I did not want to spend too much. They may work in another circuit design, but not this one.

The good news is that I did make progress.

The IRF likes voltage, I experimented with 24vdc and finally settled with a 18 vdc 3A ps (old laptop ps)

The IRF350 can take a beating and is rated for 9A 150w, so I added an additional bias potentiometer to increase gate bias. The only issue was with the current diode setup, the idle was at 250ma.

Also the current rx loop causes a -db on the recieve.

So I changed the TX/RX to relay actuated, using a common relay, 2N2222A, 22uF, and 100p circuit. I removed the diode and L2 rx loop.



Now I have 15mA of standby idle and draw about 1.7A with a 1w DK from my Uniden 810 and 2.4A from my DX-959, also no more rx attenuation.

The uniden with a 1w DK was swinging to 3w pep

The dx-959 with a 12 w DK was swinging to 44w pep

Not earth shattering, but out of one $5 fet certainly has potential.
 

Attachments

  • DSC04020 (640x480).jpg
    DSC04020 (640x480).jpg
    202.7 KB · Views: 237
  • DSC04021 (640x480).jpg
    DSC04021 (640x480).jpg
    227.7 KB · Views: 275
A quick pic of the amp, the case acts as a nice heatsink, although a TO-3 heatsink would be a little more efficient. Not taking it apart any more.
 

Attachments

  • DSC04025 (640x348).jpg
    DSC04025 (640x348).jpg
    143.2 KB · Views: 330

dxChat
Help Users
  • No one is chatting at the moment.
  • dxBot:
    Tucker442 has left the room.
  • @ BJ radionut:
    LIVE 10:00 AM EST :cool:
  • @ Charles Edwards:
    I'm looking for factory settings 1 through 59 for a AT 5555 n2 or AT500 M2 I only wrote down half the values feel like a idiot I need help will be appreciated