99v IRF520 Low watts

[Spike5221]

After the irf520 per (cb tricks) with low carrier 1-2 watts, only swing to 10-13 pep when turned up to 8 watts it will swing up around 40-45 why am I not getting full Modulation swing from a low Carrier?

 

[LeapFrog]

why am I not getting full Modulation swing from a low Carrier?

I don't want to be offensive so please don't be offended.. but...
In a properly configured radio you aren't supposed to see full swing output from 1-2 Watts, that is a separate modification and not one I recommend unless you have a lot of time on your hands and an oscilloscope.

"But, LeapFrog, it's as easy as adding a diode and changing a resistor" Yeah it sounds real easy in theory, but putting it into practice is a whole 'nother story.

Not too many CB Techs that I've seen online do that mod, and even fewer get it to be somewhat clean when it's done.
Just my 2 Cents.

I can tell you from my experience with some EKL modifications, the bias voltage was low, so I had to make a custom "companion component" or use a dedicated bias voltage rail to get it where I wanted. Then the power came up.

I've never performed an EKL mosfet conversion on a Galaxy EPT3600xx, so I couldn't tell you if the bias is adjustable after the mod, or if it is fixed.
But yeah if you have proper bias at the gate of the Fet, and it isn't doing what you want then it sounds like you are asking about a "swing mod".

Edit: Just to add, the first "export radio" I ever owned was a Galaxy 66v and it had a 1n4001 rectifier diode for the "swing mod", I splattered into my TV so hard, the flat screen was giving me talk-back.
After I removed the non-factory stuff, and re-soldered the clipped resistor the radio no longer interfered with my TV.

 

[Spike5221]

the meter swings up and seems like it hits a wall, doubled checked everything is right all parts are exactly what was called for ,the little cap on soldered side removed ,872 regulator new, other 2 small 1869 regs. new, is it just inferior mosfets? irf520n

 

[LeapFrog]

the meter swings up and seems like it hits a wall, doubled checked everything is right all parts are exactly what was called for ,the little cap on soldered side removed ,872 regulator new, other 2 small 1869 regs. new, is it just inferior mosfets? irf520n

I don't believe so.. I think the radio is doing what is supposed to do, as you increase the carrier the peak output will come up with the carrier, as long as the limiter/AMC is intact, that is the "brick wall" you are seeing, the AMC is attempting to reign in the radio so it doesn't over-modulate and destroy the adjacent channels with splatter.

If you want to be heard on 3 channels (interfering with the adjacents) then by all means disable the AMC for more "swing" but just know it is wasted power.
Disclaimer: I'm not advocating for anyone to break any regulatory rules, always follow the regulations of your location.

The IRF520n is a rugged part, but I like the FQP13n10 personally.

Sorry if I came across as offensive, just calling it as I see it.

 

[Spike5221]

I don't want to be offensive so please don't be offended.. but...
In a properly configured radio you aren't supposed to see full swing output from 1-2 Watts, that is a separate modification and not one I recommend unless you have a lot of time on your hands and an oscilloscope.

"But, LeapFrog, it's as easy as adding a diode and changing a resistor" Yeah it sounds real easy in theory, but putting it into practice is a whole 'nother story.

Not too many CB Techs that I've seen online do that mod, and even fewer get it to be somewhat clean when it's done.
Just my 2 Cents.

I can tell you from my experience with some EKL modifications, the bias voltage was low, so I had to make a custom "companion component" or use a dedicated bias voltage rail to get it where I wanted. Then the power came up.

I've never performed an EKL mosfet conversion on a Galaxy EPT3600xx, so I couldn't tell you if the bias is adjustable after the mod, or if it is fixed.
But yeah if you have proper bias at the gate of the Fet, and it isn't doing what you want then it sounds like you are asking about a "swing mod".

Edit: Just to add, the first "export radio" I ever owned was a Galaxy 66v and it had a 1n4001 rectifier diode for the "swing mod", I splattered into my TV so hard, the flat screen was giving me talk-back.
After I removed the non-factory stuff, the radio no longer interfered with my TV.

the plans call for R249 removed so wouldn't that make it swing somewhat? I even tried the NPC volting and no difference still the same just like it swings up fast then hits a wall and stops, the bias is set @ 3.6 driver, 3.5 finals

 

[Spike5221]

the NPC actually cleaned the audio up good reports clean sounding

 

[LeapFrog]

That's interesting because I don't see R249 mentioned anywhere in the EKL Fet papers.

Try this on your test bench, of course use a dummy load:
re-install R249 (Limiter) or don't,
& add a 1N4148 diode in series @ R238, banded end towards PC board (left side if I remember correctly).
See what that does for the "swing".
Again i'm not advocating this mod, but you want something to try, then go ahead.

You could also play with the value of R238, increasing it.
Or even install a 1N4148 and 100 Ohm (play with the value try 220 Ohm)
in series with each other placed in parallel at C-189 (I forget which way they go in).

I don't recommend this mod, but you asked..
I don't know but to me 3.5 Volts seems a little low, i'm not largely into SSB radios, & I don't own any that have a MosFet driver... (even though 3.6 V is "close", I seem to remember running 3.7 w/ the 520)


Have you performed a dual tone linearity test for SSB?

If your RF section is under-biased, I suspect SSB will not look correct on the 'scope.

Good luck and 73.
-LeapFrog in the Sand Pile

 

[Spike5221]

That's interesting because I don't see R249 mentioned anywhere in the EKL Fet papers.

Try this on your test bench, of course use a dummy load:
re-install R249 (Limiter)
And instead add a 1N4148 diode in series @ R238, banded end towards PC board (left side if I remember correctly).
See what that does for the "swing".
Again i'm not advocating this mod, but you want something to try, then go ahead.

You could also play with the value of R238, increasing it.
Or even install a 1N4148 and 100 Ohm (play with the value try 220 Ohm)
in series with each other placed in parallel at C-189 (I forget which way they go in).

I don't recommend this mod, but you asked..
I don't know but to me 3.5 Volts seems a little low, i'm not largely into SSB radios, & I don't own any that have a MosFet driver...


Have you performed a dual tone linearity test for SSB?

If your RF section is under-biased, I don't suspect SSB will look correct on the 'scope.

I'm sorry the 249 was for the NPC. which also calls for 100ohm in series with a 1n4001 at c189, I'll try messing around with it..Thanks leapfrog

 

[LeapFrog]

Hey no worries brother.

I just want to leave you with one more tid-bit, I feel it is my duty, because I don't know when or where this radio will be used..


From the comment section, by OP of video:

The technician cannot do anything about it. The resulting signal I discussed occurs inside the mixer, and the technician can't take the integrated circuit apart. As I said in the video, the resulting signal occurs at a frequency that is in the designed pass band of the radio. The radio amplifies it because it occurs in the same frequency range it was designed for. The technician could possible attenuate the effect a little by retuning the radio so that it amplifies poorly in the 10M band, but it would still be there a bit. If that is what the technician wants to do, then they need to at least know about the problem and that is where this video comes in

The only good thing about this is that the offending signal is down 50 dB relative to the 27 MHz carrier, or at least it is on the one I measured. If no amplifier is used to make this unwanted signal worse, it is relatively a minor signal.

This inherent design flaw is why my EPT3600xx radios sit on the shelf unused.
To emphasize, if you run "barefoot" it shouldn't be too big of an issue, but add a little helping hand to the mix and your station can stick out into 10 meters amateur band plan, I know a few guys that love to Fox-Hunt signals, so I'm not sticking my neck out.

 

[LeapFrog]

You could try upping the bias of the driver to 3.6 V and the bias of the final to 3.7 V, see if the power comes up, or if the final section just gets hotter!

If anything pops, i'm not responsible.
Experiment at your own risk, I assume no responsibility for what does or doesn't happen.


The last NPC-RC mod I did was almost two years ago, and it was on an early Connex 33.
If I remember correctly it had a MosFet final upgrade, but the driver Tr was still a factory BJT.

I added components to the driver section, mimicking the SSB radios layout, to bias the driver to function more linear, and when the radio was done it would modulate to 135% on the 'scope, cleanly, and it didn't bleed channels. A local bought it from me but was disappointed when it wouldn't "swing" the way he wanted it to, I seem to remember telling him to examine the driver section if he wanted it to "swing" differently.

 

[Tallman]

The VGth is between 2 and 4 Vdc with 3 Vdc being listed as typical.

 

[LeapFrog]

I don't know who your supplier is but if the 520's are knock-offs or imitations that could play huge role here.

http://www.vishay.com/docs/91017/91017.pdf

 

[Handy Andy]

There are many many ways to make something work...

​

 

[LeapFrog]

Reminds me of the factory carrier control circuit Uniden employed in several radios, minus the capacitor obviously.
D63, R194, R228 almost look like the junction in your diagram.

Nevermind the comments placed in the B.O.M., that is for something I'm working on involving a Darlington configuration.
I could inject my own commentary here, but without any data to add, I will not.

 

[Handy Andy]

Most people that do the "NPC-RC" mod (Reduced Carrier) Forget some very important points...

In your design? I've drawn that in to help you. (I'll keep it thumbnails until you wish to go BIG on this) Look for RC and RA - as well as a Diode - it can be a 1N60 for the sake of symmetry...

​

IN another thread in a forum not too far away - we talked about Symmetry and balanced carrier - in your design you just removed a key component - a feedback path - that helps keep the symmetry.

When some do that NPC mod - they REMOVE (which they shouldn't) the 560 ohm resistor - in doing so you lose 1/2 your ability to retrieve and balance the symmetry in an otherwise 100% modulated (or less) envelope - simply due to the regulator side is now relying on the 8VDC constant to replenish the supply the diode blocks the line from returning to the Gain Cell.

The above posts' variants allow for that Gain Cell to recoup lost energy faster to help maintain symmetry when negative peaks are not present.

I'll let you have the ball back - good job!