Print

[TANAKASAN]

I'll have another go at explaining this.

One of the most likely explanations for your instability is the layout of your amplifier, in particular the connections to the FETs. The connections need to be VERY short, certainly below four or five cm, and the last thing before each FET should be the resistor. If you look at the pictures of the prototype amplifier you will see all the FETs in a line connected by big thick copper followed by a short wire to the PCB, this is a better approach because it means lower inductance in the connections.

A FET has a very high gate capacitance. This means that to drive them you need low impedance low inductance connections.

Tanakasan

[KF6QEX]

add an extra 4,7ohm resistor in series with the existing 12ohm resistor very close to gate?

Yes. An extra 4.7 Ohm resistor.
You'll cut the trace and the gate lead as close to the mosftet as possible, and insert the resistor.

If you don't have a chip resistor, cut the leads of the resistor and solder it as close to the gate as possible.

The objective is not so much to increase the total resistance as it is to present a resistance as close to the gate as possible.

As for the wiring layout, it might even make a difference if you try to shorten the existing wires and maybe cutting the board holding the mosfets shorter.
 

 

[W8JI]

Do not change parts values just yet. There is no need to add any resistors.

First understand anyone can have good luck and make one of something work, even when the next one might not work.  Because one amplifier with a poor layout works, it does not mean the next one with the same layout will work.

You should make the leads very short to the FET's. You also should have a big ground plane. It is very important you do this, especially the path length from the ground at the sources of the FETS back to C1 and C14 ground, and C5 ground. These capacitors and R2 and R3 gounds must be to a large groundplane with short connections.

What resistance do you have for R2 and R3? Are they 0.18 ohms or 180 milliohms times 6 in parallel??

On the gates, you have R4 and R5 that are 6x   12-ohm resistors. That 2 ohms is the same purpose as the suggested 4.7 ohm.  If you wanted to increase resistance you would remove some resistors and the value would increase. I would not do that until you have better connections.

All of the FET leads and connections to transformers need to be short.

R10 and R11 are feedback resistors, make sure the phase is correct with T3.

The bias circuit looks OK for FET's.

[KF6QEX]

R4 and R5 that are 6x   12-ohm resistors.

There are 12 mosfets,  push-pull in two groups of six
So really there is a 12 Ohm resistor in front of each gate ..well...almost an inch in front

 

[TANAKASAN]

Folks, can we stop for a moment and have a look at the schematic. There are some safety issues here which are being ignored.

For the bias supply he takes 110V DC and drops it down to 12v using a single 6.8K resistor and a 12v 1W zener diode. This is then fed through a 1A 5V regulator. Assuming 200mA bias that resistor is dissipating 20W and if the zener fails the regulator will receive a 110V input.

This is NOT a safe design.

Tanakasan

[1431]

i sort all connections and i think the problem with oscillation stop. I adjust the current at 200mA without driving power, and with 3w input driving power with 106vdc power supply the output was 200w. The xurrent was 6A. But the manufacture claims that with 110v and 2w driving power the linear gives 500w output R.F power. do you have to propose any change to give the linear more output power? Maybe in R.F transformers? (to change the number of turns?). Thank you very much

new photos

http://tzitzikas.webs.com/fet20.JPG
http://tzitzikas.webs.com/linear20.JPG
http://tzitzikas.webs.com/linear21.JPG
http://tzitzikas.webs.com/linear22.JPG
http://tzitzikas.webs.com/fet21.JPG
http://tzitzikas.webs.com/pompos20.JPG

[TANAKASAN]

Changing the number of RF transformer turns will not increase the output power, just change the impedance of the output. The only ways to increase the power is to install more FETs on each side and/or increase the power supply voltage.

Isn't 500W enough?

Tanakasan

[1431]

500w is enought! But the manufacture (prototype) and the other radio amateur who construct this linear, claims that their linear gives 500w r.f output power at 110vdc voltage (13A current) and with only 2w driving power. But the problem is that my linear gives only 200w r.f output power at 106vdc voltage (6A current)with 3w driving power. I can't understand what is the problem and my linear gives only 200w. All connections are checked ok. But the modulation is wonderful.

[1431]

W8JI wrote <<R10 and R11 are feedback resistors, make sure the phase is correct with T3.>>

What do you mean if the phase is correct with T3?
The T3 constructed by me by using the prototupe photo:
( http://tzitzikas.webs.com/T3.jpg ).
Do you suggest to invert the 2 connections at the first coil (1 turn coil)??

[W8JI]

W8JI wrote <<R10 and R11 are feedback resistors, make sure the phase is correct with T3.>>

What do you mean if the phase is correct with T3?
The T3 constructed by me by using the prototupe photo:
( http://tzitzikas.webs.com/T3.jpg ).
Do you suggest to invert the 2 connections at the first coil (1 turn coil)??

I take it this is for a pirate radio AM transmitter?

[TANAKASAN]

"I take it this is for a pirate radio AM transmitter?"

This is why I asked which band(s) the linear was intended to operate on. Take comfort Tom, with that design it will be Darwin in action pretty soon now.

Tanakasan

[1431]

Hi. the mosfet that i am using (IRFP360) are no name. Do you believe that the linear gives less power than the prototype because the mosfet are no name instead of the prototype which is using International Rectifier mosfet?

[KM3F]

Why is anyone trying to defend a pile of parts that are not to even resonable engineering standards?
Just because one person may have gotten away with it does not excuse the issue.
The builder of this does not understand circuits and basic RF theory and as such gets into a problem  by doing such foolishness at RF frequency.
Would you cut up a cat, it dies, then ask a vet what you did wrong?
I understand these days, the country this is comming from is not in very good shape.
Everthing is going out of control!
Do we need to add to it?
Some of these people come here and expect to do the same things they did in there home country and create problems.
Then wonder why they feel they are picked on.
Sorry for the caustic attitude but this is all I can see out of this thread after some of the best have offered comment and it meets a brick wall.

[KF6QEX]

Why is anyone trying to defend a pile of parts that are not to even resonable engineering standards?
Just because one person may have gotten away with it does not excuse the issue.
The builder of this does not understand circuits and basic RF theory and as such gets into a problem  by doing such foolishness at RF frequency.
Would you cut up a cat, it dies, then ask a vet what you did wrong?
I understand these days, the country this is comming from is not in very good shape.
Everything is going out of control!
Do we need to add to it?
Some of these people come here and expect to do the same things they did in there home country and create problems.
Then wonder why they feel they are picked on.
Sorry for the caustic attitude but this is all I can see out of this thread after some of the best have offered comment and it meets a brick wall.

I'm quoting this here so that I have it to use it as a future response on other topics
For the record I disagree with all of it.

[1431]

i sorted all connections.
When i gave 3watts of driving r.f power, it gave to output only 190w at 106VDC (6A current). Two radio amateurs who have construct this linear claim tha it gives 500w
r.f power at 110vdc.
But when i tried to give 4watts of driving r.f power the ferrite Cores (43 material)
of transformer T3 broken! Which do you think is the problem??

some pictures
http://tzitzikas.webs.com/linear20a.JPG
http://tzitzikas.webs.com/ferrite_broken1.JPG
http://tzitzikas.webs.com/ferrite_broken2.JPG

the feerite cores that i use are:
http://uk.farnell.com/fair-rite/2643626402/ferritecore-196ohm/dp/1463423
from manufacter:
http://www.fair-rite.com/cgibin/catalog.pgm?THEAPPL=Suppression+Components&THEWHERE=Cable+Component&THEPART=Round+Cable+EMI+Suppression+Cores#select:freq1
Round Cable EMI Suppression Cores
Broadband Frequencies 25-300
MHz (43 material)
some radio amateur told me that maybe this cores are improper for these frequences. But the designer of this linear
claims that we must use 43 material cores for the transformers.
Thank you.