KF1Z 's new 75 meter class-E

Post your photos here and write-ups you may have

KF1Z 's new 75 meter class-E

Postby kf1z » Mon May 28, 2007 7:56 pm

Well, thought it was about time to share a few pics of my new endeavor.
Yes, still plenty more to do, but, it's making RF.

May even get on the air tomorrow... don't want to wear out the dummy load :D


Image
Image
Image
Image
User avatar
kf1z
Senior Contributer
 
Posts: 558
Joined: Sat Mar 18, 2006 7:07 am
Location: Ely, NV.

Postby steve_qix » Tue May 29, 2007 8:01 am

Looks good ! How well attached to the ground plane are those source busses? I have found it takes sevaral screws, placed VERY NEAR to the actual source connection to the source bus to ensure a really good connection, else parasitics result.

Also, check for heat generated in the RF bypass capacitors at the DC end of the RF output transformers.

But, the layout looks pretty good !

Regards,

Steve
Visit the class E web site at:

http://www.classeradio.com
User avatar
steve_qix
Class E Guru
 
Posts: 434
Joined: Fri Jan 31, 2003 11:03 am
Location: Townsend MA

Postby kf1z » Tue May 29, 2007 8:43 am

Yes, the source busses have 2 screws, both right at the source lead...

No heating so far int the dc bypass caps....

Very little warmth in the heat spreaders....

Still seem to be lacking a little RF output....
Maybe 20 watts less than expected..... but getting there....
User avatar
kf1z
Senior Contributer
 
Posts: 558
Joined: Sat Mar 18, 2006 7:07 am
Location: Ely, NV.

Postby steve_qix » Tue May 29, 2007 8:25 pm

Hi Bruce,

Run as much series C as possible (up to 500 or so pF). Reduce the coil to accomplish this. I've found efficiency goes up under most circumstances.

Regards,

Steve
Visit the class E web site at:

http://www.classeradio.com
User avatar
steve_qix
Class E Guru
 
Posts: 434
Joined: Fri Jan 31, 2003 11:03 am
Location: Townsend MA

Postby kf1z » Tue May 29, 2007 8:48 pm

You know, I tried that Steve,

The first try putting RF through this thing, I cut off one turn from the series inductor from the old rig.

Was about 4.6uH....

That really didn't seem to work out well....

Good waveforms and all that, but only about 63% efficiency!

Now, back with the 10 turn, 5.5 uH inductor, up to 88% or so.....

Hard to tell easily just how much series C I'm running, but it's close to 500pf at 3.600 mhz
(Little vacuum variable tuning cap)

Interesting though, the old rig wouldn't tune at all below about 3.775 or so.
But this one seems to play from 3.600 to 4.00 mhz...

Sure the efficiency varies a bit (better at 3.600mhz)....


Does seem like I have to run quite a bit of loading capacitance though....
Not sure about that.....

I'll learn about what I'm doing one of these days! :D
User avatar
kf1z
Senior Contributer
 
Posts: 558
Joined: Sat Mar 18, 2006 7:07 am
Location: Ely, NV.

Postby frank carcia » Wed May 30, 2007 8:24 pm

yup inductor looks big and if it works better at the low end of the band that is even a better clue that it is too big. I extend the kapton beyond the spreader a bit so there is a greater distance to flash around the the ends.
Have you measured the shunt C or just run the formula? the formula only works if both surfaces are perfectly flat. You have to lap them to come even close. might try a bit more shunt c to see if things improve say 250 pf each phase. How thick is your kapton? I do 3 mil.
frank carcia
Senior Contributer
 
Posts: 765
Joined: Fri Jan 31, 2003 10:48 am
Location: Enfield Ct. / Niantic Ct.

Postby kf1z » Wed May 30, 2007 8:57 pm

The inductor is 5.5uH, as Steve specified for the 300-500 watt amp. (with 2 modules of 5 FETs).....

I did extend the Kapton a bit around the edges... varies from about 1/16" to 1/8".

The shunt C is a measured value.....
The Kapton was supposed to have been 2 mil.... (my micrometer is "missing" so I couldn't verify though.

I lapped the copper to 1200 mesh... and the heatsink to 600.
(lapidary grade diamond laps and diamond based cloth backed 'paper' on a machine tool steel plate)


Well, I do WANT the rig to be able to operate down in the 3.600 mhz range, AND up to 3.890....

If anything now, maybe I can try lowerin the L a bit.... I tried to kee the leads between the output transformer, and the coil, and to the tuning cap VERY short.....
So not really able to just spread the turns out to test..... (#4 solid copper wire)

Well, I guess I could let the tuning cap float on the chassis to test, and if better results are obtained, remount it......

Maybe I'll try that....
Beats trying to remove 1/2 turn .....
User avatar
kf1z
Senior Contributer
 
Posts: 558
Joined: Sat Mar 18, 2006 7:07 am
Location: Ely, NV.

Postby frank carcia » Thu May 31, 2007 7:43 am

Sounds great on the shunt. Reducing L Just put a screw and nut with big washers between two turns and short a turn out to see what happens. This way you can go back easily if it degrades. (brass works the best.)
5.5 uh you must also include the leakage l of the transformer and strays due to wiring.
You say it is working better at the low end how about 3.5 MHz for another clue.??
frank carcia
Senior Contributer
 
Posts: 765
Joined: Fri Jan 31, 2003 10:48 am
Location: Enfield Ct. / Niantic Ct.

Postby kf1z » Tue Jun 12, 2007 6:41 pm

Another source of a mysterious problem seems to have been caused by the shunt caps themselves....

Didn't think about the copper expanding at a different rate than the aluminum.....

The copper blocks must have heated a bit, and shifted position...

I loosened the four screws each, and retightened, and was able to get the two modules to draw within about 1/4 amp of each other when everything is cold...

Every once in a while, I key up (first transmission...all cold) one module is drawing about 2 amps, and the other 4.5 to 5 amps.....

If I unkey, and key again... they start to equalize....

kind of irritating, because I'd like consistant results..... one way or the other.....
Not this working well one moment, and not the next silliness.... :)



I can't PROVE yet that the heat-spreader caps are the source of my possed rig, by I'm fairly sure.....

Makes sense though.... 1/2" thick copper probably increases in thickness by about 0.0005" with a temperature change of 40 degrees F.....
Not sure of that figure.... I'd have to look it up ....

Even if the heat-spreader was aluminum, there'd be some change in pressure between the heat-sink and spreader......



Well, at any rate.... not sure I'm going to keep this configuration....

I'm thinking I'll go back to using "store-bought" shunt caps, as soon as I can get my mits on some ATC ceramics......

Don't get me wrong..... when it's working....it works WELL ...

Nice clean drain waveform, and reasonable efficiency (as well as I can measure it anyhow..) Still seems like room for improvement, even if my scope needs adjustin'....


So, that's where I'm at with that.....

Just got ahold of a 650pf vacuuum variable I'm going to put in place of the 500pf that's in there....
Mostly 'cause when I get near the 500pf limit of this one.... it get flaky...


Down to about 4.6uH now on the series inductor, by the way..... from 5.5uH....
Seems to be working A-ok....
User avatar
kf1z
Senior Contributer
 
Posts: 558
Joined: Sat Mar 18, 2006 7:07 am
Location: Ely, NV.

Postby frank carcia » Tue Jun 12, 2007 9:10 pm

I,ve been running heat spreader shunt caps since '96 with no problems. copper on aluminum 3 mil kapton.
Current imbalance could be due to drive not timed correctly???
frank carcia
Senior Contributer
 
Posts: 765
Joined: Fri Jan 31, 2003 10:48 am
Location: Enfield Ct. / Niantic Ct.

Postby kf1z » Tue Jun 12, 2007 9:57 pm

Well, the drive is just out of a J-K flipflop same drive as my previous rig...

But, it isn't stable... probably the way I mounted the heat-spreaders...

Once it's on, and warmed a bit.... no problems at all....

Usually any problems crop up the first time keying up..... then all is fine..


Had an imbalance that wouldn't go away , one module at about 75% of the current of the other, (along with ringing on ONE module...the 'low' one)

Since I retightened the heat-spreaders, it's been acting much better..


Not sure.... but I suspect the mounting... though I really don't know why it's a problem, unless not tight enough...

As you can see in the picture above.....

4 - #8 screws in each, with nylon shouldered bushings in each hole for insulation.


I've check, and rechecked ALL connections etc....
And can't find any (mechanical) reason for an "intermittent" problem like this...
User avatar
kf1z
Senior Contributer
 
Posts: 558
Joined: Sat Mar 18, 2006 7:07 am
Location: Ely, NV.

Postby frank carcia » Wed Jun 13, 2007 7:17 am

I use nylon step washers on my 160 meter rig so I can't be very tight. The 14 FET screws hold the spreader to the heat sink with no additional hardware. It has never been a problem. I would look at both phases of your drive at the same time to make sure there isn't a time when both phases are on. I prevent this by using a driver transformer. Other guys use rc time constants to insure no cross over. Your layout looks good so I don't think it is the final.
The problem with FETs is they turn off slower than they turn on so you actually need a little less than 50% drive. I have not looked at the driver chips yet because my stash is still in the bag. fc
frank carcia
Senior Contributer
 
Posts: 765
Joined: Fri Jan 31, 2003 10:48 am
Location: Enfield Ct. / Niantic Ct.

whatadummy

Postby kf1z » Wed Jun 13, 2007 9:27 am

Well, I wish I could say this was a first for me...... but it aint.....

I had hooked up the wrong module to the wrong meter....
You know, the left module was connected to the right meter....

no big deal, just had to remember which was whhich....

Today I decide to switch them around....
Took all of 3 minute.....


Fired the rig up..... and Lo and Behold..... the SAME module was reading low in current........
Hmmm.....

Problem is in the METERING.... :evil:

One of the ammeters is INTERMITTENT!!! an internal connection......


Wouldn't you know........ :roll:

May not have as many bugs as I thought.......

I'll change out the meter. and report backk


whatadummy.....
User avatar
kf1z
Senior Contributer
 
Posts: 558
Joined: Sat Mar 18, 2006 7:07 am
Location: Ely, NV.

Postby frank carcia » Wed Jun 13, 2007 9:33 am

It is the simple problems that humble you.
It would be great if it is only the meter. Imaging how much circuit hacking you could have done if they weren't backwards.
It looks like your transformer is still 1:1 so don't know how you will need to modify the tank to work into that low Z source. My 22 FET is almost 500 pf series at 1 KW and about 1800 pf load with 1:2. The Tank coil wants to be around 6 uh with my wiring. fc
frank carcia
Senior Contributer
 
Posts: 765
Joined: Fri Jan 31, 2003 10:48 am
Location: Enfield Ct. / Niantic Ct.

Postby kf1z » Wed Jun 13, 2007 10:01 am

yessiree....

things are looking up..........VERY small difference in curerent between modules now...

I hate intermittents!!!!!


at any rate.....

What will making my transformer ratio 1:2 insted of 1:1 do for me?

Just change the amount of loading capacitance?....

I know it will increase the voltage, and thereby increase the impedance...


But would it change efficiency any?




Right now I've got 4.6uH and about (I'd have to disconnect the Series cap to measure it) 450pf....
And about 2700pf loading...

even if I push the limit, and tune so I JUST see ringing appear, and the sharp cut-off where the fets turn on again , I'm seeing about 250-260 watts rf out at 320watts dc input........


I'm only concerned with the efficiency, because I KNOW it should be better!
User avatar
kf1z
Senior Contributer
 
Posts: 558
Joined: Sat Mar 18, 2006 7:07 am
Location: Ely, NV.

Next

Return to Your Photos and Articles

Who is online

Users browsing this forum: No registered users and 1 guest

cron