The AM Forum
March 28, 2024, 09:17:41 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: 1 [2]   Go Down
  Print  
Author Topic: Class E PAs - FET Evaluations at 40 Meters  (Read 30891 times)
0 Members and 1 Guest are viewing this topic.
Tom WA3KLR
Contributing
Member
*
Offline Offline

Posts: 2120



« Reply #25 on: April 18, 2007, 09:09:20 PM »

Some more peripheral data -
Here is the results of 10 simulations I did more than a year ago with a 75 Meter Class E single-ended PA matching 50 Ohms.  An ideal switch is performing for the FETs, with 0.1 Ohms series resistance.  Shooting for 1500 Watts out.  Each of the 10 simulations is with a different setting of the tank series capacitor. 

The graph is created from the DC input and r.f. output data of the simulations versus the tank capacitor setting.  The “properly tuned” setting is the 2637 pf. data points.

With the range of capacitor values used for the simulations, the drain efficiency is always rising as the tank capacitor is increasing in value.  At the same time, the output power is going downward.  Best efficiency when detuned to the high capacitance side.

Does anyone have such a plot on one of the real PA’s to compare?

* classe6_anal_eff.pdf (6.32 KB - downloaded 359 times.)
Logged

73 de Tom WA3KLR  AMI # 77   Amplitude Modulation - a force Now and for the Future!
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #26 on: April 18, 2007, 09:34:30 PM »

wow, Very interesting. 11 - 11n90s on a side puts me at about .1 ohm per phase.
my loading cap is right around 2000 pF and my efficiency is about 88 %.
Looks to me your graph confirms I need higher C2 and I know I need less L.
I wonder what your Drain Z was? I'm running a 1:2 transformar and the drain is sitting at around 48 volts for 1200 watts out.

I find my rig likes to be a bit on the high C side of peak.The waveform looks the best and allows for the shunt caps to drift lower as they heat up a bit.

Logged
Tom WA3KLR
Contributing
Member
*
Offline Offline

Posts: 2120



« Reply #27 on: April 21, 2011, 05:36:19 PM »

Thursday, April 21, 2011 Update

Three newer power FETs have been added to the old list of FETS used in the 40 meter drive evaluation.  They are:

ST STY60NM60 RDSon 0.05 Ohm,  Id 60 Amps, Qg = 178 nC.

Fairchild FCH35N60 RDSon 0.079 Ohm,  Id 35 Amps, Qg = 139 nC.

ST STW20NM60FD RDSon 0.26 Ohm,  Id 20 Amps, Qg = 37 nC.

These 3 additions are ranked solely by using the Rule-Of-Thumb formula I came up with 4 years ago which used lab data on 9 power FETs and the resulting formula of the gate drive efficiency data curve.  I have not acquired samples of these FETs and have not run them.  These 3 newer FETs theoretically are easier to drive than the older 9 FETs.

FET type     Lab Drive Efficiency    Data Sheet-R.O.T. formula

STY60NM60                                    82
FCH35N60                                      40
STW20NM60FD                               40

FQA18N50V2        33.2                   31.7
FQP18N50V2         30.3                   28.5
FDH15N50             26.6                   25.1
FCA20N60             23.3                   24.9
IXFH28N50F          21.6                  20.6
FCP11N60F           14.0                   15.6
STW20NK50Z       11.7                   11.8
IRFP450B              5.4                     5.6
FQA11N90            2.5                     2.3

Logged

73 de Tom WA3KLR  AMI # 77   Amplitude Modulation - a force Now and for the Future!
W1DAN
Contributing
Member
*
Offline Offline

Posts: 899



« Reply #28 on: April 23, 2011, 10:58:24 AM »

Tom:

Excellent work!

Thanks for sharing your findings.

I wonder if the efficiency number would change at all when you try push-pull?

When I built my RF deck it was single phase. Going push-pull seemed to run better. I did not do any numbers though.

Dan
Logged
Tom WA3KLR
Contributing
Member
*
Offline Offline

Posts: 2120



« Reply #29 on: November 18, 2011, 10:23:59 AM »

Friday, November 18, 2011 Update

The latest FET I spotted in September appears to be faster switching than previous FETs.  So this could be a real good choice for 40 meters.  It is made by IXYS and the data sheet I have (March 2011) is labeled “Advance Technical Information”.  It is available from both Digikey and Mouser for $11 each.  I have not ordered any.  

But the part is:

Polar3 HiPerFET  600V, 64 Amps, Rds(on) = 95 milliOhms, Qg(on) = 145 nC.

IXYS IXFK64N60P3  TO-264 package
and     IXFX64N60P3  PLUS247 package.

My “Rule-of-thumb ease of drive number” based on this part's data sheet calculates to 58, which means it is relatively easy to drive for it’s size compared to most other power FETs.

* 11/19 - Attached is the IXYS overview sheet on the design of the new P3 power FET family.

* Ixys PBNPOLARP3_final.pdf (573.87 KB - downloaded 277 times.)
Logged

73 de Tom WA3KLR  AMI # 77   Amplitude Modulation - a force Now and for the Future!
W4AMV
Member

Offline Offline

Posts: 672


« Reply #30 on: November 18, 2011, 10:41:35 AM »

Tom, wonderful work. For reference, I tried to find a schematic as a reference to your BB. Is that in another thread? Thanks! Alan
Logged
Tom WA3KLR
Contributing
Member
*
Offline Offline

Posts: 2120



« Reply #31 on: November 19, 2011, 03:45:47 PM »

Alan,

Attached is the schematic for the test bed (in Bob Pease style).  It was not posted before this; only described in words on page 1 of this thread, Reply #3.


* 40 m fet testbed1.jpg (729.58 KB, 3356x2516 - viewed 614 times.)
* 40m fet testbed1.pdf (324.6 KB - downloaded 252 times.)
Logged

73 de Tom WA3KLR  AMI # 77   Amplitude Modulation - a force Now and for the Future!
W4AMV
Member

Offline Offline

Posts: 672


« Reply #32 on: November 19, 2011, 05:00:50 PM »

Perfect and thanks!
Logged
G3UUR
Member

Offline Offline

Posts: 141


« Reply #33 on: November 22, 2011, 05:18:25 AM »


One interesting combination that jumps out at me is a pair of FETs with almost identical current and Rds(on) ratings; #3 and #8 in the efficiency rating.  #3 was 2 in the drive ranking and #8 was 8 in drive; same impedance but different in drive.  This is exactly the type of thing I was looking to identify.

#3 is the FDH15N50 and #8 is the IRFP450B. The FDH15N50 benefits from being several generations newer power FET design which is focusing on drive efficiency.


Tom,

Ciss and Crss are very different for these two devices. The IRFP450B has more than 1.5 times the Ciss of the FDH15N50 and the Crss is nearly 4 times as great. I'm not sure what your voltage gain on the PA is, but in any event I'd imagine the contribution of Crss to the overall input capacitance only amounts to some 20%. That pretty much explains the difference in drive requirements.

At the level of input capacitance of the devices you're experimenting with, I'd have thought that you'd have been getting into the region where the resistive part of the input impedance would start to have an effect. You're probably aware that tubes have an input resistance at VHF which is determined by cathode inductance, grid-cathode capacitance and gm. Well, MOSFETs have a similar impedance inversion and Ciss, gfs and the source inductance. Ls, do a similar thing. The value is much lower, sooner in MOSFETS as you go up through the frequency spectrum because of the larger gate-source capacitance.
It's probably not quite low enough to worry about at 7Mhz, but is worth considering if you decide the tune out the capacitance as siggested by AB1GX. I didn't have driver ICs when I was playing with switching MOSFET PAs in the late 1970s and early '80s, so I used to tune out the input capacitance with the drain choke of the driver stage and use a gate resistor of suitable value to keep the Q low. I could drive the MOSFETS of the time to 90% efficiency on 1.9 and 3.6MHz with that arrangement without too much trouble, and that was at 13.5V. Mind you, I was only working at 25 - 30W most of the time. My biggest PA was 400W and used push-pull IRF630 MOSFETS in linear mode, though that was later changed to 100W class D for AM. 

It's good to see someone is keeping up with all the new devices and thoroughly evaluating them. Have you tried any of them in Class F yet?

73,

Dave. 
Logged

Dave,G3UUR
Vintage AM from the East of England
Tom WA3KLR
Contributing
Member
*
Offline Offline

Posts: 2120



« Reply #34 on: November 25, 2011, 01:48:20 PM »

Hi Dave,

I can't say that I am keeping up with all of the new devices.  I manage to spot some of them.  

I don't see any reason to try Class F work.  I would like to study Class C with FETs since the drain voltage swing is lower and the load change tolerance is better.  

What I am wondering now is, since the switch-mode power supply FETS I tested 4 years ago did fairly well at 40 meters, is to do testing at 20 meters next.  Here, the square-wave driver i.c. will be critical.
Logged

73 de Tom WA3KLR  AMI # 77   Amplitude Modulation - a force Now and for the Future!
G3UUR
Member

Offline Offline

Posts: 141


« Reply #35 on: November 26, 2011, 06:13:27 AM »

Hi Tom,

Switching MOSFETs in class C have always worked fine for me, but since I always drive them hard of the gurus may claim they're really in class D. I've never used class E because I can achieve equally good efficiencies with much simpler broadband matching circuits and low-pass filters on 1.8 and 3.6MHz. I'd probably be tempted to try class E on 29MHz if I had trouble getting reasonable efficiency up there. The problem on 10m is the very low input impedance - it's barely better than a bipolar device up there, so the device needs to be tuned and matched properly.,
Even on 20m, I suspect the driver ICs will run out of steam unless you use smaller devices with lower Ciss. It'll be interesting to see how you get on. I'm sure you'll keep us posted and I'll look forward to seeing what you come up with.

Happy Holidays!

Dave.
Logged

Dave,G3UUR
Vintage AM from the East of England
Pages: 1 [2]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.079 seconds with 19 queries.