Modulator question

Anything goes here, but abusive posts will be deleted.

Modulator question

Postby K9ACT » Sat Aug 07, 2010 12:31 pm

I am having trouble finding specific information on designing the modulator output filter.

What parameters do I need to know to determine the values of the filter components.

All I know at the moment is that the Class D RF deck puts out 55W with 24V and 2.5A.

For the sake of this discussion let's assume this to be the HV supply voltage. It will actually be 48V but I don't know the current until I get the supply.

In tube talk that would be about 10 Ohms load impedance but I see refs to modulator carrier level and I am not sure just what that means.

I would like to make air core coils for experimental reasons but I have no idea what to plug into the Butterworth calculators to get the values for L's and C's.

Jack
K9ACT
Class E Enthusiast
 
Posts: 146
Joined: Sun Jun 27, 2010 12:03 am

Re: Modulator question

Postby kf1z » Sat Aug 07, 2010 3:35 pm

Jack,

If your deck is going to run at 24volts, at 2.5 amps...

Then your filter design impedance is 9.6 ohms.

you want the "corner frequency " to be around 15khz.


So for a 4 element filter... (Butterworth T )

L1 = 77.96uH
C1= 2.04uF
L2= 188.21 uH
C2= 0.85uF



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

Re: Modulator question

Postby steve_qix » Sat Aug 07, 2010 9:55 pm

The filter is function of the load resistance presented to it, in this case, by the RF amplifier. It doesn't matter about the input voltage, power supply voltage or any other parameter.

I use a butterworth calculator, and plug in the number of poles (always inductor input), and the terminating resistance (the actual DC resistance of the rf amplifier(s) ). Works quit well !

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

Re: Modulator question

Postby K9ACT » Sat Aug 07, 2010 11:45 pm

kf1z wrote:Jack,

If your deck is going to run at 24volts, at 2.5 amps...

Then your filter design impedance is 9.6 ohms.



I was having a problem understanding what "run at" meant but more re-reading of Steve's stuff sort of straightened me out. The key was the 2.5 x carrier level. This started making sense when I divided the HV by 2.5 to learn what that carrier level will be. The carrier level on the PDM Generator now makes sense. This is a difficult concept for an old tube buzzard to get a handle on.

>So for a 4 element filter... (Butterworth T )

L1 = 77.96uH
C1= 2.04uF
L2= 188.21 uH
C2= 0.85uF

Exactly the numbers I then came up with using http://www.pronine.ca/butlf1.htm what a wonderful gadget.

Thanks for the confirmation.
K9ACT
Class E Enthusiast
 
Posts: 146
Joined: Sun Jun 27, 2010 12:03 am

Re: Modulator question

Postby K9ACT » Sun Aug 08, 2010 12:02 am

steve_qix wrote:The filter is function of the load resistance presented to it, in this case, by the RF amplifier. It doesn't matter about the input voltage, power supply voltage or any other parameter.


But how do you determine the load resistance without knowing the voltage and current?

>I use a butterworth calculator, and plug in the number of poles (always inductor input), and the terminating resistance (the actual DC resistance of the rf amplifier(s) ).

Glad you mentioned that last part as I was reading your schematic backwards and it looked like C2 was the first pole.

Another confusion (several); your article talks about 6 poles but the schematic only shows 4.

When I tried plugging your numbers into my filter calculator, it would not allow 4 poles. 3 or 5 but not 4. I went back to the Pronine web site and found the one in my last posting which allows 4 poles and corroborates your numbers exactly so I was able to use it to come up with numbers for any value of R. I don't know what that other calculator was but I learned something in the process.

After doing this I sort of lost interest in the air core inductors. I have two FT114-43 cores. Do you think they would handle 50W carrier level modulation?

Jack











Works quit well !

Regards,

Steve[/quote]
K9ACT
Class E Enthusiast
 
Posts: 146
Joined: Sun Jun 27, 2010 12:03 am

Re: Modulator question

Postby steve_qix » Sun Aug 08, 2010 7:42 am

I don't know if the FT114-43 will work or not in this application. The biggest problem you will have to deal with is core saturation. Magnetics has a core calculator that works very well, and you can use this to see if the core will saturate or not. It also fill figure out how many turns you need, etc.

One of the major drawbacks of using non-air inductors is the change in inducatance VS current. To this end, I have usually speced HI FLUX, MEGAFLUX or X-FLUX inductors. These are some of the most stable materials available.

So, when using a non-air core inductor(s), be SURE to calculate the inductance variation (the Magnetics calculator will do this) based on the absolutest highest positive modulation that the modulator will deliver. I usually tripple the DC current (that's 200% positive) and do all the calculations based on that number. So, if my DC current is 8 amperes, I do my calculations based on 24 amperes. You probably get away with 2.5x the DC - 150% positive, but definitely NOT the unmodulated DC. If the inductance changes a lot with current, or if there is saturation (the inductance will change a LOT under saturation), your filter effectiveness will change with modulation, and you will likely produce spurious products at the switching frequency, depending on the positive swing at the time.

Filter poles - I use 4 poles with lower power transmitters (400 watts and down), and 6 poles for anything higher power. No particular reason for choosing 400 watts as the threshold.

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

Re: Modulator question

Postby K9ACT » Sun Aug 08, 2010 10:00 am

steve_qix wrote:I don't know if the FT114-43 will work or not in this application. The biggest problem you will have to deal with is core saturation. Magnetics has a core calculator that works very well, and you can use this to see if the core will saturate or not. It also fill figure out how many turns you need, etc.



To evade learning a new trick, the calculator I use

http://www.dl5swb.de/html/mini_ring_core_calculator.htm

provides a Max Flux value and then calculates actual, based on freq and voltage saying nothing about current.

Using this, the 114-43 is ok up to 90v but I am not sure what voltage to use. Seems like the max voltage it could see is the supply voltage but I suppose I am missing something here.

The Magnetics software site is a bit intimidating. Can you tell me the specific program you are using for this if the above will not provide the needed info.

js
K9ACT
Class E Enthusiast
 
Posts: 146
Joined: Sun Jun 27, 2010 12:03 am

Re: Modulator question

Postby kf1z » Sun Aug 08, 2010 10:31 am

I'm not sure what supply voltage you are planning on using?

Even for your small class-D setup, I would suggest a supply of 100 to 130 vdc.

Then operate the transmitter at 35 to 50 volts at carrier.

In your first post you mentioned a 48v supply...
Am I correct in assuming this is the TOTAL voltage of the supply?
If that is the case, then your carrier level would need to be less than 20vdc.
User avatar
kf1z
Senior Contributer
 
Posts: 558
Joined: Sat Mar 18, 2006 7:07 am
Location: Ely, NV.

Re: Modulator question

Postby K9ACT » Sun Aug 08, 2010 10:56 am

kf1z wrote:I'm not sure what supply voltage you are planning on using?

Even for your small class-D setup, I would suggest a supply of 100 to 130 vdc.


But this is a two fet guy rated at 100W and I assume this means PEP. Perhaps I am confused here.

>Then operate the transmitter at 35 to 50 volts at carrier.

I think that would be well over 100W carrier.

>In your first post you mentioned a 48v supply...
Am I correct in assuming this is the TOTAL voltage of the supply?

Roger. I just ordered it on Ebay so it will be my opening gambit.

>If that is the case, then your carrier level would need to be less than 20vdc.

I came up with 16V which should give 40W carrier and about 240 PEP if I understand all this.

js
K9ACT
Class E Enthusiast
 
Posts: 146
Joined: Sun Jun 27, 2010 12:03 am

Re: Modulator question

Postby kf1z » Sun Aug 08, 2010 11:31 am

No, the 2 fets (FQA11N90) can run 100 watts CARRIER, not PEP.

Example:
If your supply was going to be 100vdc....
And you operated the transmitter at 30 volts, 3 amps....AT CARRIER.
that's 90 watts INPUT to the rf deck.... that's about 80 watts RF output, assuming 90% efficiency.

And leaves enough voltage headroom for 150% positive peak modulation.

10 ohm impedance.


=============

Now, if your supply is 48vdc....

Then yes, 16vdc at carrier also leaves you enough headroom for 150% modulation.
But will be less effiecient.

16 volts at 2.5 amps would be 40 watts input... about 30-35 watts RF output.

And that will make your load impedance the pwm filter will see around 6.5 ohms.


So it's possible..... :D

And will probably work ok for a low power "get started" transmitter.
User avatar
kf1z
Senior Contributer
 
Posts: 558
Joined: Sat Mar 18, 2006 7:07 am
Location: Ely, NV.

Re: Modulator question

Postby Gilly » Mon Aug 09, 2010 6:12 am

Hi

OK on the PWM Output filter values for Jacks TX...

For L2 of appox 190uH thats heaps of turns to be wound even on a CH777060 core....

Why do you have to use a cutoff of 15Khz as on my calculator if you use a cutoff of 30Khz @ 160Khz the -db is well under 60....and the number of windings are much less.....

I have just made Jays Class D TX with produces 100Watts output 40Volts @ 3Amps = 13 Ohms load...
L2 in this case is 254uH ...... I'm just not going to wind that many turns even on a big core....

If I use a cutoff at 45Khz thats L1 = 38uH and L2 of 85uH .... much easier to achieve......


Thanks


Wayne
Gilly
Class E Enthusiast
 
Posts: 167
Joined: Tue May 15, 2007 6:59 pm
Location: Mt.Eliza Victoria Australia

Re: Modulator question

Postby Gilly » Mon Aug 09, 2010 6:23 am

Sorry made a mistake...

45Khz is only appox -40db...

The max cutoff I could use is 25Khz that would give around -60db @ 160Khz........
L2 in this case is 152uH still a large no. of turns.....

Is there a way around this like maybe use 2 cores in series...?

Why use a cutoff of 15Khz......?

Maybe a Class H Modulator.......


Wayne
Gilly
Class E Enthusiast
 
Posts: 167
Joined: Tue May 15, 2007 6:59 pm
Location: Mt.Eliza Victoria Australia

Re: Modulator question

Postby Gilly » Mon Aug 09, 2010 7:36 am

Sorry Jack I don't mean to take over your post.....

Its late here and I keep jumping the gun on these posts...

I'm playing around with the turns ratio on the Class D to achieve 30Volts @ 4Amps = 7.5 Ohms that makes the PWM filter Inductors lower in valve so less turns etc: ......


Wayne
Gilly
Class E Enthusiast
 
Posts: 167
Joined: Tue May 15, 2007 6:59 pm
Location: Mt.Eliza Victoria Australia

Re: Modulator question

Postby K9ACT » Mon Aug 09, 2010 9:04 am

Gilly wrote:Sorry Jack I don't mean to take over your post.....

Its late here and I keep jumping the gun on these posts...

I'm playing around with the turns ratio on the Class D to achieve 30Volts @ 4Amps = 7.5 Ohms that makes the PWM filter Inductors lower in valve so less turns etc: ......


Wayne


Not my post. Love the extra inputs.

Your problem never occurred to me as I can't get to 100W with my supply leaving any headroom.

Even at 7.5 Ohms, there are lots of turns on an air coil. I am going with the smaller ferrite I discussed earlier till I find out it won't work.

Sounds like changing the output link might be the fix.

Also, I need to correct a previous posting. The carrier level voltage for a 48v supply should be 19, not 16v. This give me 40W at 2.5 A or 7.6 Ohms.

js
K9ACT
Class E Enthusiast
 
Posts: 146
Joined: Sun Jun 27, 2010 12:03 am

Re: Modulator question

Postby K9ACT » Mon Aug 09, 2010 11:19 am

Wayne,

Try adding 1 turn to the secondary. I just did that and the R went down to 5.5 Ohms, power and current went up and efficiency stayed about the same.

As a point of interest, I started by removing a half turn and it went all to hell. A full turn and it came up somewhere between hell and where I started. I then added a half turn and got nothing useful again and then adding the full turn made it all nice.

Must be some bino core law I do not know. Even though one can have half turns, they do strange things.

My Elmer worked up a Spice model for this RF deck that is fun to play with but we can not find the exact model for the 11n90 and he uses something close.

If you or anyone else is interested, I would be happy to send the model. It runs on LTspice and I don't know if .asc is universal or unique to this program.

js
K9ACT
Class E Enthusiast
 
Posts: 146
Joined: Sun Jun 27, 2010 12:03 am


Return to General Chit-Chat

Who is online

Users browsing this forum: No registered users and 1 guest