The AM Forum
March 28, 2024, 06:41:04 AM *
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 [3]   Go Down
  Print  
Author Topic: 350W 3 band Class D FET AM Transmitter  (Read 71691 times)
0 Members and 1 Guest are viewing this topic.
M0VRF
Member

Offline Offline

Posts: 172


« Reply #50 on: May 06, 2018, 05:27:50 AM »

Added for further 'Clarity'

Nothing like LTSpice espec when you have the correct model!

J.


* 60_40.jpg (191.22 KB, 1919x917 - viewed 1431 times.)

* 50_50.jpg (189.99 KB, 1912x915 - viewed 1249 times.)
Logged
vk3alk
Member

Offline Offline

Posts: 238


« Reply #51 on: May 06, 2018, 06:11:18 AM »

Thats interesting..

The LTC6992-4 only goes to 100% positive and 5 % negative....
When at 50:50 DC: you.. I think....can see the top and the bottom of the sinewave rounding off a little.....
When at 60:40 DC you definitely only see the negative rounding off.....because the bias at the MOD input is less...

Can you run the spice program with a LTC6992-1......which is what Nigel has used...
You will probably see negative clipping when at 60:40 DC.....with the same drive level...
This is because the audio input is too high....for that DC...

Once the audio drive input has been set you should not have to adjust it again....that is for a given DC....
If you vary the RF output power level by say reducing / increasing the voltage at the drain of the Modulator stage ....you will not over modulate  Smiley
So you should not have to adjust your audio input .....
The recommended DC is 60:40....which gives you plenty of headroom for positives over 100%....


Wayne

Logged
M0VRF
Member

Offline Offline

Posts: 172


« Reply #52 on: May 07, 2018, 11:03:06 AM »

The modulator will not modulate from 40% to 100%.

It will from 50% to 100%.

I'm not willing to waste anymore time on this, there are better designs that will and I'll stick to those.

Thanks for your input.

I'd rather draw a line under this as have other technical issues to resolve.

I was hoping the author would comment, but has chosen not to.......

Cheers

J.


* both.jpg (222.04 KB, 1916x917 - viewed 1265 times.)
Logged
g7klj
Member

Offline Offline

Posts: 4


« Reply #53 on: June 20, 2018, 10:54:00 PM »

Quote
OK, It's the O/P from the LTC, so if you set the carrier to less than 50% the bottom of the sine will clip.

Sadly, you didn't DC couple your 'scope on the output waveform. If you do it will be obvious (? lol), or rather, its obvious to me after many years of faffing around with this crap that the modulator cannot ever output NEGATIVE DC (which is what you would require for *symmetrical* mod in the 60/40 case. Even if it could, your PA transistors won't thank you for supplying negative drain volts!)

It looks to me as if your circuit has the expected behaviour -- you say its not 100% positive modulated @ 60:40 because you are not increasing the input once you see the negative peaks clipping -- or are you?

If peak supply is 40V, and you set carrier @20V, now you will see "symmetrical" modulation -- and you do, right?

So now, for argument's sake (silly example) set the resting carrier voltage to 10V. The modulator will clip the neg peaks very soon (coz 0v is only 10V away) but the positive peaks should not clip until they hit 40V. It will "overmodulate" like mad in the negative direction (just to prove the point), but increasing drive should increase the pos peaks til they get to the supply volts, minus a little tiny bit.

Is this what you are seeing? If it is, then its expected behaviour, and its your outboard audio processing (or make a NEGATIVE ONLY peak clipper) to smooth off the negs just before the modulator hits 0v.

J, (is it John?) am I on the same page as you. Perhaps I missed something but I did take the time to look a couple of times at your posted images.

Summary: if you offset the carrier for more than 100% peak positive modulation, your audio chain has to take care to avoid the carrier getting shut off on the negative peaks. I do this with a peak limiter that only works on the neg peaks. Some people connect a diode to the drains from a very low voltage constant source. It still clips in the negatve direction on a sine, but the carrier can't shut off due to the diode conducting 2 or 3 volts to the drains when the modulator pinches carrier.

Anyhow, I hope you get some more responses because I have what I think should be a fairly common and logical statement on here for three weeks, and not one bugger has replied!

BTW: if you do have a modulator that takes care of this asymmetrical modulation automatically, I want to know about it :-)


Steve
Logged
M0VRF
Member

Offline Offline

Posts: 172


« Reply #54 on: June 21, 2018, 02:22:04 AM »

Many thanks Steve, truly, for looking into this. The author still hasn't bothered to reply which is somewhat unacceptable. That and another thread has been started 'Detracting' from this.

I've got some time this weekend to muse over your thoughts and will spend some time in my lab.

It's with the LTC chip where the issue lies. I've built the same circuit as published and it doesn't work as expected.

Cheers.

JohnB.
Logged
vk3alk
Member

Offline Offline

Posts: 238


« Reply #55 on: June 21, 2018, 07:31:01 AM »

Hi John

Not sure why you are having troubles with this circuit....

Mine is disconnected at the moment but can set it up again fairly quickly if required....

It is an easy circuit really but could you do these things....

1.   set the duty cycle to 50:50 that's a DC voltage of 500mV to the input of the LTC...
2.   apply 10 volts to the drains and check the carrier / output from the modulator is 5 volts..thats
      a DC of 50:50
3.   Now apply an audio tone of whatever frequency say 1Khz at 1 volt PEP to the input of the LTC
      using your audio oscillator / generator ..... the output should be symmetrical....
4.   check both waveforms input and output using your CRO ...... both should not be clipped and
      should show a sinewave....one at 1 volt PEP and the other 10 volts PEP....
5.   Now change the Duty Cycle from 50:50 to 60:40 ... thats a DC voltage of 400mV...at the input
      of the LTC ....the output should be 4 volts .....
6.   both waveforms should be similar apart from the voltage levels.
      The negative going part of each waveform should be clipped....and the positive should not .....
      But now there is headroom....input has 100mV and the output has 1 volt.....
7.   Now if you further decrease the Duty Cycle you will see on both waveforms more clipping of
      the negative going part of the waveform....
8.   Going back to a Duty Cycle of 60:40
9.   Increase the audio level input to 1.1 volts PEP
10. Both waveforms will show more clipping of the negative going part of the waveform but....the
      positive part will reach the upper limit ... 1 volt for the input and 10 volts for the output...

If that's the case then all is going as it should...


Wayne

PS:  hope you understand what I have written ...
Maybe a change from 1 volt to 1.1 volts might be a bit fiddly.
Changing the DC from 60:40 to 70:30 or lower might be easier for you to adjust etc: ...
Anyway see how you go...

PS..PS.. Use a dummy load on the output of the Modulator not a TX




    
Logged
M0VRF
Member

Offline Offline

Posts: 172


« Reply #56 on: June 21, 2018, 08:19:03 AM »

Thanks Wayne, I've got an Agilent E series sig gen with LF out so no problem adjusting any audio or RF level.

As I have explained in previous posts I'm unable to obtain the full O/P voltage unless I use 50/50 duty cycle.

Any other duty cycle will clip prematurely.

Surely whatever duty cycle you use the O/P waveform should still clip symmetrically, well at least it does on my UCC35702 PWM circuit that I've used for years.

I only built this one for its simplicity.

I will report back!

Regards

JohnB.
Logged
vk3alk
Member

Offline Offline

Posts: 238


« Reply #57 on: June 21, 2018, 08:44:10 AM »

OK John....

Could you do me a favor ...
Was reading your posts again and you had the upper and lower gate resistors back to front....
Do you mind just for my sake changing or checking those resistors again..
The upper FET has to have the 4.7 Ohm and the lower FET 16 Ohm......
Also the charge pump diode has to be a fast recovery like a uF4007..
That's the one going from 12+ to pin 6 of the IR2110....
You used a 1N4148 instead...
I know you said it worked for you but just for me could you change / check its a uF4007 or similar....


Thank you


Wayne




Logged
g7klj
Member

Offline Offline

Posts: 4


« Reply #58 on: June 21, 2018, 10:18:55 AM »

Thanks Wayne, I've got an Agilent E series sig gen with LF out so no problem adjusting any audio or RF level.

As I have explained in previous posts I'm unable to obtain the full O/P voltage unless I use 50/50 duty cycle.

Any other duty cycle will clip prematurely.

Surely whatever duty cycle you use the O/P waveform should still clip symmetrically, well at least it does on my UCC35702 PWM circuit that I've used for years.

I only built this one for its simplicity.

I will report back!

Regards

JohnB.

Hello again John!

No sweat. Take your time. Its damn frustrating when things don't work out as expected.

I also have one of the ucc-chip modulators and it does have the expected behaviour: when you set the carrier level less than 50:50, the neg peaks clip before the pos ones. As they should. Otherwise, the universe would explode from failing to obey the laws of physics.

LOL. Keep your end up. PM me if you need to.

Steve.
Logged
M0VRF
Member

Offline Offline

Posts: 172


« Reply #59 on: June 23, 2018, 01:27:38 PM »

Hi Steve, please forget about everything other than the LTC chip, this issue have nothing to do with the diode (4148 is fine at the voltages I'm using BTW) or the gate Res (just a cheap way of adj dead time).

As shown in my previous posts I clearly show the LTSPICE sims and it's clear that 60:40 won't go to 100% without clipping, sim it yourself and see.

I won't be using it and have reverted to my designs using the UCC device.

I'd rather not waste anymore time.

Cheers BTW.

JohnB.




Logged
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #60 on: July 16, 2018, 07:20:09 PM »

Might make sense to email Nigel - he's usually very ready to comment and help.
He well may be traveling now that it is summer?

Haven't heard him on the air in more than a month, fwiw.


                      Btw, I've heard him on the rig, and it sounds fine, and the modulation waveform is excellent.



                     _-_-bear
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
M0VRF
Member

Offline Offline

Posts: 172


« Reply #61 on: July 17, 2018, 03:41:25 AM »

Sure it will sound fine but doesn't work as expected and sure I emailed him numerous times before even posting on here Bud.

I was trying to be polite but now fed up and have given up due to lack of help.

Cheers


JB.
Logged
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #62 on: July 20, 2018, 02:16:10 PM »

Bud??

Hey, look, I was just trying to help.
I've spoken to Nigel on the air, and emailed him in the past.
He's been fine to deal with.

IF he's not responding then there is a good reason.
Perhaps one of them may be that your email gets "eaten" as a SPAM so he doesn't see it?
Possible.
Or, as I said, he is away.
Or, perhaps he is SK.

Also, this is not a KIT, and this is not a hand-holding, spoon feeding society - although people do try to help.

Regardless, take a pill and calm down.

                                  _-_-bear
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
Pages: 1 2 [3]   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 0.093 seconds with 19 queries.