The AM Forum
March 28, 2024, 06:14:59 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: 1KW - PWM - CLASS - E - 75meter - TRANSMITTER BUILD  (Read 29543 times)
0 Members and 1 Guest are viewing this topic.
K9MB
Member

Offline Offline

Posts: 359


« Reply #25 on: January 03, 2022, 08:27:36 PM »

One of the things we're looking at is to stack several MLC capacitors in parallel.  Possible 3 .047uF devices, or maybe a .068, a .047 and a .033 in parallel.  Still researching parts.

This is to increase the current handling capability of the RF bypass capacitors at the DC side of the RF output transformers.

Steve- Isn’t this cap in parallel with the PWM output LPF cap and form the total value of the output capacitance?
I am asking because I am planning to use WIMA high current, low inductance caps for the filter and it has never been clear to me whether the requirements for esr and rise time for these 0.15 uF caps has to be better in HF ESR than the WIMA puse caps.
In this case, you have two 0.15 UF caps in parallel across the PWM output LPF circuit, or 0.33uF total- right?

Please clarify the requirements for these caps- and whether high speed WIMA film caps can work well here. If so, it seems like straining at a gnat to try to stack these MLCC caps when bigger ones could be bootstrapped near the board. Please comment and also correct me in this matter. Thanks, Mike
Logged
KK6IRZ
Member

Offline Offline

Posts: 20


« Reply #26 on: January 04, 2022, 05:31:20 PM »

Hi Steve and all else
Please keep in mind that I am only doing 80meters so I hope the .22uf will be ok.
However, if the idea is to increase current, keep overall capacity at .15uf, fit nicely on Doug's boards, and be available for purchase then maybe the C4532NP02E683J230 would do the trick.

the pads on the board are made for a chip 2525 and these are 1812.
I already checked the layout and it would fit nicely with 2 of those C4532NP02E683J230 parts side by side on the pads. That would be .136uf pretty close to the target .15uf and I have a bunch in stock and I can see they are also available for purchase at Mouser.

What are everyone's thoughts on that possibility?

Thanks
Jeff
Logged
KQ6F
Member

Offline Offline

Posts: 124


« Reply #27 on: January 04, 2022, 08:32:04 PM »

Jeff -

The cap you are talking about does indeed become a part of the last stage in the LPF filter.  But it's effect is minimal as far as the filter characteristics are concerned.  Small variations in its value will have essentially no effect.  As Steve has pointed out, it's main purpose is to provide RF suppression at its point in the circuit.  As such, it's important to have a low ESR rating.  Distributing its capacitance between different caps is a plus since that divides the current surges.

Rod KQ6F
Logged
KK6IRZ
Member

Offline Offline

Posts: 20


« Reply #28 on: January 04, 2022, 10:07:01 PM »

so very interesting
So it sound like the .068 doubled up would meet the intended of .15uf value before the .22uf cap was used due to non-availability of the .15uf.

I have about 100 of the C4532NP02E683J230 and may try two of them in parallel

tks
jeff
Logged
steve_qix
Contributing
Member
*
Offline Offline

Posts: 2599


Bap!


WWW
« Reply #29 on: January 05, 2022, 02:10:07 PM »

One of the things we're looking at is to stack several MLC capacitors in parallel.  Possible 3 .047uF devices, or maybe a .068, a .047 and a .033 in parallel.  Still researching parts.

This is to increase the current handling capability of the RF bypass capacitors at the DC side of the RF output transformers.

Steve- Isn’t this cap in parallel with the PWM output LPF cap and form the total value of the output capacitance?
I am asking because I am planning to use WIMA high current, low inductance caps for the filter and it has never been clear to me whether the requirements for esr and rise time for these 0.15 uF caps has to be better in HF ESR than the WIMA puse caps.
In this case, you have two 0.15 UF caps in parallel across the PWM output LPF circuit, or 0.33uF total- right?

Please clarify the requirements for these caps- and whether high speed WIMA film caps can work well here. If so, it seems like straining at a gnat to try to stack these MLCC caps when bigger ones could be bootstrapped near the board. Please comment and also correct me in this matter. Thanks, Mike

Hi Mike,

Any good cap can be used in the PWM filter.  We have not seen any issues in the field with various capacitors in this part of the circuit.

The reason that we can't put a lot of extra capacitance at the board is that the overload shutdown circuit, which is fairly fast and is located after the PWM filter, will start reacting to high audio frequencies from the modulator being fed to too much capacitance at the RF amplifier.  This will cause false overload triggers at high audio frequencies which is not desirable.

It was kind of a balancing act between getting sufficient bypass capacitance to do the job, without putting too much capacitance there where false overload triggers would be an issue.

Regards,  Steve
Logged

High Power, Broadcast Audio and Low Cost?  Check out the class E web site at: http://www.classeradio.org
K9MB
Member

Offline Offline

Posts: 359


« Reply #30 on: January 05, 2022, 04:38:38 PM »

One of the things we're looking at is to stack several MLC capacitors in parallel.  Possible 3 .047uF devices, or maybe a .068, a .047 and a .033 in parallel.  Still researching parts.

This is to increase the current handling capability of the RF bypass capacitors at the DC side of the RF output transformers.

Steve- Isn’t this cap in parallel with the PWM output LPF cap and form the total value of the output capacitance?
I am asking because I am planning to use WIMA high current, low inductance caps for the filter and it has never been clear to me whether the requirements for esr and rise time for these 0.15 uF caps has to be better in HF ESR than the WIMA puse caps.
In this case, you have two 0.15 UF caps in parallel across the PWM output LPF circuit, or 0.33uF total- right?

Please clarify the requirements for these caps- and whether high speed WIMA film caps can work well here. If so, it seems like straining at a gnat to try to stack these MLCC caps when bigger ones could be bootstrapped near the board. Please comment and also correct me in this matter. Thanks, Mike

Hi Mike,

Any good cap can be used in the PWM filter.  We have not seen any issues in the field with various capacitors in this part of the circuit.

The reason that we can't put a lot of extra capacitance at the board is that the overload shutdown circuit, which is fairly fast and is located after the PWM filter, will start reacting to high audio frequencies from the modulator being fed to too much capacitance at the RF amplifier.  This will cause false overload triggers at high audio frequencies which is not desirable.

It was kind of a balancing act between getting sufficient bypass capacitance to do the job, without putting too much capacitance there where false overload triggers would be an issue.

Regards,  Steve


Thank you for the explanation, Steve. It makes sense. I knew that you have used Orange drop caps there in the past, so the WIMA caps I have would be even faster than the orange drops, so it should work.
Interesting that the rise time spec is so fast that the bulk of the output filter capacitance is not “visible” in this time frame. Anybody who has done anything in RF or high speed buses knows that a trace to a “ground” or low impedance- looks like anything but. 😉

Just about the time I think I have peeled this onion to the middle, I learn that I know nothing much about it. It is great to have you as a resource.
Logged
KK6IRZ
Member

Offline Offline

Posts: 20


« Reply #31 on: January 05, 2022, 06:16:46 PM »

quoted by steve_qix 1/5/22
Hi Mike,

Any good cap can be used in the PWM filter.  We have not seen any issues in the field with various capacitors in this part of the circuit.

The reason that we can't put a lot of extra capacitance at the board is that the overload shutdown circuit, which is fairly fast and is located after the PWM filter, will start reacting to high audio frequencies from the modulator being fed to too much capacitance at the RF amplifier.  This will cause false overload triggers at high audio frequencies which is not desirable.

It was kind of a balancing act between getting sufficient bypass capacitance to do the job, without putting too much capacitance there where false overload triggers would be an issue

Thanks for the info Steve
So the 2 caps I suggested again sound like a good choice to try.
With two .068uf in parallel the total capacity of .136uf is still higher than the low side tolerance of the .15uf original caps.
I may need to cherry pick a bit to ensure the .068uf caps are on the equal to high side of tolerance but seems like it is possible to be a good choice.

I am curious Steve if the issues that were seen connected to false overload trips could be related to operators trying to modulate beyond 100% and using the higher .22uf caps?
Also Steve were this seen issues while operating on 160 or 80 meter?

Thanks
Jeff

Logged
K9MB
Member

Offline Offline

Posts: 359


« Reply #32 on: January 05, 2022, 08:39:01 PM »

I do not think there would be a difference between 160 and 80 meters for the bypass capacitors.
I suppose that if the Xc was marginally high, it would, but that should not be a factor in a choice between 0.15uF and 0.22uF.
Of course, the shunts on the drain bus are much more important to limit spikes that can toast the 11N90s. Those are indeed and specific and also bus length specific, in terms of number of devices.

For the overload shutdown circuit, high rise times are essential, so the larger value slows it down too much and affects function of the shutdown circuit.
Logged
steve_qix
Contributing
Member
*
Offline Offline

Posts: 2599


Bap!


WWW
« Reply #33 on: January 05, 2022, 09:44:29 PM »

To answer Jeff's question - we actually haven't seen many false overload shutdown triggers at all.  Usually if the overload (properly adjusted,  of course) triggers, there is a real reason for it.

The way the overload shutdown works is with a window comparator.   The current is constantly compared to the voltage, even as the voltage moves around.  So, if the transmitter is modulating very high in the positive direction, the voltage is higher and therefore the allowable current is higher - because the current is compared to the voltage. 

This differs from essentially every other overload shutdown scheme I've seen, where typically only the current is sensed, and since the current varies with modulation, false triggers are highly likely if the overload shutdown is to actually be able to protect the transmitter in any real sense.

Logged

High Power, Broadcast Audio and Low Cost?  Check out the class E web site at: http://www.classeradio.org
KK6IRZ
Member

Offline Offline

Posts: 20


« Reply #34 on: January 05, 2022, 10:49:15 PM »

Hey Steve sounds good
So it sound like the theory is if to much capacity is used at the bypass it has higher possibility of a false trigger but reality on the transmitters out there is that if everything is adjusted correctly there is little cause for alarm for possible false triggers using .15 or the .22uf caps,

The real issue in the field is the current spikes on the bypass caps and parallel caps would provide added capability in that area of the circuit.

With all this information I think I will try the .068uf in parallel
It will double the current capability and still keep the value close to original design.

thanks all
Jeff
Logged
KK6IRZ
Member

Offline Offline

Posts: 20


« Reply #35 on: February 25, 2022, 08:57:02 PM »

Hi guys sorry it has been awhile. I have been finishing off a homebrew Tube rig, pair of 8005s modulated by a pair of 811s.
I was playing with the digital phase splitter driver and has a couple curiosities.
Being that I am using the PA boards from Doug should I really include a phase adjustment on the driver? I of course have control of duty cycle and it seems to be working fairly smooth although I did notice I may need something to indicate that both are exactly the same of does some variance cause little issues?
I would be interested in your thoughts on this section of the transmitter

Thanks

KK6IRZ/KA1WLZ
Logged
vk3alk
Member

Offline Offline

Posts: 238


« Reply #36 on: February 27, 2022, 07:00:25 AM »

Hi Jeff

As far as the Duty Cycle goes I would adjust it as best you can using your CRO as an indicator....
Its hard to get them exactly the same as the waveforms will be slightly different....
Doing the best you can .. to get close enough.. is good enough I would say....
I don't think an indicator would be of benefit really just use your CRO etc:

If the two phases from your Flip Flop are 180 degrees apart that should translate to the gates...
By using your CRO monitoring both phases at the gates you could simply move them together and check the phases are 180 degrees apart....

Does this answer your questions ?

I'm sure Steve will reply ....


Wayne


Logged
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #37 on: February 27, 2022, 08:18:04 AM »

The goal when adjusting the duty cycle is to maximize efficiency, with the side effects of higher output and lower harmonics. A couple things need to be understood…  efficiency and power both come when both banks equally share the load and together, they’re on all the time (each half the time).  This isn’t at a straight 50% drive on time…or this would be easier.  The gates are capacitive.  They take a certain time to charge and turn on, and another amount of time to discharge and turn off.  The point of drivers like ixxd614’s is to pass big currents quickly to turn them on faster…the duty cycle being a bit below 50% is giving one side time to discharge and turn off before the other side turns on. In other words the mosfet essentially stretches the drive pulse, its output is on longer than its input.

So when adjusting duty cycle, you want to measure efficiency and watch power levels, too.  Power in (DC) vs power out.  If the duty cycle is too high, there will be overlap, and a lot of heat. If too low, power will drop.  Efficiency MAY peak when there’s a little under lap, but the power will drop a little.  Some people will say how great this is it’s so tweaked and so efficient….but it also has more distortion/harmonic output. Target highest efficiency then rock to each side for max power, which may or may not require any change at all. It may also be instructive to us the add function on your scope with the two drain currents…if there’s gap or overlap, you’ll see more variation.  I suggest you play with this at reduced power…where you can deliberately misadjust it to see what happens without risk of blowing it up.

Ed
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
steve_qix
Contributing
Member
*
Offline Offline

Posts: 2599


Bap!


WWW
« Reply #38 on: March 20, 2022, 06:00:14 PM »

Ed, I don't think anyone could have said it better !!!!!

That is it exactly.  The drivers and the MOSFETs all add a little to the "on" time of the waveform.  So, you have to use a duty cycle (as measured at the gates of the MOSFETs) of less than 50% - usually 40% works very well.
Logged

High Power, Broadcast Audio and Low Cost?  Check out the class E web site at: http://www.classeradio.org
KK6IRZ
Member

Offline Offline

Posts: 20


« Reply #39 on: April 11, 2022, 11:17:56 AM »

OKIEDOKIE!!
I have the duty cycle under control and can smoothly adjust it for both wave forms out of the Flip Flops. The cable going from my driver box over to Doug's boards are exactly the same length and I am not seeing any added phase shift at the inputs of the boards.
Looks like I will not need to worry about variations in phase so I only am using duty cycle control with no phase control.
So Far so good. Now I just need to get the PWM boards from Steve so I can finish my layout.
I am sure I will have more questions but so far things seem to not be giving me much of a hard time.

Thanks
For All The Input.

Jeff
Logged
KK6IRZ
Member

Offline Offline

Posts: 20


« Reply #40 on: June 28, 2022, 11:36:48 AM »

Hello all and Hello Steve QIX

Steve I have been sending emails hoping to get your boards and parts for this project.
I hope you are ok because I have not received a response in some time now.

I am hoping you might see this post.

Are your PWM 3 board set still available?

Thanks
Jeff
KA1WLZ/KK6IRZ
Logged
K9MB
Member

Offline Offline

Posts: 359


« Reply #41 on: June 30, 2022, 11:49:53 AM »

Hello all and Hello Steve QIX

Steve I have been sending emails hoping to get your boards and parts for this project.
I hope you are ok because I have not received a response in some time now.

I am hoping you might see this post.

Are your PWM 3 board set still available?

Thanks
Jeff
KA1WLZ/KK6IRZ

I am not sure what PWM version you want, but I have some REV C boards. They require a few mods to get them equivalent to REV F. If Steve has some REV F boards, it is more convenient, but REV C boards will work fine with Mods, Steve has told me in the past. I was able to get a set of REV. F boards to save making mods.
If you are interested in the REV Cs, I can make sure I have them.
73, Mike
Logged
KK6IRZ
Member

Offline Offline

Posts: 20


« Reply #42 on: June 30, 2022, 10:31:57 PM »

Hi Mike
Well I will hold a bit longer to see if we hear from Steve but if not yes I would be interested. I was hoping to get the 3 boards and parts combo package from Steve but if not I can still make it work.

Thanks
Jeff
Logged
K9MB
Member

Offline Offline

Posts: 359


« Reply #43 on: July 01, 2022, 10:38:15 AM »

Hi Mike
Well I will hold a bit longer to see if we hear from Steve but if not yes I would be interested. I was hoping to get the 3 boards and parts combo package from Steve but if not I can still make it work.

Thanks
Jeff

I agree, Jeff. I got the C boards as kits and then got sidetracked by life and when I got going again, the versions had passed through C,D, to F.
They are essentially the same boards with new mods that can be added to tue Cs, but the Fs made it cleaner.
I Used kit parts plus added more to get my boards stuffed.
I have some extra parts like the current sensors ICs, BPT, FETs, and other things if you cannot get complete kits. I Always buy extras to lose and some to destroy- you know. 😬😉
73, Mike
Logged
steve_qix
Contributing
Member
*
Offline Offline

Posts: 2599


Bap!


WWW
« Reply #44 on: July 06, 2022, 07:27:45 AM »

Hello all and Hello Steve QIX

Steve I have been sending emails hoping to get your boards and parts for this project.
I hope you are ok because I have not received a response in some time now.

I am hoping you might see this post.

Are your PWM 3 board set still available?

Thanks
Jeff
KA1WLZ/KK6IRZ

I am missing some key parts for the PWM boards :-(   When I moved from Townsend to New Hampshire, somehow I lost all of the LTS 25 NP Hall Effect devices !!!  This represents a couple-thousand dollars worth of parts.  I know I have them somewhere.  These aren't available anymore.

Let me know if anyone can find some of these parts - the system won't work without it.
Logged

High Power, Broadcast Audio and Low Cost?  Check out the class E web site at: http://www.classeradio.org
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #45 on: July 06, 2022, 09:20:50 AM »

I am missing some key parts for the PWM boards :-(   When I moved from Townsend to New Hampshire, somehow I lost all of the LTS 25 NP Hall Effect devices !!!  This represents a couple-thousand dollars worth of parts.  I know I have them somewhere.  These aren't available anymore.

Steve, isn't there a drop-in replacement from LEM?  CAS 25-NP
https://www.lem.com/sites/default/files/marketing/lem%20leaflet%20cas%20casr%20cksr.pdf

IF (yes, big IF) they work, Digikey has a couple hundred in stock at $17.07/single piece

Ed
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
K9MB
Member

Offline Offline

Posts: 359


« Reply #46 on: July 06, 2022, 11:37:33 AM »

Hello all and Hello Steve QIX

Steve I have been sending emails hoping to get your boards and parts for this project.
I hope you are ok because I have not received a response in some time now.

I am hoping you might see this post.

Are your PWM 3 board set still available?

Thanks
Jeff
KA1WLZ/KK6IRZ

I am missing some key parts for the PWM boards :-(   When I moved from Townsend to New Hampshire, somehow I lost all of the LTS 25 NP Hall Effect devices !!!  This represents a couple-thousand dollars worth of parts.  I know I have them somewhere.  These aren't available anymore.

Let me know if anyone can find some of these parts - the system won't work without it.

I bought 4 of these last year.

https://www.ebay.com/itm/322226781590

They are only $5.22 each.
If you need a couple, I can send them along. I found one in a kit I got from you and put that one in my overload shutdown board, so I just kept these for spares.
It took about 6 weeks to get here from China.
I can send a couple for you to test before you order any of them to make sure they work correctly.
 No payment- Just give me a credit or replace them when you find that big hoard under something you forgot about- been there…😉
I can send out right away. 73, Mike
Logged
w9jsw
Two shots of Whisky
Contributing
Member
*
Offline Offline

Posts: 592



« Reply #47 on: February 05, 2023, 09:50:29 AM »

Looking for the rest of the story?
Logged
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 0.066 seconds with 19 queries.