A Clean Canvas - 24 Pill Class E Rig Construction

[K1JJ]

here's the official music video to Tom's rig:

http://www.youtube.com/watch?v=xeX9zoWSut8

I should do a mixdown of this , bet I get plenty of LOL's.  

Then sushi afterwards !

then a ambulance ride! get down!  

HAHAHAHA!!  Gawd that video was funny enuff to bring tears to my eyes, Derb.  I haven't heard that in years.  Yep, that's the official Rico Suave Class E song. Just like when they play a song when the WWF or MMA fighters come out.    

I can see it being played in the background when Rico gets fired up for the first time on the air...

T

[K1JJ]

Are these digital panel meters OK for using with RF?  At only $9 each, they're cheaper than analog meters and pretty cool looking for monitoring the four module's current and voltage.

Is anyone using them now or at work?


0-100A Blue 3.5" Digital meter:

http://cgi.ebay.com/3-5-Inch-Blue-LED-Digital-AMP-Panel-Meter-DC-100A-Shunt_W0QQitemZ370301414274QQcategoryZ25411QQcmdZViewItemQQ_trksidZp4340.m8QQ_trkparmsZalgo%3DMW%26its%3DC%26itu%3DUCC%26otn%3D20%26ps%3D63%26clkid%3D8360502073947355137

0-199V Digital Meter:

http://cgi.ebay.com/NEW-3-Digital-Blue-LCD-DC-0-199-9V-Volt-Panel-Meter_W0QQitemZ270490090047QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item3efa773a3f
T

[KF1Z]

I thought about those too...

However.....

My handheld digital VOM  will NOT read correctly trying to measure the DC from the PWM modulator....

Maybe I needed more bypassing or some such...

But it has always been said to use an analog meter... so I've stuck with them

I pay $8 each for my panel meters....

I don't care what brand they are....

I'm not going to pay more than $20 each for a panel meter for any reason.. !

Heck I'm not building a nuke reactor here!


If you decide to try those Tom, please let me know how they work!

That would make an E-Rig look super spiffy!   

[w3jn]

NOT good to read this, John. I built the N3ZI DDS to be a VFO for a Class D TX. Would this shift be heard over the air? Or wuld this be unacceptable as a L.O. in a RX?
Also reading feedback from others here who have built the N3ZI. Noisey and not enough ooomph???  Norcal still hard to get?

Good vid Tim
FRED

Fred, prolly up to a hundred or two watts the N3ZI kit oughta be OK and won't cause too much neighborly trouble.

It needs a better frequency reference.

[steve_qix]

I would use analog meters.  You can see a whole lot more  

Here's a schematic.  A lot of people have asked for it.

http://www.classeradio.com/class_e_24_fet.pdf

For Printing:

http://www.classeradio.com/class_e_24_fet_top_half.pdf

http://www.classeradio.com/class_e_24_fet_bottom_half.pdf

I tried posting a .jpg of the schematic, but it was not readable - too many components to shrink to 8 inches wide!

[K1JJ]

Steve,

The schematic makes everything clear, thanks!

Questions:

What are the voltage ratings for C-shunt (.15u) and C-Bypass (1500pf) ?

Could the FOUR 12V regulators for the digital drivers be substituted by using ONE 10A 12V regulated supply?

I see that the input tranzorbs have been eliminated. Are they used on the input of the phase splitter board if used with a ricebox?

BTW, reading the "official" site archives back in 2008, Frank/GFZ came up with a nice PC board to hold the digital driver chip and RF MOSFET leads and associated parts. Has anyone fabricated this board? Any idea how to knock one out easily?  This board  wud complete the board set and get rid of the clap-trap mess on the RF heatsink.  I attached it below.


BTW, the link above on AMfone for class E had me wondering where all the "meat" was.  The REAL good construction info and threads are located at the "Official" Class E Website:  

http://www.classeradio.com/

We should probably have a link to get there to avoid confusion to others, as it did me. I spent a day reading the "official" site and will continue to until I digest the latest dope. For newcomers, this is probably the best way to get up to speed. Then pick out a schematic or rig config and start writing down what is needed to pull it off.


T

GFZ  universal digital driver/final board:

[steve_qix]

Steve,

The schematic makes everything clear, thanks!

Questions:

What are the voltage ratings for C-shunt (.15u) and C-Bypass (1500pf) ?

Could the FOUR 12V regulators for the digital drivers be substituted by using ONE 10A 12V regulated supply?

I see that the input tranzorbs have been eliminated. Are they used on the input of the phase splitter board if used with a ricebox?

Hi Tom,

The shunt capacitors (1500pF) are very specific parts, so the voltage rating (500 volts) is part of the specification.  No other parts should be used other than the ATC100C series 1500pF caps.

The bypass caps - I use 150 volt or better.  The current rating is the key.  I use 2 or preferably 3 orange drops in parallel.

Right on the transzorbs - but there IS  a transzorb across each group of drivers to protect them in the unlikely event of a failure getting back into the drivers.  You don't want an avalanche (sp?) condition.

Regards,

Steve

[steve_qix]

The TransZorbs are across the driver power supplies to prevent, in the unlikely event of a failure, any high voltage getting back into the driver power supply and cascading through the system.

On the power supply question, you could use a single supply for all of the drivers.  Those simple switching regulators are very nice; small and efficient which is why I use them.  Usually, big supplies are either poorly regulated, or are analog (linear), and waste a lot of power in the form of heat.  I have seen some good switchers out there - maybe that's what is proposed here 

Regards,

Steve

[K1JJ]

OK on all Steve.  Then I'll use a single 12V switching supply for the DD Vcc.


Today I stopped down at the local scrap yard and found a few sheets of brass at $2 /pound. These make great capacitor bank connections as well as RF MOSFET connections.

It's a small step, but here's the BEER CAN main cap bank ready for action. A little here, a little there. After a while it adds up to big db...

This is 35,000 ufd at 150V for Rico Suave.  


Hmmm.....  Now that I look at it, should a bar be soldered across at the center - shorting the the five center stubs together  - or is it correct as-is?

T

[WA1GFZ]

How about using  computer power supply for a driver source?

[Opcom]

I like this project, hope to follow it to completion here.

[K1JJ]

How about using  computer power supply for a driver source?

Good idea.  Though I picked up a switched 12V, 20 amp supply for $38 on e-Bay tonight.  It will power the digital driver as well as some other low power 12V points on the other boards.

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=270487051020&ssPageName=ADME:L:OU:US:1123

I've been ordering parts and digging out what I'll need. There sure is a lot to building this rig.

I've been looking into making the digital driver and RF final area as compact, RF-friendly and stable as possible. I think eliminating those driver transformers and wires is a good step. They will be replaced with a QIX phase splitter board and the digital drive 1XX414 drivers on the heatsink.  Short leads and fat conductors.  

I'm cornvinced 24 pills using robust devices like the 11N90 will put out more power than we are accustomed to when the layout is tighter.  This includes the output transformer layout and lead length. We'll have to see.  Thanks for your layout advice and new ideas, Frank.

T

[WA1GFZ]

There is a whole new world to observe out there when you can see 1GHz on your scope. Open up your rice box and look at the final amplifier strip. There is a good reason for a clean layout.

[K1JJ]

I wanted to keep this project updated as it progressed. The new 24 pill PDM rig is now about 60% finished. The sequencer, shutdown board, PDM generator, PDM power board, 135V power supply and control wiring are all set. The RF modules and interconnecting wiring need to be wired.

Notice there are two PDM power boards and room for a second shutdown board. The reason is that I may add another 24 pill RF set of modules to make it a 48 piller in the future. Depends on how stiff the on-air competition continues to be....

Notice the RF unit has 10A meters mounted (viewed thru the panel Plexiglas) and is fully digital driven. There is a small A/D board mounted that will permit direct drive from the ricebox RF driver with 1 watt.  The ferrite core transformers are mounted and held down with plexiglas.

I hope to have something working in a few weeks or so.

73,

T


1)  24 pill RF deck with digital drive, output transformers and 10A old buzzard meters.

2) Two PDM power modulator boards

3) Sequencer and Overload board ready to be mounted in top modulator cabinet

[K1JJ]

1) Top view of power supply wiring. There is a sheet of plexiglas mounted above the filter caps that supports the relays and associated components.

2) Lower cabinet ready for RF deck tenants

3) RF deck. Full digital drive.  Huge heatsink.

[K4TLJ]

N3ZI has a new improved DDS VFO. http://www.pongrance.com/super-dds.html
More expensive but much better specs than the old one.

[KF1Z]

N3ZI has a new improved DDS VFO. http://www.pongrance.com/super-dds.html
More expensive but much better specs than the old one.

If you want one.....

I'll send you one for the shipping cost....
Built, but needs improvements   

I'll never use it ....

[K1JJ]

Thanks for the generous offer, Bruce.  (and the VFO info, Terry)

Do these VFOs generate phase noise that may be heard down the band with a big QRO signal? What is the disadvantage of using one?

I understand it is important to run them at X2 freq to keep the VFO signal from interfering when in receive mode.

The reason I went with the A/D converter is to be able to keep transceive capability, which I like. Separates can be a pain to zero beat, etc.  But I'm always looking for new dope to change my ideas.

T

[KF1Z]

Thanks for the generous offer, Bruce.  (and the VFO info, Terry)

Do these VFOs generate phase noise that may be heard down the band with a big QRO signal? What is the disadvantage of using one?

I understand it is important to run them at X2 freq to keep the VFO signal from interfering when in receive mode.

The reason I went with the A/D converter is to be able to keep transceive capability, which I like. Separates can be a pain to zero beat, etc.  But I'm always looking for new dope to change my ideas.

T

Yeah they generate a ton of crud...
AND only generate 200mV +/- pk-pk output.

so would need filtering and amplification to drive, well, anything really...
At the very least.

They are a very simple kit, not a lot of parts, so they are good for just general use aound the shack, when you need a quick signal for testing, but I would not recomend them for a QRO rig.

bruce

[KF1Z]

Running the VFO at 2X  (or more) does give the advantage of not interfering with RX.
And any drift, or what have you is also divided by half, making it at least appear more stable on freq.

Makes getting 2 square-waves out of phase a tad bit easier too...
Not much, but a little 

[w3jn]

I think the older N3ZI DDS has too much phase noise for a big rig Tom.  Not sure about the new one.

It seems to me that most of the troubles guys have with Class E rigs center around them using riceboxes for drivers.  They have problems with key up spikes, forget to turn the power down, or whatever, and BAM a bunch of fried FETS.  Steve's design is robust enough that barring drive issues there's not much that can fry a FET otherwise.

My suggestion:  bit the bullet and build a 7 MHz VFO.  Use the Franklin design - it's more stable than anything else I've built - 10 hz/hour drift using a tube.  Cut a hank of good stiff B&W coil stock, find a good quality variable cap, and firgure a series polystyrene cap to get the tuning/resonant range you want.  I think you have some of those National dial drives from your telescope projects - this would be an oustanding place to use one.  Hang a 7474 divide-by-2 off the output and yer done.

[WA1GFZ]

Tom,
Since you have an HP606 I suggest you check the reactance of an orange drop cap at RF. Put it in series with a 10 ohm resistor. You are trying to bypass a 5 ohm source so you need lower than 1/2 ohm reactance. I would put some ceramic caps in parallel with the OD cap. Those RF chip caps work well at RF much more stable than dork knobs. If you want to save some money get the ones without foil and lay them between the source and drain leads.

[K1JJ]

I think the older N3ZI DDS has too much phase noise for a big rig Tom.  Not sure about the new one.

It seems to me that most of the troubles guys have with Class E rigs center around them using riceboxes for drivers.  They have problems with key up spikes, forget to turn the power down, or whatever, and BAM a bunch of fried FETS.  Steve's design is robust enough that barring drive issues there's not much that can fry a FET otherwise.

My suggestion:  bit the bullet and build a 7 MHz VFO.  Use the Franklin design - it's more stable than anything else I've built - 10 hz/hour drift using a tube.  Cut a hank of good stiff B&W coil stock, find a good quality variable cap, and firgure a series polystyrene cap to get the tuning/resonant range you want.  I think you have some of those National dial drives from your telescope projects - this would be an oustanding place to use one.  Hang a 7474 divide-by-2 off the output and yer done.

John,

That sounds like a great idea.   I could use the A/D converter already in there to take  1 watt and convert it to digital drive. Do you have a working schematic of a suggested Franklin?  I wonder if it will need a buffer before the A/D?

Will the A/D converter generate its own phase noise or is this not an issue?

What is the reason for using a tube vs: a signal transistor or FET?   



Frank:  What's an OD cap?  What orange drop bypass caps do you mean, the bypass caps at RF deck's modulation input (.15's?)    Or the 1500 pf ATC caps at the RF drains?     I already have quality ATC caps coming for the drain bypass.  I only have 0.15 orange drops for the input modulation bypasses, however.

T

[WA1GFZ]

the OD is orange drop and they are not a good bypass at RF. Check them you will see. The only high value film cap that was worth a darn was made by Sanders Associates. I put a bunch of .005uf discs in parallel with the OD to get a HF good bypass
Do a reactance sweep with your HP606 you will see. I bet you see the reactance take off above about 800KHz.
Build a VFO with a FET with a second one as a source follower

[K1JJ]

the OD is orange drop and they are not a good bypass at RF. Check them you will see. The only high value film cap that was worth a darn was made by Sanders Associates. I put a bunch of .005uf discs in parallel with the OD to get a HF good bypass
Do a reactance sweep with your HP606 you will see. I bet you see the reactance take off above about 800KHz.
Build a VFO with a FET with a second one as a source follower

OK, OD is orange drop.... duh...

So you're saying the 0.15 cap at the bottom of each RF output transformer  (where the PDM power comes in)  needs to have additional bypass caps for the higher RF freqs.  How about a 1000pf doorknob there in addition to the 0.15 orange drop?   I'd hate to spring for expensive ATC caps there - I need 8 of them for the full job.

T