The AM Forum

Hi,
I have started my first homebrew amp project!
The goal is to have an amp that will let me participate in more AM nets, and also possibly get on 160m too.

The two amps I have now don't do 160m and have not been recommended for AM use. (Collins 30L-1, 30S-1)
Will be using a Collins 32V-2, 32S-3 or a Harris RF-1310 exciter to drive the amp.

I read many things on the forum before picking a design.
Did not go w/ 813's driven by 813's.
The costs and complexity for 4 tubes, and the fact that I could only find chinese sockets & caps killed that option.

Went with a simple one tube design.

This is all way over my technical level anyway, but that is how I hope to learn, with help from as many elmers as I can get.

So here is the pile'o parts I have collected so far.

- EIMAC 3CX3000F7, w/ chimney, homebrew filament transformer.
- Big ROTRON blower from a Gates something or other
- Dahl plate transformer: 240 in, 2600, 4000, 4500 out.
- Components to build the rectifier
- A bunch of torroids to make the filament choke
- A selection of big tank coils and big band switches
- A selection of Mica and ceramic caps
- Two - 6 Mfd 6000v oil filled caps
- A 10H choke from an old Gates
- Six 100k ohm 225 watt Ohmite bleeder resistors
- A big aluminum cabinet on wheels to put it all in.

I have not bought the two tuning and loading vacuum variable capacitors yet, am still learning how to do all the calculations for each band using the "match.exe" program provided with the ARRL Handbook.

Plus I know there are 9 million little parts and pieces to collect for all the safety and control circuits.

But at least I have started!
My goal is to have it working before I retire in 2017 @ age 67.

73
Frank
KJ4OLL

Looks like a nice project, Frank.

You might shoot for about 40 uf for filtering in the HV supply.  

Test the blower. If it is loud, consider locating the amp far away or running a Variac to slow it down or get a quieter blower.  Little or NO background noise is important on AM.  On SSB we can use a noise gate effectively, but on AM a gate is unnerving.  Loud blower noise on AM is distracting and will intermod with your audio.

The 3CX-3000A7 is a FB tube making -40 to -45dB 3rd order IMD figures. Very clean.  You will need a very clean driver that is at least -45  3rd order to do it justice.  This is a big challenge since most rice boxes or old AM rigs will be hard pressed to do better than -35 or so.

What are the figures on your Harris RF-1310 exciter?  I think the 32S3 has RF NFB, so you may be in good shape once modified for AM use.


Good luck with it and keep posting as it progresses.

Tom, K1JJ

Frank, look here for some ideas.

http://w2dtc.com/w2dtc-plasma-amp-page.htm

and here

http://w2dtc.com/index-amplifiers.htm

here also
http://w2dtc.com/w2dtc-class-a-rf-amp-page.htm

I have a Henry 2000d plasma generator amp that runs the 3cx3000a7, am in the process of multi-banding it, have gone through a few design permutations so far, think I have settled on  Vac cap for C1, roller for my inductor and bread slicer for c2, you may want to investigate using the same, as the c2 position in a pi-l net hasn't as much voltage as the c1. Keep us posted. 73 Ed AB3HT

There's only one problem with using the Harris RF-1310, it only does 100mw of power, so assuming you are going with grounded grid, that's not anywhere near enough drive. That exciter was meant to be used with the RF-110A amplifier, which used a pair of 4CX1500Bs driven by a 4CX250B. Also, I'm not sure about the 32S-3, but the 32V-2 is a 100 watt rig, so you're going to definitely want to cut back on the power output somehow.

Thanks for the tips & comments!
I'll add more filtering for the HV supply, would like to understand more about how the capacitance value is arrived at/calculated.

The blower is going inside the cabinet, and the intake/exhaust can be run through a muffler, if needed.
But the blower is pretty quiet in testing (w/o any back-pressure, however).

Forgot about the Harris RF-1310 100mw max drive, I won't be using it to drive this amp.
(But I am always looking for a Harris RF-100A and the single-phase power supply for sale within driving distance of my QTH...hint hint)

Interestingly, I might try to use the RF-1310 for QRP - I recently restored an HP 230B RF amp, it should be a hoot to drive the ancient Hewlett Packard amp w/ the Harris exciter!
http://www.youtube.com/watch?v=X_bTy0RIo2Q

The w2dtc web site was my inspiration for using the 3CX3000F7, GREAT!

I have a BIRD 30db in-line attenuator to use between the exciter and the amp input.
However, the amp that w2dtc built did 100 watts in---1500 watts out, so possibly I won't need the attenuator.

Open to suggestions for other options to drive the amp with so as to obtain the low IMD - what would you use?

73
Frank
KJ4OLL

If the choke will be used then 12uF is more than adequate for both filtering and regulation. OTOH around 25-30uF will get the job done by itself. At 4-5KV more is not better and just costs more.

Carl

I still lust after that Harris Amp RF110A. That is POWER..
Good luck with your project. Does the 3CX3000 have handles??? That's where it's happening these days.

I have the Boonton 230 which was renamed after HP bought them. A good 10-500mc coverage at up to about 5W out.
It and a HP-608C was used for antenna tuning out in the back yard back when that was the "easiest" things to move outside. They havent been used for almost 20 years.

Wasnt there a Harris amp that took a 100mw input and wound up with an 8122 to drive the final stage?

Carl

Much will depend on how much current the transfomer can supply under full load and how hard he pushes the 3CX-3000.   My power transformer is fair and I ocassionally push things hard, so have about 110 uf in my supply.  Still, it sags a little and detracts from IMD figures.   I feel we can never have enuff filter capacitance if it is available, especially to smooth out dynamic, syllabic fluctuations as generated by ssb.

I prefer capacitor-only filtering for my own use here.  I have a big fixed choke that gets switched in, but never use it except when testing a new rig.  Maybe a swinging choke would suit me better.


T

Big cap + little transformer = smoke

Hi,

I have been reading about HV power supply chokes, WOW, that gets into some esoteric territory!
So far I seem to understand that if I use the choke, I need to know if it is a "swinging" type or non-swinging.

It also seems that there can be quite a challenge to match the choke and capacitors so as to obtain the correct resonance, apparently somewhere at or slightly over 120Hz.

Can you tell anything from the photos as to which design GATES choke is?

The capacitors might go w/ the choke, it looked like Fair Radio parted out a GATES,
and I bought the choke and caps from them at the same time (appeared on their site simultaneously)

(http://i180.photobucket.com/albums/x257/fish1_07/am/power_supply_components2_zps4c9b2229.jpg)

(http://i180.photobucket.com/albums/x257/fish1_07/am/gates_choke1_zpsfc50fe8b.jpg)

(http://i180.photobucket.com/albums/x257/fish1_07/am/plate_transformer_info_zps7fcd1511.jpg)

73
Frank
KJ4OLL

That is a straight smoothing choke, no resonating cap needed. The resonated choke was in the ham and lower commercial tiers such as Henry as a way to save money and weight.

Carl

You may want a swinging choke if you power both the mod & RF sections from the same supply.  It is gapped little or none at all, such that the inductance varies inversely proportional to the current drawn, such as in the case of a class 'B' modulator.

With it in line it gives more smoothing when you need it for quiet passages, less for when things are closer to the edge in terms of modulation percentage.  This also lessens the voltage 'droop' caused by increasing heavy mod demand.

The lineup traditionally is raw B+ to a swinging choke, first cap, (this is where the mod section typically gets it's B+), to the smoothing choke, then another cap.  The RF amp demand hooks here.

Smoothing chokes have a more pronounced gap, so they exhibit little change in inductance when current varies within design limits.

The above is seldom used today, as bigger caps are available to keep things in line when modulation is heavy, but the old way was best for fidelity and low distortion requirements of BC rigs.

73DG

Sounds like he's building a linear, so there won't be any modulator, but a plate modulated 3CX3000A7 would definitely be cool. Create your own band openings with that much power.  ;D

Looking at the calender it appears to say 2013. Where is anyone going to find a swinging choke suitable for single phase choke input for a decent price?

That Dahl xfmr is going to be low resistance secondary and should work without a choke. Swinging choke supplies are popular with AM folks because of literature which says they are better for regulation and necessary for class B modulators. They also make things easier on the power transformer. Older high resistance transformers can benefit. That's not what is being used here.

Since this amplifier is to be used on AM it should be just fine without parts that are very difficult to obtain. Capacitor input supplies have been the norm for amateur power amplifiers for a lifetime.

A step start should be used.

A 30L-1 will work in AM service if you ditch the 811s and install 572Bs. I'd stay away from rocket science suggestions like only putting (3) 572Bs in. Those tubes already have a relatively poor filament emission/plate dissipation ratio. 4 work fine. Even the Asian tubes will work. I had one in service on AM for a year before selling it to a friend. 125 watts carrier and no more.
Some late Asian tubes have a bayonet locator tang which is in the way when plugging into a 30L-1. One tube socket screw must be removed in order to temporarily re-clock the socket about a dozen degrees. It can be put back after plugging the tube in.

The 30S-1 makes a FB am linear amplifier especially with a few tweeks. The screen power supply modification which uses a 125 volt isolation transformer in series with the original screen winding really wakes the amplifier up. The 4CX1500B fits right in for more plate dissipation but the factory blower will not deliver enough air for AM. The easy way out is to get a commercial transmitter cabinet exhaust fan and mock up an air handler box. Fit the box over the existing hot air exhaust so the air is sucked out. This will keep the tube nice and cool. If you keep the amp near your mike people will think you already have the 3CX3000 : ) The above is tried and true. Your amp should not wilt at legal limit AM. Remember to keep the air inlets and tube clean. The air inlets are located in the cabinet recesses in the pedestal. 

If you like to waste power run your +20 dBm exciter into the 30L-1 and then the 30S-1. Output will land up with 5 or 10 watts from the pipe. I did this on the AM QRP net once.   

Interesting you should mention the 30L-1, mine has 4 Svetlana 572B's and the Young Kim power supply upgrade,
plus I did all the mods that Dennis Brothers said to do.

But I have not used it for AM as the 32V-2 I normally use has almost 100 watts of carrier.

Does the additional 25 watts provided by the 30L-1 result in significantly better signal reports? 

No, 100 watts to 125 watts will not even be discernible.  We generally need at least X4 power, which equals about 6 dB, to make it worthwhile in the real world.  This is about like comparing a Yagi to a dipole at the same height. (5 dB)

X2 = 3 dB
X4= 6 dB
X10 = 10 dB


The only difference may be that the low level exciter and 30L-1 may be able to generate a cleaner signal than the 32V-2.   Run some IMD and THD tests to see.


The difference between 100 watts and what the 3CX-3000 can put out is quite meaningful and worth the effort.

T

Are you using any software to analyze your network? If so, let us know what you are using.

Tnx.

Philip
KA4KOE

Software used so far:

I downloaded this amplifier design software:
http://www.duncanamps.com/psud2/index.html

Took a while to find a system to run it on @ my QTH, (all my computers are LINUX),
but I installed VirtualBox to run windoze7 under LINUX.

Now I have to figure out how to option the psud2 program configuration to describe the variables in my amp setup, which are very different from the psud2 program defaults.

I also used "match.exe" from the ARRL handbook CD to calculate the capacitance and inductance values for the tank circuits.
Am about 30% done on those calculations.
Need son: "Mr. math wizard" to come home from engineering college (UF) to check my work.

This is about as far as I have delved into the HAM software arena.

I don't have any software to analyze the antennas, I just use an MFJ 259B to get the tuning close for each band.
Then I use an AGILENT spectrum analyzer w/ tracking generator and a return loss bridge to get the tuning exactly where I want it.

Generally the extra 25 watts carrier will not change on the air results. The 32V-2 most likely already has a good percentage of relatively clean modulation.

There are other circumstances where the use of an amplifier such as the 30L-1 will make a difference. One would be use with a Johnson Ranger.

Yes the Ranger will make almost half what the 30L-1 will do but it isn't all that simple. The Ranger's modulator is out of suds when used at full power. It is running into gain compression.

When configured with the low power mod which leaves the modulator on high B+ but only asks it to make about 1/4 of it's design power, gain compression is no longer an issue.

Take this and pipe it into the 30L-1 and look at peak power verses the originally configured Ranger. The increase will be a little under 6 dB which now makes it a worthwhile increase.

The same could be said for a few other transmitters whose modulators remind me of an early Chrysler Hemi with a tiny carburetor.

Those old calculations in ARRL publications which show AM is useless are skewed to promote that other phone mode. You won't find that junk in the Orr handbooks.

MMANA GAL is a great free antenna modeling program. You just have to input the proper gozeintas to get correct gozeouttas.

Philip

All you ever need to know for amp design in one place.

http://www.ifwtech.co.uk/g3sek/

Thanks for the link!
It looks like I will have to become a member of the Radio Society of Great Britain
to gain access to some of the content I need on the site.

Send an email to Ian, he may help.

Carl

Hi,
Made some progress this weekend - rectifier is coming together!

(It might be a rectifier, or possibly some type of incendiary device....time will tell)

I copied a DELCON T210 design.

- Used drafting tools to lay out the pins on a sheet of fiberglass.
- Made a simple fixture for directing the swaging forces on to the pin shoulder to avoid bending them.
- Installed components, almost ready to apply some voltage!

(http://i180.photobucket.com/albums/x257/fish1_07/am/rectifier/fixture_tools_pinned_board_zps90c64526.jpg)
(http://i180.photobucket.com/albums/x257/fish1_07/am/rectifier/soldered_rectifier_components_zps53769112.jpg)

73
Frank

The wires at the bottom of the picture seem to be a bit thin.  You might want to use wire with heavier insulation as there is high voltages at those points, except for the ground point.

Fred

Thats a neat rectumfire board you made there.  Can I ask where you got the pins? 


C

I looked up the specs on the white wire - insulation is rated at 5KV, but it is only AWG-15, so I'll see if I can find
something better today @ Skycraft - that is where I found the pins too.
The red wire is AWG-13 / 20KV.

http://www.skycraftsurplus.com/

Frank

Even #20 wire is fine for short runs considering that the average 3-4 KW plate choke has several feet of #24-26 wire.  If I want more insulation capability I slip a Teflon sleeve over it which is dirt cheap at the local electronics surplus shop

All the thick wire gets you is that its often harder to dress neatly.

Carl

Just dont use it for your antenna  :o

Hi,
More progress, as it is raining so much here that I am unable to do yard work,
and thus am forced to putter about in the workshop.

Using a piece of 1/4" steel plate to bolt the main PS components to.

The idea is to produce a modular unit that can be inserted into the cabinet w/ a small crane.

Laying out the components on the plate so I can drill and tap the plate for the fastners:

(http://i180.photobucket.com/albums/x257/fish1_07/am/positioning_components_on_ps_mount_zps3858993d.jpg)

Facing the 1" Teflon rods to square them up for bolting to the plate and the rectifier:

(http://i180.photobucket.com/albums/x257/fish1_07/am/facing_rectifier-standoff_zps9bba4540.jpg)

Drilling for 3/8x16 tap:

(http://i180.photobucket.com/albums/x257/fish1_07/am/center_drill_rectifier_standoff_zps356035c8.jpg)

Cutting internal threads in the teflon:

(http://i180.photobucket.com/albums/x257/fish1_07/am/cutting_threads_rectifier_standoff_zps3c1bb763.jpg)

Mounting plate drilled, tapped and painted:

(http://i180.photobucket.com/albums/x257/fish1_07/am/power_supply_mount_plate_zps437c1896.jpg)

All components bolted in place:

(http://i180.photobucket.com/albums/x257/fish1_07/am/bolted_to_plate_zps1af4e6c7.jpg)

Still need to add the 100K 225W bleeder resistors, I think they will fit nicely on the rectifier standoff legs.

Frank
(Hand still sore from all the tapping - the plate is crappy, high-carbon steel,
many hard spots tried to break my taps....)

You can move to the head of the class now ;).

73DG

Hi,
A little more progress on the homebrew!

Trying to make a filament choke for 160M.

Need 180uH for this choke.

I read that it might be possible by using toriods, thus obtained some type 43 material and did a test.

23.6uH for the first test with about 16" of toroids.


TYPE 43 TOROID TEST
(http://i180.photobucket.com/albums/x257/fish1_07/am/filament/c8b8ac02-9ac9-4153-9c63-b436a42d8cd4_zps69bb05c4.jpg)



This choke would have to be too long to fit in the cabinet using this method to obtain 180uH, so on to plan "B".


Obtained some type 43 material 5/8" x 7" rods, put 7 of them in some heat shrink tubing, wound 30' of #12 varnish
coated wire on the lathe.

TYPE 43 CORES TEST
BIFILAR WINDING
(http://i180.photobucket.com/albums/x257/fish1_07/am/filament/30_ft_no12_wound_on_lathe_zpsb62e515c.jpg)

TEST CHOKE
(http://i180.photobucket.com/albums/x257/fish1_07/am/filament/test_filament_choke_zps8887e093.jpg)

SEVEN CORES
(http://i180.photobucket.com/albums/x257/fish1_07/am/filament/seven_type43_ferrite_seven_inch_long_test_choke_no_2_zps616de04b.jpg)

NOW WE ARE GETTING SOMEWHERE
(http://i180.photobucket.com/albums/x257/fish1_07/am/filament/more_than_half_way_to_180_zps44e6c6b6.jpg)

Based on these results, I am going to make a 14 (seven in each bundle) core choke and use a longer piece of #12

73
Frank

I think toroids usually refer to donuts, you're using rods.  I think you may get more inductance using a toroid core but it may be tough to wind #12 on it.

But, keep trying you're making progress. You're up to 98uh, another 80uh and you're there.

Fred

You really dont need that much L in the choke. With a much lower amount just reduce the input cap to compensate and part of the choke will be part of the circuit.
I use a 33 mix rod, 1/2" diameter and 7.5" long filled with bifilar #12 for the typical pair of 3-500's which has been in all the HB's for decades. For the 3X3 you need 50A wire capability so use 2 seperate rods with 2 #12 in parallel on each.

Using the plate choke as part of the network is also common in the output pi net and is how many commercial amps got by with 90-110uH on 160 or 50uH on amps that stopped at 80M. It also reduces the maximum C needed in the Tune cap which equals smaller size and less money.

Carl

Thanks very much for the tips and ideas Carl!

Hi,
Made a little more progress on the amp components!

POWER SUPPLY
I have never worked with plate transformers before, is it typical for one to make this much racket?

http://www.youtube.com/watch?v=GmAQbfU8U1E

FILAMENT TRANSFORMER
Getting very close to the output voltage needed from the new secondary winding that I made for the filament transformer.
Need about .3 to .5 of a volt additional, so will be rewinding the core again, when I obtain more enameled magnet wire.

http://www.youtube.com/watch?v=zxV4qEl4Cwo

73
Frank
KJ4OLL

Frank,

   This rig of yours looks serious!

The video of that power transformer being ramped up might be a concern. That 60 hz buzz might just be some loose lamination(s) that vibrate as the core flux density rises, or it could be indicative of core saturation. If it is core saturation, the primary excitation current will rise very quickly after a point. It would be normal to have the current rise pretty much linearly with the applied voltage. If it is linear up to say 200 volts, and then goes up very quickly from there as you hear the buzz, then you have core saturation. This would be bad, and has to be avoided.

The filament transformer looks like it took a bunch of work. Adding more secondary turns at this point could mean almost starting over as you might have to remove the entire secondary winding.  :'(
One thought about the transformer. If the primary had less turns, then the secondary voltage will rise. One way to remove turns from the primary would be to add more primary turns over the secondary, and phase the new turns to 'buck' the original primary turns. Then you have to worry about core saturation.  ???

Jim
WD5JKO

How about using a Variac that can be configured for overstep to make up the last filament voltage fraction?

You need about 5% step up which is not too bad....

Also, with ther Variac you can ramp up the fil on start up.


Another possibility - if the fil xfmr is 120V primary, then a 6.3 fil transformer as an added boost in series with the primary might do the trick.

T

Hi,
If I put an amp meter on the input and look for non-linear current draw, sounds like I should see if the core is saturating - I'll try that.

I was thinking the same thing about the variac - the ancient Gen Rad only goes up to 135, but I have a Powerstat that goes to 140, I'll test it and see what effect it has on the secondary.

I have rewound the filament transformer secondary many times to get this far, it is not a big deal to do it again, except not much room remaining for many more turns.

The filament transformer secondary is wound with four bonded strands of #12 to get the current capacity, but possible less conductors would work.

I found a chart w/ Google that provides information about capacities:
- four strands of #12 gives about 245A for "chassis wiring" and 150A for "power transmission".

Possibly rewind it with one strand of #8 instead?
The current rating for a short run of #8 is 73A for "chassis wiring" , and 24 for "power transmission"
(51A needed for the EIMAC filament).

Thanks for the comments and ideas!
73
Frank
KJ4OLL


 

Frank,

  I suggest you refer to some of the information from Patrick Turner:

http://www.turneraudio.com.au/powertranschokes.htm

 It appears that you intended a turns ratio of 120/7.5 or 16, but due to sag you are getting closer to 19 (135/7). This must be due to a combination of copper loss, and transformer leakage inductance. If you rewind, perhaps split up the primary such that half of it is under the secondary, and the other half is over the secondary. This should increase the coupling since the leakage inductance will be reduced.

How many turns per volt? If it is over 2 Turns/Volt, perhaps you are short on core area?

So as a guess maybe you have 80 primary turns, and 5 secondary turns (1.5 T/V). If you were able to snake in 8 more primary turns (10% more) into your existing transformer without taking it all apart, and then phase it to 'buck' the original primary turns, then you in effect will have 72 primary turns. If so, then the turns ratio will 72/5 or 14.4. Take 120/14.4 and the secondary voltage would be 8.33V (unloaded). This approach would  increase the turns per volt up to 1.67..might be closer to saturation.

Jim
WD5JKO

Hi,
I put an amp meter on the primary of the plate transformer, can't seen to see any excessive current spike as I turn up the variac.
So hope it is not saturating, just plate vibration.

At maximum 240VAC in to the plate transformer is sinking 1.1A.

I was expecting more DC voltage output from the rectifier.
But possibly having a 10H choke and the 100k bleeders has some significant effect on the rectifier output voltage, loading it down.

I have read that 1.414 x the AC voltage in should yield the DC out of a full wave rectifier. Probably unloaded?

But the readings I get are only about 1.14 times the AC input.


POWER SUPPLY TEST
240vac in to 240 primary tap
secondary 2600 volt tap measures 2760VAC to the rectifier, 3100 VDC out of rectifier

240vac in to  220 primary tap (The transformer does not seem to like this, very noisy, just did it for a test)
secondary 2600 volt tap measures 3000vac to the rectifier, 3400 VDC out of rectifier

240vac in to 240 primary  tap
secondary 4000 volt tap measures 4400vac to the rectifier, 4900 VDC out of rectifier

73
Frank
KJ4OLL

If the audio recording was accurate, it sure sounded like a mechanical buzzing rather than a growing hum. Mine hums, but does not sharply buzz, if the input volts (115 rating) gets above 130 or so via the 'variac'.

That transformer is a candidate for a dual variac of that size. It is the 6.5KVA size, no?

The variac I am using for the power supply is a 7.8kva, 1 phase,
240-280 28 amp.
The plate transformer makes noise at full voltage, but in the video, the noise is amplified by the big metal cabinet, so it sounds louder than it would be in open air.
I found that the balancing resistors on the rectifier get hot, so I'm going to install a cooling fan, then let the PS run for a while, see if the plate transformer gets hot.
If the plate transformer does not heat up much, I hope that means the core is not saturated, and the noise is just plates vibrating.


The variac used for the filament transformer is a 1.3kva, 1 phase
135v, 10 amp. No noise here, but at 135v on the primary, the core might saturate.

Will try to rewind the secondary so that a lower primary voltage yields the required secondary 7.5 vac @ 51A.

I looked over the information from the Patrick Turner web pages, WOW, that guy covers every single detail!

73
Frank
KJ4OLL

That's what you get with a choke input filter with a load sufficient to activate the choke properly.  Just about any ARRL handbook or similar reference will give you the "DC out vs. AC RMS in" formula for a choke input filter:

The handbooks will also give you a chart showing the effect of increasing capacitance in a capacitor input filter (no choke), the output approaching 1.35 DC out/AC RMS in, at very light loading resistance and very large filter capacitance and dropping to 0.75 with very heavy load and little capacitance.

Also there are neat charts showing approximate wattage vs. core cross sectional area from a typical E/I laminated transformer.   A 500 watt transformer needs about four square inches, i.e., 2 x 2 in. 

..and tons of other neat stuff.  Oh, even though you have only 100k bleeder, the class A current of driver amplifiers plus the full carrier current of the class C final in an AM transmitter will sufficiently load the choke such that you get only about 1.1 DC/AC rms output.  CW operation, however, is another story.

Thanks for the explanations Rick!
I have a new ARRL handbook, but it is intimidating, vast amounts of information, and I am new to homebrew!!!!
73
Frank
KJ4OLL 

How about trying some rubber donuts to set the trans. on? Sort of like a compressor in the condensing unit out back. The piece maybe sits on 1" thick rubber donuts, and on top of its flanges are half -compressed conical springs, held in place by the bolt heads. Just a thought because that hum might still be annoying after you deal with its reduction, if the cabinet amplifies it so much. I like the blower there, saw only part of it.

Hi,
I found some dense rubber at the local surplus place (Skycraft) and made shims to go between the
plate transformer and the chassis, that helped the vibration noise, thanks for the idea!

Last weekend my son (KJ4WLW) and I worked on the plenum box for the blower/filament transformer/choke/PA mount.

(http://i180.photobucket.com/albums/x257/fish1_07/welding_and_forming/DSC00042_zpsb7c7c4e7.jpg)
(http://i180.photobucket.com/albums/x257/fish1_07/welding_and_forming/DSC00045_zpsdde5078f.jpg)


It was fun to be able to mock-up where the plenum will eventually fit in the amp, and imagine it
operating someday!

(http://i180.photobucket.com/albums/x257/fish1_07/welding_and_forming/DSC00048_zps43ab595b.jpg)

Worked on the PA mount assembly this weekend:

(http://i180.photobucket.com/albums/x257/fish1_07/am/3CX3000A7_socket/ready_to_flow_zps20187798.jpg)

(http://i180.photobucket.com/albums/x257/fish1_07/am/3CX3000A7_socket/pa_assembly_ready_zps29421362.jpg)

Next is finishing the plenum and mounting the blower.

73
Frank
KJ4OLL

Hi,
It has been a while since I updated the AM amplifier project.

Was "out of commission" for a couple months w/ medical stuff, all back to abby-normal now!

Was able to finish all the TIG welding on the plenum, ugly but at least I did not destroy it (I hate welding Aluminum)

Mounted the PA assy:

(http://i180.photobucket.com/albums/x257/fish1_07/am/PA_mounted_zpsd68b89f4.jpg)

And started working on mounting the tank circuit:

(http://i180.photobucket.com/albums/x257/fish1_07/am/tank_circuit_zps60b204dc.jpg)

Hope to get the filament transformer and choke mounted in the plenum next, too hot today, I am beat!

73
Frank

Hi,
More progress, will soon be able to power up for the first time and grab fire extinguisher.......

(http://i180.photobucket.com/albums/x257/fish1_07/am/band_switch_tuning_loading_controls_zps468b596a.jpg)
(http://i180.photobucket.com/albums/x257/fish1_07/am/tank_circuit_installed_zps1a37d47a.jpg)
(http://i180.photobucket.com/albums/x257/fish1_07/am/control_board_12_13_13_zps317fca74.jpg)

I have no idea what to drive this thing with!!

Was going to use the 32V-2 as exciter.
But many said that would be a waste, suggested a Flex.

After reading:
http://amfone.net/Amforum/index.php?topic=35503.0

Decided Flex might not be the best for the clean  3CX3000A7.

Can't I just build something as a simple exciter w/ good numbers, or is that impossible for homebrew?
 
All the 3CX3000 needs is about 25-30 watts in to make a nice big carrier.
Thanks,
Frank
KJ4OLL

The 32V looks great in the photos and already expects to be the exciter. Using anything else would upset it.

That TX can be setup to run at lower power out. There are plenty of connections available for this without hacking the rig.

The best method would be something that allows the modulator to run at full plate voltage while reducing the final B+ to about half. That gives 1/4 output power in a class C final. Leaving the modulator at full B+ and reducing mic gain should provide more modulator dynamic headroom and make for less audio distortion. Some of that distortion lands up as unwanted RF bandwidth. Every effort should be made for a clean signal when high power is involved. Simply padding the RF of the Collins is a bad idea. This is known as the Timtron Low Power Mod. I learned it at The Timtron Instutute of Technology.

Provision should be made for switching to normal power level out of the 32V for final touch up of the linear amplifier. It will need steady carrier tweeking at peak power which will be at least 4 times the carrier power out the pipe.

There are lots of ways to go.
The 32v is one of the best sounding stock radios.

I always disliked running a tube rig into an amp though, the solid state stuff makes it easy to tune up, just adjust the power level out, switch modes, no drift, digital readouts, but I have no idea which ones are clean and good sounding.

The flex 5000 can have very good audio out, but has the hash.
A ranger can sound very good and do lower power out, or you can make any tube rig work well with an outboard modulator,  or you could home brew a 6146 rf deck and a hifi modulator for it.

RF wise, the 32v is going to be clean, no digital hash or mixer signals getting out, and for real hifi, driving the grids with an audio amp into the reverse transformer will be very clean if the modulator is run at full voltage and the rf is run lower.
At 30 watts out, the 807's would likely be in AB1 at the power levels needed to get plenty of modulation.

I have an outboard variac on mine, as I use it as an RF exciter for my 813 rig.
Easy to do, and the 32v has most stuff on the rear terminal strips.

All you have to do is make a variac supply for the RF section and feed it into the rear terminal strip.

FT102's used to be all the rage, and they have mods to make them very hifi, all mode, but you need to dip and load...

 

Frank
Looking thru your posts and the great pictures. Congrats! I hope it flies and makes nice powerful RF. Some people on this board just have the gift of building and make neat stuff.
The output P.E.P. and CW carrier capabilities will be interesting.
I do not remember seeing what your expectations are for this amp.
Fred

Hi Fred,
None of the old store-bought amps I have do 160, and also are not designed for AM.
So I am hoping to use the homebrew for AM on 160 and 80, possibly 40.
And SSB on 160.
And to heat up the shack;)
Power out will conform to rules & regs.

73
Frank

Hi,
Almost finished with the control board!

There are three steps to power the creature up:

1- Power on the blower, which activates the vane switch.
   If the vane switch "makes", the next stage is available for power up.

2- Power up the Filament stage.
   If the filament powers up OK, voltage is available for the step start phase.

3- Power up the step-start phase, bringing the plate transformer & HV on line.

I finished "Phase-2" today, CLACK is working for the filament.

So This part of the project will be done pretty soon.

The next project will be the PTT circuit and the "dead man" interlock.

I have a question - what technology/design should I be looking for in the parts bins @ the Orlando Hamfest next week in terms of a PTT antenna relay?

Thanks, & 73
Frank
KJ4OLL

Very nice metal work!

N2DTS,
Thanks, been doing various metal bashing hobbies for 50+ years - now trying to learn more about amateur radio electronics - really fun technologies to experiment with!
73
KJ4OLL

Well, just be careful with the high voltage.

Hi,

Made some more progress on the amp!

Finally am at a point where something can be powered up, even if it is only the blower circuit...took a while just
to get to this point........but momentum is necessary, even if it is in very small increments.

www.youtube.com/watch?v=jcsdbGfauLw


Going to try some .004 thick Titanium foil to increase the "sail area" of the vane switch.
But have to keep it far enough away to avoid arcing.
(Not sure how far that should be @ 5000v on the plate???)

73
Frank
KJ4OLL

This is a really great project! I hope you will take more pictures and publish all the schematics so I can learn from them. The tank circuit -those are some serious looking parts.
I have been collecting stuff too for a 3CX3000A amp. Have most of it. My transformer's only rated 1.6A but it ought be OK.

"The 3CX-3000A7 is a FB tube making -40 to -45dB 3rd order IMD figures. "
Is that figure for GG service or grid driven? It's not clear in the datasheet I have. The reason for asking is that grid driven service is nice because only 150W is required for max QRO, so that a quite small exciter of 35W PEP will do for ham limits, while 400W is required for QRO GG service and only a 100W PEP exciter is needed or ham limits. Or is it a silly question?

"Does the 3CX3000 have handles??? That's where it's happening these days."
No handles but, well anything more than a hand-full is hard to manage! 3CX3000 is just the right size, even if it has to be wiggled a bit to loosen it up

"..a plate modulated 3CX3000A7 would definitely be cool. Create your own band openings with that much power."
Create your own ionosphere opening too!

8-) enjoying this topic!

Yes, pictures please. Schematics  too!
  I have a 3cx3000a7 project that has stalled,   this may be   the what  I  need  to get back to building-thanks.
73,  Ed

Hi,
Thanks for the encouragement!
I have posted many photos on QRZ , look under KJ4OLL.
(Not sure if I should post many photos here, costs the site storage $$$)

I'm no engineer, so this project would not be possible without help from many.

Ken, W2DTC posted most of the core technical information.
I am copying most of that Ken did.
http://w2dtc.com/3cx3000f7.htm

Don, K4KYV gave me a super T/R relay sequencing design.

A retired broadcast engineer I work with, Joe Lang, designed the control circuits, power supply, tuner.
As I see Joe every day at work, it is easy to review new ideas and plans.

For instance, when Joe viewed the recent video of the blower test, and saw that the vane switch did not work
as I planned, he gently informed me:
"IDIOT - put the vane switch directly in front of the blower outlet inside the plenum"

Mark, N2ZAB, gave me more ideas and tips.
And of course, this forum has provided much more guidance and advice!

The amp is a Class AB2, Cathode Driven, as shown in the specs here:
http://www.cpii.com/docs/datasheets/78/3CX3000A7.pdf

Joe's control circuit design is really superb, I will post the schematics someday.

However, as the control circuit schematics were drawn up freehand, then the inevitable corrections were added,
and now it looks like....

Well, imagine you had a live chicken.
You put ink on the bottom of the animal's feet.
Simultaneously administer a large dose of LSD, and also energize electrical terminals installed in the brain.

The resulting marks on the paper would be the current schematic appearance......

73
Frank
KJ4OLL

Hi,
Too many other things keeping me away from this project!
(Trying to bring a 30S-1 back to life has been a challenge)

Found time to get the vane switch moved into the plenum box, so now it receives airflow
direct from the blower outlet, vane switch now "makes" the circuit properly.

https://www.youtube.com/watch?v=wdkBEc9HHeM

Also installed the Manometer!

https://www.youtube.com/watch?v=MNo1uH9-q38

73
Frank
KJ4OLL

Hi,
I'm stuck with a weird airflow problem that I'm sure Kelly Johnson (aeronautical engineer) would have known the answer to.
(His fellow engineers said that Kelly could see air, as he was so adept at pointing out flow problems and solutions)

The vane switch is the first safety switch in the control circuit of this amp.

The vane switch is positioned in front of the blower outlet, inside the plenum that contains the filament transformer, choke,
PA base, etc.

The amplifier is started in steps.

Operating the first switch (step-1) turns on the blower, and via the vane switch, makes +24VDC available for the second stage to start.

If the blower fails for any reason, air flow to the vane switch decreases, the vane switch opens the circuit, dropping all the relays and contactors.

Operating switch two (step-2) pulls in the filament contactor and starts heating the filament w/ about 53 amps of 7.5 VAC.
This step also makes +24VDC available for the plate circuit.

Operating switch three (step-3) pulls in the first step-start contactor, beginning flow of AC to the HV power supply, charging the capacitors.
A timer expires and the final contactor activates, providing direct 240VAC line to the plate transformer primary, plate is
now at 4900 VDC.

You can hear each "clunk" behind the blower noise in the video, and see the stupid vane oscillating:

https://www.youtube.com/watch?v=87HIDXmHSvE

I have been experimenting with this all day.
I can get the vane into a position and angle so as to stop oscillating with the plenum open.

Found a vast number of methods that are completely ineffective in stopping the vane oscillation with the plenum sealed up.

There is a vane position and angle that stopped the oscillations, but when I put the plenum cover back on, the pressure and flow inside the plenum changes, and the vane starts oscillating again.
All the relays and contactors drop out.

When I put the plenum cover on, something happens to the airflow inside, and the vane starts oscillating (even though the vane was stable w/ the plenum cover off), causing the vane switch contacts
to open and drop the Filament circuit. (And everything else)

(http://i180.photobucket.com/albums/x257/fish1_07/am/blower/plenumcover_zps30022ea7.jpg)

So there must be some kind of standing wave or pulsing that starts when it is all sealed up.

Should I try a baffle to block or redirect airflow near the vane?

Or a little duct inside the plenum that focuses pressure directly on the vane?

Or some better method of performing the "blower on/off" logic for the subsequent stages?

Or just buy an ALPHA and call it a day?

73
Frank
KJ4OLL

Where the blower mounts,  it would appear the plenum is not airtight.

If that's the case,  there is your problem.   The vane needs to be in the only exhaust,  otherwise you can see all kinds of problems.
--Shane
KD6VXI

Hi,
The blower and plenum are sealed up pretty tight, Manometer reads .7 WC.

Recent info from a guy in the commercial HVAC biz indicates that vane switches are generally a pain to get working and to keep working,
so I am changing the design.

Deleting the vane switch, replacing the function with a "DWYER PRESSURE SWITCH, DIFFERENTIAL"

This is a pressure diaphram that acts on an included electrical switch.

Not dependent on direct airflow, operates like the manometer, just responds to pressure in the plenum.

73
Frank
KJ4OLL

Frank,

I agree that the 30L-1 isn't a good choice for AM: it uses 811's.

IIRC, the 30S-1 uses a 4CX-1000A, which runs Class AB₁, so I don't understand why it wouldn't be appropriate for AM. Is there an issue with the power supply?

73,

Bill, W1AC

Hi Bill,
You are correct, the 30S-1 works fine on AM, so long as I do my part to keep within it's capabilities.
I had read the 30S-1 plate transformer was marginal, and mine was the original from 1959-60 (amp S/N 194).

But after the recent:
- filament transformer arcing (replaced w/ new).
- brittle and corroded HV leads falling off the plate transformer (replaced w/ new)
- Screen Supply rectifier leaking AC upstream (replaced all the "tophat" diodes)

Now the 30S-1 is back to abby-normal, and when driven with the Flex 3000 in AM mode, seems to be happy at 150-200 watts.

73
Frank
KJ4OLL

Hi,

Finally got all the control circuit pushbutton switches installed and connected.
(spared no expanse on the labels.....)

(http://i180.photobucket.com/albums/x257/fish1_07/am/schematics/frnt_pnl_switches_zps6a13641f.jpg)

The blower & filament circuit is working! (filament = 58 amps of 7.5 volts)

(replaced the goofy vane switch with a Dwyer 1910-1 Differential Pressure Switch - problem solved)

(thanks to Mike WA8HNS for the idea!!!!)

https://www.youtube.com/watch?v=mW4yR21eYvg

Plate & HV next.

I have been told that because the EIMAC 3CX3000A7 has been sitting for so many years, the tube
must be run through this cycle before applying full B+:

1- turn on blower & filament
2- bring up about 800 volts of B+
3- run it for about a day to let the getter work.

After this, the plate should accept full B+ w/o arcing.

One lesson learned so far is that this thing puts out lots of noise & heat.
Probably going to have to roll it outside of the shack when operating!

73
Frank
KJ4OLL

Looking back at that project, the intake duct on the blower is possibly presenting a higher than necessary impedance for the amount and velocity of the flow.

An air horn on the blower open to the interior of the cabinet would probably help. Or just nothing. I'd ditch the duct on the intake, and put one on the tube if necessary, or divide the cabinet.

Hi,
Did not see any difference in Manometer pressure w/ or w/o the duct in place.
The amp gets really hot inside the cabinet, and the duct brings cool air in to the plenum from under the cabinet.

Made a baffle to fit inside the plenum so as to direct air on to the filament choke, it was operating at 160-170 degrees,
due to the voltage drop across it.

Amp is starting to look like it might actually work someday soon!
https://www.youtube.com/watch?v=hdcv2YPhMhI

Next is to build the Cathode drive circuit, then the PTT/timing circuit.
Finally, a tuner will be built into the amp cabinet, using these components:
(Except the water bottle...)

(http://i180.photobucket.com/albums/x257/fish1_07/am/amp_tuner_parts_zpsfad87544.jpg)

73
Frank
KJ4OLL

I wondered what the extra space might be used for. A matcher is perfect. I like the use of a VFD for the blower. No doubt the falling price of used VFDs will be increasing the price of 3 phase blowers including 400Hz ones.

I'm working with a lot of ABB VFDs lately.

http://www.youtube.com/watch?v=boq8JMIRNqc#t=20

Will the matcher be a pi network type circuit? Hopefully a low pass design. hehe next video, show that HV cranking up to 4800!! 

Good job on this progress. I like the amplifier's tank circuit component size. It looks like an assembly, but I can not tell. Can you provide more info on that very interesting area?

How about a Pi-L,  and use the roller for the L.   That would give good harmonic suppression as well as allow a large range of impedances to be matched.

Saves the cost of a cap,  too.

--Shane
KD6VXI

WOW, pretty wild use of VFD's!!!

Exactly correct on the 3 phase - an old, beat up single Phase Rotron for $300 or a NIB 3 phase Rotron for $100, plus $50 for the VFD.
Seemed like a simple decision.


The tuner can be configured either "Pi" or "L", depending on jumpers. (Kind of like the old Collins 180S-1)

(http://i180.photobucket.com/albums/x257/fish1_07/am/schematics/tuner_schematic_zps28fd8c70.jpg)


The tank circuit is a Collins 185J-1.

(http://i180.photobucket.com/albums/x257/fish1_07/am/schematics/collins_tank_circuit_dec_18_2013_zps8673ea4f.jpg)

When I bought the thing, Google did not know anything about it.

I took a chance, as it was cheap.

Planned to strip it for parts for the amp project.

When my elmer looked at it, he said ""do not canibalize it - this is a complete tank circuit, ready to plug & play in your amp"

Collins forum folks were also baffled, but there was one 1940's similar design that was auto-tune that they did locate.

Looks much the same, but with three servo motors on the front.

I suspect this one was originally auto-tune, a previous owner fabricated the front panel and added manual-tune knobs.
 
http://s180.photobucket.com/user/fish1_07/library/collins_mystery_tuner?sort=3&page=1

The really interesting thing is that it operates at B+ - thus the PA direct connection.
Here is the tank at 4019 VDC, using a current-limited HP 4kv power supply.

(http://i180.photobucket.com/albums/x257/fish1_07/am/hv/4kv_on_tank_test_zps71854ccf.jpg)


I'm running the plate transformer secondary on the first jumper position 2.6kv.
If that does not produce adequate power from the exciter drive, the next jumper is 4kv.
If that does not do it, the final option is 4.8kv.
I fear it.

73
Frank
KJ4OLL

No worries! the plate-direct-to-coil thing is or was common on some units. What voltage did the 185J-1 run at?

I agree going to 4KV and higher is scary but it may be needed to try different voltages to get the best performance or lowest distortion.
A wise man advised me to use a 3" tall loop of #30 wire as a fast acting fuse in the HV line. If there is an arc or flashover, only the wire is vaporized.

Those shafts are going to be hot hot hot.. Will you use a long insulated section and a grounded knob-shaft at the front or some other scheme? Things like that worry me when the voltage is so much. I mean, that it could some how get across the insulated shaft section.