[AMRadio] Re: How can we rid the HF bands of AM ?

Patrick Jankowiak recycler at swbell.net
Fri Aug 12 00:11:48 EDT 2005

> From: Geoff <w5omr at satx.rr.com>
> Subject: Re: [AMRadio] Re: How can we rid the HF bands of AM ?
> To: Discussion of AM Radio <amradio at mailman.qth.net>
> Message-ID: <42FAC124.3040806 at satx.rr.com>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
> Patrick Jankowiak wrote:
>>Hey Geoff,
>>You are right, I won't be totin' the Tucker KW or even the BTA-250.. 
>>For me the AM is a hobby and a love for doing things in one of the old 
>>and respected ways. No General yet, maybe soon. I have alot of fun 
>>just fiddling with them.
> You like fiddlin', and that's fine.  But, for me (and I think some 
> others here agree) there's nothing finer than saying "rig here is 
> homebrew" and getting that "man, that's a good sounding rig" report in 
> return.  The feeling is indescribable.

I can relate to that. First time we cathode-modulated Dennis W5FRS's 
homebrew 250TH amp with an Altec 1570B tube amp, we were ecstatic over 
the reports. I think the one that got me was "sounds just like a 
broadcast station". We did tweek the amp quite a bit to get the 
modulation linear, It was originally running AB2, and we had to increase 
the bias and drive, and load it a little less, but it sure did sound 
sweet. Some pice are here:

>>The Transmitters/restorations are on hold as I am doing a major 
>>clean-out of my lab. Each weekend we have been hauling 800 lbs of junk 
>>to the landfill, it is -most- politically incorrect. nothing of 
>>interest here, unless someone wants old 9-track computer tapes and 
>>stuff like that.
> You never know what someone will buy on ebay.  At worst, it hangs around 
> your shop another week.  At best, you have something that someone else 
> thinks is an antique relic and is 'just the part' they need to finish 
> out their collection.  *shrug*  hey, it could happen  ;-)  
> I tried to sell a 450TH, appearing as if it was unused, no scratches on 
> the base - asked in here for $50 for it, in the metal spring-loaded 
> shipping frame, in a box.  No one bit.  Put it on ebay for 7 days, and 
> started it off at 9.95 and it sold for some 3X (plus shipping) what I 
> was asking for it originally, in here.
> The worst than can happen is, it hangs out for a week.

A 450TH is a very specialized tube today. Sometimes it takes just the 
right buyer. $50's cheep cheep for a new one, I think.

>>For what was lost, well that is true to a degree and 3 or 6 dB does 
>>not make that much difference at the other end, but I guess I am 
>>looking at the old Tucker rig and knowing its audio and RF power 
>>supplies are good for 3KW DC each. COL Tucker built it for military 
>>communications from Dallas to Austin. I mean, the thing was/is capable 
>>of who knows how much audio input on top of the KW DC. I'm thinking of 
>>the peak power. It just seems a shame to idle it out along the 
>>wireless in 2nd gear. The upside is that it will probably last forever.
> Think of the money you'll save on spares ;-)

Well you have that right.. once all the bugs are worked out.

I did go through a few of 5R4's in the modulator bias supply before I 
figured out that one of the 304TH's had an intermittent short from 
filament to grid, but only with HV applied. Oddly, whatever happened 
'blew up' the filament in the 5R4 before any fuses or overloads could 
trip, and then the bias supply fuse blew.. This was in a supposedly good 
304TH I bought from ebay. Due to the scarcity and high costs of 304TH's, 
I have changed to 3-500Z's now, and I need to re-do the bias supply for 
0-30V instead of 0-300V! I think a 10:1 voltage divider/bleeder drawing 
about 100mA at 300V should do (The modulator bias power supply is rated 
200mA). That way the rig stays more original, and the grid supply now 
has a 300 ohm resistance!

>>The status on that beast is that there is a parasitic oscillation in 
>>the final's grid circuit, between the 4-1000 control and screen grids, 
>>at about 140MHz. Plate tuning, current, and neuralization setting do 
>>not matter or vary much at all with this issue and the grid current 
>>and screen current interact alot. I found that a 50 ohm resistor in 
>>series with the grid helped. Maybe I need to increase that value or I 
>>don't know.. The screen grid is heavily bypassed directly at the G2 
>>terminals on the socket, and an RF choke is optionally in series with 
>>the screen to the chassis connector to no effect. This oscillation 
>>happens no matter if the grid drive connector is shorted, open, or 
>>terminated, so this tells me it is in the control grid circuit 
>>somewhere. Nothing's been changed from the original design there, so 
>>it is further mysterious.
> Just for grins and giggles, try putting 2 50 ohm resistors in series to 
> ground, and then the grid to the center of the two 50 ohm resistors.  
> It's helped other rigs.

I will try that. I am not sure of the effect it will have with the grid 
tank feeding a low resistance load of 100 ohms, but it will point me in 
the right direction.

I'm looking at these conditions (they represent much more power than I 
will legally use, but should be good for figuring. There were no figures 
for 2500V or less.)

Plate Modulated Class C Amplifier [CCS]
Plate Voltage ................................. 3000 V
Grid No. 2 Voltage ............................ 500 V
Grid No. 1 Voltage ............................ -200 V
Peak Grid No. 1 Voltage........................ 340 V
Grid No. 1 Current ............................ 12 mA
Plate Current ................................. 600 mA
Grid No. 2 Current ............................ 145 mA
Driving Power ................................. 12 W
Power Output (approx) ......................... 1390 W
Maximum Frequency ............................. 110 MHz

My question, is from these, how do I figure the impedance of the load 
presented by the grid circuit to the grid tank?

Using the 4-1000A curves for 2KV on the plate and 500V on G2, what I see 
in the curves is with -200V bias, the peak grid volts hits +140V, and at 
that point, the grid current is 250mA. that's 560 ohms. Of course this 
is not linear as I found out. It peaks to 1.1K at about 100V and falls 
off either side of +100V.

I put this in excel, and posted this:


I am curious why the grid resistance has such an odd curve with varying 
positive grid voltage (as calculated from the constant current curves 
e/i=r, right?). So the grid resistance varies wildly over any positive 
portion of a drive signal.

Looks like a case for excellent drive regulation requirement.

>>Also, I need to rig a 220V 30A outlet and power cord for it. I keep 
>>popping the 120V breaker when running up the final and modulation 
>>sides together (the unit has two 120V plugs presently).
> Seperate 120v circuits?

Yes, indeed COL Tucker apparently did not want to install a big ugly 
240V outlet, so he apparently used a pair of 20-amp 120V outlets 
connected to 240V in the breaker box. Of note in this rig, is that 
neutral is bonded to ground. period. That is, there is no ground prong 
on the plugs and it is not possible to separate neutral from ground in 
this rig without a major rewiring job. So, I will have to take care that 
I do use a 4-wire grounding 240V plug, and see that I really do have a 
good ground back to the breaker panel. Should present no problem if done 

>>As far as audio, I'm willing to help anyone who wants better quality 
>>sound from their AM rig. I'm not a great RF expert at all but I am OK 
>>with increasing the fidelity/reducing distortion in audio amps.
> But.. one works hand in hand with the other.  In fact, John/WA5BXO and I 
> have been discussing (at length, and credit to John for sticking with me 
> and my dumb questions ;-)) audio, it's effects on the transmitted 
> carrier, how much voltage is neeed to fully modulate the carrier and 
> what is needed in the way of a reactor to sucessfully couple the audio 
> to the B+ line to the final amplifier.  This conversation encompassed 
> power supplies, through the modulator and the final.  All components are 
> cruicial.

I picked up a modulation reactor some time ago. There might be room for 
it if I get rid of the filament transformers for the 673 rectifers and 
put in solid state. I kind of dread building two 15000PIV 1A bridge 
rectifiers, but I need to do it anyway.

> John and I also talk, from time to time, at length about asymetrical 
> audio.  I've achieved it, somewhat.
> Sometimes technology is overwhelming.
> For less than $100, I bought not only a digital camera that takes 
> amazing 3.2megapixel pictures, but it also has a 'movie' mode.
> No audio, but the movie mode is pretty awesome.
> "Hel-l-l-l-l-o test. one, two, three, four."
> http://w5omr.shacknet.nu:81/~w5omr/pictures/2005-08-08/000_0049.MOV
> The 'noise' you see is a box-fan at full speed.  It was over 100° here, 
> that day.

That is very impressive, I always strive to be rid of power supply noise.

>>At this moment I'm limited to NVIS type antennas, but I have a tower 
>>to put up. another project. I hope a tower will get me better 
>>shortwave listening due to a higher antenna (and give me a place to 
>>put a 6M beam for FM).
> Mike/WA5CMI has been on 6m lately.  He's in Garland, not too far from 
> you.  Maybe y'all could hook up and check out 6m together.
> he's probably monitoring either 50.4 or 50.125 - give a holler for him.

I have to get an antenna up at least 30 feet to get over my 'horizon' 
here in our little valley.

> As for the general... that's a cinch - cup o'tea - piece of cake.  Take 
> it.  Get it over with.  You'll feel better, and you can then put them 
> rigs on the air.  ;-)
> ---
> 73 = Best Regards,
> -Geoff/W5OMR

Best regards,

More information about the AMRadio mailing list

This page last updated 17 Dec 2017.