Central Electronics Broadband "No Tuning" Grid and Tank Circuits...

[W1AEX]

This is something I have wondered about for a long time and I figure this is probably the only forum that has someone who knows the answer. The short question is, has anyone ever figured out how CE built their broadband "no tuning required" grid and tank circuits that appeared in their 600L "813 amp" and rigs like the CE-200V? Some of those rigs are still on the air and appear to be working fine, so I assume it was a pretty robust design. I just figure that this was a design that was way before its time, and it surprises me that it never went further during the tube era. There are some beautiful pictures of a restored 600L here:

http://www.isquare.com/personal_pages/600L.htm

Note the bluish "blobbed" components that I assume are the broadband components in the tank circuit.

Back in the 70's I spoke with a ham who had a 200V on the air with a coupler that had failed on 75 meters. He attempted to open up the broadband tuning network but it was apparently epoxied to prevent entry. He never figured out how to fix the original circuit, but made it work on a narrow slice of the band with conventional components. The components within the broadband couplers appear to be securely sealed to prevent inspection by the typical user.

It just seems to me that if CE designed and successfully manufactured these devices in the 60's, then why aren't we doing it now with bigger and better components in tube amplifiers?

Rob W1AEX

[KD6VXI]

Funny, KM1H and I where just talking about this.

He told me that in addition to Dentron, Heathkit on the SB230 and Alpha on the 8874 based amplifiers had broadband inputs (no switched LC networks or Pi, broadbanded!!!).

Having NOT worked on them, I can't say yea or neah.... BUT!!!!  They are there, iffin ya can find some schematics on them.

This was because someone had done a 3CPX800 conversion to an ML2500, and he was inquiring about the inputs....  I didn't know, until then, that anyone other than Dentron had used them... Guess CE did as well!

--Shane

As soon as I hit reply, I realized you're probably talking about the OUTPUT, when stating "tank", and the input on the grid?

I've also seen a sweep tube amplifier that had a broadbanded output for "ten" meters.  IIRC, it was branded as a BLJ amplifier, but man, this has been YEARS, so no guarantees.  It was turn on and operate.  Some of the KLV or RM Italy tube amps (I don't mess with that offshore crap) I seem to recall did a EL-509 or something similiar, completely widebanded.

Edited to add the last two paragraphs after I realized my reading comprehension had gone to pewpew

[WA1GFZ]

Well, A pi network is a low pass filter and an impedance transformer so I bet the network is a low Q low pass filter. A Q of around 7 should do it. the impedance transformation is done with a ratio between input and output cap values. A two stage Q of 7 filter each stage would work well. You would only get into trouble if the VSWR got too high taking out a cap.
Collins made a SSB rig for the Air Force with a pair of 4CX250s It had a variable inductor and fixed caps switched in to cover HF
I think a pair of bread slicers was cheaper than a pile of fixed caps with a multipole switch.

[KD6VXI]

Collins made a SSB rig for the Air Force with a pair of 4CX250s It had a variable inductor and fixed caps switched in to cover HF

I have the same type of output in my Harris RF-103.  Made it VERY easy to "channelize".  Mine has 10 memories on the front panel, all controlled by a 10 turn pot.  Set the pot where the roller inductor is where you want it,  and the Q is now fixed.  Takes a few seconds to QSY, really an easy amp to operate.  Single 3-1000Z.

Works <<ok>>....  I've often thought about the SWR issue, but Harris has an antenna coupler to minimize that....

--Shane

[W1AEX]

Heh, I see from the time stamp on your post you were up very late. I can't keep a single thought straight at that hour!

The broadband couplers at the input and the output are very curious devices to me. At BAMA I did find a service manual for the CE-100V and CE-200V series, but the broadband couplers are treated as "black box" devices with no adjustments required. Likewise, even more interesting to me is the 600L amplifier with an 813 in the final. The input is shown with a broadband coupler just like the output tank circuit. The schematic shown here, which shows the couplers as little voodoo boxes, is for the broadband 813 amplifier, which covered 160-10 meters. When you think about it, that's an amazing piece of engineering! According to the manual, the 813 was capable of 500 watts input on CW without much concern about SWR. In fact, they mentioned that you could run it into a 3:1 SWR without worry, and a breaker system was present to prevent damage if a no-load condition was present. So... the question is, will this remain a mystery forever, or has someone figured out the details of their designs? It appears that CE is still around under the name Tusa Consulting, with a little history appearing here:

http://www.tusaconsulting.com/ce.html

For all I know, the design was patented at some point. I'm not sure what that means as far as replication by individual homebrewers, but it sure would be nice if the designs are not allowed to slip into obscurity as time goes by.

[k4kyv]

I wonder if there is someone still around who used to work at C-E, who would be willing to reveal the exact design details on that broadband tuning circuit.  I'm sure the patent has expired by now, and there is not much demand these days for tube type output circuits, so there would be no commercial advantage for anyone in keeping it a secret.

I suppose it could still have a state-of-the-fart application with class-E.

[Pete, WA2CWA]

There were several articles written on the 100V and 600L in Electric Radio Magazine over the years. Wes Schum, W9DYV, the founder of Central Electronics is still around and active. He checked in to the vintage sideband net several times over the years. He and I actually had a 100V to 100V sideband contact on the net one time.

Unfortunately, the designer of the broadband couplers, Joe Batchelor, W4EGK, passed away in 2004.
http://www.mindspring.com/~cacutts/radio/ba/misc/k4egk.html

And, as is detailed on Nick Tusa's web site:
"Tusa maintains a close association and friendship with Mr. Schum. Both are jointly involved in a variety of projects, which include several new Comtronics products, the expansion of Schum's high voltage test set product line, continued Central Electronics equipment support and development of high-power broadband amplifiers using the Batchelor broadband coupler scheme."

[WA1QHQ]

There was an article in electric radio a number of years ago reviewing the 600L amplifier. The author did a write-up on the broadband tank design. My recollection on this is hazy since it was a number of years back but I do recall him saying that the transformer was a very difficult component to design and get to work correctly, lots of tweaking. I think the ER article is probably the best reference on this subject.

[Steve - WB3HUZ]

He told me that in addition to Dentron, Heathkit on the SB230 and Alpha on the 8874 based amplifiers had broadband inputs (no switched LC networks or Pi, broadbanded!!!).

And Henry and most every other GG amp made in the last 20+ years. It's SOP for GG amps, both for impedance matching and more importantly, IMD reduction. BUT, these would not work in an output scenario for two reasons. First, they generally have a low Q - around 2 - thus the broadband nature. But they provide very little lowpass filtering. Second, the impedance matching they are doing is very minimal, usually from 50 Ohms to some like 50-100 Ohms. In an output scenario, you are looking at going from impedances in the thousands of Ohms to 50.

There was a QST article in the 40's or 50's for broadband link coupling between multiple stages in a TX. IIRC, it included the output state. It boiled down to a form of stagger tuning that created a gentle double humped response across the required passband. The passband ripple was small enough the tuning/coupling was good across an entire HF band.

[W3SLK]

Mark said:

There was an article in electric radio a number of years ago reviewing the 600L amplifier. The author did a write-up on the broadband tank design. My recollection on this is hazy since it was a number of years back but I do recall him saying that the transformer was a very difficult component to design and get to work correctly, lots of tweaking. I think the ER article is probably the best reference on this subject.

You are right and I need to find the article. I remember them talking that each one of those output filters was individually made for that particular rig. Man I think that would have killed production right then and there!

[WA1GFZ]

Steve, say it was a pi L to soften the impedance blow. Another option would be to increse the resting current on the 813 to get closer to class a so you don't need a heavy flywheel. It would be interesting to know the efficiency at the middle and ends of a band.

[WD5JKO]

I have a 600L, and this is an amazing, and heavy amplifier. I tested mine on 75m AM, and it would put out 100w carrier and be able to modulate 100% upward very cleanly. It only takes a few watts to drive it, so it is a huge power mismatch to the 100 watt class 100V/200V exciters. This amplifier is a better match to the CE 10B and 20A exciters. Getting 100w AM out of an 813 in linear mode is amazing, especially since the modulation is linear all the way up to 100%. Sure it blushes a bit, but between the regulated G2, G1 supplies, and a mule of a plate supply, the 600L squeezes a lot of watts out of the 813.

The 813 runs class AB1, and the RF output is a near perfect sine wave. Whatever the RF couplets are, I don't think they use any lumped capacitance (no pi or pi-L) unless it uses distributed capacitance between turns to get there.

Nick Tusa can still provide CE parts including the couplets. Rumor has it that hurricane Katrina heavily damaged his stock of CE parts and gear.

Regards,
Jim
WD5JKO

[WA1GFZ]

Jim,
What is the resting current of the 813 without drive? I wonder if the second harmonic is well attenuated? Also how well does it work at the band edges?

[WD5JKO]

Francis,

    My 600L is not hooked up right now. I do recall however that it operates with low idle current. Since mine was putting out a nice sine wave, I suspect  harmonic attenuation was decent. My recollection is that if the waveform is asymmetrical, then we have elevated second harmonic, and if it is symmetrical and flat topped we have elevated odd order harmonics. The 600L had a nice clean sine wave up to maximum carrier output which was about 350 watts CW.

Jim
WD5JKO

   Here is a post from Nick Tusa relevant to the subject:


Post from July 16, 2004 on the Central Electronics Reflector:

[Central-Electronics] 600L Broadband Coupler Availability   NICKTUSA at aol.com
Replacement 15M broadband couplers for the 600L amplifier are now available. 
The cost is $95, postage paid, with your "dud" coupler in exchange or $115
outright.  I am collecting dead couplers for future recycling once my stock of
NOS coil forms is exhausted.  These newly designed couplers have a flatter
frequency response than the original units and are constructed with modern Teflon
wire. 

New couplers are also available, from stock, for the 80, 40 and 20M bands,
with replacement 160 and 10M units planned for release within the coming weeks.

73..Nick  K5EF
Dominic F. Tusa
Tusa Consulting Services, Inc.
19 Augusta Drive
Covington, LA 70435

985-892-7348 Tel
985-892-2075 Fax

www.tusaconsulting.com
www.polaristowers.com

[W1AEX]

Interesting stuff. Must be fun to play around with that amp Jim. From the anecdotal information, it sounds like the couplers are fairly complex, especially if each one is kind of unique in its settings. Ah well, maybe I'll pass this thread on to The Mythbusters producers and see if they can figure it out!

 

[KM1H]

He told me that in addition to Dentron, Heathkit on the SB230 and Alpha on the 8874 based amplifiers had broadband inputs (no switched LC networks or Pi, broadbanded!!!).

And Henry and most every other GG amp made in the last 20+ years. It's SOP for GG amps, both for impedance matching and more importantly, IMD reduction. BUT, these would not work in an output scenario for two reasons. First, they generally have a low Q - around 2 - thus the broadband nature. But they provide very little lowpass filtering. Second, the impedance matching they are doing is very minimal, usually from 50 Ohms to some like 50-100 Ohms. In an output scenario, you are looking at going from impedances in the thousands of Ohms to 50.

There was a QST article in the 40's or 50's for broadband link coupling between multiple stages in a TX. IIRC, it included the output state. It boiled down to a form of stagger tuning that created a gentle double humped response across the required passband. The passband ripple was small enough the tuning/coupling was good across an entire HF band.

That particular thread got a little distorted over here.

I was relating only to those those GG amps that used a broadband resistive load, with and without a toroid. The earliest Im aware of doing it that way was the NCL-2000 but that is grid driven.

Getting back to CE I modified my 100V many years ago in the late 60's to cover 160M and duplicated the 80M driver and 6550 stages format as close as possible. I was able to get 100W over most of the band after several tries and finally left it that way.

On another forum (probably boatanchors@qth.com) there was someone selling the CE info and claiming that only he had the original info. Not owning a 100-200V any longer I didnt bookmark it.

Carl
KM1H

[Pete, WA2CWA]

There was a QST article in the 40's or 50's for broadband link coupling between multiple stages in a TX. IIRC, it included the output state. It boiled down to a form of stagger tuning that created a gentle double humped response across the required passband. The passband ripple was small enough the tuning/coupling was good across an entire HF band.

I'm doing this from memory, since I just did a 100V for Grant 2 weeks ago, but I believe the manual includes the details for the broadband coupling for each band and interim stage. It does not give construction details for the broadband coils for the output stage. As far as I know, Nick Tusa is the only person who has the information to recreate those coils as they were originally designed.

[Pete, WA2CWA]

On another forum (probably boatanchors@qth.com) there was someone selling the CE info and claiming that only he had the original info. Not owning a 100-200V any longer I didnt bookmark it.

Carl
KM1H

I'll assume you mean boatanchors@mailman.qth.net since qth.com doesn't have a boatanchor reflector. I did a search on all the posts on the boatanchors reflector and the last 100V post was February 2008 (someone trying to sell or give away a VOX Coil for a 100V/200V). The last time there was a 100V/200V discussion on boatanchors was July 2004. On the Central Electronics reflector, the only name that comes up for parts and information is Nick Tusa. Might have been some other reflector or maybe the Yahoo or Glow Bug groups.

Since I've never had a problem with the output broadband coils in my 100V or 200V, I never screw with them.

[nq5t]

I'm doing this from memory, since I just did a 100V for Grant 2 weeks ago, but I believe the manual includes the details for the broadband coupling for each band and interim stage. It does not give construction details for the broadband coils for the output stage. As far as I know, Nick Tusa is the only person who has the information to recreate those coils as they were originally designed.

Yes, and I found my 100V manual set, the day your new one arrived.  Never fails :-)  And did finally manage to get the HV caps replaced -- what a chore.  It's like doing laparoscopic surgery to get in there.

I do have the 160 meter kit for my 100V, which includes a non-potted non-secret output coupler -- haven't installed the kit yet.  The output coupler is a two terminal device -- See pic and schematic.  The circuit looks a bit odd -- but it is what it is.  I haven't looked at how it fits in with the rest of the output circuitry.  There is no external connection to the top of the parallel tuned circuit unless that stud at the top of the photo meshes with some sort of clip when the coupler is plugged in.  There's no mention of anything of that sort in the installation instructions.

Grant/NQ5T

[KM1H]

The coils in my 100V were not potted. I simply took the 80M ones into National and with some help from Engineering and their test equipment I was able to duplicate for 160. The final test was bringing the 100V in on a Saturday.

I also fried the 10 and 20M coils (thanks to a TH-6 and a fried BN-86 from the NCL-2000) and had those rewound. National had fancy coil winders available for engineering use and the new ones worked perfect over both bands.

Maybe the person selling the info was Nick, I didnt pay much attention to details or the forum.

Carl
KM1H

[WBear2GCR]

Carl,

Do you still have them?
Can you post jpegs?

Grant, that looks something like an AC coupled autotransformer perhaps using the cap values to tune/peak the affair?

Thanks for the post of the jpegs and schematics.

This was an SSB only rig?? I ask because if that is the output coil, it does seem a tad small in size??

            _-_-bear

[nq5t]

This was an SSB only rig?? I ask because if that is the output coil, it does seem a tad small in size??

USB, LSB, USB w/Carrier, LSB w/Carrier, DSB (no carrier), AM (obviously with carrier), CW, PM, FSK.

The 160M coupler and other parts for various mixer/Osc circuits are from Nick Tusa.  I purchased them several years ago, but never found a Round To-it.  The ceramic form is 2 7/8" x 1".  The radio is nominally up to about 180W PEP input and about 100W out.  I get around 90W out on CW on 20M.

I did a complete electrical restoration of the transmitter around 2001 or so (the physical shape was excellent from the outset), including a "new" (rebuilt) PTO from Nick on a swap basis.  It plays really well.  Sadly, my pocketbook can't handle the current freight on the matching 600L amp -- or I'd have one in a heartbeat.  Of course, a donation of one would be happily accepted ;-)

The transmitter has very effective high SWR sense and shutdown curcuitry.  And if that's properly adjusted, it should keep one out of trouble.

Grant/NQ5T

[KM1H]

I sold my 100V and 200V around 1985 along with all documentation. Only kept the 75A4's and still have one of them.

Carl
KM1H

[WA1GFZ]

So it must be a low Q parallel tuned tank with a series tuned link. I wonder if you can measure the inductors and read the cap vlaues? cool idea.

[Gito]

Hi

Maybe a stupid Idea of me,comparing this broad band RF amplifiers to an audio amplifier,

An tube audio amplifier has a fixed in impedance and a fixed load,like 6 Kil.ohm to 4 Ohm,it has an out put trans former that has A broadband frequency response from 50 hz to 20000 Hz,a good design output transformer has a flat response.
So it's possible to design a broad band audio transformer.

Now a transmitter has a fixed output impedance and a fixed output load for instance 807 transmitter has a 6 kilo ohm output impedance and an antenna load of 50 Ohm,

Maybe we can use a broadband toroidal balun transformer ( ferrit  ,for Rf)  as an output" transformer" ( coupler),with enough turns in the primary,that has enough Reactance (milli henrie) .and has 6 kilo ohm(primair) to 50  Ohm(Secundair ) ratio .
Since if the turns ratio is right,the Transmitter is correctly  loaded .

So the point as can we make a broad band transformer using balun(powder iron toroidal inductors) as an broadband output coupler(RF) for the final RF tube

as some of the antenna tuner used balun  coupler and has broadband response.(frequency range 1.8 to 30  mc)

Gito.