6MJ6/6LQ6/6JE6C 1 KW HF Amp ?
Perry Scott
that I ACTUALLY USE THE HF BANDS. Go ahead and flame me, I love mail.
I rediscovered vacuum tubes after my curiousity took me into the realm
of "macho" HF transmitters where transistors are still too expensive for
the interesting power levels (What I really want is to get my signal 3
dB over the noise on 80 and 40.) I may even move into VHF after I get
HF figured out.
Are there any schematics out there for a 1KW HF amplifier that uses the
6MJ6 sweep tube ? I'm having a little trouble designing a plate tank
for these beasties, due to the huge inter-electrode capacitances
generated by 4 or 10 6MJ6s in parallel. The TUNE and LOAD values are
radically different than say, a vanilla amp using a 3-500Z at 3 KV,
which makes finding components quite difficult.
I need at least one of three things:
1) I understand there is an Ameritron HF 1KW amp that uses 4? 6MJ6
tubes. If you have one of these, I'll send you a SASE for a
schematic, a dime for the copier, and an "attaboy" certificate
not-quite-suitable for framing.
2) I've been told that there may be a 6MJ6 linear amp schematic in one
of the old ARRL Handbooks. If you've got one, the SASE offer still
applies.
3) The manufacturer's specs for the 6MJ6. There seems to be a serious
lack of these, at least for the mfgr (Sylvania) that has their name
printed on the tubes I have. I've been guessing Cplate = 23pF,
Vp(max) = 1000V, and Pdiss = 30W, which are radically wrong. The
full specs would be welcome. An error of a few pF multiplied by 4 or
10 is the difference between something that works and something that
gets unsoldered. Again, the SASE offer applies.
The 6MJ6 has several aliases. It also answers to 6JE6C and 6LQ6, so
that complicates things a little.
Thanks to all for helping me reduce my spurious emissions. The Rube
Goldberg amp I've got is permanently QRT until I design a tank that
actually gets a plate current dip.
73, de KF0CA
aka
Perry "meets FCC spectral-purity requirements" Scott
..{ihnp4,hplabs}!hpfcla!perry-s
York David Anthony @ WKTD, Wilmington, NC
> Are there any schematics out there for a 1KW HF amplifier that uses the
> 6MJ6 sweep tube ? I'm having a little trouble designing a plate tank
> for these beasties, due to the huge inter-electrode capacitances
> generated by 4 or 10 6MJ6s in parallel. The TUNE and LOAD values are
> radically different than say, a vanilla amp using a 3-500Z at 3 KV,
> which makes finding components quite difficult.
Why do this? (The merely obvious answer might be to avoid
building a 3-4 kv supply for proper vacuum tubes). Frankly, I can't
**stand** my Julian Betts AB10SG 10 kw AM transmitter (commercial
broadcast) which uses dual 4CX10000D's in a screen modulated circuit;
multiple sweep tubes have got to be at least as difficult to construct
networks and tune.
I have no idea what the continuous plate dissipation of the
6MJ6 tube is, but suspect that 30-40 watts is not unreasonable. For
400 watts plate dissipation, you'd need 10. At wholesale prices
this works out to $111.27 (from Richardson Electronics).
(The 6146A, with lower interelectrode capacitance, is $8.95/tube).
Now, a 4-400C is $120, a 3-500Z is $126.
Either amplifier will be infinitely easier to neutralize than
the sweep tube amplifier; although I'm sure the latter is possible.
Of course, an air-system socket is required for the 4-400C, and the
glass chimney is highly desirable.
I went down this path once in modifying a RCA BTA-1M transmitter
(for drivers) and was not impressed with tube life. In constructing
any circuit, it is not unusual to lunch receiving tubes...they
are not designed for momentary overloads. Improper loading
and loss of excitation can waste a sweep tube in a hurry. On
the other hand, the 4-400C and 3-500Z have some thermal reserve
and are rated for overdissipation during tuning.
Although the initial cost is somewhat higher, you can
probably find 4-400C's pulled out from your local AM radio
station that have plenty of life left for amateur use. Under
ICAS conditions, a 4-400C or 3-500Z (new) should last 5000
hours with a proper filament voltage management program.
Just out of curiosity, what are the operating con-
ditions for 10 sweep tubes in parallel?
York David Anthony
DataSpan, Inc
Strawberry Jammer
>In article <788...@hpfcdc.HP.COM>, pe...@hpfcdc.HP.COM (Perry Scott) writes:
>> Are there any schematics out there for a 1KW HF amplifier that uses the
>> 6MJ6 sweep tube ?
> I have no idea what the continuous plate dissipation of the
>6MJ6 tube is, but suspect that 30-40 watts is not unreasonable. For
>400 watts plate dissipation, you'd need 10. At wholesale prices
>this works out to $111.27 (from Richardson Electronics).
>(The 6146A, with lower interelectrode capacitance, is $8.95/tube).
>Now, a 4-400C is $120, a 3-500Z is $126.
The only way I have seen "sweep tube linears" work is in SSB Grounded Grid
service - AB1 I think. You are quite right about the plate dissipation,
but the PEAK dissipation and allowable/achievable peak current is extremely
high. Thus the 4 tubes for 1kw (PEP!). About 30 seconds CW ot "tune up" on
one of these and you buy new tubes though.
For the unititiated, one of the "problems" with SSB is that the peak power is
far higher than the average power. The 2:1 ratio applies ONLY to a two tone
sine wave, uncompressed voice can easily be 10:1 or 20:1. Thats 20KW peak
for 1kw average! Most amplifiers are limited in their peak output as well as
their average output, so a 2KW PEP amplifier may only run at 3-400 watts
average power with a normal voice.
Ah yes but we can compress the voice to get a better ratio. True but ANY
change to the voice signal adds distortion and makes it hard to understand, or
if the "compression" takes place in the amplifier (called "clipping" now) it
produces "splatter" or spurious signals. Just listen to 20M SSB to hear lots
of examples of both.
There are compressors that have been designed that will compress 10-20 db
without "too much" distortion (define too much ...), but they are expensive.
The "sweep tube linear" is a reasonable solution if (a) you don't run a
compressor AND (b) you NEVER run CW or any continuous transmission for more
than about 1 second at a time. This includes tune up!
> I went down this path once in modifying a RCA BTA-1M transmitter
>(for drivers) and was not impressed with tube life. In constructing
>any circuit, it is not unusual to lunch receiving tubes...they
>are not designed for momentary overloads. Improper loading
>and loss of excitation can waste a sweep tube in a hurry. On
>the other hand, the 4-400C and 3-500Z have some thermal reserve
>and are rated for overdissipation during tuning.
Tube life is not exactly great, but for most ham stations to put 200hrs of
actual transmitting will be several years. Try a clock hooked up to you XMIT
switch if you don't believe me. Thus tubes lasting 100hrs or so are quite
reasonable. The sweep tubes are designed for momentary overloads with a thick
anode which will absorb a lot of heat for a few hundred
milliseconds.
> Although the initial cost is somewhat higher, you can
>probably find 4-400C's pulled out from your local AM radio
>station that have plenty of life left for amateur use. Under
>ICAS conditions, a 4-400C or 3-500Z (new) should last 5000
>hours with a proper filament voltage management program.
An excellent source if you run CW and have a limited budget BTW. The cost of
the 3-4KV power supply is high though and not easy to duplicate by work
arounds. I used 4 1940-50 "console" radio supplies to get 3KV at 700Ma peak
once, but it was bulky and those old radios don't cost 50 cents each any
more!
> Just out of curiosity, what are the operating con-
>ditions for 10 sweep tubes in parallel?
Never tried it, must be almost like a solid state amp. - 50 ohm output!
--
*Mike Waters AA4MW/7 wat...@dover.sps.mot.com OR wat...@cad.Berkley.EDU*
"His super power is to turn into a scotch terrier."
Avatar
In article <15...@unccvax.UUCP> d...@unccvax.UUCP (York David Anthony @ WKTD, Wilmington, NC) writes:
+In article <788...@hpfcdc.HP.COM>, pe...@hpfcdc.HP.COM (Perry Scott) writes:
+
+> Are there any schematics out there for a 1KW HF amplifier that uses the
+> 6MJ6 sweep tube ? I'm having a little trouble designing a plate tank
In the 1971 issue of the Radio Amateurs Handbook you will find a circuit
that uses 4 sweep tubes of this type. Actually, any of the common types
can be used. 6LQ6, 6LF6, 6CE6, 6MJ6..ad absurdum. BUT- remember! These kinds
of amps are very seldom operable in any mode but SSB (drive it light, the
glass envelopes won't take much!) or light-duty CW.
+> for these beasties, due to the huge inter-electrode capacitances
+> generated by 4 or 10 6MJ6s in parallel. The TUNE and LOAD values are
Nope. Forget it. More than 4 tubes of this type is gonna be a kludge, and
worse yet; the harmonic content of your output is going to be about 2 dB
below your intended signal. Can't really tell you why, it just works out
that way. In a prior incarnation, (God, it was hell on 11 meters! :-) ) I've
serviced a few of this kind of amp. One of them, the "Golden Falcon", used
10 6LQ6's, the tube life was very short with AM use, about 4-6 months! If
run in SSB service, the tubes would last a year, max. The damn thing had a
s-cage blower in it, and on-the-air it sounded like a truckload of turkeys
crashing thru a gay-bar at happy hour! VERY BLOODY-WIDEBANDED! LOTS OF TVI!
..despite all attempts at adjustment, running at full power, the design is
not exactly "good engineering practice".
+> radically different than say, a vanilla amp using a 3-500Z at 3 KV,
+> which makes finding components quite difficult.
+
+ Why do this? (The merely obvious answer might be to avoid
+building a 3-4 kv supply for proper vacuum tubes). Frankly, I can't
He's right. Don't be lazy, build a good power supply so you can run a
proper tube, the sweep tube linears are mostly junk! Besides, you cannot
run them with packet, AMTOR, RTTY or really hot CW anyway...
+ Just out of curiosity, what are the operating con-
+ditions for 10 sweep tubes in parallel?
+
It borders on desperation. ;-)
--
-Avatar-> (aka: Erik K. Sorgatz) KB6LUY +-------------------------+
Citicorp(+)TTI *----------> panic trap; type = N+1 *
3100 Ocean Park Blvd. (213) 450-9111, ext. 2973 +-------------------------+
Santa Monica, CA 90405 {csun,philabs,randvax,trwrb}!ttidca!ttidcb!sorgatz **
Perry Scott
I picked up the amplifier at a flea market from a CB'er. Four tubes were
feeding six. At 4 watts input, I got about 350 watts out. The problem
is that my 100-W xmtr can't go that low ! I've already fried the input
matching network.
Re: operating conditions
For class C, 1 KW in, Rplate = 500 at 990volts (max for the 6LQ6). I
found some specs, and it's true that it could handle the momentary overloads
of SSB, since it can dissipate 300 W for 20 seconds. However, the pi
network caps are huge - about 10X anything I've ever seen.
I found some specs after calling all the TV repair shops in town. I've
also found a couple schematics for 4-tube amps (500W SSB). Thanks to all
the responders.
I found a cheap source of 3CX250Bs in the chip fab here at the plant, so
I'll probably go that route. Path of least resistance and all that.
Perry Scott, KF0CA
Strawberry Jammer
>I picked up the amplifier at a flea market from a CB'er.
Your first and biggest mistake right there! Every CB "linear" that I have
ever seen was (a) GROSSLY overrated as to its power capability and (b) had
a spurious output that had to be seen to be believed. There used to be one
operated mobile in Clearwater Fla. that put out more power on 2M than
11M. We used to hear it override repeaters all over town.
For CB use they are illegal, for amateur use they are JUNK! I have been
offered many at hamfests and flea markets and their owners get insulted
when I tell them that the case is worth 50 cents, the connectors about 10
cents and I would throw the rest in the trash!
Four tubes were
> At 4 watts input, I got about 350 watts out. The problem
>is that my 100-W xmtr can't go that low ! I've already fried the input
>matching network.
Be sure to measure the freq. of the output! Bet most of its not where you
think it is! And the front panel says something like 2KW linear too right?
At 350 actual watts, that is higher power than most such "linears" put out
though.
>I found a cheap source of 3CX250Bs in the chip fab here at the plant, so
>I'll probably go that route. Path of least resistance and all that.
A much better choice - you won't regret it.
--
*Mike Waters AA4MW/7 wat...@dover.sps.mot.com OR wat...@cad.Berkley.EDU*
Goldenstern's Rules:
1. Always hire a rich attorney
2. Never buy from a rich salesman.
Perry Scott
properly designed. Most of these "CB" amplifiers give sweep tubes a bad
name because they are typically tuned for maximum smoke and have poorly
designed pi tank networks. At least this particular amplifier does - no
dip in plate current because the TUNE cap is too small.
According to DeMaw, W1FB, in QST 2/80, you can make a sweep tube
amplifier that develops 1KW input, because they have a 300W intermittent
dissipation spec. He warns that FM, AM, and CW could easily exceed the
continuous power ratings (30W), so some caution is needed. The second
and third harmonics are way down, on the order of 30dB, so it is
possible to make one that works well.
The biggest difficulty I've found so far is in designing a proper
tank with the parts that came with the amplifier. Because sweep tubes
can only take about 1000V, the plate resistance is very low - on the
order of 500 ohms for 1KW dc input. This requires larger TUNE caps
than the other classic designs, say using a 3-500Z at 3KV. The real
problem is that 500 ohms falls in the never-never land between the high
impedance matching networks used for tubes and the low impedance networks
used for transistors, so finding pre-designed networks is difficult.
Another problem appears to be designing around the high combined plate
and cathode capacitances. It isn't particularly difficult, although the
sweep tube doesn't work well at VHF because of this.
On the other topic (using 3CX250Bs):
Where is a cheap source for sockets and chimneys ? I looked in the AES
catalog and nearly fainted when I saw the prices. The socket and chimney
costs as much as the tube !
Perry KF0CA