Amp Supply Company

[Steve - WB3HUZ]

Another thing to keep in mind is that with a 100 Watt driver, that third 3-500 doesn't buy you any more power output. But each tube is dissipating less, so that's another reason why they can use the fan and no chimneys.

[ke7trp]

Carl. Its not fuss. Its just discussion

Not all amp supplys had PD tranformers.  They where put together in runs of available parts.

Tom.. My idea was to copy the heathkit design using a tube from bottom deck to the top of the pins to blow air across the heat sinks. If I do this, I will post pictures. Since the duty of my transmitter is low, I prob dont need anything more. Its designed as Eimac states in the tube data sheet.  I have two huge blowers that are back to back blowing up the tubes.

C

[WQ9E]

I have two huge blowers that are back to back blowing up the tubes.

C

I will have to try this blower technique sometime, in the past I have relied upon making mistakes to blow up my tubes 

Are you planning on using the air pipes like Heath used in the Chippewa?  Post pictures if you add them please.

Rodger

[ke7trp]

Yes.. I want to add the air tubes.  Not sure when. The thing is that this transmitter runs low power. I am in a neighborhood on a 50 ft wire.  90% of the time, its going to be loafing along under legal limit.  Only the 10% of the time when I hear Tom, steve and the others out east, will it go to 4kv  

This transmitter has low, oh my god and holly ****, power levels

C

[WA1GFZ]

3-500s make great modulators.
50 feet of wire, don't get too close to the mic

[W3GMS]

Talking about plate voltage on 3-500Z's, my mentor Harry W3FDY (SK) had a pair of 3-500Z's in grounded grid push pull operation and ran 5KV on the plates of the tubes.  He did have a twin squire cage blower that pressurized the underneath side of the chassis.  It produced lots of air under backpressure conditions.  I can't speak on the life of the tubes, but I do recall that Harry always bought quantities of everything!  That amp was used on SSB as a replacement for his old push pull 250 class C rig.   
Regards,
Joe, W3GMS     

[ke7trp]

I have seen them many times with two and three tubes. 4500 to 5k. Big numbers out. Thats why I thought of using 4 of them in my big amp.

C

[Steve - WB3HUZ]

It comes down to drive power after two (or at most three).

I have seen them many times with two and three tubes. 4500 to 5k. Big numbers out. Thats why I thought of using 4 of them in my big amp.

C

[K1JJ]

Yep, a driver is the issue.  It is difficult to get a driver that will do, say 200-300w pep out and still be as clean as the main amp. The varying input load to the main final GG tubes will make IMD suffer somewhat in the driver. It's just like when driving modulators -you need a low impedance driver to handle the varying grid load in a linear fashion. In contrast, the final amp sees a steady antenna load.

One good idea is to use a single (or even pair) of 3-500Z's as a driver. Put them in GG, 1500V on the plate and load them VERY heavily so that the resonant plate dip is very broad.  This will produce a driver that has about -45db 3rd IMD. That is a sensational driver.  It requires about 40w pep to do 300w output as a driver. An FT-1000D is very clean running at 40w pep as an exciter, being a 200w radio. This is a case where this ricebox gets cleaner at lower power - I tested it.  Some do not do this and may even get worse at low power.

Four 3-500Z's wud be a neat amp, since you already have the tubes. If someone did not have the tubes already, I would opt for a single bigger power grid triode designed for ssb service. There are many around in the 3CX-3000A7 class to choose from.

T

[KM1H]

Tom, according to Eimac the SK-406 chimney is correct for the 3-500 and 4-400. The 3-400 uses the shorter SK-416.

Getting back to the LK-550 it uses dual fans which provide more than sufficient air under any permitted condition. If anything its a bit noisy which is why the 2 speed circuit was added. This can be controlled from the front panel, an internal sensor or by the PTT line either by model revision or a simple mod. Ive yet to see any unsoldered pins or sockets needing replacement. I believe the earliest model fans werent too reliable so rumors tend to circulate forever even after the problem was fixed.

Id say the 550 is fine at 500-600W AM, the huge external Dahl plate transformer is certainly up to it as is the rest of the PS. The PS is a FWB instead of the half wave doubler used by most others, this adds to the regulation. The FWB is used in all the LK series amps and is another plus to their quality.

A vacuum relay was optional on the LK-500 and standard on the LK-550, 750 and 800 series. Ive replaced one in mine in 24 years, VOX CW and contesting tend to run up the cycle clock

There are a couple of weak points, easily fixed, that Id be glad to discuss with anyone seriously contemplating buying one. Drop me a PM.

Carl

[ke7trp]

I need to test out this LK800 this week.  Its the military version used at the monitoring station that was shut down. Probably very little hours.

Tom. The 1000D does 350 watts turned up. 300 rock solid clean.  I know guys that rag chew on them turned up like that and have for years with no ill effects. It has 400 watts worth of final transistor and the PS is large enough for 100% duty at 200 watts RTTY.  

C

[W1VD]

There's 'clean' at -35 dB (3rd) and then there's 'really clean' at -45 dB (3rd). When Tom says 'clean' I think he usually means 'really clean' ... at least in my conversations with him.

Perhaps when one says 'clean' there should be clarification, like a number in parenthesis, after the word 'clean'. There's a whirl of difference between -35dB and -45 dB IMD performance ... assuming higher order products are well behaved.

[K1JJ]

Yes, agreed, Jay.

A stock FT-1000D running at full recommended output (200w out) does about minus 30 to 32db 3rd IMD. This is acceptable but you really don't want to be less than that.  If the power is cranked up above 200w the IMD figures will surely deteriorate from there.

Running barefoot, nobody will bother us, but if a 300-400w FT-1000D was driving a big amp that was a clean -45db 3rd, the cranked up FT-1000D would probably limit the overall 3rd to less than -27db 3rd or so.  These are rough numbers, but that's the idea.

The rule goes like this.... if you are running 5db more than the average QRO station in a given band area, then you better be 5db cleaner to blend in with the pack - otherwise you will be singled out and beat up...   It's easier to get away with it in the AM areas cuz most expect wider signals, but on ssb with 3khz spacing, they will call out the lynch mob if the IMD is much below -28 3rd AND the average power is 5db or more above average QRO hams.  SSB is a good acid test for a driver/ linear system.


The way to check the FT-1000D is to run it on LSB into a dummy load and listen carefully to a RX that has a 12" wire stub for an ant. (Or use a good spec analyzer/ two tone test) Run thru programmed taped voice material and then listen to the RX up and down the band at various power levels. If you listen carefully and watch the s meter, you will see the difference between 40w pep and 200w pep.  Add the linear, normalize the S meter reading by shortening the ant stub and tune around again. The linear almost always makes the situation worse.  The FT-1000D is very clean at 40w pep out - I'd estimate upwards of -40db 3rd.

Remember that if a linear is, say, -40db 3rd and the driver is -30db 3rd, then the overall signal will always be LESS than -30db 3rd. The final output signal's IMD is always worse than the weak link in the chain.



T

[KM1H]

Most SS rigs have an IMD sweet spot at 70-80% of advertised PEP. As you continue down in power the IMD climbs and then drops again around 30-40W. The ESSB crowd run Class A in the xcvr and then a very linear IPA into a tube with handles.  Ive heard claims of high 40's 3rd total system.

Carl

[K1JJ]

Most SS rigs have an IMD sweet spot at 70-80% of advertised PEP. As you continue down in power the IMD climbs and then drops again around 30-40W. The ESSB crowd run Class A in the xcvr and then a very linear IPA into a tube with handles.  Ive heard claims of high 40's 3rd total system.

Carl

Carl,

What kind of IPA amp and config are these guys using to produce a clean driver?     I see the common SS amps are not much better than -30db or so.  I once tried a pair of grid driven 4CX-150's driven by two low level stages that had a lot of RF negative feedback, a commercial amp,  which did OK. 

The class A  ricebox at -70db?  3rd is a good start.

Is there some popular IPA lashup these guys use today?

T

[KM1H]

Tom, it looks like the L4B is a common choice with the idle current cranked up, seems to hold together thanks to the chimneys and some PS mods. Guys are claiming around -42 3rd.

For the same chimney reason the AL-82 is another choice.

Id think a 3 hole Alpha 76CA/PA would do well also as it has 1200W PD already or you could go with a pair of 3CX800's swapped in. Some use the 2 holer and modify to fit the 800's but power is less due to smaller iron. Alphas have been showing pretty cheap as owners and buyers are afraid of 8874 replacement cost. Find one without or soft tubes and go from there. Either one of those tubes puts you in the mid -40's 3rd as is and you can experiment with bias. The 3 holer comes with a EI or the Dahl transformer as an option. Mine (no mods as its only used on CW/SSB on the 2nd station) will run cool at 1800 key down with 80-90W drive from the TS-940, thats within 8874 specs and it doesnt get near as much use as the LK-500 on the main TS-940.

Carl

[K1JJ]

Carl,

Very interesting.  I didn't know they were using 3-500Z's etc as drivers. Actually a good idea, something like a quasi-class A routine which needs big plates to pull off.

I have my 3-500Z grids at ground, so with only 1500V on the plates they JUST idle, but I would like them to idle somewhat harder. Right now when I load them very heavily, they show a little more color, but I could use some more. How would you get them to idle higher... ie, put a positive bias on them?  The logical way seems to insert  a regulated negative 10 VDC (or whatever)  1A supply in series with the cathode leads to give the grids a positive bias... 

Any other ideas to do it simpler? 

BTW, I draw air in from outside the house using a squirrel cage. The fan runs slowly with barely any noise.  Excellent cooling in the spring and fall, but I freeze in the winter.

T

[KB2WIG]

' BTW, I draw air in from outside the house using a squirrel cage. The fan runs slowly with barely any noise.  Excellent cooling in the spring and fall, but I freeze in the winter. "

draw from da inside in winter??


klc

[KM1H]

Id just go to the tube curves Tom and see what works at the plate voltage desired.

At 1500V you are already at -46 at zero bias according to Eimac.

With the low cost and ease of construction a pair of YC-156's might interest you as the next stage Just dont expect to run a pair above 20M without getting creative.

Carl