Questions about 160 amp skemat

[K5WLF]

I'm considering building this 160 meter amp which uses a pair of 572Bs. Would there be any problem with substituting a pi output network for the tuned-link output that's shown in the schematic? Would there actually be any advantages? It seems that there'd be a better energy transfer to the antenna with the output going through a wire rather than "air coupled", but I've been wrong before and I'm new to building RF amps.

For now, I'll be driving it with my Yaesu FT-897D. The article doesn't have a power spec for the amp, but refers to "a kilowatt with key down". Figuring that's PEP will probably give me somewhere between 250-300 watts real AM carrier. Any input on that assumption will be appreciated.

The pi network calculator I'm using needs to know the plate impedance. Dividing the Ep of 2600 by a presumed 115 mA Ip at 300 watts output, I came up with 22600 ohms. The calculator goes a bit nuts and subs a Q of ~22 rather than the usual Q of 12 for the pi network. Am I figuring the Zp wrong or is that the correct value?

I've attached the schematic and the entire original article. Thanks in advance for any help y'all can offer.

ldb

[WQ9E]

A Pi Net is more convenient for a band switched amplifier but since this one is a single band unit there is nothing wrong with staying with the original link coupled design.  The old transmitters with separate plug in coils for each band generally have more efficient output circuits than those found with a band switched tapped Pi Net but the Pi Net is definitely more convenient for multi-band usage.

Check and make sure that your FT-897D can directly switch the 120 volt AC relay line.  I have an FT-897 (a great portable radio!) but I have not used it with an amplifier so check the switching specs and you may have to go with a different relay setup for the amplifier.

You may have to adjust the component value of the input network (L1 and C1) depending upon how good of a match they provide to your 897.  This amp was designed during the vacuum tube/pi net transceiver days and if it presents to much of an input mismatch your FT-897 will reduce its power considerably.  If you have an MFJ or similar antenna analyzer this comes in very handy when building an amplifier.  You can use it to adjust both the input and output networks of the amplifier; the output adjustment is done with power removed and a resistor from the 572B anode to ground to simulate the normal plate load of the amplifier.  Adjust the output link for 1 to 1 SWR with the MFJ connected to the amp output.

For the power supply, use 1N5408 rectifiers instead of those specified and lose the parallel resistors and capacitors across the rectifiers.  They are not needed with modern rectifiers and cause more harm than good.

I believe Doug DeMaw (or someone else) published a 160 meter conversion for the Heath SB-200 amp (another twin 572B amp) in several of the early/mid 70's handbooks and he probably used a pi net output.  You might look that article up also, especially if you decide to switch to a pi net output.


 

[KM1H]

Here is all you ever need to know about amp design.

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

Im a huge fan of the pi net output simply because it allows lower voltage tuning caps and is fully adjustable over a wide range of antenna VSWR. I use only coax fed antennas and on 160M a full size dipole will vary to around a 3 or 4:1 VSWR from one end of the band to the other. Efficiency is better IMO than link coupling, especially in a hard wired single band amp and it also allows a space saving toroid to be used.

Carl
KM1H

[W1AEX]

Just a quick note on the FT-897D as an AM driver. They have a fairly obnoxious tendency to overshoot when initially keyed up. Mine will shoot up to 100 watts out before dropping back down to whatever power it is set for. This can be a huge problem with grid driven amps, but will most likely only be an annoyance for you with the pair of 572's in gg operation. A few guys have said that they overcame the overshoot by applying ALC voltage with a regulated supply to the ALC input on the back panel. I never tried that, but it might be worth investigating.

Looks like a nice project!

[W2VW]

A Pi-net or link coupled tank for a single band won't make too much practical difference except as noted by Carl.

Your equation is missing the coefficient used for class of operation. The denominator is multiplied by the proper number. That plate impedance number is higher than it should be.

Anyone considering building a homebrew amp should base tube selection on available plate iron. What do you have?

A pair of 572Bs would be good for 125-150 watts carrier at full modulation at best. There are much better tube choices if you don't already have them.

[w3jn]

Rob, my FT1000MP has a huge overshoot using the hand microphonium, but not if you mash the "stab to talk" switch on the front panel of the rig.  Conclusion - the sound of the PTT switch snapping in, being picked up by the mike, creates the overshoot.

[W1AEX]

I wish it was a simple acoustic issue, but the problem is present even when keyed through software in the non-voice digital modes. It appears to be caused by the clumsy way that the ALC is used to control power output at lower settings. It's not the most annoying thing about that the 857/897 series of rigs, but it definitely got my attention the first time I set it up to run RTTY with an amp!

In the Yahoo user group for the 897D, if you do a search for "overshoot" there are quite a few threads about the problem.

http://groups.yahoo.com/group/FT897/

[w1vtp]

Just a quick note on the FT-897D as an AM driver. They have a fairly obnoxious tendency to overshoot when initially keyed up. Mine will shoot up to 100 watts out before dropping back down to whatever power it is set for. This can be a huge problem with grid driven amps, but will most likely only be an annoyance for you with the pair of 572's in gg operation. A few guys have said that they overcame the overshoot by applying ALC voltage with a regulated supply to the ALC input on the back panel. I never tried that, but it might be worth investigating.

Looks like a nice project!

Yes on the overshoot with the FT-857 too.  So every time you key the TX those grids are going to be jolted awake.  My 857 also suffers from low modulation in AM -- with the mic turned up to 100%

Al

[KD6VXI]

Just a quick note on the FT-897D as an AM driver. They have a fairly obnoxious tendency to overshoot when initially keyed up. Mine will shoot up to 100 watts out before dropping back down to whatever power it is set for. This can be a huge problem with grid driven amps, but will most likely only be an annoyance for you with the pair of 572's in gg operation. A few guys have said that they overcame the overshoot by applying ALC voltage with a regulated supply to the ALC input on the back panel. I never tried that, but it might be worth investigating.

Looks like a nice project!

Yes on the overshoot with the FT-857 too.  So every time you key the TX those grids are going to be jolted awake.  My 857 also suffers from low modulation in AM -- with the mic turned up to 100%

Al

The ALC trick is what the guys using the FT-1000 do. 

Interesting on the 857 that wouldn't make full mod...  Mine has MORE than enough..  I have the DTMF mic w/ rig controls and I'm set at 35 for AM Mic Gain.

With the stock mic, it was turned up near max...

--Shane

[N2DTS]

A pair of 572b's has 300 watts of plate dis.
I built an amplifier with four 813's with 500 watts total plate dis.
I get 175 watts carrier out, I could run it at 200 but that is really pushing it.

So figure, with the 300 watts plate dis, you are going to be safe at about 100 watts carrier output.

You will expend a LOT of power in heat to get a boost from 25 watts to 100 watts.

Put four 572b's in and have 600 watts of plate dis and run cool at 200 or 250 watts carrier.

Brett

[K5WLF]

Thanks to all who've responded to my query. Seems like the projected low output on AM might make this one not the ideal project. Going to four 572Bs, as Brett suggested, is one solution. I'm not against that, but not against one higher-powered tube either. I'm open to suggestions on that. An 833A is less than half the cost of a matched quad of 572Bs. Just a thought.

Dave -- I don't have any plate iron yet. The Hammond transformer that's spec'd in the article is still available, as are all the parts listed. The plate impedance seemed high to me too. Should I have divided by 2 since there are two tubes in parallel?

Rodger -- I'd planned to replace the 120 VAC T/R relay with a 12 VDC unit. The implementation as drawn seemed pretty clumsy to me. I have an MFJ-269 that I'll use to get the input and output networks tuned.

Rob -- I was aware of the overshoot problem with the FT-897D and had heard about the ALC trick. What I didn't realize until you pointed it out was that the effect of the overshoot would be lessened in a GG configured amp.

Carl -- On my midget city lot, I'll be using a shortened coil-loaded dipole so tunability will be even more important to me.

I haven't started collecting parts yet, so my options are open. All ideas and plans will be appreciated. I'd love to find a good 833A design. That tube would look great behind a window in the panel.

Thanks again for all the info so far.

[KM1H]

I built an amplifier with four 813's with 500 watts total plate dis.
I get 175 watts carrier out, I could run it at 200 but that is really pushing it.

Are those tantalum or graphite plate 813's?  If graphite, at what power out do they start showing a small bit of color?  My experience with the graphite version is that they dont show color until around 200-225W dissipation in Class C AM.

Carl
KM1H