Amps
[Top] [All Lists]

Re: [Amps] TL922A recommended mods?

To: amps@contesting.com, 4cx250b@muohio.edu
Subject: Re: [Amps] TL922A recommended mods?
From: Vic K2VCO <k2vco.vic@gmail.com>
Date: Tue, 14 Aug 2012 09:01:24 -0700
List-post: <amps@contesting.com">mailto:amps@contesting.com>
Jim,

The clunky relay is actually dangerous, because it is so slow that it will 
hot-switch with 
semi-QSK CW. This is the reason for the spark gap that is installed in the 
stock model, 
and (in my opinion) for the warning in the manual that operating CW with the 
higher plate 
voltage available in the SSB position of the switch will cause serious damage 
to the 
amplifier!

I replaced this relay with the usual Jennings RJ1A-type relay, plus a coaxial 
reed relay 
that I got from Allen Bond, K4MGS. The reed relay also had an extra set of 
contacts that I 
used to increase the bias on the tubes to cutoff when not keyed. I used a 
simple circuit 
similar to that on Measures' web site to operate the relays. I have a 
transceiver that has 
a reasonable delay between the amplifier key line and RF output (a K3) so I 
didn't use any 
special sequencing circuitry.

I left the original relay in place, cut and labeled the wires leading to it. 
These wires 
go through a harness to the back of the amplifier and the power supply area, 
and I managed 
to use them for the new circuitry.

I added a simple step-start to protect the power switch. There is plenty of 
room near the 
terminal board where you select the primary taps on the transformer.

It is also true that the filament voltage is high, which is bad for tube life. 
I put a 
resistor (I THINK it was 15 ohms) in series with the filament transformer 
primary to 
reduce it to 4.95v from about 5.25v (in my case). This also slightly reduced 
the voltage 
on the pilot lamps, which are annoying to replace.

The existing bias cutoff circuit is such that it will burn out the transformer 
in the 
event of a grid-filament short, and should be changed. This is a simple 
modification. I 
used a different bias arrangement, so I didn't do this.

I replaced the VHF choke in series with the plate RFC with a 15-ohm glitch 
resistor. I 
added several additional bypass capacitors on either side of it to help keep RF 
out of the 
filter capacitors, especially on 160M.

The amplifier was unable to match a 50-ohm load on the bottom of 160M. I added 
some 
capacity to the 160M padder to make it work, but the tank circuit really 
requires more 
inductance on 160. I think it was originally designed for 1.9 mHz. I didn't 
want to fool 
with it, since I am not a 160M nut (yes, this was the case BEFORE I replaced 
the VHF choke 
and added the bypasses!)

The above are the modifications that I think are required, or at least helpful 
and 
harmless. The following are slightly more controversial.

Some people have connected a wire from the tuning capacitor rotor to ground 
(this is 
probably the 'bizarre' one that you mention). If you look at the way the tank 
circuit is 
built, you will see that the path from the rotor back to the grids is long and 
depends on 
several sheetmetal junctions. I put a strap from the bracket that the capacitor 
was 
mounted on to the main chassis near the grids.

I increased the bypassing around the filament choke. I thought 0.001uf was 
inadequate for 
160M.

I find the arguments of W8JI about grounding the grids directly convincing! So 
I took out 
all of the components between the grids and ground and grounded each grid 
terminal 
separately to its nearby ground post with a copper strap (it would have been 
better to 
connect the strap directly to ground right next to the terminal, but I didn't 
bother). I 
won't go into details about this, but W8JI's website discusses it. There is 
another 
'bizarre' mod that involves strapping the terminals together. I can't see how 
this would 
help anything.

Since I'm a CW operator, I set the bias for a total of 100 mA ZSAC for two 
tubes instead 
of 200 mA, for a little bit less dissipation and more efficiency. I would not 
do this if I 
operated SSB, since it would increase IMD.

I have been using the amplifier on CW for about a year (I always set the CW/SSB 
switch the 
SSB position!) and it produces 1300 watts or more on all bands except 160 and 
10, where it 
is less efficient. I believe more output would be obtainable on 160 if one 
wanted to add 
inductance to the tank circuit. It's completely stable on all bands.

On 8/14/2012 6:35 AM, Jim Garland wrote:
> A friend has a recently acquired lightly used TL-922A that is unmodified and
> works perfectly on all bands. He has asked me to replace the clunky T/R
> relay with a quieter & faster vacuum relay, and also to install a solid
> state relay to take the burden off the power switch.  I'm not familiar with
> this amplifier, but from looking at on-line postings it appears to be an
> amplifier that owner's can't keep their hands out of! I've never seen such a
> hodge-podge of proposed modifications, some of which strike me as bizarre,
> others of which purport to improve on the original design, but which appear
> to be mostly preventative maintenance rather than solutions to actual
> problems. Still other mods probably work just fine, but are hard to install
> and require complicated circuitry.  So, my question is whether anyone can
> recommend a simple way to swap out the noisy antenna relay and take the
> burden off the main power switch? I'm particularly interested in the
> mechanical layout, routing of wires, etc.  I've not looked inside the
> amplifier yet, but from photos it appears to be rather difficult to work on.
>
>
> Thanks!
>
> Jim W8ZR
>
> _______________________________________________
> Amps mailing list
> Amps@contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
>

-- 
Vic, K2VCO
Fresno CA
http://www.qsl.net/k2vco/
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps

<Prev in Thread] Current Thread [Next in Thread>