That would help out...
Originally Posted by M0ELS
one of the other things you need a grid current meter.
This will aid you in finding where you are loosing efficiency.
Hi Gary, I dont have another choke to try, but will try and source one locally in the UK via ebay or elsewhere. I never thought that the higher bands would be/could be so critical to get working. Attached is where I am at now as I have placed the big coil into the vertical position and brought the 10m coil much closer in to the plate capacitor.DSCF4514a.jpg
According to the RCA specifications document on the 813, the tube is usable up to 120 MHz (76-percent power) and at full ratings up t0 30 MHz.
However, those frequency ratings would have to be obtained under VERY specific circumstances including special circuit design. Frankly, the majority of amateur radio operators have had problems getting satisfactory operation on the 10-meter band and, for many, even on the 15-meter band, especially in grid driven designs.
In the older ARRL Single Sideband Manuals, there are single band linear amplifiers described for bands up to the 10-meter band. However, these are grounded grid and not grid driven.
There are detailed information on the net that describes a method that gets very good efficentcy on all bands. It use some very interesting circuits to accomplish this and those circuits would make multiband operation exteremly difficult. On a single band it would work very well.
On the plate choke it would be best to have one that you can substitute. If there was some way to make a parallel tuned circuit for 10 meters that could withstand the stresses of being in the plate circuit. Then you could use that in place of the plate choke since it would be a high impedance to the 10 meter signal but a very low resistance for the plate voltage. This would be for a test purpose only, unless all you want to work is 10 meters. The isolation of this would be very important since cross coupling would cause problems. To make this circuit you would need to have the componets that would make that circuit and some way of tuning it for proper operation. Again a GDO would be very helpful, but if you had a GDO you wouldn't need to do this since you would know if the plate choke is causing the problem or not.
A plate choke is just a bunch of wire wound on a insulating form. You could probably make one and see how it works. The thing you are trying to do is to see if it changes anything. An inductance of 25uh would have a reactance of about 4400 ohms. That might be too low but using the same application with a 100uh choke the reactance would be about 17,500 ohms.
You can try this and see if it helps.
I have reworked the lower freq coil and now getting 400w out on 15m, which I think is near its optimum range. Changing the 10m coil size and amount of turns, make not much difference now. I was given another plate choke to try, which will happen over this weekend.
Good to hear you are making much progress.
Looking forward to hearing hos things are progressing.
Since grid-driven 813s are not even close to being linear, converting to g-g/cathode-driven with tuned inputs whose Q is 2 to 3 would be a good idea because it would make the amp usable on SSB. However, IME good 28MHz efficiency is expecting a bit much from an 813. The good news is that the 813 is quite unlikely to be able to oscillate at VHF.
Originally Posted by M0ELS
Last edited by AG6K; 05-05-2012 at 08:27 AM.
If you do not have an input circuit its no wonder you are having low power as the frequency increases. If you want to run swamped grid about 50W or more will be needed and the 813 requires the bias to be for AB2 if you want efficiency. IMD will be in the low -20's which is typical in much of EU anyway
The grid driven 813 also usually requires neutralization on 10M and it is also quite prone to parasitics with the layout you have. The sockets should be recessed with the tubes internal grid leads below the chassis.
Finally, are those NOS tubes or unknown condition?
Thanks for the replies. We may be close to europe, but not part of them sofar...hi hi. The sockets are recessed below the chassis and the tubes unknown RCA types. I am going to look into having a tuned input circuit for just 10m for now. The space on the chassis is limited so it may mean a rebuild of the amp.
Ok, Been busy with the amp and umpteen changes to the 10m coil, a change in the plate choke with no real difference to 10m pwr out. I dont think there is much more I can do. I tried a trick I saw on the AMP reflector which was to connect the plate cap to the centre of the 10m coil, but this killed off the output. I have not mentioned that the 10m band switch tap position is 2 turns from the low freq inductor side. Any changes to the tap position kills the output......