Wire Substitutions

[W9ZSL]

This is probably one of those dumb questions.  Looking through many old schematics I've run across numerous references to wire used for coil winding including single and double silk and single and double cotton.  I'm not about to do a reproduction and dig around to find those components so what is the modern equivalent?  This refers to the input tuned circuit of the 813 amp.

It seems to me that the thickness or composition of the insulation would have very little effect on the impedance / ultimate resonant frequency of the coil as much as it being wound with stranded or solid core wire.  From what I've gathered, a few turns of #18 stranded on the cold end of the input coils as a link would do the job.  The coil itself should be solid.  Of course I acquired some B&W stock that will absolutely work but more compact alternatives are always on the table.

Can't be too careful or exacting.

[WBear2GCR]

 I suspect that the cotton and silk was to keep the coil via the wire being a sort of litz working at higher freqs...

Of course I am too young to know this for certain... also makes thinner wire easier to wind.

[WD8BIL]

It also establishes the space between windings.

[K1JJ]

Might as well do it right.  If you can find some Teflon insulated wire, you can use that for most anything. It has a big breakdown voltage rating and will not melt from soldering or high circulating currents. The coil turns difference to Litz, etc., is insignificant. You will probably need to add or subtract a turn anyway due to stray capacitance, layout or lead length.

When examining wire at a flea market, the test:  Teflon is shiny, slippery and a flame will not melt it.  Many times a seller will not know it is Teflon and group it with the cheap plastic insulated wire rolls.

The mark of a class rig is being wired with Teflon wire. The Collins round emblem S-line rigs are an example. Most of my homebrew rigs use Teflon. Well worth it.


eBay Teflon wire:

http://www.ebay.com/sch/i.html?_from=R40&_trksid=p2050601.m570.l1311.R1.TR6.TRC2.A0.H0.Xteflon+wire.TRS0&_nkw=teflon+wire&_sacat=0

T

[W9ZSL]

My main concern is for spacing between turns with cloth v.s. teflon or plastic.  I bookmarked that link at eBay.  I scored several B&W coils that match a home-brew input turret assembly for an 813 amp.  The only drawback by using them is they will take up more space than if I used 3/4 or 5/8" phenolic forms wound with magnet wire.

[WD8BIL]

I wouldn't worry too much about the little difference in spacing. As Tom said, a turn +/- or lead length can be played with to get it to work well. At HF frequencies it ain't gonna matter much.

[W9ZSL]

I'll use the B&W coils on hand.  I'm going to assume that the #18 stranded wire link on the cold end can be substituted with solid?  #1 is the schematic using the B&W stock.  #2 is the schematic of the amp I'm going to build.  I want to use the B&W input coils from diagram one and substitute them in diagram 2.

The first diagram is of an all purpose amp and includes a string of voltage regulators.  The input switch is 3P5T.  I'm not going to use this thing for SSB...CW and AM only. Schematic #2 is the one I'm going to build substituting the B&W coils and DP5T switch for the turret.  Comments?

[K1JJ]

The screen supply source shows a 100W 40K power dropping resistor connected to the 2.5 KV plate power supply to power the regulator tubes.  That sounds like a lot of heat and waste.  I would build up a separate small 50 mA  @ 500 V screen supply. (or whatever the screen voltage will be)  You can use the same regulator tubes and circuit.

Here's my calculations and reasons not to use the present screen source:  If we have a 2500V supply and a 40K dropping resistor and need 2KV to drop across the 40K power resistor leaving 500 volts for the regulators, that's  2KV/ 40K = 50ma.     50 ma X 50 ma  *  40K  =  100 watts of power resistor dissipation alone.   This does not include the heat from the regulator tubes and the loss in the 813 as screen current, though a separate supply would have this extra too.  Total power waste is close to the plate dissipation of the 813 plate.  I think the 40K should be of higher resistance to reduce the current through the Oa2s while still dropping 2KV across the 40K. R4 is only 2K so will have little effect on this.  And, if you do not key the HV supply somehow, the way the circuit is wired, the 40K will be sucking down the same power even in standby.

My calculations might not be based exactly on your parameters, but they show the potential problem.

To answer your wire question:  There is no significant difference between using solid or stranded for the coil on HF frequencies.


BTW, a single 813 in linear service is good for maybe  65 watts of carrier output....  a little bit more than a barefoot Ranger.  Since it is still early in the design phase and you are going through a great effort to regulate and do things right, consider two 813s in parallel.  130 watts of carrier is really a more reasonable level for AM these days.

T

[WBear2GCR]

Are we talking tiny receiver, RF/IF type transformer/choke windings, or HV power windings, like tank coils or RFC for plates??

[W9ZSL]

I'm talking about the tuned INPUT circuit of a single 813 class C amp.

[K1JJ]

I'm talking about the tuned INPUT circuit of a single 813 class C amp.

??  Referring to the first diagram above which you intend to build...  you do mean class B or AB2  linear amplifier, right?   It has regulated screen and grid voltage circuits.  Class C would not require this extra regulator effort.

You COULD run it class C  if you wanted, but that would be for CW or plate modulated AM only.  (Or "special class C/AB "  AM carrier linear mode)

T

[W9ZSL]

I'm building the SECOND amp using the first amp's input coils since I have the B&W stock on hand. 

[K1JJ]

Got ya.  Yes, the second amp is a better choice for class C. It uses pure grid leak and a low voltage screen supply.

What is the 6Y6G doing.... is that a screen overcurrent protection circuit?

How do you plan to PTT key it on transmit?  Connecting the filament transformer center tap to ground with a PTT relay works well. (center tap of T2)

Be sure M3, the plate current meter is well insulated and has a Plexiglas panel protector.  Better choice would be to connect it in series with the fil CT to ground and mentally subtract the grid and screen current.  

The screen circuit needs a meter. The grid circuit has one OK.

They are using a lot of shielded cable in the DC/ AC  grid, screen, plate and fil  circuits. Not really needed.

One 813 in class C is good for an easy 250 watts out -  quite ample.

Good luck with the project.

T

[Ralph W3GL]

What is the 6Y6G doing.... is that a screen overcurrent protection circuit?

Good luck with the project.

T

Tom,

It's called a "CLAMP" circuit...

Remove the RF drive and it shuts down the amplifier.  

I built that circuit back in 1950, used it on 75 AM with screen modulation (had separate switched in modulated screen voltage) and drove it with a signal shifter on cw...

Put it behind an standard 8 3/4" panel with power supply on the same chassis...  

Still have that box in my garage with most of my old HB gear...

Naw, not that circuit, the one I reference below that was  shown in his earlier posting , looking for simple 813 amp...
________________________________________

http://amfone.net/Amforum/index.php?action=dlattach;topic=40258.0;attach=48401;image

[Ralph W3GL]

Sorry, hit the wrong key...

[KA2DZT]

The 6Y6 clamp tube circuit only works if you use a screen dropping resistor for the screen voltage.  Using a fixed separate screen supply and the clamp circuit is out.

If you use some fixed bias along with grid leak bias you may not need the clamp tube.

I use my version of the circuit in my HB 813 and it works well.  I use all grid leak bias.

Fred

[Ralph W3GL]

Yeah Fred, I used the circuit in the simple 813 design likewise, no fixed bias, works fine as I said when I was trying to fix my post above...

[K1JJ]

The 6Y6 clamp tube circuit only works if you use a screen dropping resistor for the screen voltage.  Using a fixed separate screen supply and the clamp circuit is out.
Fred

Yep, that's what confused me. That circuit does not have a screen dropping resistor, thus how would it do anything other than draw some big current from the screen supply with loss of grid drive.  

Otherwise, I can see how it would work as a clamp tube once the voltage dropped across a dropping resistor.

I've never used a clamp tube - rather,  always use fixed grid bias to solve the loss-of-grid-drive problem...

Ralph, your circuit looks more like it, tnx.

T

[W9ZSL]

I really don't want to key the amp which is why I chose this particular schematic.  It was originally from the November 1954 QST and reprinted in the '57 handbook.  Yes, the 6Y6 is a clamper tube and the article states, "...is used in the screen circuit to reduce the input to the 813 to a safe value when excitation is removed, OR IF STAGES AHEAD OF THE 813 ARE KEYED."  I may simplify the input circuit further by using a single tapped coil.  I can always add a fixed bias supply, but would like to follow this diagram with the exception of the input coils.  This version calls for individual 3/4 and 5/8" forms wound with magnet wire.  I can substitute the B&W coils or wind one tapped on a 3/4" form.

This particular schematic shows a rotary inductor in the Pi Net; I have one (G.E.) which I can certainly use and probably will rather than mess with switching, though that is an option because I have an Air-Dux also.  I have LOTS of meters including one for the filament, grid and screen...all matched, NOS Simpson 3".  I'll probably put the 300 MA Plate in the filament CT.  I also will include Plate Volts & Plate current meters for both RF and Modulator ( 2 x 811As).

I can always go with the "simple" 813 amp Ralph, but modified with a Pi Net as I mentioned earlier.  It would eliminate the need for a screen supply HOWEVER, since I'm going to Plate Modulate, the best bet is the schematic below.  Besides, if I don't build it, what the heck am I going to do with all those .001 bypass capacitors?! (LOL).  I'm definitely NOT going with the amp with the string of voltage regulator tubes in the screen.

[w1vtp]

<deleted comments>

[K1JJ]

Sounds good on the meters, etc.

Can you post the screen power supply back from where the "+SC" connection connects ?   There needs to be an additional power resistor in series to make the 6Y6G clamper work.  

In addition, there needs to be a way to modulate the 813 screen with a series choke or resistor.  This assumes you will be plate modulating this as a class C stage.

I would recommend PTT keying the 813 with the fil CT technique. Use a 50K resistor to ground and short it out with a relay.  Keeping the 813 keyed all the time with all voltages applied, even though there should be little to no idling heat in class C, can produce receiver noise and possibly other problems.... Keeping a live final stage always keyed is not a common practice.  

T

[W9ZSL]

I can provide any screen voltage through a 15 Hy choke.  Still not sure about keying.  As far as it's concerned, I'm still wrestling with how to excite this thing.  If I can use my Kenwood TS-440 dialed down, great.  I also have a Heath DX-20 which would be a hoot, but without modifications, that is no-go.  I can build a 6AG7 oscillator followed by a 6V6 which was the original idea.  I have plenty of relays for keying anything.  When it comes to building stuff, I'm good at that but when it comes to design, not so much!

[Ralph W3GL]

.  If I can use my Kenwood TS-440 dialed down, great.

This is the way to go.  Use the 440 in CW mode and just set the output level to what is required to drive the 813.  Plug your PPT relay in parallel to your key line and your good to go...

You will have to watch the 440's output for a short time due to a typical Kenwood trait called "Carrier Creep" ..  

I use a 440 to drive a final as do several other hams on the air, one being  Nick, KG2IR and another is Jay, N3WWL
.
Good luck with your project...

[K1JJ]

Here is a proven 813 circuit that is plate modulated by a pair of 813s.  Use it as a reference. It has most of the circuits we have discussed.

What tubes do you plan to use for the plate modulator... and do you have a modulation transformer yet?

http://www.amwindow.org/tech/htm/813/813.htm


You can use a 15H screen choke for the screen modulation, but will still need a power resistor in series for the clamp tube to work. The choke does not have enough DC resistance to drop the voltage needed during a fault. It probably originally used a power resistor as both the clamp tube dropping resistor and as a screen modulation impedance.  Check out the circuit  you have to see what they did there.

T

[W9ZSL]

I have a UTC NOS S-22, 250 watter; 811A mod tubes.