208V->230V autotransformer reversed for 230V->208V is OK right?

[Opcom]

208V->230V autotransformer reversed for 230V->208V is OK, right?

Can't think of any reason why not.

I want to use it to power a 208V step down filament transformer, and it would be much simpler to mount it to a panel and forget about it than have to drill holes and use up a variac for just doing 230V -> 208V. Only going to use at 20% its power level. (space isn't an issue, but this unit isn't very big either).

When the transformer was found, I didnt believe the 3.1KVA rating at first. Then the light came on that it's an autotransformer, so its own portion of the load power @ full blast is only around 300VA through the 0-208V winding.

So, I presume the step up portion of the winding (from 208 to 230V, the top 22V) is rated about 13A, and the remaining portion, the 208V section, has only to provide the 300VA of power/excitation for the 22V@13A section.

Reversed, the current through the 208V section seems to be the same, as would be the current through the 22V step-up section. The 208V section should have the same current for a given load either way, but any difference through the 22V section depending on how it's wired is supplied by the 230V or 208V line, depending on whether it's step up or step down.

Am I trying to think too much about it?

[KA2DZT]

The xfmr will work fine the way you are trying to use it.

Fred

[Opcom]

Thanks! I don't know why I just couldn't see that.

[WD5JKO]

Patrick,

   We miss you and Colonel Tucker on 75m....

On that auto transformer I see that the tag says 85 degrees Celsius Rise.

That means to me that the core rises 85 degrees above ambient when running at full rated load.

Some transformer makers rate for a 45 degrees C rise which usually means they design in a lot of excess capacity.

This one is either being spec'd to the limit, or they weld the lamination's together.

Welded  lamination's cause high Eddy currents, and the transformer gets hot as heck, even when lightly loaded.

Welding is a cost reduction step sometimes taken.

Jim
Wd5JKO

[Opcom]

Jim,

Colonel Tucker has seizures on 75M. He's unstable on 40M but careful tuning and lots of drive helps. Much of this is probably the RF deck layout around the plate area plus the neutralization -was discussed a few years ago, and I thought it was fixed but then worked on the grid circuit.. I think what it really needs is some grid swamping to fix things up. Previously I had a 3000 Ohm NI resistor from grid to GND and it was very stable, but the national MB-40 would not reach high enough frequency for 10M, so I took it out. Maybe time to put it back and eventually do something else besides the MB-40.

For whatever reason, the PI network's coil is next to the loading caps, instead of next to the tuning cap. The distance between them is really excessive, maybe 10" to 12", with the tube anode in the middle.

The 'proper' alternative is to completely re-arrange the RF deck and I don't want to drill more holes in the front panel plus cut a square hole for the PI coil turns counter. - or put in a 'ham band' coil, but the re-arrangement is still needed.

See the long strap from the plate tuning cap (right) to the PI roller coil (left). Tube's in the middle. That tube likes VHF and it's really sensitive. I'm still open to suggestions but still against moving the plate coil. There's no guarantee of that fixing it, since the loading cap bank is to the left side of the PI coil, even farther away from the plate, at least 8".

In the pictures, disregard the very long HV lead to the plate choke. That was fixed when the choke and bypass cap were moved as close to the tube as possible.

sorry for the long tale on Col. Tucker, but he been sick.

--------------------

You got me thinking about the 80 deg C rise. I will have to look at the transformer, take the cover off and see if the core is welded. The 208V load is less than 500VA so it is light duty. If things don't work out I could find a generic transformer and set it up to buck the 230 down to 208. Or do like Bubba and just use a big Ohmite resistor in there.

[WD5JKO]

Pat,

   You might consider loading that MB40, or just backing out those Mod's back to a passive grid with resistor loading. The plate side, as I understand it, has been hard to tame on 75m. I know you moved the plate choke, but I wonder if that plate choke has insufficient inductance on 75m, or is too close to being resonant?

   I hear that 40m has been opening up more and more, usually mid afternoons. Maybe that is an option for Colonel Tucker as he is now?

Jim
Wd5JKO

[Opcom]

My best option is taking your advice that it needs a resistor across the grid (to gnd).
I need a tank or something for a voltage step-up since it's class C, -150v bias, and drive is by a solid state transmitter.
I could live without 10M but it seems cheezy to let that go since it originally had it.

40M has been open, but since the crowd moved to 80M, I have not been on and other projects have pushed their way into the operating area and required some RF to drive them, so a lot of things are disconnected right now. I seem overwhelmed by tasks that have cropped up this year, so it's going to be a while before I'm back even with an amplifier & TS-430S.

The plate choke is a Raypar RL-100. Its instructions say to avoid the vicinity of 25 mcs. where the4 coil's first series resonance is located (22.5 to 26.9 mcs.). I don't think its ever been a problem but I'm not an expert on these.

On un-modding it, Intractible problems prevented me from making it work as found or fixing the original guts (a theme throughout this TX but that's another 5 or 6 stories).


The original grid circuit was a Barker and Williamson JCTI 10-80 grid tank used with a custom variable cap value, see below. This is a 'turret style' unit with the bandswitch in the center and coils around it.

It was ruined in a few ways: age-related decay of the plastic coil supports. It just fell apart when removed for repair, but it had also been irreversibly modded by the builder in the 1950s.

He used a larger tuning cap to resonate those coils on the next lowest band, and changed the coils like this:

orig
coil -> band switch position
80M -> 2726KC -Texas State Guard HF freq.
40M -> 80M
20M -> 40M
15M -> 20/15M via additional switch
10M -> 10M

- so it was a big mess including the tuning cap which is the wrong value assuming I could have fixed/replaced the coils.

The remedy was an MB-40. This worked well but the TX was impossible to make stable, even after taking the advice of many people.

A 3K non-inductive resistor was used across the MB-40 grid side to calm things down. It worked decently but the MB-40 would no longer tune up to 10M frequencies.

I removed the 3K resistor. 40M was stable as long as enough drive was used and it was very carefully tuned and loaded, but it would act up sans drive if the slightest thing was wrong. You might remember me on 40M a few times hurriedly saying that something was wrong here and turning it over. That was when the TX was acting like it wanted to have a big crap out.

The 4-1000 is very sensitive. If I'm unwilling to re-layout the plate area in a sensible manner, then I need to put a grid swamping resistor back in. I tried a 15K and a 3K. The 3K made me furnish 40-50W of drive but it seemed very stable.

------
On the transformer, I inspected and was happy to find no welded core. The core is only what can be seen in the picture. The rest of the enclosure is empty room for electricians' wires, wire nuts, etc. Next time I'm out there I'll hook it up to 230V and see what current it wants to take & if it gets hot, etc. I agree it'll be fine unless there's something else unusual about it.