SB-220 HV Transformer Failure

[KN4SK]

About a year ago I replaced the tubes (with a new matched pair) in my dad's SB-220 because one went out. I also replaced the high voltage capacitor stack, bleed resistors and diodes, installed a rectifier board and a new cooling fan, all from Harbach.

All has worked well until my dad called saying "smoke is coming out of the linear". Tearing into it, everything I installed looks fine visually and I tested the resistors and diodes from the HV cap stack.

Below the HV transformer I can see all kinds of "stuff" that appears to have come off of it, maybe the water/humidity-proofing varnish dip melting off?

http://www.teamhokies.com/hosted-pics/2015-11-24%20SB-220%20HV%20Transformer.jpg

Anyway, clearly a HV transformer failure and I found the Hammond part number replacement (PWDP13034, stocked at pwdahl). My question is, is there something that may have caused the failure aside from degradation over 40-50 years of use? In other words, I don't want to replace this and blow the next one. Looking at the schematics and the amp itself, there's nothing jumping out at me.

Thanks all,
Ray, KN4SK

[N2DTS]

That happens to transformers when they are run past their limits.
If you get them hot enough long enough and nothing else craps out, they self bake.

Its an SSB amp, so it should be limited to about 200 watts carrier I would think, more for short transmissions.
Everything has a maximum rating and a duty cycle rating.

[WD5JKO]

 Interesting dilemma here.

  Was wondering about the Antekinc line of toroidal transformers. The one linked below is 800VA rated, and has two 800v wingdings. In parallel with full wave voltage doubler rectifier, this could make 2KV + loaded (about 2250 idle). There are two 6.3v@5A filament windings. What else would an SB-220 need? Might be an ALC winding, but that would be optional.

  Not bad for $105, and a in stock part.

http://www.antekinc.com/an-8t800-800va-transformer/

Edit: I was thinking SB-200 with the part linked. For the SB-220, there is also an Anatek 950VAC 1500VA which would make around 2.7KV idle. For filaments, another transformer would be required.

http://www.antekinc.com/as-15t950-1500va-950v-transformer/


Jim
Wd5JKO

[K4RT]

Is one or both transformer windings open? 

[WD8KDG]

If only a transformer issue, another option WZ1M Gary, if he still rewinds those things.

Craig,

[w1vtp]

That happens to transformers when they are run past their limits.
If you get them hot enough long enough and nothing else craps out, they self bake.

Its an SSB amp, so it should be limited to about 200 watts carrier I would think, more for short transmissions.
Everything has a maximum rating and a duty cycle rating.

If only a transformer issue, another option WZ1M Gary, if he still rewinds those things.

Craig,

I agree with Brett on the SB220.  This is not an ideal amplifier to be running on AM mode.  The PS is a voltage double affair and the transformer is really good only for SSB service.

Craig has made an excellent suggestion.  Contact Gary for a transformer rewind

http://members.tripod.com/tubes_tubes_tubes/transformerrewindingservice/

I just looked at his page (see above) and the price for a rewind is $262.  Looking at your picture: why is that transformer cocked at that angle?  Did the amplifier get jarred?

I owned a SB-220 in a former life and loved it - built it as a kit.  It will always have a special place in my memories. But only in SSB service

GL, Al

[KN4SK]

Thanks to all for the replies.

For some info on operating conditions: My dad drives it with an FT-1000, 200W output. He leaves the SB-220 in the same mode (CW I think) all the time and almost never turns the drive down from full power on the FT-1000. He will literally tune it up under full power drive. Sparks fly in the tuning caps now and then. We use it for mobile and rover contesting where it gets unhooked/rehooked up to antennas several times per day, sometimes into an open circuit, etc. because it wasn't checked carefully.

In other words, he basically abuses the dickens out of this thing. It's a testament to the robustness of this amp that the tube failure last year was the first issue he's ever had in ~40 years of use. I've preached the gospel about taking better care of it, but that's not going to happen.

It gets used in CW and SSB mode for county hunting and contesting (including outdoors during summer/winter). No AM or other modes.

I haven't unhooked the transformer to see what windings are open. I've got an MSEE and have been an RF designer (or manager, gasp) for my career so I'm pretty fluent in RF, but not so much in magnetics. I guess this transformer is wound as indicated on the schematic? I'll ohm it out when I get into the workshop tomorrow.

Thanks for the info on the re-wind, I hadn't considered that. I'm also curious about the Antekinc part, though my dad doesn't mind spending $300 on a transformer and my inclination is to stick with something that has proven its mettle.

Thanks again!

[N2DTS]

Hard to tell from the pic but a transformer overload is normally a long term thing, not a short term overload.
Short term overload takes out fuses, ckt breakers, diodes, meter shunts.
Long term overload melts tubes, cooks transformers and chokes.

When you replace weak links with strong ones, something else fails in a different way.
I have four 14kv 1 amp (250 amp surge) diode bricks in my power supplies, full wave center tapped, and when I forget to remove the shorting clip and key it up it blows the 15 amp breaker.

[w1vtp]

Thanks to all for the replies.

For some info on operating conditions: My dad drives it with an FT-1000, 200W output. He leaves the SB-220 in the same mode (CW I think) all the time and almost never turns the drive down from full power on the FT-1000. He will literally tune it up under full power drive. Sparks fly in the tuning caps now and then. We use it for mobile and rover contesting where it gets unhooked/rehooked up to antennas several times per day, sometimes into an open circuit, etc. because it wasn't checked carefully.

In other words, he basically abuses the dickens out of this thing. It's a testament to the robustness of this amp that the tube failure last year was the first issue he's ever had in ~40 years of use. I've preached the gospel about taking better care of it, but that's not going to happen.

It gets used in CW and SSB mode for county hunting and contesting (including outdoors during summer/winter). No AM or other modes.

I haven't unhooked the transformer to see what windings are open. I've got an MSEE and have been an RF designer (or manager, gasp) for my career so I'm pretty fluent in RF, but not so much in magnetics. I guess this transformer is wound as indicated on the schematic? I'll ohm it out when I get into the workshop tomorrow.

Thanks for the info on the re-wind, I hadn't considered that. I'm also curious about the Antekinc part, though my dad doesn't mind spending $300 on a transformer and my inclination is to stick with something that has proven its mettle.

Thanks again!

First of all, Ray.  It's a mistake running the amp in the CW mode during SSB as that is tapping down the HV a little.  Not sure why Heath did that except to say the transformer is, in my opinion, a underpowered thing and tapping down the winding (lower HV) might have been a stress management thing..  He should be running it in the SSB tap for SSB and CW tap for CW.  

Having said that, Gary would be using modern materials which should make the transformer tougher.  I recommend staying with the original transformer since anything else would probably not fit.  If it's being used for CW & SSB service, it should last a long time

GL Al

[Opcom]

Not everything has a duty cycle, unless that includes a 100% duty cycle. The SB-220 at this point would be a good candidate for use with an adequate external power supply and maybe adding more cooling in the space previously taken up by the old power transformer. Those amps are not very well cooled.

[WZ1M]

I believe the SB-220 primary voltage is 115/230. With todays house voltages approaching 122/244 will push the VPT (volts per turn) higher causing the transformer to go deeper into saturation. This means more current draw and more heat. The original volts per turn on the full primary is 1.0550 which is pretty high to begin with. Not many turns in the primary of these transformers. It only takes 5 turns for each primary to bring the VPT back down to a reasonable level for a much cooler running transformer. When these transformers were built, most of them were filled with epoxy. Bad idea in my opinion. Keeps the heat in. Yes, I am still winding these things. Done 4 this year.
Regards,
TRS
Gary

[k3msb]

Hi Ray

Looking at your picture, the power transformer looks to be mounted at the front of the chassis near the two switches.

I've worked on several SB-220 amps for people in the local radio club, and the power transformer is always located in the right rear of the chassis (viewed from above)!    Am I interpreting your picture correctly?

As Jim WD5JKO mentioned, those toroidal transformers may be worth looking into.  I've been considering going that route for my first ART-13 power supply.

[KN4SK]

Looking at your picture, the power transformer looks to be mounted at the front of the chassis near the two switches.

I've worked on several SB-220 amps for people in the local radio club, and the power transformer is always located in the right rear of the chassis (viewed from above)!    Am I interpreting your picture correctly?

OK, I wrote down "HV Transformer" on my note pad but I just re-checked the schematic and it's the Filament Transformer". Geez, bozo over here! Good news is that it's only $150 from Harbach.

Hopefully this isn't an indication of anything amiss with the tubes, in terms of why it went out?

[N2DTS]

Maybe after all these years it got an inter-turn short and cooked up?

[KN4SK]

I did pull the tubes and ohm out the filament to grid and to filament to plate (and grid to plate) and they were all open circuits. The filaments measure 0.1 - 0.2 ohms DC resistance.

I wouldn't think any kind of oscillation would cause problems in the filament circuit (I have minimal tube experience).

I think I'll just order up a new transformer and drop her in. I'm guessing old age took her out....

[KD6VXI]

Check the bias zener as well.

While in there,  install a large filter C on the bias supply.   Will help keep it stable.

Ground the grids.   Helps with stability.

I ditch the zener,  or use it for ssb/CW.   Then add a string of diodes for AM.   Takes about 20 more.   Having a .6 kva xformer,  it needs all the help it can get.


As I recall,  the fil xformer also has the - 90 to - 120 volt cut off running from. / through it.   Another fubar idea.   Use a resistor for cut off.   Rich Measures has  mod documented on his somas website.

Running the amp on CW mode just  voltage on the anode.  Causing the plate impedance e to remain the same at 1kw to 2kw input levels.

Your imd is more effected by bias not being regulated than the lower V on the anode,  from what I've experienced on the 220s and similar.

--Shane
KD6VXI   

[k3msb]

Hi Ken

If you do have a grid to cathode short, that may cook the bias transformer that is part of the filament transformer.

Can a momentary short be caused by unwanted parasitic oscillations?  Someone with more knowledge than me needs to answer that one.

When you remove the old transformer, it might be wise to see if it's the filament part or bias part that is zorched.

If memory serves me correctly, if you do have a grid to cathode short,  there is a choke under the chassis (grid choke?) and a zener diode (on the meter board?) that may be taken out.

I would do some investigating before just dropping a new one in.

Then again, as others have posted, it may have just been time to go......

[w1vtp]

<snip>

... Done 4 this year.
Regards,
TRS
Gary

There ya go Ray.  You heard it from The Man himself!  I would stick with stock design with the exception of proven mods (from Harbach, for example).  Get the transformer done by Gary and TRY to get your dad to operate it correctly.

It's a classic and deserves some respect. 

Guys, I think from reading this thread with comments, especially from Ray that we can dispense with any AM related special mods and just run it as originally designed

http://harbachelectronics.com/

I don't think it measures up to today's 2 x 3-500 amplifiers but it has my respect being a former owner

Al

[N2DTS]

Part of the 'fun?' of Heathkit was there was nothing they made that could not be improved, sometimes a LOT.

[KD6VXI]

The only am mod I spoke of was increasing the bias.   The rest are just good engineering practice.

The bs floating  grids idea has been hashed about for years.   Eimac finally stated direct grounding of grids is best.   W8JI also stated the same.   The inductors used by heath can be measured in ohms not. Milliohm......   Therefore your bias jumps all over the place.

Putting a cap across  output of  bias / center tap holds the bias value over  entire operating point of the tube, instead of letting it dance about.

The same bias mod would work for any content carrier mode....   Rtty,  etc.

The Pete Dahl usually has more B+ on it.   This can lead to Ctune arcing.   There is a replacement Ctune and when you use  newer hv cap,  your bandswitch is the next thing to go.

The 220 is a good amp.   I love mine.   With a bit of love,  it can be made better.

I'd have Gary rewind the original xformer.   I bought a filament xformer from him,  love it.   Works great,  runs cool


--Shane
KD6VXI

[WZ1M]

I can guarantee that its the bias winding that is "ZORCHED"
Regards,
TRS
Gary

[KN4SK]

Thanks again for the replies. I did ground the grid the last time I was in there working on this thing. No other mods that I didn't note already in this thread. Dad is putting a new tower up next week and wants the linear back on the air so I'll be buying a new one from Harbach. Thanks again all.

Ray, KN4SK

[KN4SK]

I found a post by K5JV that states he saw a leaky cap in the relay bias circuit cause a filament transformer failure:

http://lists.contesting.com/_amps/2009-08/msg00543.html

I think this would have to be C4, a 20uF electrolytic cap to ground, rated for 150VDC, run at 120VDC nominal. I'd have guessed it would take a substantial leakage current to overheat a transformer that's this physically large. The post seems credible but it doesn't make sense to me, intuitively, that this cap could leak enough without showing visible signs of damage.

I suppose the I ought to pull that cap and test it. I've got a variac and a 34401A so generating the 120V with a single diode rectifier/filter cap and measuring leakage would actually be easy.

[K4RT]

I suppose the I ought to pull that cap and test it.

The smoke & transformer cook-out suggests that the transformer took one heck of a beating, but the windings might still be intact - although no longer in a condition to provide useful service.  The xfrmr may need to be rewound or replaced, but also check associated PS components, eg caps resistors, chokes, for damage or failure.