800vac supply on PT secondary

[unclerny]

Hi Forum, I'm trying to figure out an issue with an old Sunn I have and it's blown fuses if powered up too fast since I've owned it.  I decided that since it's been repaired so many times by several different techs, I'll gut it and start from scratch.

One of the first things I figured I should do was verify that the PT wasn't he source of the main issue.

It has 6.5vac on the filament winding with 110vac on the primary, unloaded.

The B supply has 800vac across it unloaded.  Considering the amp uses a GZ34, plate to plate of 550vac, I think I'm onto the root of the problem.

Can anyone comment on their thoughts about this?  Do I need a new PT or am I missing something?

If I install the GZ34 my B+ right out the gate is over 600vdc, that can't be good.

[6G6]

Since they copied the Dynaco design, they kept the GZ 34 rectifier.
A lot of them used two of them.

600V unloaded isn't too extreem, because it will be lower unloaded.
I think they should run around 525V loaded.
It could be higher, now that you may have 120VAC going in, instead of 110.

Has it ever been recapped?

[DummyLoad]

Hi Forum, I'm trying to figure out an issue with an old Sunn I have and it's blown fuses if powered up too fast since I've owned it.  I decided that since it's been repaired so many times by several different techs, I'll gut it and start from scratch.

One of the first things I figured I should do was verify that the PT wasn't he source of the main issue.

It has 6.5vac on the filament winding with 110vac on the primary, unloaded.

The B supply has 800vac across it unloaded.  Considering the amp uses a GZ34, plate to plate of 550vac, I think I'm onto the root of the problem.

Can anyone comment on their thoughts about this?  Do I need a new PT or am I missing something?

If I install the GZ34 my B+ right out the gate is over 600vdc, that can't be good.

assuming that would be end to CT of 400V? in other words 400-0-400 so then each plate of gz34 gets 400V x 1.3 = ~ 530V - 600V unloaded would not be unreasonable.

rebuild the bias supply, the PS, and rewire the OT and output tube sockets, install the output tubes, put a load on OT, and see what you get then. i'd bet around 550V or so.

what model sunn are you working on? schematic available?

[ernie_jr]

i have rebuilt several sunns and even made a copy from scratch. your voltages sound ok unloaded. these amps are easy
to rebuld. all my rebuild i have done using a hoffman board type layout. sounds like it probaby needs a cap job. also, the
resistors, especially the 15k and 33k in the power section tend to drift after this many years.. i would replace them with
some new metal oxide ones. add a grounded power cord for safety and you should be good to go.
good luck,
ernie

[PRR]

Which "Sunn"??

The big Sunns were big Dynacos, except where Dyna ran 450V-475v Sunn pushed 520V-540V to force 20% more power. Difference between stay-at-home leisure amp and go-out working amp.

To get clean 525VDC without absurd cap costs, some Sunns deviated from a simple big first filter cap charging to nearly the peak AC voltage. Instead they have a small first-cap, a small(ish) choke, and a small second cap. The ripple on the second cap is quite acceptable. The ripple on the first cap is astonishingly large. The choke averages that. The DC is significantly lower than peak AC. But un-loaded, the ripple vanishes, the average DC gets very close to peak AC. Much higher than the loaded voltage.

600V un-loaded does not seem wrong. You do have to buy caps for that cold-start 600V surge. Two 350V in series, minimum. Or considering that these pioneering amps turned out to have lasting value, it might be time to break the bank and find 600V oil-caps which will last a lifetime not mere decades.

I've just simulated a Sunn Model T, a four-6550 amp supposed to idle at +519VDC and roar at +481VDC, and probably pulling over 500mA at roar.

First: go with what you know. When caps are cheap, use one uFd per mA, or maybe half that. A lot of 80mA 2-6V6 amps use 40uFd. So a simple one-cap supply should use 250uFd-500uFd at Sunn T power demand. That's brutal, we see why Sunn looked at other ideas.

Sunn T uses a mere 20uFd for the first-cap. Less than 1/10th the "required" size. They get 150V pp ripple, yowsa!

But after some choke and another 20uFd, it is 13V pp ripple, 2.5%, clean-enuff for pentode mode.

To get 525VDC at reasonable idle, 480V at full roar, I had to use 420V AC (each side). Close to your 800V AC reading. Maybe Sunn's chokes are better than I simmed.

When I try for "no load" (3mA; any less confuses the CPU), I get 587V DC, in sight of your >600VDC observation. 

> it's blown fuses if powered up too fast since I've owned it.

In simulation, current can be quite high. This is a design angle which is too-too-tedious to do by hand, easy on PC, so it may have been mis-judged long ago. What model "Sunn", and what size fuse?

[unclerny]

This is a Sunn Solarus.  There's a schematic floating around the internet somewhere and I have one I'm willing to send out but I don't know how to post it to this forum or it's library.

TO be clear, unloaded means there are NO leads of the secondary connected to anything.  And yes 400-0-400 with my variac at 110vac and a current limiter in line.

I'm terribly concerned that I get this right as the amp used to run for a while, pending how hard I'd run it, and then it would loose volume, down to a whisper so to speak.  I've heard that trannies can go bad like this once in a blue moon and this amp came to me from a barn it had sat in for years.  Rats nest in it and chicken crap and straw on it.  It was a mess.

I have a new multi-cap for it and when I'd put in the GZ34 I'd have all the tubes out of the amp, except the GX34 of course.  Seeing that the RCA manual says 550 vac plate to plate on that tube I don't want to eat up tubes.

One major change I'm making to the amp is getting rid of the 6AN8.  I'll change the PI over to a regular long tail pair with 12a?7.  Otherwise I'm removing the tremolo and the reverb return circuit is also on the board with the tremolo so that goes too.  This will free up a 9 pin socket for whatever I think of.  Probably an overdrive circuit like the old Dumble's used.

OK, the cats out, it won't be much of a Sunn when I'm done.  I don't care, had it for years with the old ultra-linear OT and although the amp had pristine cleans if I turned it up past half it would start sounding like a solid state amp, ugly distortion.  That tranny died a few years back and I put in a 50 watt fender OT and now the amp sounds like a Bassman.  Driven it was much better and the cleans had that tiny vintage Fender edge to them.  I felt it was an improvement as I didn't need a distortion pedal any longer to create my tone but just to add edge which is always preferred too me.

thanks guys

[sluckey]

it's blown fuses if powered up too fast since I've owned it.

That may be why later models had a cheap circuit breaker rather than a fuse.

Seeing that the RCA manual says 550 vac plate to plate on that tube I don't want to eat up tubes.

Are you sure? I believe that rating is 550vac plate to ground or 1100vac plate to plate. Most all Fender amps that use GZ34 use voltages much greater than 550 plate to plate.

[rzenc]

Hi

Philips data sheet cotes the use of an iron capable of delivering 2x550VAC, resulting in 640VDC @ 160mA's with 60uF as first input cap, which is the same conddition described by telefunken too.
Unfortunately, the Telefunken datasheet is too big to upload here...however I could post the Philips datasheets

There is another set of input voltages with choke first instead of cap.

Hope this helps

Best Regards

[PRR]

> lose volume....  I've heard that trannies can go bad like this once in a blue moon

Well, maybe, but you say it ran "for a while" before going mute, so why are you fretting a start-up condition? And "lose volume" is, in my experience, much more likely to be a simple loose signal-path contact than any obscure transformer problem. Transformers usually go dead, sometimes emit smoke.

> RCA manual says 550 vac plate to plate on that tube I don't want to eat up tubes.

Huh? No, that's a "typical", NOT a "max".

At that kinda current, you are allowed 475V RMS Per Plate. That would be 950V end-to-end.

> I'm terribly concerned that I get this right as the amp used to...

What is the worst that can happen? You already replaced half the guts. If it burns, you can replace the other half.

Bring up the power output stage, couple an iPod to the driver and a speaker to the output. Run it night and day, soft and loud. Someplace it won't burn the smoke-alarm. Either you incidentally fix the problem (by wholesale modding), or it smokes, and -now- you know where the problem was.

[unclerny]

Ok thanks guys, I really need to get my head around data sheets better.

PRR, you are the go to guy!!  Thank you, I'll do just as you suggest.

Tim

[unclerny]

wow has it been since 2009 that I've looked into this?  I guess so.

I'm really sorry that I haven't been on this amp and forum.  It's been rough since the crash and I'm working several jobs to keep my house.  I'm sure many of us are feeling the same crunch.

Well low and behold someone brought in another Sunn (sonic 2) doing the same damn thing.  It seems to run fine with the power tubes out with expected voltages.  It'll run with them in if I use a current limited but here I get voltages well below what I'd expect even under the limiter which is telling me something.  I know this is right there in front of me and I'm not seeing it.

I also feel the main power switch is bad because I can see it going on and off here and there and flipping the switch solves it usually.  I've inspected both amps VERY thoroughly and I'm at a loss.  My gut instinct says that one or both power tube sockets, maybe rec socket, is bad and once the load comes up enough... POP.  I had a problem like this with an Ampeg once and it was doing this same kind of thing.

I'm at a loss and I could seriously use some encouragement in some direction.

[unclerny]

okay, fresh morning.  fresh look.

With the 6550s in it there's no way to power it up w/o a POP.

I put in 6L6GC and can manage to get it to come on for about a minute before I get an arch in the rec tube and POP.

Back to square one.

[unclerny]

replaced all three octal sockets.

with Weber bias-rite on with 6L6s I can get it to fire up and then watch as the voltages keep dropping AND the current as well with no end in sight of it stabilizing.

voltages at first are as high as 550vdc with 35m.a. draw.  over the next 20 seconds it will drop to 300vdc and about 22m.a.

okay, this should be a major clue.

Please, someone tell me they've seen this?

Tim

[unclerny]

yep has standby.

what would pull the voltages down as the amp runs longer?  It's stable with the power tubes out so I'd put this on the power amp or the power supply.

the bias circuit appears fine and has new caps.  usually when an amp over runs like this that's where to look but this one seems to be serviceable.

At a loss.

Tim

[HotBluePlates]

the bias circuit appears fine and has new caps.  usually when an amp over runs like this that's where to look but this one seems to be serviceable.

New caps for the bias supply, or the entire power supply?

[unclerny]

oops, that wasn't very clear was it?  new caps in the bias.  I'm waiting on our order from CE for the main multi-cap.  I only go this way about it because the other amp from a year ago has new everything but sockets and still does the same thing with el34 in it.  If it's not one thing it's another.

I'll be back next week with a new cap set in this thing.  Let's see what happens.

Thanks, Tim

[ernie_jr]

If you use a stock replacement filter cap, you will soon run into the same problems. It only has a 525 volt rating and the amp exceeds that with 120 volts in.  Follow Doug"s fender filter cap replacement of using 2 100 mFd at 350 volts in series giving 50 mfd at about 700volts. This will solve your filter cap problems. If you want to keep the stock filter can in place, parallel the caps giving 40 and 50 mfd for the 3rd and 4th nodes. It will help with the bass response a bit.
Ernie

[unclerny]

Hi Gang, I wanted to post what's happened with that Sunn amp now that I finally got it running.

I installed the multi-cap as mentioned above.  This stopped the sudden fuse pops that I was getting on start up.  I could now watch the plate voltages drift down slowly until the current limiter starts glowing more than I fell safe and also by this time the plates are down to 300vdc so something is shorting somewhere.

I did many little things like removing the coupling caps to the PA and changing the speaker jacks as they had a very slick switching jack on the extension and I figured this many years down the road it could be failing.  This solved nothing.

I decided I'm down to a bad OT.  All we had around the shop was a hammond 1750w 100w 4ohm xmer suitable for a Twin or similar.  I figured it's paid for and what do I have to loose.  I swapped them out.  Of course the new xmer is not ultralinear so I just tweaked the power supply to match a Fender PA, it's just sitting there begging for it.

I put in fresh 6L6s and put it on the current limiter and what do you know it settled much closer to what I'd expect.  I took it off the limiter, fired it up and checked the bias and it's perfect.  I was looking at several failed parts all with their own problem.

The plates sit on 500vdc at 45m.a. and it hums right along.

With the volume on 7 it's a plexi sounding little beast.  Now I have to start on the first Sunn again and start replacing sockets and probably the power switches as those really arch more than I care for.  I can put some carlings on the back and no one will be any wiser.

That's it.  Another amp with more time in it than it's worth several times over finally breaths life again.  HEHE

Thanks, Tim

[Ritchie200]

Check for arcing - carbon traces between pins 2 and 3 on the output tube sockets.  If it is UL, there are huge flyback spikes if you push the amp into distortion.  Is the original transformer a Dagnall?  That would explain a transformer failure....

Good luck!
Jim

[unclerny]

Hi Ritchie, Both Xmers have been replaced.  The OT is a 100w Fender xmer made by Weber.  The PT is a Mercury Axion.  I have already budgeted for new octal sockets.

I hope to have time this week to work on it but I see John just got in several boxes of vintage guitar parts, bodies, necks and misc electronic parts we'll have to sort through this morning.

THank you for the response
Tim

[SILVERGUN]

I noticed that you had mentioned that you like more "edge" in an amp...have you considered going ss rectifier?...Doug offers a plug in that will fit the tube socket
It will enable you to use larger filter caps earlier in the ps, and could possibly make a large improvement in overall response. It will however raise the B+ slightly, so keep an eye on that.....if your rebuiding the PS anyway consider just a heavy duty bridge rectifier mounted directly to the chasis(as a heat sink), like a 1000 volt: http://www.surpluscenter.com/item.asp?catname=electric&item=22-1206

Here's how D^mble did it, with caps in series to increase power handling (like ernie mentioned):   http://www.el34world.com/charts/Schematics/files/dumble/97fullschem.pdf

[unclerny]

Hi Silvergun, yes I've been installing ss recs on these amps.  They seem to make the amps tighter, closer to that crunch sound Marshall made famous and then lost the secret.  Sorry not a fan of modern production Marshalls.

Tim

[SILVERGUN]

Yeah, I couldn't agree more....

Just f.y.i....I'm coming from a background in welding machine repair (lots of current, and when we blow fuses, everyone in the building knows it), and that's where the ss rectifier recommendation comes from....I tend to like everything bullet proof, and the more components that I don't have to think about going boom, the better.....that's why ---overrated everything, from PTs to resistors to rectifiers.....for me, the bigger the better, especially in power supplies........I hate that manufacturers are trying to save 5 cents on every component, and you wind up with amps that sound like modern Marsh@lls.
I am actually going to build one of those Humble's just to see all those beautiful caps in one place

So when I read that your tube rectifier was arcing, it was an obvious solution.....not that it would have fixed your amp, but, just give you one less thing to think about...I'm sure those rect. tubes dont appreciate that kind of abuse......you know,,,,,now that i think about it,,,,even if you just put the ss plug in while your troubleshooting, your rect. tube will thank you!

[unclerny]

That last thought about using ss rec during troubleshooting is a good idea as long as you still use a current limiter when you're not sure about shorts.  Once you have no shorts then testing with the ss rec will be safer.  Plus you know that if the amp can handle the ss rec then it certainly can handle a tube rec.  There's a bit of reverse thinking there since a tube rec is softer.

anyway, the day is done, long drive home.

thanks, Tim