New amp build from old pa - transformer question...

General discussion area for tube amps.

Moderators: pompeiisneaks, Colossal

MikeR670
Posts: 85
Joined: Sun Jul 10, 2016 7:56 pm
Location: B.C. Canada

New amp build from old pa - transformer question...

Post by MikeR670 »

Hi all

I'm mapping out a build of a SE amp with single 6550 and 12ax7 preamp. I have a large pp output transformer that I'm going to try to use in it. Can anyone explain how it would be wired when being used in SE format?

Found lots of schematics but nothing clear on how it's wired. Some show the center wire to ground, some to the power tube grid and some just show it left sitting, capped off?
User avatar
trobbins
Posts: 194
Joined: Tue Jun 01, 2010 8:51 am
Location: Australia

Re: New amp build from old pa - transformer question...

Post by trobbins »

Are you trying to clone an amp circuit that has a designed output transformer spec/requirement?

Do you know what the specs are of the PP output transformer?

Are you aware that the power rating of the output transformer will end up being significantly less when used in a SE amp, compared to when that output transformer was used in a PP stage?
User avatar
jjman
Posts: 753
Joined: Sat Oct 06, 2007 2:33 pm
Location: Central NJ USA

Re: New amp build from old pa - transformer question...

Post by jjman »

Some use the CT as one end. Many say it's taboo. Most talk about an "air gap" being needed. I think the issue is "saturation." (A bad kind.) Don't do it unless you know how to employ it that way.
If it says "Vintage" on it, -it isn't.
User avatar
Phil_S
Posts: 5936
Joined: Tue Oct 23, 2007 10:12 pm
Location: Baltimore, MD

Re: New amp build from old pa - transformer question...

Post by Phil_S »

First, let me say this isn't meant to be all that technical, so please give me a little latitude if I don't say it all that precisely. The other posts are getting at the problem with core saturation. The air gap in a SE build mitigates the problem. It may be a combination of how the lams are stacked and the gap, not sure. When you use a PP type transformer (no air gap) you'll need a much bigger transformer to avoid saturation. I'm not sure of how big. I'm guessing you'll see 15-20W SE 6550 and you'd want a PP rated for about 4x that or more. I've read you only use half the winding (CT to one outer leg.) If you think it's big enough, go for it, but I think I'd keep my expectations low.
MikeR670
Posts: 85
Joined: Sun Jul 10, 2016 7:56 pm
Location: B.C. Canada

Re: New amp build from old pa - transformer question...

Post by MikeR670 »

trobbins wrote: Thu Mar 30, 2017 10:52 pm Are you trying to clone an amp circuit that has a designed output transformer spec/requirement?

Do you know what the specs are of the PP output transformer?

Are you aware that the power rating of the output transformer will end up being significantly less when used in a SE amp, compared to when that output transformer was used in a PP stage?
Not sure on specs, it was designed to run four 6550 tubes in a Garnet G400PAR pa unit. Schematics here: http://ampgarage.com/forum/download/fil ... &mode=view

I am wanting to heavily reduce output because I'm building a practice amp and only want a 8ish watt output. I know that a pp trans is not the greatest idea in a SE, but I've been told that if it's sufficiently large I may not get saturation or blocking distortion issues. I want a clean low wattage sound, that's what led to me an SE with single 6550C......
MikeR670
Posts: 85
Joined: Sun Jul 10, 2016 7:56 pm
Location: B.C. Canada

Re: New amp build from old pa - transformer question...

Post by MikeR670 »

Phil_S wrote: Fri Mar 31, 2017 12:52 am First, let me say this isn't meant to be all that technical, so please give me a little latitude if I don't say it all that precisely. The other posts are getting at the problem with core saturation. The air gap in a SE build mitigates the problem. It may be a combination of how the lams are stacked and the gap, not sure. When you use a PP type transformer (no air gap) you'll need a much bigger transformer to avoid saturation. I'm not sure of how big. I'm guessing you'll see 15-20W SE 6550 and you'd want a PP rated for about 4x that or more. I've read you only use half the winding (CT to one outer leg.) If you think it's big enough, go for it, but I think I'd keep my expectations low.
Oh you read my mind :)...yes the tranny ran four 6550s in an old pa. It was supposed to be a 400 watt unit. The tranny weighs 12 lbs, but I don't know how to test it's specs.

Schematic: http://ampgarage.com/forum/download/fil ... &mode=view

I'm looking for a clean sound all the way up, and only about 8 watts output....
MikeR670
Posts: 85
Joined: Sun Jul 10, 2016 7:56 pm
Location: B.C. Canada

Re: New amp build from old pa - transformer question...

Post by MikeR670 »

MikeR670 wrote: Fri Mar 31, 2017 12:59 am
trobbins wrote: Thu Mar 30, 2017 10:52 pm Are you trying to clone an amp circuit that has a designed output transformer spec/requirement?

Do you know what the specs are of the PP output transformer?

Are you aware that the power rating of the output transformer will end up being significantly less when used in a SE amp, compared to when that output transformer was used in a PP stage?
Not sure on specs, it was designed to run four 6550 tubes in a Garnet G400PAR pa unit. Schematics here: http://ampgarage.com/forum/download/fil ... &mode=view

I am wanting to heavily reduce output because I'm building a practice amp and only want a 8ish watt output. I know that a pp trans is not the greatest idea in a SE, but I've been told that if it's sufficiently large I may not get saturation or blocking distortion issues. I want a clean low wattage sound, that's what led to me an SE with single 6550C......
Sorry, didn't answer your first question: I'm not cloning any particular design, I think I'm cobbling pieces together of circuits I've seen.....my old pa has a bunch of 12ax7s, some 12au7s and four 6550Cs - so I'm wanting to build using from items I have. It does also have a little OT which was a 12au7 reverb driver....I even considered building an SE with two parallel 12au7s, using that little tranny, but I think it would not give 6 or 8 watts of clean sound.....
User avatar
lumox0013
Posts: 235
Joined: Wed Nov 19, 2008 4:10 am
Location: independance ,MO.
Contact:

Re: New amp build from old pa - transformer question...

Post by lumox0013 »

I would sell that one and order a kit for what you actually want. good luck
Live , Love , Learn
User avatar
trobbins
Posts: 194
Joined: Tue Jun 01, 2010 8:51 am
Location: Australia

Re: New amp build from old pa - transformer question...

Post by trobbins »

The estimate path would be:

6550 datasheet shows max signal plate current of circa 300mA in fixed bias PP pentode mode.
Likely peak current through half a primary winding (other half winding is in cut-off so is not passing any current) is circa 600mA.
The output transformer (OT) core may well be moving towards saturation region at 600mA in the turns provided by half the primary, but lets assume that any such distortion is not too gross.

The OT appears to be 3k PP impedance for the 4 and 8 ohm output levels. It may be worth your while confirming that with a turns ratio test.

If you use half the primary winding, then its 'impedance' loading will be 3k/4 = 750 ohm. That may be too low for your new SE design.

If you want to use the full PP winding, which will present 3kohm loading to your SE, then the max peak current should drop to 25% as the winding turns with that current have doubled. So 150mA peak. SE bias would nominally be set to up to half that peak level, so 75mA idle bias. That gets pretty close to the 6550 datasheet SE reference levels for a 400V plate and 225V screen design that uses a 3kohm loading and 87mA idle - so not too bad an outcome.
R.G.
Posts: 1223
Joined: Tue Dec 02, 2014 9:01 pm

Re: New amp build from old pa - transformer question...

Post by R.G. »

PP vs SE>
An inductor with a voltage across it lets current ramp up forever, until limited by something other than the inductor. Needless to say, unless the "something" stops the current increasing before things melt, you get ugly smells, smoke, flames and other unfortunate effects. So for audio use, you want to keep it from the "smoke and flames" level. Really, any use I guess.

Transformers are magic electrical seesaws that convert voltage and current across an isolated boundary. We pay for this by having to feed an unavoidable inductor that's always found on the primary. The inductor has limits. It can only swing so many ampere-turns in one direction before the magic seesaw inside quits being linear. This limit is bidirectional; if it can swing S ampere-turns with the current in one direction before failure of the magic seesaw (,which is what we call "saturation"), it can then swing to -S ampere-turns in the other direction.

A p-p transformer is set up so that an audio signal in one polarity can pull it to +S, and the other polarity can pull it to -S, so it gets a total of 2*S signal size before seesaw failure. It can only do this because in the absence of signal, it sits at a magnetic field of zero.

A single ended transformer can only be pulled in one direction by its output tube(s). So to keep the audio signal balanced, not cut off on one polarity or the other, the single ended transformer has to have a "DC" magnetic bias of 1/2*S. That is, it's magnetically balanced right in the middle of one swing. It can't go below 0 magnetic field, because its output tube can't pull it there. It can't go above S magnetic field because the core saturates and the transformation seesaw quits working.

So for a given quantity of magnetic core material, >> you only get to use 1/4 of the magnetic field "goodie" to transform audio signals.

Generally, you want more than that, and you also want good bass response, so there is a whole series of finagling in the the transformer design that introduces air gaps (air cannot saturate) and makes the core physically bigger, and then adds more turns back in to get back the low frequencies you lost when you introduced the air gap and on and on and on.

>> SE and PP tranformers are fundamentally different<<< PP transformers are designed to run at an average zero magnetic flux, SE are designed to run at + 1/2* S magnetic flux. If you are trying to run a PP transformer as SE, you have some issues.

1. You have to find what "S" is, and arrange to make the ampere-turns be S/2. As a practical matter, you set the DC current to some level, run some signal into the output tube, then increase the signal in the transformer primary while watching the transformer secondary. When something starts flattening off, you're seeing seesaw failure. You then change to a different DC current and do it again. At some point, you'll get crudely equal flattening on both + and - sides. That's the right opeating point for the PP transformer in SE mode.

The problem is, this point is not necessarily good for the output tube's current, voltage, or impedance operation. Maybe, maybe not. You may have to tinker the power supply voltage is dget a usable circuit. Or change output tubes. But then, maybe you'll get lucky.,

You can finesse this a bit by noting that there is a whole half of the primary you're not using, the other side of the PP. It is possible to simply suck DC current back through the unused half of the PP transformer, and pull its average operating point back down to nearer zero magnetic field, or anywhere in between. In fact, this doesn't even need to be a tube, just a current sucker. resistors don't work all that well because they load down the transformer works, but power transistors and MOSFETs can do the job. There may even be a patent on disconnecting signal from one tube in a PP and using this as an "SE mode".
User avatar
Phil_S
Posts: 5936
Joined: Tue Oct 23, 2007 10:12 pm
Location: Baltimore, MD

Re: New amp build from old pa - transformer question...

Post by Phil_S »

lumox0013 wrote: Fri Mar 31, 2017 2:31 am I would sell that one and order a kit for what you actually want. good luck
Agreed! Holy moly! 4x 6550 OT? You can probably get at least $100 for it on eBait, quite possibly more. Sell it. Get something else appropriate to the build.
MikeR670
Posts: 85
Joined: Sun Jul 10, 2016 7:56 pm
Location: B.C. Canada

Re: New amp build from old pa - transformer question...

Post by MikeR670 »

trobbins wrote: Fri Mar 31, 2017 9:24 pm The estimate path would be:

6550 datasheet shows max signal plate current of circa 300mA in fixed bias PP pentode mode.
Likely peak current through half a primary winding (other half winding is in cut-off so is not passing any current) is circa 600mA.
The output transformer (OT) core may well be moving towards saturation region at 600mA in the turns provided by half the primary, but lets assume that any such distortion is not too gross.

The OT appears to be 3k PP impedance for the 4 and 8 ohm output levels. It may be worth your while confirming that with a turns ratio test.

If you use half the primary winding, then its 'impedance' loading will be 3k/4 = 750 ohm. That may be too low for your new SE design.

If you want to use the full PP winding, which will present 3kohm loading to your SE, then the max peak current should drop to 25% as the winding turns with that current have doubled. So 150mA peak. SE bias would nominally be set to up to half that peak level, so 75mA idle bias. That gets pretty close to the 6550 datasheet SE reference levels for a 400V plate and 225V screen design that uses a 3kohm loading and 87mA idle - so not too bad an outcome.
Thanks! This is encouraging - I've just started reading the two Morgan Jones books so might help me to understand how to map it, and build. If I try using the full pp winding, how would that be physically wired in the circuit?

If an SE OT runs one secondary to the tube plate, and the other to b+, how would you wire in the pp OT with its two secondaries and center tap? I've seen different schematics all doing it in different ways, some seem to take one secondary to plate, the other to b+, and the tap to ground?
MikeR670
Posts: 85
Joined: Sun Jul 10, 2016 7:56 pm
Location: B.C. Canada

Re: New amp build from old pa - transformer question...

Post by MikeR670 »

Phil_S wrote: Fri Mar 31, 2017 10:54 pm
lumox0013 wrote: Fri Mar 31, 2017 2:31 am I would sell that one and order a kit for what you actually want. good luck
Agreed! Holy moly! 4x 6550 OT? You can probably get at least $100 for it on eBait, quite possibly more. Sell it. Get something else appropriate to the build.
I have this hankering to build something from my parts, but if it doesn't work out at least selling them off is still a good option. I guess I'll learn plenty along the way and might be in a better position to choose new parts to suit what I'm after....
MikeR670
Posts: 85
Joined: Sun Jul 10, 2016 7:56 pm
Location: B.C. Canada

Re: New amp build from old pa - transformer question...

Post by MikeR670 »

R.G. wrote: Fri Mar 31, 2017 10:27 pm PP vs SE>
An inductor with a voltage across it lets current ramp up forever, until limited by something other than the inductor. Needless to say, unless the "something" stops the current increasing before things melt, you get ugly smells, smoke, flames and other unfortunate effects. So for audio use, you want to keep it from the "smoke and flames" level. Really, any use I guess.

Transformers are magic electrical seesaws that convert voltage and current across an isolated boundary. We pay for this by having to feed an unavoidable inductor that's always found on the primary. The inductor has limits. It can only swing so many ampere-turns in one direction before the magic seesaw inside quits being linear. This limit is bidirectional; if it can swing S ampere-turns with the current in one direction before failure of the magic seesaw (,which is what we call "saturation"), it can then swing to -S ampere-turns in the other direction.

A p-p transformer is set up so that an audio signal in one polarity can pull it to +S, and the other polarity can pull it to -S, so it gets a total of 2*S signal size before seesaw failure. It can only do this because in the absence of signal, it sits at a magnetic field of zero.

A single ended transformer can only be pulled in one direction by its output tube(s). So to keep the audio signal balanced, not cut off on one polarity or the other, the single ended transformer has to have a "DC" magnetic bias of 1/2*S. That is, it's magnetically balanced right in the middle of one swing. It can't go below 0 magnetic field, because its output tube can't pull it there. It can't go above S magnetic field because the core saturates and the transformation seesaw quits working.

So for a given quantity of magnetic core material, >> you only get to use 1/4 of the magnetic field "goodie" to transform audio signals.

Generally, you want more than that, and you also want good bass response, so there is a whole series of finagling in the the transformer design that introduces air gaps (air cannot saturate) and makes the core physically bigger, and then adds more turns back in to get back the low frequencies you lost when you introduced the air gap and on and on and on.

>> SE and PP tranformers are fundamentally different<<< PP transformers are designed to run at an average zero magnetic flux, SE are designed to run at + 1/2* S magnetic flux. If you are trying to run a PP transformer as SE, you have some issues.

1. You have to find what "S" is, and arrange to make the ampere-turns be S/2. As a practical matter, you set the DC current to some level, run some signal into the output tube, then increase the signal in the transformer primary while watching the transformer secondary. When something starts flattening off, you're seeing seesaw failure. You then change to a different DC current and do it again. At some point, you'll get crudely equal flattening on both + and - sides. That's the right opeating point for the PP transformer in SE mode.

The problem is, this point is not necessarily good for the output tube's current, voltage, or impedance operation. Maybe, maybe not. You may have to tinker the power supply voltage is dget a usable circuit. Or change output tubes. But then, maybe you'll get lucky.,

You can finesse this a bit by noting that there is a whole half of the primary you're not using, the other side of the PP. It is possible to simply suck DC current back through the unused half of the PP transformer, and pull its average operating point back down to nearer zero magnetic field, or anywhere in between. In fact, this doesn't even need to be a tube, just a current sucker. resistors don't work all that well because they load down the transformer works, but power transistors and MOSFETs can do the job. There may even be a patent on disconnecting signal from one tube in a PP and using this as an "SE mode".
Hopefully once I work through the Morgan Jones books I'll be able to understand this detail level, I'm a long way from that at this point. I'm planning to learn the necessary theory and couple that with build attempts along the way.....thanks for your input, much appreciated! - just a bit over my head for now......
MikeR670
Posts: 85
Joined: Sun Jul 10, 2016 7:56 pm
Location: B.C. Canada

Re: New amp build from old pa - transformer question...

Post by MikeR670 »

MikeR670 wrote: Sat Apr 01, 2017 6:23 pm
R.G. wrote: Fri Mar 31, 2017 10:27 pm PP vs SE>
An inductor with a voltage across it lets current ramp up forever, until limited by something other than the inductor. Needless to say, unless the "something" stops the current increasing before things melt, you get ugly smells, smoke, flames and other unfortunate effects. So for audio use, you want to keep it from the "smoke and flames" level. Really, any use I guess.

Transformers are magic electrical seesaws that convert voltage and current across an isolated boundary. We pay for this by having to feed an unavoidable inductor that's always found on the primary. The inductor has limits. It can only swing so many ampere-turns in one direction before the magic seesaw inside quits being linear. This limit is bidirectional; if it can swing S ampere-turns with the current in one direction before failure of the magic seesaw (,which is what we call "saturation"), it can then swing to -S ampere-turns in the other direction.

A p-p transformer is set up so that an audio signal in one polarity can pull it to +S, and the other polarity can pull it to -S, so it gets a total of 2*S signal size before seesaw failure. It can only do this because in the absence of signal, it sits at a magnetic field of zero.

A single ended transformer can only be pulled in one direction by its output tube(s). So to keep the audio signal balanced, not cut off on one polarity or the other, the single ended transformer has to have a "DC" magnetic bias of 1/2*S. That is, it's magnetically balanced right in the middle of one swing. It can't go below 0 magnetic field, because its output tube can't pull it there. It can't go above S magnetic field because the core saturates and the transformation seesaw quits working.

So for a given quantity of magnetic core material, >> you only get to use 1/4 of the magnetic field "goodie" to transform audio signals.

Generally, you want more than that, and you also want good bass response, so there is a whole series of finagling in the the transformer design that introduces air gaps (air cannot saturate) and makes the core physically bigger, and then adds more turns back in to get back the low frequencies you lost when you introduced the air gap and on and on and on.

>> SE and PP tranformers are fundamentally different<<< PP transformers are designed to run at an average zero magnetic flux, SE are designed to run at + 1/2* S magnetic flux. If you are trying to run a PP transformer as SE, you have some issues.

1. You have to find what "S" is, and arrange to make the ampere-turns be S/2. As a practical matter, you set the DC current to some level, run some signal into the output tube, then increase the signal in the transformer primary while watching the transformer secondary. When something starts flattening off, you're seeing seesaw failure. You then change to a different DC current and do it again. At some point, you'll get crudely equal flattening on both + and - sides. That's the right opeating point for the PP transformer in SE mode.

The problem is, this point is not necessarily good for the output tube's current, voltage, or impedance operation. Maybe, maybe not. You may have to tinker the power supply voltage is dget a usable circuit. Or change output tubes. But then, maybe you'll get lucky.,

You can finesse this a bit by noting that there is a whole half of the primary you're not using, the other side of the PP. It is possible to simply suck DC current back through the unused half of the PP transformer, and pull its average operating point back down to nearer zero magnetic field, or anywhere in between. In fact, this doesn't even need to be a tube, just a current sucker. resistors don't work all that well because they load down the transformer works, but power transistors and MOSFETs can do the job. There may even be a patent on disconnecting signal from one tube in a PP and using this as an "SE mode".
Hopefully once I work through the Morgan Jones books I'll be able to understand this detail level, I'm a long way from that at this point. I'm planning to learn the necessary theory and couple that with build attempts along the way.....thanks for your input, much appreciated! - just a bit over my head for now......
One question: if the fact that the ot is well oversized when running just one SE 6550 maybe still doesn't help, is there an option to run it with two power tubes in pp form, but reduce power so the output could still reach my goal of around 8 watts with good clean headroom? I realise the 6550s will be no good for that in pp setup, but I maybe I could substitute for two different smaller power tubes, and build around that?

I've been keen on a SE build because of the reduced output and possibly nicer tones, but this is my first build after all....
Post Reply