Gapped vs Ungapped Transformer

Hi,

What is the big difference between Ungapped and Gapped transformer?
When are we using one and not the others?

Thank's

A gapped transformer can handle DC on the primary.

If a DC current has to go through the transformer (for example single ended/class A  or push-pull class B)  it will magnetize the core in one direction. When the audio signal arrives the core will saturate, resulting distortion in the signal. To prevent this happening a gap between the winding and the core is created.

If no DC has to go through the transformer then this gap is not required (in places where the DC on the output is blocked by a capacitor or eliminated by servo).

transformers with an air gap are used for Single Ended power amps,

these amps usually have only one power tube,

this tube will be fed DC  current by way of the output transformer primary,

now in a push-pull amp, current also goes thru the output transformer,

but it enters the winding by way of the center tap,


this means the current will wind clockwise thru one half of the pri,

and counter clockwise thru the other half,

this means that the flux set up from the dc will be canceled,

this leaves all the flux capability of the core for the ac signal,

since the SE transformer has DC flowing thru the pri in one direction only,

what you have is an electro magnet just like you made in science class,

since the core has to support the DC AND the AC flux, it will need to be bigger.

a series air gap is just like a resistor in an electrical circuit,

only instead of limiting current, the gap limits flux,

it does this by lowering the perm of the core.

if you lower the perm, the DC bias will not saturate the core as quickly as with no gap.

but we will have fewer henries for the same amount of turns on the same core,

this means we have to make the core bigger to compensate,

this is the negative tradeoff we make for SE transformers, more weight and cost,

but some people like the SE sound, which features more odd harmonics than push-pull

this is because the p-p circuit tends to cancel odd harmonics.

this is also why SE amps tend to sound crappy at first,

we are so used to p-p circuits as they are more efficient and cheaper,

99 percent of all the amps you listen to in your life will be push-pull.

but once you get used to SE, the push pull will lack some character.

there is an optimun air gap for each core and each value of DC current,

an optimun air gap will split the reluctance evenly,

50% energy stored in the core, 50% in the air gap,

small cores such as 625 EI have a natural air gap of about 0.0003 inches,

so you can get away without a gap for small SE amps,

air has much less perm than 4% 29M6 silicon grain oriented steel.

and air has tons less perm than Supermalloy 80% Ni cores,

this means that an air gap will drop the perm rapidly in an Ni core,

example:

29M6 4% Ni core has an un gapped perm of about 8,000

an Ni core has a perm of about 80,000,

an 0.001 inch gap in the M6 core drops perm from 8,000 to 2,500,

but the same gap in an Ni core drops the perm from 80,000 to 3,500,

so there is no point in wasting a good Ni core by inserting a gap,

this is because you end up with close to the same perm as the M6 core,

but at tens times the cost for the expensive Supermalloy,

so you will rarely see a gapped Ni core,

a 50/50 core with 10 x 10m stacking might be found in an old WeCo xfmr,

but that is the only time i have seen  one.

Cheers CJ.

I stand corrected on push-pull.

also note that the gap is actually in the core, not between windings and core. Stack all E's together one way, all I's together and put a piece of tape between the two when you put them together to for the standard EI core shape. This is the gap. It's made bigger or smaller depending on amount of DC current expected to go through the winding.

besides lowering saturation  in a SE xfmr, the series air gap will:

1. increase exciting current
2. lower inductance
3. stabilizes the core's permeability
4. lowers harmonic distortion
5. in a polarized transformer or inductor, inductance may decrease, but it can also be increased by certain optimum gaps

also, Q is the quality of merit of an inductor or transformer,

usually denoted by

Q=2 pi f L/DCR, this just means inductance divided by the DCR = quality,

you want a hi Q of 10 or better for most filter circuits,

now, bandwidth of filter/resonance of filter = 1/Q.

so if we increase Q for a given core, we  narrow the bandwidth, = better selectivity,

perm and Q for a transformer or inductor are related by:

uQ = ue/Qe,

perm times Q = effective perm times Q.

when you gap a core, effective perm is lowered un-gapped perm,

so Q goes up and the filter works better.

the optimum gap for raising Q occurs when copper loss = core loss,

formula is

Q opt = 1/2  root of Re/R1, where Re is core loss and R1 is copper loss,

now sometimes tubes get mis-matched or steve vai hits a not that looks like dc,

so a gapped push pull might help, this gets Shaib off the hook, 

so since the world ain't perfect, engineers compensate and cover their azz,

here is a Freed 6l6gt outie, 14 x 14 stacking, creates a "semi" gap, just in case,

Wow thank's to CJ ! and to others who complete this !

Thank's

CJ said:

now sometimes tubes get mis-matched or steve vai hits a not that looks like dc,

so a gapped push pull might help, this gets Shaib off the hook, 

Click to expand...

Didn't want to stir up.  ;D

The theory is true.

Just discovered this topic while searching on the subject,
and guys, i really need to say :
Thank You !
CJ  , even in english ,
i understand your explanation  way better than my science teachers in college !!
thanks for the knowledge !

no problem,

i am still struggling with the  u x Q=ue x Qe  part,

turns out they (Grossner book) are talking about the Q of the core, not the inductor or transformer,

it says that if you drop the perm, the Q goes up,

but this is counter intuitive, as if you raise perm, you raise inductance, which means if you keep the DCR constant by using the same coil and you raise inductance, then you increase the Q by way of Q = XL/2 pi f

what i think he is talking about is as you gap a transformer core, you tilt the hysteresis curve, and when you do this, the BH loop area inside the curve (core loss) actually goes down as the curve squishes down a bit so that the sides of the BH curve get closer together,

so if you combine inductor Q and core Q, you get

1/Q = DCR/XL + XL/R-e, where R-e is the shunt eddy current resistance of the core,

i need to get some more books, Reuben Lee has a couple of good ones,

cj

> more books, Reuben Lee

Reuben also published in the trade rags.

Distortion in HiFi amps
http://www.americanradiohistory.com/Archive-Radio-Enginering/30s/1937/Radio-Engineering-1937-06.pdf  mag page 16 PDF page 18

Power Amp Output Circuits (Radio Frequency!)
http://www.americanradiohistory.com/Archive-Radio-Enginering/30s/1934/Radio-Engineering-1934-07.pdf  mag page 10

Bleeder resistance
http://www.americanradiohistory.com/Archive-Radio-Enginering/30s/1934/Radio-Engineering-1934-08.pdf mag page 18

Is there a way to test whether a Xformer is gapped besides looking at it?

I have a Carnhill that has a board soldered to it that makes it impossible to see, without a lot of work.

Can I just connect a 9V battery at the primary and see if there's dc on the secondary?

is there a part number on the transformer?  what kind of appliance is the transformer mounted in?

transformers block DC from primary to secondary so no good on the battery test,

why do you need to know if it is gapped or not? maybe going to rip it out of what ever it is in and use it for something else?

Details. Okay

Well, this was labeled as a VTB 2291, but when I measured the coils, it didn't match the OTHER VTB 2291  I bought at the same time.  It appears to be either a VTB 1148 gapped or VTB 9057 ungapped according to the resistances I'm getting. I'd like to know which, because it sounds pretty great in my G1176 and I have another channel.

do you have a LCR meter?  (inductance, capacitance, resistance)

signal generator or variac will work also,

lower inductance will be the gapped transformer,

CJ said:

but some people like the SE sound, which features more odd harmonics than push-pull

this is because the p-p circuit tends to cancel odd harmonics.

this is also why SE amps tend to sound crappy at first,

we are so used to p-p circuits as they are more efficient and cheaper,

99 percent of all the amps you listen to in your life will be push-pull.

but once you get used to SE, the push pull will lack some character.

Click to expand...

Great post, CJ, much of the details flying above my altitude...   

However...as far as I know...

Push-pull only cancel odd harmonics when using negative feedback.  Otherwise P-P is likely to have more odd harmonics than single-ended, due to potential cross-over distortion.  S-E inherently have fewer odd harmonics and mostly second and even harmonics... 

As far as people liking single-ended sound, everyone who loves the Neve 283-family of amps, like 1073 etc. is hearing a solid-state single-ended amp, rich in even harmonics.  The sound is often attributed to the iron, but it's also the single-ended design IMO. 

But I'm still learning.. ...

tommypiper said:

As far as people liking single-ended sound, everyone who loves the Neve 283-family of amps, like 1073 etc. is hearing a solid-state single-ended amp, rich in even harmonics.  The sound is often attributed to the iron, but it's also the single-ended design IMO. 

But I'm still learning.. ...

Click to expand...

What you say is true. However, you do not have to use a gapped transformer in a single ended design - the alternative is to capacitively couple the transformer to the single ended stage in which case the transformer does not need a gap.  In general terms, this means the transformer will have better performance than its gapped equivalent. The RED47 does this for instance

Cheers

Ian

ruffrecords said:

What you say is true. However, you do not have to use a gapped transformer in a single ended design - the alternative is to capacitively couple the transformer to the single ended stage in which case the transformer does not need a gap. In general terms, this means the transformer will have better performance than its gapped equivalent. The RED47 does this for instance

Cheers

Ian

Click to expand...

Sorry for reviving an old thread but I am trying to go back and read as much as I can about this stuff.

@ruffrecords - what do you mean by better performance in this sense? Lower distortion?