Output Transformer Questions

[Andy Le Blanc]

appologies....

lewilson had a question about how load effect the primary impedance
and to follow yours ......

"So the question remains: exactly what is happening when you run the 4 ohm tap into an 8 ohm cab on the Express with the 6.6K OT?"..........

when your describing an output transformer the secondary or load determines the reflected impedance of the primary ........the
ratio of the primary to the secondary is representative of not only the impedance but also ratios of current, voltage, and the number of turns of the primary and secondary windings
the working example has been transformer with a primary impedance of 6.6Kohms
and a secondary of 8ohms this gives Z1/Z2 of 825/1
this is equal to the turns ratio squared (T1/T2)2 or 28.7/1
this changes of course with matching taps to maintain the primary impedance
for example 6.6k / 4ohm gives a Z1/Z2 of 1650/1 which leads to your statement

"Andy - I really don't understand your data? 3.3K is the primary impedance of a 6.6K OT with its tap set at 8 ohms - with a 4 ohm load applied. Likewise, 3.3K is the primary impedance of a 6.6K OT with its tap set at 4 ohms - with an 2 ohm load applied. Likewise again, 3.3K is the primary impedance of a 6.6K OT with its tap set to 16 ohms - with an 8 ohm load applied. Etc., etc.."

so with a 6.6k primary in mind we have the ratios for loads

16 ohm........ 412/1 turns ratio of 20.3/1
8 ohm........ 825/1..........................28.7/1
4ohm.........1650/1.........................40.6/1

which means for example that when you connect an 8 ohm load to a
transformer with 1650/1 impedance ratio you end up with a 13.2K pri. imp.
other relationships of transformer action are

V1/V2=T1/T2
I1/I2=T2/T1
Z2=V2/I2
Z1=V1/I1

now to return to skyboltone's statement

"I wonder why tube manufacturers/ design engineers recomend specific load impedences? I wonder why they post performance curves based on those impedences?"

for given voltage conditions each device (a power tube) has a sweet spot
where it will produce a maximum output with minimum distortion products
this is seen when graphically plot a load line on you plate family curves
the load determines the avalible power output as a proportion to plate dissipation and it is these currents (I max., I min., and I bo "quiesent")
that are use to find % harmonic distortion
slight changes in load can as you know can affect the over all feel and
tone and as long as the tube is held with in its safe operating limits a wide
range of loading is possible, the load may not be optimal in some applications
as its presented in the tube manuals but if it sounds goods ....

[skyboltone]

appologies....

<quoteing me>"I wonder why tube manufacturers/ design engineers recomend specific load impedences? I wonder why they post performance curves based on those impedences?"

for given voltage conditions each device (a power tube) has a sweet spot
where it will produce a maximum output with minimum distortion products
this is seen when graphically plot a load line on you plate family curves
the load determines the avalible power output as a proportion to plate dissipation and it is these currents (I max., I min., and I bo "quiesent")
that are use to find % harmonic distortion
slight changes in load can as you know can affect the over all feel and
tone and as long as the tube is held with in its safe operating limits a wide
range of loading is possible, the load may not be optimal in some applications
as its presented in the tube manuals but if it sounds goods ....

I want to jump back in here and offer my own appologies. My experience with 6CA7s and EL-34s goes back to my hi-fi days and particularly Dynaco. I had seen so many designs use the 6.6K primary that I just assumed that this was the design center value for OT primary. Turns out I'm dead wrong. Mullard does show a set of curves for 7K but they also show more data at 3.5K. Ditto the other manufacturers that give much data at all. So where I thought that the 3.3K was a gross mismatch is just plain wrong. (Hangs head) It seems that there is a very broad set of load conditions for the EL-34 and it turns out to be a whatever floats your boat situation. Again, forgive my assumptions based on experience. The data prove me full of it.

Dan

[rooster]

OK, let me be brief and restate my words here.

The ratio of a 6.6K OT is 825:1 - for whatever the taps might be, 2-4-8-16-32-etc. And I hope we can all agree on that/this.

So, since the ratio is constant and is as follows, 6.6K:8 = 825:1. 6.6K:4 = 825:1, 6.6K:16 = 825:1.....

...it would then follow that changing the 1 in the ratio to 1/2, in the ratio 825:1, we would have 412.5:1/2 - which would still manitain our ratio of 825:1. Please agree with this..

So to finish, whatever else I may have mistated, it will follow that plugging an 8 ohm speaker load into the 16 ohm tap on a 6.6K OT that has a ratio of 825:1, we would then have a primary impedance of 3.3K. It is just this simple, a ratio that is constant with the OT.

Are we agreed on this?

For further example, look at the classic swap, the Twin Reverb OT swap from 4 6L6s to two 6L6s. Here the primary impedance goes from 2K+ to 4K+, and likewise, the secondary impedance goes from 4 ohms to 8 ohms. This is no different a swap than with our 6.6k OT except we are switching tubes that have a different impedance. The OT doesn't care, all it wants to do is maintain a ratio.

Are we agreed on this?

[Andy Le Blanc]

correct ......here an illustration of what happens when you change load...
its not an el34 but it paints the picture.......

[morcey2]

OK, let me be brief and restate my words here.

The ratio of a 6.6K OT is 825:1 - for whatever the taps might be, 2-4-8-16-32-etc. And I hope we can all agree on that/this.

So, since the ratio is constant and is as follows, 6.6K:8 = 825:1. 6.6K:4 = 825:1, 6.6K:16 = 825:1.....

No. A transformer has no real impedance, just an impedance ratio. If a 6.6k transformer only has an 8-ohm tap (with an 8-ohm load), then the impedance ratio is 825:1 If it has a 4-ohm tap, the ratio becomes 1650:1. The 16-ohm tap has a ratio of 412.5:1.

...it would then follow that changing the 1 in the ratio to 1/2, in the ratio 825:1, we would have 412.5:1/2 - which would still manitain our ratio of 825:1. Please agree with this..

There is no difference at all between 825:1 and 412.5:1/2. They are just different ways of saying the same thing, just a description of a 6.6k tranny w/ an 8-ohm load connected to the 8-ohm tap.

So to finish, whatever else I may have mistated, it will follow that plugging an 8 ohm speaker load into the 16 ohm tap on a 6.6K OT that has a ratio of 825:1, we would then have a primary impedance of 3.3K. It is just this simple, a ratio that is constant with the OT.

Are we agreed on this?

Yes

For further example, look at the classic swap, the Twin Reverb OT swap from 4 6L6s to two 6L6s. Here the primary impedance goes from 2K+ to 4K+, and likewise, the secondary impedance goes from 4 ohms to 8 ohms. This is no different a swap than with our 6.6k OT except we are switching tubes that have a different impedance. The OT doesn't care, all it wants to do is maintain a ratio.

Are we agreed on this?

I think you have your cause and effect backwards on that statement. When a pair of tubes are pulled, the load on the secondary of the OT must be doubled to maintain the correct impedance as seen from the power tubes. I think you have a pretty good grasp of the end result, but are a little confused on how the transformer in involved.

Like Andy said, most of this can be seen if you plot the loadline. Then it breaks down to "I want the plate-to-plate load to be X and I've already got a Y-ohm cab that I want to use, so I need a X:Y OT that can handle Z watts."

Matt

[rooster]

morcey 2 - OK, I guess what you are thinking is that the secondary has only one tap, and we are just adding different loads to it? I think this is the only way your formula will work out.

The thing is, if the secondary were actually static, then we would be really messing with the pwr tube big time as we changed the load. And this is not the case. The impedance of the tube pair, in this particular case, is centered around 6.6K at all loads, 4-8-or 16 ohms. ....And it should follow here that if I WERE to lower the speaker load by 1/2 to the appropriate secondary tap, say, place a 4 ohm load into the 8 ohm tap, the impedance of the primary would be 1/2 of what it was, too. It has to be because the ratio is constant. Yes?

Now, I can also go up, say put the 4 ohm load into the 16 ohm tap, and this will change the ratio. Or put the 4 ohm load into a 2 ohm tap if one existed. This would change the ratio, too. And some of the numbers that were tossed around would be correct, actually. But what I am talking about is keeping the ratio what it was from the designer, but using a different pwr tube pair. That's my goal because it allows the best transfer of energy.

So I think you need to consider what the secondary is doing in an OT, and why it has taps at all. The secondary is designed to keep the ratio constant at all loads. If it didn't do this the pwr tubes would not deliver the goods, so to speak.

[lewilson]

An nice way to find out the exact load would be to measure the real impedance of the tube in circuit.

[morcey2]

morcey 2 - OK, I guess what you are thinking is that the secondary has only one tap, and we are just adding different loads to it? I think this is the only way your formula will work out.

The thing is, if the secondary were actually static, then we would be really messing with the pwr tube big time as we changed the load. And this is not the case. The impedance of the tube pair, in this particular case, is centered around 6.6K at all loads, 4-8-or 16 ohms. ....And it should follow here that if I WERE to lower the speaker load by 1/2 to the appropriate secondary tap, say, place a 4 ohm load into the 8 ohm tap, the impedance of the primary would be 1/2 of what it was, too. It has to be because the ratio is constant. Yes?

Now, I can also go up, say put the 4 ohm load into the 16 ohm tap, and this will change the ratio. Or put the 4 ohm load into a 2 ohm tap if one existed. This would change the ratio, too. And some of the numbers that were tossed around would be correct, actually. But what I am talking about is keeping the ratio what it was from the designer, but using a different pwr tube pair. That's my goal because it allows the best transfer of energy.

So I think you need to consider what the secondary is doing in an OT, and why it has taps at all. The secondary is designed to keep the ratio constant at all loads. If it didn't do this the pwr tubes would not deliver the goods, so to speak.

All I'm saying is consider each different ohmage (real word?) tap separately. It sounds ot me like you understand it, but you were having trouble explaining it.

You've got the job of the secondary taps backwards. It's supposed to keep the primary load constant over all impedance ratios. Another way of saying it keeps the primary load constant with changing secondary loads. The constant in the whole thing is the primary load. Sortof.

Say you want a 3.3k primary load out of the aforementioned transformer. In that case, the transformer just became a 3.3k transformer with 8,4,and 2 ohm secondary taps. Forget that it was ever defined as a 6.6k tranny. It's a 3.3k tranny now.

Again, the secondary taps on the transformer's jobs are to make sure the primary load stays the same with different secondary loads. The ratio changes so that the primary load doesn't.

Matt.

[lewilson]

the ratio stay the same
A 6.6 ot with a 4-8-16 ohm output could become a 3.3k ot but it would be a 3.3k with 2-4-8 ohm loads if I understand correctly
turns ratio squared times the load. The turns ratio is always constant
20 to 1 or 28 to 1 or 40 to 1

[Andy Le Blanc]

heres a bit more to help out...... sorry its math

[rooster]

Andy and Matt - (I guess we're the only ones here at this point?), I wanted to say that you are both gentlemen and I have enjoyed, and learned from this post, as well. Thank you both.

FWIW, for my second Express clone, I am using the smaller 25 watt Hammond OT that has a 7600 ohm primary. Where currently, with the Tonesluts OT @ 6600 ohms I will just keep my taps mated to the given values/loads, i.e. 4-8-16 ohms. (Glen is a big influence here for me regarding this OT, of course.) However, I do think that this newer OT's performance will be better served by treating the taps as 2-4-8. But we'll see. I am really curious to hear the Express as a 6V6 amp amp, as well, so it may be that I never find out how this OT sounds with EL34s..... I am actually just hoping with this OT choice that I can enjoy the sound and feel of the Express circuit at a tad lower db level, and still have some clean volume, too. Like I said, we'll see.

Anyway, good luck with your projects.

[morcey2]

the ratio stay the same
A 6.6 ot with a 4-8-16 ohm output could become a 3.3k ot but it would be a 3.3k with 2-4-8 ohm loads if I understand correctly
turns ratio squared times the load. The turns ratio is always constant
20 to 1 or 28 to 1 or 40 to 1

I think we're saying the same thing, but in different ways. We have 3 logical trannies in one physical one.

Tranny one is 20:1 turns ratio.
Tranny two is 28:1 turns ratio.
Tranny 'C' is 40:1 turns ratio.

If we want a 6.6k primary impedance, you need to pick the right ratio, an thus the right secondary tap. Each ratio doesn't change, we pick a specific ratio for the loads involved.

heres a bit more to help out...... sorry its math

We don't need math anywhere near amps! They lose they're tone if they see anything math related!
(this coming from a geek that picked up a calculus book from a thrift store just cause I wanted one.)

PostPosted: Thu Jun 05, 2008 10:43 am Post subject:
Andy and Matt - (I guess we're the only ones here at this point?), I wanted to say that you are both gentlemen and I have enjoyed, and learned from this post, as well. Thank you both. Cool

FWIW, for my second Express clone, I am using the smaller 25 watt Hammond OT that has a 7600 ohm primary.

Thanks for the kind words.

I used that same tranny in a Plexi6V6 and it sounds awesome. The only reason I picked that one (1650F) over the 1645 (6.6k, 20W) is so I could use KT66's if I wanted to. Never did that though.

I'm building an express for a friend of mine using a 1650H and 272HX. I'm also building one for myself with a MPS 6.6k 20W tranny. Mine will be 6V6 only but the other one will be able to run either EL34 or 6V6s.

Anyway, I'll shut up now.

Matt

[Andy Le Blanc]

ah..... but the fun has just started...... this has a very practical application
you can test your own trannies...... you need a signal generator and a multi meter
I have a fender 25w replacement.... nsc041318 with a published 8.5k pri.
and an intended 8 ohm sec. .... this gives a impedance ratio of 1062.5 : 1
and a turns ratio of 32.6:1.....

1) connect your signal generator to the secondary and measure
A) Hz
B) V

2) measure at primary (CAUTION: there can be high voltage)
A) V

this gives enough shreds to do it yourself

with my tranny I got........ @1K Hz....... 3V sec......... 89V pri.
so thats 89 : 3 or a turns ratio of 30 : 1 and impedance of 900 : 1
So @1K Hz this tranny with an 8 ohm load presents 7.2K at the primary
it is quite differnt from the published spec at that freq. by 15%

now I only have a low freq. HP generator so I cant test above say 1.3k
but if you find a product the sounds the balls you can plot your own data
and use it as a basis of requirement for other models

you can get a ball park figure on a mystery tranny

and if your tranny is muddy or lacks definition or bass you can find out
if its an issue of loading

and with the load line information make much better design
choices for load, plate voltage, and point of operation

[lewilson]

I think it would be nice to check the actual impedence of the output tubes. I know with preamp tubes the z is different depending on gain. Higher gain higher z. its not usualy a problem with guitar amps becouse the preamp tubes are working into a much higher impedence. But with output tubes it seems like it be usefull to see the actual impedence in the circuit. I know the tube manuals have this info, but it would be fun to check them. I do it with preamp tubes sometimes, but never output tubes, I think the process would be the simular. Anyone done it?

[RJ Guitars]

I was digging for info on a good output tranny for a quad of 6V6GT tubes in a Dumble ODS type of circuit and came on this thread.

I'll propose a hypothetical transformer use with the intent of understanding how the frequency response changes when using a transformer rated for 60-15Khz at a 6.6K primary impedance and 8 ohms secondary impedance is actually used at a 3.3K primary impedance.

Q1 - If I understand things correctly, with 3.3k on the primary side I can plug an 4 ohm speaker into the 8 ohm jack and maintain the ratios of input to output. Does that make sense?

Q2 - Assuming I have the ratios working in the correct direction in my assumption cited above, a tranny with a 6.6K primary with 4,8,16 ohm outputs becomes a 3.3K primary with 2, 4, 8 ohm secondary taps.

Q3 - Can anyone tell me if this type of use impacts the frequency range or cause other issues in the ultimate sonic performance?

I appreciate any insight available on this topic.

rj