Use an OT for higher Z

[uneumann]

The recent threads about OT design showed me how little I really understand about these components.
Here is a practical problem that may help me (and others) develop some better instincts about OT behaviors.
I want to use a 5F6a OT (4K:2ohm) as a 8k:4ohm OT or a 12k:6ohm OT or even a 16k:8ohm OT.

Since the OT is designed for 4k:2R, what problems arise when I use it for higher primary Z cases?
From the prior discussions, I understand I lose some bass response as the primary Z increases, since the primary inductance is designed for an expected 4k impedance - did I get that right?
What other effects/issues are likely?

I'm thinking about the OT performance only - not the circuits driving the OT.
I assume I can get the PP tube voltages/currents to match up to a given primary Z.

Apologies if this was covered elsewhere. I did do a search but may have missed something.

thanks

[Old Tele man]

Transformers are impedance "ratio" devices, so whatever LOAD you hang on the secondary (output) side is "ratioed" UP (by the TR) into the PRIMARY (input) side of the transformer. So,the two 'key' words are:

LOAD(Z) and TURNS-RATIO(TR)

Also, the higher the reflected input-Z becomes, the greater the chance of "arcing" being induced (flyback) into the primary circuit (OT winding AND tubes) if/when secondary LOAD goes 'open' suddenly.

[Mick Bailey]

+1.

The thing to bear in mind is that the transformer doesn't have an intrinsic impedance. Also, within the constraints of a guitar amp there is flexibility in the OT design such that a range of matching scenarios can be met with a transformer of a given turns ratio so long as the current capacity of the windings is not exceeded - this is a critical factor and impacts on the use of the transformer in different amps, perhaps more than any other factor.

[Helmholtz]

The recent threads about OT design showed me how little I really understand about these components.
Here is a practical problem that may help me (and others) develop some better instincts about OT behaviors.
I want to use a 5F6a OT (4K:2ohm) as a 8k:4ohm OT or a 12k:6ohm OT or even a 16k:8ohm OT.

Since the OT is designed for 4k:2R, what problems arise when I use it for higher primary Z cases?
From the prior discussions, I understand I lose some bass response as the primary Z increases, since the primary inductance is designed for an expected 4k impedance - did I get that right?
What other effects/issues are likely?

I'm thinking about the OT performance only - not the circuits driving the OT.
I assume I can get the PP tube voltages/currents to match up to a given primary Z.

Apologies if this was covered elsewhere. I did do a search but may have missed something.

thanks

Could you be somewhat more specific about your intentions?

Transformers have limits on currents, voltages and power. I can't think of a problem using a 45249 type OT in a PP amp of up to around 50W requiring 8K:4ohm. Especially as this OT has a high primary inductance of around 190H at rated power (Marshall 50W models having 70H to 100H) and because currents will be lower and primary voltage will not rise significantly.

[uneumann]

Could you be somewhat more specific about your intentions?

Transformers have limits on currents, voltages and power. I can't think of a problem using a 45249 type OT in a PP amp of up to around 50W requiring 8K:4ohm. Especially as this OT has a high primary inductance of around 190H at rated power (Marshall 50W models having 70H to 100H) and because currents will be lower and primary voltage will not rise significantly.

OK - So, let's say I'm keeping a 450v B+ in all cases. So, I'm not too worried about driving the OT when it reflects a higher primary Z than it was designed for. Maybe I should be, but that's part of my question. The OT was designed for 4k:2R. If I use that same OT for 16k:8R, I expect lower primary peak currents and somewhat higher voltage swings (depending on bias) since the load line becomes more horizontal. But I don't see either of those as potential stresses for the OT. Or am I missing something? Am I likely to damage the OT.

The real question is however, about OT performance when used at different impedance than it's designed for. From prior discussions, I've come to expect I'll lose some bass response since an increase of 2X or 4X primary Z will cause the LF limit to increase by 1-2 octaves since the inductance is fixed by the designer for use as a 4k:2R OT. The 5f6a OT may have a huge L, but still, the roll off will be higher - right?

Are there other issues that arise when using an OT at a much higher primary Z (or lower for that matter) than it was designed for?
What about HF performance? Power loss/efficiency? Something I'm not aware of?

[Malcolm Irving]

… I want to use a 5F6a OT (4K:2ohm) as a 8k:4ohm OT or a 12k:6ohm OT or even a 16k:8ohm OT.

Since the OT is designed for 4k:2R, what problems arise when I use it for higher primary Z cases?
From the prior discussions, I understand I lose some bass response as the primary Z increases, since the primary inductance is designed for an expected 4k impedance - did I get that right?

I'm thinking about the OT performance only - not the circuits driving the OT.
I assume I can get the PP tube voltages/currents to match up to a given primary Z.
…

Yes, that's right. A good model of a transformer, beyond the 'ideal transformer' model, is to include the magnetising inductance in parallel (i.e. shunt) with the primary of the ideal transformer. While the load reflected onto the primary changes in proportion to the load on the secondary, the inductance stays the same.

The low frequency -3dB point, roughly speaking, is when the reactance of the shunt magnetising inductance is equal to the reflected load impedance on the primary. (At this point half the current supplied to the OT is wasted into the magnetising inductance and only half gets through to the load).

So if the reflected load is 4x higher, the -3dB frequency will be 4x higher (so that 2.pi.f.L can be 4x higher).

[Malcolm Irving]

Perhaps I should clarify why I said ‘roughly speaking’ in the above. If the output tubes are current sources (which pentodes are approximately) then it is fair to say that half the current is wasted into the shunt magnetising reactance at the -3dB point.

But if the output tubes were voltage sources, it wouldn’t matter what shunt was across the load, as the voltage across the load stays the same. Triode mode takes the tubes closer to being voltage sources but still a long way from it.

For a more exact calculation of the -3dB point, the output impedance of the tubes would need to be taken into account.

[Helmholtz]

The real question is however, about OT performance when used at different impedance than it's designed for. From prior discussions, I've come to expect I'll lose some bass response since an increase of 2X or 4X primary Z will cause the LF limit to increase by 1-2 octaves since the inductance is fixed by the designer for use as a 4k:2R OT. The 5f6a OT may have a huge L, but still, the roll off will be higher - right?

Right, but I thought this was clear already. Also the bass response of any OT gets worse at lower power as primary inductance decreases at lower magnetizing currents. To calculate the -3dB lower corner frequency at lower power you would need to measure the primary inductance at the respective lower primary/magnetizing current values or just measure the frequency response in the designated amp at different output levels.
The upper limiting frequency is determined by the leakage resonance and is typically around 20kHz for this kind of transformer, i.e. far beyond guitar reproduction requirements.