OTs and Bandwidth

[dorrisant]

How much does the build of an OT affect the bandwidth that is passed. In particular, I have a prototype amp (Ultra Phonix style), that uses a Twin OT where the power tubes are 50-60 watts (SR PT). I am finding that the bottom end seems comparable to the same type of amp with a 50 watt OT and an added cathode driver stage. Drop the CFD and the bandwidth narrows. Change the OT from Super Reverb spec to Twin Reverb spec and the bandwidth widens again.

Anyone else experience this?

This seems to jive with changing the OT sometimes... the most noticeable is when I change a VHT D-Fifty stock OT to a SR clone from Mercury Magnetics. If you say/think there is no difference then you may not care to read further. But, to me there is a difference.

Comments? Opinions?

[mhuss]

The build will affect the frequency response. I don't know by actual measurement, but I'll bet any OT meant for a guitar amp will have less bandwidth then, say, a MacIntosh OT. Extending the HF range means minimizing the inductance and inter-winding capacitance - this is why some high quality OTs will use two parallel windings ("bifilar") for each tap. Extending the low end means a bigger core (to avoid saturation). Both of these options cost more money and/or add weight, typically not spent on guitar amps.

[R.G.]

How much does the build of an OT affect the bandwidth that is passed?
Some. To the extent that the OT is what is limiting the bandwidth passed, it's big. To the extent that other things are limiting bass or treble to less than the bandwidth of the raw OT, then it will be hardly noticeable if at all. And to the extent that feedback is not being used, it will be more noticeable, as feedback can extend the bandwidth of a circuit to wider than the no-feedback bandwidth.

A subtlety is that the driving impedance on the primary of the OT has an effect on the bass cutoff. The lower the impedance that drives a primary inductor, the more the bass cutoff is extended. The higher the impedance that drives the primary, the more the bass cutoff point is increased, decreasing bass.

From your description of what you're doing, I think you're seeing a combination of effects, not just the OT, although changing the OT brings out the differences.

The bandwidth of an OT was fairly well explored in the Golden Age. For better bass response >> of the OT itself << you make the primary inductance of the OT as big as you can, limited by the other constraints, such as budget, weight, marketing, use of feedback, yada, yada, yada. For better treble response, you make the leakage inductances as small as you can, within the other constraints. At really ear-bleeding treble ranges, you start worrying about capacitances between windings. Guitar amp OTs do not do this last set of considerations in anything I've ever seen. In the Golden Age, the "figure of merit" for an OT was the ratio of primary inductance to leakage inductance. This needed to be as big as possible, often up to 1000 - 10,000.

[dorrisant]

Ok... Then it stands to reason that a HiFi OT would be better for a topology like a Dumble, where the PA section is more Hifi than most other amps. Thanks to you both for the confirmation.

What I have experienced is that either adding a CFD or upgrading the OT to a much physically larger specimen has increased the bandwidth available. I've done this on plenty of builds and rebuilds, enough to hear proof of it. Combining both yields amazing results. I was just wondering which one was doing this more and if there was a way to compare the increase. No huge explanation needed though. It does make me wonder about the cost effectiveness of either.

I know some have complained about the higher price tag we see with Mercury, but this is one of the benefits that seems to jump out of the amp when upgraded. As you point out, the circuit may cut off the bass that would allow one to notice this difference, so some circuits will not benefit as much as others. In the case of a Dumble circuit, it is as apparent as the nose on your face.

[bepone]

there is a big diference in bass quality in various OT's.. it is function of lamination quality (permeability) and number of the turns.. i like more inductance (nuber of the turns) because lower register is stronger in comparison with lower inductance where is muddy,

[R.G.]

I think that it's even more demanding to design a bass-specific amplifier OT than a power transformer. If you do the right thing, design for drop-D tuning on an electric bass and realize that a bassist really needs 5 to 10 times the power than a guitarist, you get some truly big transformers that also have to be at least modestly hi-fi. So yeah - bass ability is going to be greatly variable.