Plate Load/OT/Speaker Impedance Mismatch

[EETStudent]

I have a concern regarding a 5F4 style build Im putting together. I have some iron from a Tweed Pro laying around that Im considering using. The output transformer is 6.6K, single 8 ohm tap.

For the 5F4, Ill be using a pair of Weber 10F150s for a speaker load of 4 ohms. I dont know the math, but given the OT primary and secondary impedance along with the speaker load, what sort of mismatch am I looking at here? Should I consider purchasing a 4 ohm OT for the project, or can I get away with this? Im looking for a technical answer

[EFK]

I'm not excessively technical but 6.6K at 8 ohms into a 4 ohm load means the amp is working into a halved, or 3.3K, load. Whether or not this is acceptable is going to depend on a lot of things like choice of power tubes, the plate voltages involved and bias etc.

[R.G.]

EFK is correct; it will depend on a largish number of things.

However, tube amps are known to not die when shorted; also, a mismatch of 2:1 in either direction is usually non-fatal unless there are other issues lurking inside the amp. Many people use 2:1 mismatches for the change in tone.

[pdf64]

What power rating are the speakers?
If ok you could try with just 1 speaker to present 6k6 to the tubes.
You could then connect the other speaker (in parallel) and see what effect that has on the amp, in terms of tone/volume.
The other speaker needn't be in the cab; rather it'd be best for it to be remote, eg in another room so you can't actually hear it, thereby eliminating it's sound reinforcing effect.
3k3 shouldn't upset the operating conditions of the output tubes.
My experience is that lower impedances sound punchier, higher impedances sound more 'harmonicy', so the mismatch may lose some of the tweediness.
If monitoring the B+ ripple, I often notice that it increases with a lower than nominal load, so you could compensate by upping the caps in the 'Pi' filter.
The lower load will alter the NFB ratio, so you may chose to reduce the 56k feedback resistor to compensate; if so, consider something around 39k (=56 x 0.7, in line with the voltage ratio between 4 and 8 ohms).
Whatever, consider fitting screen grid resistors http://www.webphix.com/schematic%20h..._5f4_schem.pdf
Pete.

[EFK]

One particular time I would be careful going with a lower load is if the amp is cathode biased, and especially so if it is cathode biased and running higher than typical plate voltages for cathode bias (400V +). The lower impedance *may* allow the cathode current to increase more than it might otherwise, and without enough voltage sag to compensate, you may end up over-dissipating your tubes. Not that this is warranted in a 5F4, but since you mentioned "5F4 inspired" I thought you possibly may be considering running it in cathode bias. Lots of people seem to do so in older Fender-inspired circuits.