Choke filter

[pjd3]

I'm curious if anyone has ever experimented and tried say, over rating the HV voltage while implementing a choke filter in an otherwise standard well known guitar tube amp model or topology - keeping everything else about the same.

My understanding is that there is much better voltage regulation with the choke filter and I'm curious what the effect there would be in the dynamics and tightness of the playing. I say "over-rating the HV voltage as there is a good deal of HV voltage dropped in this filter configuration.

Thanks

PJD3

[brewdude]

Choke input PS reduce the voltage considerably. I have built a few amps with this kind of scheme, but they were odd circuits, not really resembling any classic amps. If you wanted it to operate at normal circuit voltages for a particular amp, you will need to design around a different power transformer.
Otherwise, I suspect the low voltage will effect the sound as much as the choke “advantage”.

[R.G.]

Brewdude has it right.

A full wave center tap rectifier/capacitor input filter produces a voltage that's 0.707 times the end-to-end secondary voltage. A full wave center tap rectifier with a choke input filter produces a DC voltage of 0.45 times the end-to-end secondary voltage when loaded above the "critical current" of the choke input filter. So the output is about 36% lower for the same power transformer secondary voltage. You do, as you suspected, have to pick a transformer with a higher voltage to get a specific B+ voltage.

Choke input filters have quirks. At low currents, the output voltage is just that of the capacitor input filter, and as you load it more, it declines down to the lower voltage of the choke input filter. The current where this happens is that "critical current" I mentioned. So your caps and everything else in your circuit have to withstand the higher voltage of a cap-input filter until current ramps up over the critical current and the voltage becomes stable. To get a lower critical current, the inductance value needs to be higher, so you need a big inductor, much bigger than is typical in a guitar amp, and it has to carry high currents, so it's heavy and expensive, even more so than the typical ones in amps.

In a typical guitar amp, the capacitor input filter feeds the choke that filters only the lower currents of the screens and preamp tubes, so it's a much smaller and cheaper choke than a choke that would filter the power stage too.

People who design choke filters for some special purpose typically install a resistor bleeder after the choke to force it never to go below the critical current so they can sidestep the 34% change in output voltage if the load from the circuit goes down.

As to tone; the choke does make for a more solid voltage source after the choke. The voltage would be stable, much like a regulated voltage. So the tone change would include essentially no sag. People talk at great length about good sag, so that aspect would be gone. There would be essentially no power supply ripple, though.

Overall, the electronics world abandoned choke input filters for the reasons of cost, weight and expense. They're used only in very special cases, as regulators are cheaper and lighter.

[maxkracht]

Just tried this in another topic "Heathkit condeser checker power supply" It worked, but the choke was shaking like crazy. Possible that the old mojotone choke was noisy, needed to clamp it more/add shock absorbers, or I was doing something else stupid, but I think that's just how that circuit works.

[LOUDthud]

As to tone; the choke does make for a more solid voltage source after the choke. The voltage would be stable, much like a regulated voltage. So the tone change would include essentially no sag. People talk at great length about good sag, so that aspect would be gone. There would be essentially no power supply ripple, though.

This characterization of what a choke does in a power supply appears in many text books, but a guitar amp puts a slightly different twist on things. Ignoring the Reservoir cap for a moment, an L-C filter has the effect of taking the Average Voltage of whatever Voltage waveform is on the input side. In the case of the choke input filter where the input is just a rectified sine wave, you get the Average of those double bumps that works out to about 0.45 times the peak of the bumps. Now adding the Reservoir cap back in to form a Pi filter, the L-C filter sees a sawtooth waveform riding on some DC. The Averaging action of the choke puts the DC on the output side of the L-C filter a little less than half way between the peaks and valleys of the sawtooth (ignoring DC losses). On the output side of the L-C the sawtooth is almost completely removed, you might see a small sine wave (2x mains frequency) if you look really close on a scope.

Here's the twist. In the typical class A/B guitar amp (where the big load comes right off the Reservoir cap), when you hit a big power cord, that sawtooth wave on the Reservoir cap grows to about four times the peak to peak Voltage that it was when the amp was just idling. The peaks of the sawtooth will drop slightly, perhaps 5%, a little more if you have a tube rectifier. The L-C filter still does it's job taking the Average of it's input waveform but (and it's a big butt) that Average has dropped about 25V. The output waveform is still virtually ripple free. Many times there is a slight ringing on the output side of the L-C with a frequency of about 20Hz. You see this ringing on Fender Tweed amps where low choke resistance and smallish caps were used creating an under-damped condition.

Now consider the TW-Express (and most other TW amps). The Express has rather large filter caps compared to other guitar amps of the same power output. This means that the sawtooth on the Reservoir cap is smaller at idle and when you hit that big power cord. The class A/B nature of the Screen currents creates a big drop across that 25W 1K on power cords, but it doesn't seem to affect the preamp in a negative way. Ripple is still low in the preamp and low enough for the screen grids.

Ripple_01.PNG

[TUBEDUDE]

Just tried this in another topic "Heathkit condeser checker power supply" It worked, but the choke was shaking like crazy. Possible that the old mojotone choke was noisy, needed to clamp it more/add shock absorbers, or I was doing something else stupid, but I think that's just how that circuit works.

You might need a bigger value choke. It sounds like it's operating below the critical value for stability.
You can try shunting the choke input to ground with a small film cap, say 220nF.
It's small enough to still be a choke fed supply but will maintain the current in the choke between pulses. That might calm it down.

[LOUDthud]

Just tried this in another topic "Heathkit condeser checker power supply" It worked, but the choke was shaking like crazy. Possible that the old mojotone choke was noisy, needed to clamp it more/add shock absorbers, or I was doing something else stupid, but I think that's just how that circuit works.

Are you still using half wave rectification ? That will drive the choke "crazy". Install a solid state diode bridge rectifier.

[maxkracht]

You might need a bigger value choke. It sounds like it's operating below the critical value for stability.
You can try shunting the choke input to ground with a small film cap, say 220nF.
It's small enough to still be a choke fed supply but will maintain the current in the choke between pulses. That might calm it down.

Hadn't tried it yet, but I have my copy of merlin's power supply book open to the page which suggests the same cap. Low current and high inductance, doesn't seem to be available from suppliers I have checked. My current stash is a few assorted mH range old hammond chokes, some 3-4h stuff for 6v6's and one bigger thing for 6l6s which wouldn't easily fit in this project. Attempting to repurpose someone else's garbage, so nothing critical and a different transformer is likely easier.

Are you still using half wave rectification ? That will drive the choke "crazy". Install a solid state diode bridge rectifier.

Full wave ss bridge, nothing out of the ordinary. I literally clamped an old mojotone choke to the chassis to see if it would calm down. No dice.

[maxkracht]

Not trying to hijack this thread, I just wanted to point out the possibility of a noisy filter. I may have misread the end of this summary and thought "wild oscillations" were possible by themselves and not only as a result of "start up, switch off, or if the load current drops below the critical value". I was using a 5k resistor as a load because it was there and seemed good enough for testing purposes. Maybe this was the problem. Will try some things and report back on my original post about that power supply.

"The disadvantages are the low output voltage and high cost of the choke, and the possibly wild oscillations and flyback voltages that can occur at start up, switch off, or if the load current drops below the critical value."

IMG_8040.jpg