DF theory question

[catalin gramada]

Hello, there is a relation between damping factor and OT Ra-a load in an output stage please. May someone explain for pentode and triode mode please what effect from a damping factor point a low or high Ra-a load have ,if any ? Thanks.

[J M Fahey]

NFB has a BIG say on source/generator impedance so Damping Factor as seen by speakers.

Without NFB pentodes are straight current sources so horrible DF, the source of horrible 30s 40s early 50s boomy undefined "one note Bass" systems.
Compounded by stiff very high Q speakers which made it only worse.

Triodes on the other side have built in like it or not NFB, which provide WAY better damping and much true Bass, and overall flatter frequency response.
Couple that to better speakers, and youŽll understand why old time HiFi lovers (and current Japanese) prefer those 2A3-300B-45 based amps.

Amply surpassed decades ago but hey, "Audiophiles" are driven by perceived quality, not actual measurements or sound.

[Helmholtz]

To add:
UL operation lowers output impedance by screen NFB. Circuit properties lie between triode and pentode mode.

[catalin gramada]

Well, I was wondering if possible to get a good bass response without any global nfb ? My long trial and error experiments shows even triode strapped there was no way to keep low freq under control without additional care. A quad of kt88 was used into 2500k Ra-a load at 600v ended with exactly 100wrms. But keeping the bass under control was a challenge so local feedback was used at the driver stage and cutoff pole to the grids at 34 cps. That seems enough for bass guitar but no means really low bass.The power supply is a low impedance one able to 1.5 amps and used a hefty 500u bank caps ,so no lack of resources here. I wonder if there is something more to do to not get farty and boomy into the rumble region please ? Thanks.

[Helmholtz]

Well, I was wondering if possible to get a good bass response without any global nfb ?

In triode mode you might be able to achieve a damping factor of 2.

Why are you avoiding global NFB?
Some of the finest tube hifi amp designs use up to 20dB of global NFB (e.g. Williamson amp).
NFB not only increases the DF but also compensates the bass rolloff of the OT.

But your problem may also be insufficient acoustic damping of the speaker.

[loudthud]

I build a 4x6550 bass amp with a Pentode - UL switch. Didn't like the UL mode, switched it to Pentode and never changed it.

Here's how I measure the small signal output impedance.

[catalin gramada]

This is the circuit I playing with.

[The Dude]

So, you have plates and screens on the same PS node?

[catalin gramada]

Yes there are. At nominal power screen power dissipation at 600v with used load is under 6w within specs. No sign of oscillations with 100 stoppers. Max plate current draw 80mA around meant 48w total for plate and screens,
Also tested, hammond 1650R shows no sign of saturation at 100w rms even at 35cps

[Helmholtz]

So, you have plates and screens on the same PS node?

Well, it's triode mode.

[catalin gramada]

My only problem is not enough DF I think. Trying to boost the low bass, to shake my basement more litle bit and the speaker bass control was lost getting flabby. Unbypassed driver stage and choosing cutoff point was some which comes by trial to keep a limit control.
I know the speaker used is still capable to get lows used with a PA ss amp at same freq.

[catalin gramada]

The prototype

[Helmholtz]

Again: Why not some global NFB?

[catalin gramada]

This projects are my own experiments by curiosity. I did lots of amps using global nfb. I proposed to make a project without to get conclusions, now I ask a bit help to improve it please. My hobby have a sense by matter of sound experiments. I don.t have any ideea if changing the OT with other with different specs will make things different. As far the Ra-a used found to be best compromise for aplication at max voltage admissible: 1.9k will get 130W around with the price of fast toasting tubes, 5k makes ridiculous 55 w power, still in AB class for max admissible bias allowed. I bet harmonic distortions will come better, but who cares ?
I found in any amp the lack of global nfb makes sound more dynamical, more open. I.m just curious how far this can be done ? Maybe supplied at lower voltage and bigger current help ? Really have no ideea how the speaker reactance act in term of reflected EMF to power stage and with whatever consequences. But I.m pretty sure things may be optimised to a point...Thanks

[g1]

I found in any amp the lack of global nfb makes sound more dynamical, more open.

Unfortunately, at least for low frequencies, I think 'open' means the same thing as 'flabby'.