Best way to cure output osscilation

[Helmholtz]

My understanding is that below resonance, driver impedance drops quickly, and by 40Hz it may be below nominal.
For a driver in a cab, compared to its free air unloaded characteristics, the bass resonance freq, Q and magnitude may change; with a ported etc cab, strange things may happen to the freq / impedance chart.

You are right, for a cabinet having a bass resonance at 80 Hz or higher, the impedance may be down to nominal at 40Hz (hardly any lower according to my measurements). But as this amp will be used with a bass cabinet, resonance may be lower and impedance may be higher than nominal.

[catalin gramada]

So running DC current through secondary side will not affect the OT performance in low freq range? Thinking more about NFB circuit where partial DC current (from cathode where nfb is aplied) is running through secondary OT side.

[Helmholtz]

So running DC current through secondary side will not affect the OT performance in low freq range?

Who said so?

In principle DC on primary or secondary will degrade a transformers performance. But a few DC milliamps on the (low turns) secondary of an OT certainly won't hurt.

[catalin gramada]

Who said so?

In principle DC on primary or secondary will degrade a transformers performance. But a few DC milliamps on the (low turns) secondary of an OT certainly won't hurt.

I figured the circuit below.



So, maybe is better to choose lowest impedance secondary tap ?

[Helmholtz]

Assuming that 5mA is the total PI cathode current producing a voltage drop of 0.5 VDC across the 100 Ohm resistor and that the inductor in your drawing is the OT secondary:

- in the left drawing the OT secondary will see a DC current of 0.5mA which is close to nothing.

- the drawing on the right is no good solution at all: The 100µ e-cap acts in series with the 4 Ohm load, producing a low frequency roll-off below 400Hz. Also the cap would have to be a non-polarized (NP) type rated for AC currents over 5 A. Otherwise the speaker output could not be grounded.
To avoid these issues, the e-cap could be wired to the upper side of the secondary, so that it doesn't carry speaker current (still would need to be an NP type).

But the cap is absolutely unnecessary.

[d. spree]

Just to be clear, it's some thingie in there causing it.

or some other thingy

[catalin gramada]

And 40 cps/ 125W with Hammond 1.9k, same circuit.



Think is done. Time to put everything in a box.

[catalin gramada]

I didn't quite understand something.please. How comes to much feedback in my circuit ? Except the fact the power stage get clipping more like a SS diode I didn't found any inconvenient. I reversed the triode before from a cathode follower to regular gain stage to compensate the gain(still good 17k around output impedance).Just added 13db of nfb for more tight, controled bass reponse with a ridiculous low 235/100 ohm divider from 4 ohm ot tap. It sound good, output looks good, and still have resources, get 125w rms from a quad of el34. Should I concern about some issues regards this nfb amount, please?. Thanks

[Chuck H]

The only reason I have for sacrificing some NFB, and maybe ending up with a little more waveform distortion, would be that the more hi fi nature of a clipped wave form (and even the less obvious distortions in the sine wave form) as well as some limitation of frequency response accuracy can be a desirable aspect of a tube instrument amplifier. Otherwise guitar players would be using the same gear as car stereo amps. As was mentioned earlier and repeatedly, at some point attempting to design all the distortions and limitations out of an instrument amp becomes counter productive and even a detriment to the cause. Using NFB ratios akin to home stereo amps would probably hedge into that area.

[catalin gramada]

The only reason I have for sacrificing some NFB, and maybe ending up with a little more waveform distortion, would be that the more hi fi nature of a clipped wave form (and even the less obvious distortions in the sine wave form) as well as some limitation of frequency response accuracy can be a desirable aspect of a tube instrument amplifier. Otherwise guitar players would be using the same gear as car stereo amps. As was mentioned earlier and repeatedly, at some point attempting to design all the distortions and limitations out of an instrument amp becomes counter productive and even a detriment to the cause. Using NFB ratios akin to home stereo amps would probably hedge into that area.

Hey Chuck,
Is a clean bass amp, nothing is happen to distort over here. The reason for this amount of nfb is flabby bass reponse with the speakers around.

[Chuck H]

Build this amp for a friend to be used for bass and guitar aso (also?)

So it's a bass AND guitar amp. Many of this type do sound good for both. But certainly not a modern, high performance bass sound.

There are many incarnations of bass amps through history that don't exhibit the performance you've been chasing. It amounts to a different sound. Those amps weren't intended to make un flabby 40Hz reproduction which the speakers couldn't do anyway. It's just an apples and oranges thing. Maybe not what you're after and maybe nothing that would satisfy you or your friend. I think your performance specs are impressive in their own way and you get kudos from me for working so hard at making your design goal a reality.

[catalin gramada]

Hello,
This is how it was settled the bass preamp section, Ampeg inspired. Maybe can be more tweaked but start to get tired, in every morning seems to sound different, which of course is not. It is stable and sound pretty decent. From an electrical point one problem is the bias of PI stage which cannot be stable as time the supply drop 40-50v around between idling and full throttle. I.ll post a sketch. Please let me know if You see some horrible mistakes into the circuit which I missed Thanks.

[catalin gramada]

And one more question ,please:
Is there a difference regards nfb in effect between this two situations? It have sense in case of bypassed cathode resistor, but in the situations from my sketch I didn't get it. Is this related to output damping factor for the same amount of nfb applied? Make sense to spend one more extra resistor,why?

[Helmholtz]

And one more question ,please:
Is there a difference regards nfb in effect between this two situations? It have sense in case of bypassed cathode resistor, but in the situations from my sketch I didn't get it. Is this related to output damping factor for the same amount of nfb applied? Make sense to spend one more extra resistor,why?

[ATTACH=CONFIG]53550[/ATTACH]

I don't see a qualitative difference. Especially the damping factor will be the same if the NFB ratio is the same. Regarding quantitative results (feedback divider) you need to take into account the internal cathode impedance of the triode (roughly 1/gm, more exactly (rp+RL)/(µ+1) ).

[Chuck H]

This^^^^^^^^^^^^^^^^^^

There shouldn't be any real difference to the action of the feedback. The divided tail will affect the ratios, but I don't see why it should affect the result other than that. All things being equal. What the divided cathode resistance is good for though is when you want to use cathode bypass on the cathode you are using for the NFB input. You can use a cap to bypass the resistor connected to the cathode for those benefits and still inject NFB relative to the lower resistor. And because of the low impedance of the amps output you can pretty much have your cake and eat it too. Something like a 1.5k cathode resistance that is bypassed and then a 47R below would allow the injection of NFB with almost no detriment to bypass cathode affect. Fender did this with a couple of models.

As far as unbypassed cathode circuits I don't know of any advantage to the split resistance in the cathode for NFB input.