LM3875 Inverted Or Not?

[tom wesley]

Im thinking of building a valve buffered clone. Does it have to be an inverted setup? Am i over/under complicating things. Is non inverted possible?

 

[AndrewT]

I get the impression you don't know what you are doing.

Post a schematic and hope that someone can guide you.

 

[peranders]

You can take a peek at my project but the design isn't set yet and not explained. This will come later on. The schematics will also have variants.

My project is a single/dual E88CC tube buffer in cathode follower mode or "White Cathode Follower" mode + inverted/non-inverted gainclone (LM3886) + power supply.

 

[tom wesley]

I understand the schematic in so far as ive built gainclones (non inverting kit) and wondered if i could builb a valve buffer for it. The schematics are all a variation on inverted designs. I dont fully understand the implications of using the valve with the chip

 

[peranders]

You will get a rather low input impedance of the inverting amp so a buffer could be good to have and why not a tube buffer in order to get some tube flavored sound.

 

[tom wesley]

But is it feasible to do with non inverting amp?

 

[Robert Kesh]

But is it feasible to do with non inverting amp?

yes, though no longer as necessary to have a buffer. you would just do it for colour.

perhaps people only use inverting to justify using a buffer to themselves, and hence valves.

 

[KSTR]

Unless you have whimp of a preamp, inverted config doesn't necessarily need a pre. You can get pretty high input impedance (like 10k) when using a T-network for the feeback path.

 

[planet10]

But is it feasible to do with non inverting amp?

Thorsten Loesch's design -- one of the originally published gainclones -- is inverted. The inverted version includes everything "before" it in the feedback loop so a buffer is really required. Buffered, inverted gainclones, in my experience, are best of breed. If using a cathode follwer, for best performance requires a CCS on the tail of the CF.

dave

 

[madtecchy]

Unless you have whimp of a preamp, inverted config doesn't necessarily need a pre. You can get pretty high input impedance (like 10k) when using a T-network for the feeback path.

I have tried the t network on my inverting lm3886. The extra connection to ground seems to inject ground noise into the input... Any comments?

Ps can get 10k input impedance without t network in inverted mode

 

[KSTR]

I have tried the t network on my inverting lm3886. The extra connection to ground seems to inject ground noise into the input... Any comments?

Looks like layout/wiring issues, not a conceptual problem. T-network has a slight noise penalty in general, though. And higher than unity gain at DC with either only input coupling capacitor or T-network cap (in the GND leg).

Ps can get 10k input impedance without t network in inverted mode

Yes, that means 100k++ feedback resistor which is a bit too high for my tastes. More important is that any inverting confic should have a capacitive feedback divider (to GND rather than -Input), too, to keep it stable with open or high-Z input termination.

 

[AndrewT]

Could either of you post typical schematic to show what your are referring to?

 

[tom wesley]

Thanks for the help guys. Thats answered my question...

 

[KSTR]

Inverting amp with T feedback network

 

[madtecchy]

KSTR..

I use a computer and an external dac.. I tried the T network on an already built and in service inverting lm3886. I had no issues with it at all. I read about the t net feedback idea and i gave it a try as it is simple to implement . and with the t network computer noise is audiable.. I have no stability issues with just a single feedback resistor . 220k

 

[KSTR]

If it's runng OK (also in terms of offset) with 220k that's fine. But what happens if you happen leave the input open circuit? Unless you've taken measures against it you'd have unity gain then and the circuit would most certainly oscillate.

 

[madtecchy]

In my case leaving the input open circuit is not an issue as it is hard wired to the dac. The dac is buffered after the volume pot so the offset stays the same with diffrent volume settings too. The dac is not a diy unit.. i have had the lid off and done a little looking about i couldnt resist... KSTR you have raised some very good points and also given the solutions .. If i where to build a standalone amplifier i would implement them.

 

[KSTR]

That's simply because I'm in the (slow) process of designing an inverting nested/composite chipamp where stability at all working conditions is the paramount issue.