There are a few zero feedback buffer stages on the web and there is, of course, the excellent Nat. Semi buffer......but where are the zero feedback transistor gain stages? Some very high end companies use them....like Dartzeel. I find zero feedback the most natural sounding. I love the Nat. buffer, but I need gain. I do not want to use a transformer. Seems to me it would be easy for one of you design guys to come up with a symmetrical (for super low distortion and no coupling caps) simple one or two stage circuit that would have maybe the first stage as gain and the second as push pull followers.....or maybe just a gain stage that would have high output impedance that could be used with a symmetrical buffer circuit on the output. I think the sonic results out of such a simple circuit would be amazing and help me and the DIY community mucho. Thanks.
Here is my listening history for reference: I have listened to lots of opamps over the years including the latest Nat devices. All these high feedback opamps do not sound as good as the AD815 run in mono mode (just using one half of two devices). However, the National buffer run in highest bias mode sounds even better. However, it has no gain......hence this question. How to get the pure sound of a zero feedback buffer and have some gain too.
Here is my listening history for reference: I have listened to lots of opamps over the years including the latest Nat devices. All these high feedback opamps do not sound as good as the AD815 run in mono mode (just using one half of two devices). However, the National buffer run in highest bias mode sounds even better. However, it has no gain......hence this question. How to get the pure sound of a zero feedback buffer and have some gain too.
Last edited:
What do you mean by zero feedback? A BJT develops 1% second-order distortion from a 1mV signal at the base. If you want a buffer to pass signals greater than this with low distortion then you either need feedback or very stable distortion cancellation. You mention followers - you do realise that a follower works by using 100% feedback?
Please, look at Voltage Amplifier & Current Buffer DIY. You can use the voltage amplifier, and the gain is set by one resistor. The drawback? It uses 2SJ74 JFET's. But otherwise, it is very good...
To find the most optimal resistor values (R18 and R25 in Figure 4 in the article), it is necessary to know the max input (and output) voltage. If you are used to amplifier tweaking, it is no problem to find the values giving you your desired amplification. If you have further questions, just let me know.
Yes, no global feedback.
I would like to try this circuit on the output of a DAC (one volt RMS max) and also as a line stage with gain 3-6X.
Your circuit is running off of plus and minus 28V. Can one go down to plus and minus 12 and still have it work?
I am going to try this circuit but would love an even simpler circuit. Like the first stage somehow connected to the last follower stage.....removing the two stages in between (maybe not possible?)
I would like to try this circuit on the output of a DAC (one volt RMS max) and also as a line stage with gain 3-6X.
Your circuit is running off of plus and minus 28V. Can one go down to plus and minus 12 and still have it work?
I am going to try this circuit but would love an even simpler circuit. Like the first stage somehow connected to the last follower stage.....removing the two stages in between (maybe not possible?)
That amp only has two stages. The current mirrors are part of the first stage. The output stage is essentially a compound follower.
If you want a symmetrical amp then you get nearly twice as many components, although not necessarily twice the performance. If you want to avoid global feedback then you need either local feedback or distortion cancellation. The output buffer needs to be high quality as it has no extra feedback to linearise it.
If you want a simple circuit with no global feedback then you either need to drop the symmetric requirement, lower your performance aims, and/or perhaps use valves.
If you want a symmetrical amp then you get nearly twice as many components, although not necessarily twice the performance. If you want to avoid global feedback then you need either local feedback or distortion cancellation. The output buffer needs to be high quality as it has no extra feedback to linearise it.
If you want a simple circuit with no global feedback then you either need to drop the symmetric requirement, lower your performance aims, and/or perhaps use valves.
Yes, I should not have used the word "stages". You techies are so precise....he he.
I imagine, you could make a first "stage" without the current mirrors and maybe tie it to a single pair of followers. Or even make a single stage gain thang with no current mirrors and use a Nat Semi buffer on the output.....yes, I guess the performance (measured distortion) would be less.....but, has anyone done what I would like? All those transistors and parts are not my preference. I like super simple....but I also want the best sound possible...so I will experiment with this circuit and see (hear) what it does versus a Nat Buffer by itself.
Anyone else have a zero global feedback transistor low gain circuit? preferably symmetrical?
I imagine, you could make a first "stage" without the current mirrors and maybe tie it to a single pair of followers. Or even make a single stage gain thang with no current mirrors and use a Nat Semi buffer on the output.....yes, I guess the performance (measured distortion) would be less.....but, has anyone done what I would like? All those transistors and parts are not my preference. I like super simple....but I also want the best sound possible...so I will experiment with this circuit and see (hear) what it does versus a Nat Buffer by itself.
Anyone else have a zero global feedback transistor low gain circuit? preferably symmetrical?
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.