It is a way to get a cmos switch more linear. Not perfectly linear. It would be a possible avenue if you are OK with just switching between a few values (instead of continuously adjustable), but it would have to be incorporated in the original circuit board design - you don't want to have feedback signals doing wire runs (otherwise known as antennas) to some off-board component.
So that was my point - not that there aren't circuit solutions that would be good enough (and there probably would be endless debates about what is "good enough"), but that the noise sensitivity of the feedback circuit makes retrofitting anything pretty hard.
hypex ncore
Look at the datasheet. Noise doesn't matter anyways because it's not in the signal path if used like in figure 6.
Look at the datasheet. Noise doesn't matter anyways because it's not in the signal path if used like in figure 6.
So I assume we are back talking about the TI cmos switch, not the analog potentiometer? Can you please tell me what the datasheet says about noise and non-linearity?
The circuit in figure 6 has the cmos switch in the feedback path, where it is even more sensitive to noise than in the signal path.
The AD digital POT uses CMOS switches.
The datasheet I sent the link for has a comprehensive list of specs.
This is how it was explained to me using this technique:
"The pot is not in the audio path. its in the feedback loop of the opamp. If the pot was to cause some issue with the signal, the opamp is going to correct for it because it is IN the feedback loop. The myth of pots being bad for audio is because it is a single point contact on cermet or carbon film. If one passes to much CURRENT thru this single point contact on the wiper, both the contact and the resistive element heats at the contact point and the resistance changes locally. this can be heard and measured easily with the right gear. In our case, the feedback current is very small due the the resistors in the loop so there is no chance of this happening"
I was just hoping that as you clearly have studied them, you could let us know without each of us having to read the data sheet.
Ah, my bad. I was assuming you meant using a figure 6-like circuit in the feedback loop of the ncore, but I now realize you are talking about adding an extra opamp.
There is also the issue of the metal-cermet or metal-carbon contact point acting as a diode (remember the cat's whisker detectors in old crystal radios?).
Well I don't have personal experience with this technique. I was just told it's a very good way to control gain without putting anything in the audio signal path.
I'm thinking about trying it though.
Interesting you say that because I did investigate DIP-8 Adapters and avoided the tin versions, went for copper-Berylium versions
Don't know about the sound but I went for the least compromised option
Hi ti33er I am finding it hard to find a DIP8 Adapter that looks right can you point me in the direction of the copper-berylium version you mention
Thanks
It is not a bad way - but you will always put *something* in the signal or feedback path if you want to control gain. In this case an opamp.
If you do, please share your findings!
Which Hypex amps don't use opamps already anyways? It's the opamp used in the buffer you do this with. Works with discrete or IC.
hypex ncore
All it is, is an discrete opamp built into the board.
What's that say under features?
All it is, is an discrete opamp built into the board.
What's that say under features?
Last edited:
As I wrote, *I* would not call it an opamp as it, while being a differential, DC-coupled amp, doesn't satisfy the "high gain" criteria (op-amps typically have an open-loop gain of more than 100,000, approximating the ideal "infinite open loop gain" model), but if Hypex chooses to call it one, I guess we should go for their terminology.
D0808-42 - HARWIN - SOCKET IC, WIREWRAP, 8WAY | Farnell element14
So this was the DIP-8 I acquired for my Tag McLaren AV32RDP
It's not the correct unit for HYPEX Input Buffer (SOIC-DIP8 converter)...need to look about, but Bavmike would know better here
So this was the DIP-8 I acquired for my Tag McLaren AV32RDP
It's not the correct unit for HYPEX Input Buffer (SOIC-DIP8 converter)...need to look about, but Bavmike would know better here
Any particular reason to go for a wire-wrap model? I would have assumed a solder pin model would have shorter (and thus less noise-prone) connections.
Morning, I now have a Burson V5 based input buffer board prototype PCB being made which I shall evaluate to see if there is a sonic improvement. It will have the same mechanical layout as the Hypex ones, so will be an easy upgrade. Will keep you posted as to the progress. The standard amps, as you have all guessed uses the Hypex buffer board identical to the Acoustic Imagery Atsah 500. Colin
- Status
- Not open for further replies.