Anyone know if the TLE-2426 rail splitter would have any problems used as a "negative" virtual ground?
In this context, I'm referring to an emitter follower output stage using PNP devices powered from a single -24 volt supply.
The 2426 would have its "common" pin tied to circuit "-24V" and its "input" pin to circuit ground. So from the 2426's perspective, it would be no different than if it were connected between a +24 volt rail and ground, but from the follower's perspective, it would see a negative voltage at the 2426's output.
I'm wondering if there would be any noise or stability issues using it this way. I've already tried TI's tech support, but the guy said the 2425 is designed for +4 to +40 volts and didn't seem to understand that from the 2425's perspective it would be seeing a positive voltage so I gave up.
Thanks.
se
In this context, I'm referring to an emitter follower output stage using PNP devices powered from a single -24 volt supply.
The 2426 would have its "common" pin tied to circuit "-24V" and its "input" pin to circuit ground. So from the 2426's perspective, it would be no different than if it were connected between a +24 volt rail and ground, but from the follower's perspective, it would see a negative voltage at the 2426's output.
I'm wondering if there would be any noise or stability issues using it this way. I've already tried TI's tech support, but the guy said the 2425 is designed for +4 to +40 volts and didn't seem to understand that from the 2425's perspective it would be seeing a positive voltage so I gave up.
Thanks.
se
I think you've made your own perfectly good argument. The part only "knows" about the voltage difference between the pins. The whole thing could be sitting at 10kV or minus 50kV and it would work just the same. As long as you adhere to the AMLs on the data sheet, and the load impedance is typical, it should work. I've used the rail splitters in the past with battery supplies and found them very convenient. My only reservation is that they're a somewhat specialized part and I worry about repairability in the future if somebody can't get one.
I agree with Conrad, its rather a specialist part. If current consumption isn't much of an issue then I'd be more inclined to use two TL431s as shunt regs for this. Cheaper but admittedly more components.
The only issue I can see is about the 'noise reduction' cap. If you're using it then presumably you'll connect it to ground which is the positive rail in your case. Otherwise can't forsee any probs. TI's tech support not up to much these days is it?
The only issue I can see is about the 'noise reduction' cap. If you're using it then presumably you'll connect it to ground which is the positive rail in your case. Otherwise can't forsee any probs. TI's tech support not up to much these days is it?
Yeah, I thought my argument was fairly sound as well. But I've made what seemed like fairly sound arguments in the past only to later discover I overlooked something rather obvious.
So I just thought I'd put this out to the public for a bit more feedback before I commit to using PNP devices instead of NPN's.
As far as repairability, they're not so expensive that I couldn't buy more than enough to handle any replacements. And I'm pretty sure you'd be able to find these surplus for some years to come.
Thanks, Conrad!
se
An externally hosted image should be here but it was not working when we last tested it.
So I just thought I'd put this out to the public for a bit more feedback before I commit to using PNP devices instead of NPN's.
As far as repairability, they're not so expensive that I couldn't buy more than enough to handle any replacements. And I'm pretty sure you'd be able to find these surplus for some years to come.
Thanks, Conrad!
se
Given the growing popularity of battery powered devices, I don't know that it's a terribly specialist part.
I'll be using the TO-92 cased part which doesn't include the noise reduction pin.
The actual circuit is fully balanced so even if there were a bit of noise on the output, it would be common-mode.
It would seem not. I seemed to have been speaking in rather plain, simple English.
se
I meant specialist in the sense that its a single sourced part. Also its not very generic - looks just like a fairly low-power opamp with two resistors on the input. So if I'm thinking about number of SKUs (and that is a concern), I wouldn't stock it myself. Admittedly its really compact, that does seem to be its USP.
I consider the noise reduction pin a bit of a joke. Its a rail splitter - if you use it then it stops splitting the rails except at very LF.
I consider the noise reduction pin a bit of a joke. Its a rail splitter - if you use it then it stops splitting the rails except at very LF.
That's true. Not aware of anyone else making single part rail splitters like this. But I'll just buy more than I ever intend to use for production.
Yeah, but splitting the rail at DC is really the only thing that matters for typical applications of this device.
se
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.