So, I'm repairing a White 4001 EQ and I've come across something I've never seen before. May not shock anyone else.....
In the PSU of this EQ, the designer used TWO positive rails with positive 18v regulators to realize his +18 and -18 Rails. These halves of the circuit are identical, save the configuration of the LM342p-18 used to provide the negative rail. The pins are simply configured differently, which apparently causes the regulator to invert it's +35v input and regulate it to -18. I've never seen anything like it, and what's more, I have no idea why this was his approach. The 342 datasheet doesn't mention anything about any such configuration.
Needless to say, the negative rail is where my problem lies and I can't find an LM342 of ANY variety anywhere. I've checked with White Instruments, and apparently they don't have any. I've entertained building a sub-board using 7818/7918 regulators, but seriously, I'd rather just find an LM342.
Anybody have a solution?
In the PSU of this EQ, the designer used TWO positive rails with positive 18v regulators to realize his +18 and -18 Rails. These halves of the circuit are identical, save the configuration of the LM342p-18 used to provide the negative rail. The pins are simply configured differently, which apparently causes the regulator to invert it's +35v input and regulate it to -18. I've never seen anything like it, and what's more, I have no idea why this was his approach. The 342 datasheet doesn't mention anything about any such configuration.
Needless to say, the negative rail is where my problem lies and I can't find an LM342 of ANY variety anywhere. I've checked with White Instruments, and apparently they don't have any. I've entertained building a sub-board using 7818/7918 regulators, but seriously, I'd rather just find an LM342.
Anybody have a solution?
regulator or opamp
Do you have a schematic, Are you sure that the LM324 is a voltage regulator? Is it possibly an op amp?
Edit ~ nevermind I misread what you wrote 342 not 324.
Do you have a schematic, Are you sure that the LM324 is a voltage regulator? Is it possibly an op amp?
Edit ~ nevermind I misread what you wrote 342 not 324.
Sounds like a trick calrec used to use, two positive regulators, one at +17.5, one at +35, from a floating transformer/rectifier/cap, ground the +17.5V rail and you have +- 17.5V for the rest of the circuit.
They did it with 2N3055 and transistors, but you can do it with almost any regulator.
Regards, Dan.
They did it with 2N3055 and transistors, but you can do it with almost any regulator.
Regards, Dan.
s'ok. It's definitely the oddest configuration I've ever seen. I'll have to scan the schematic to get an image.
I can't find a National LM342 in either the 1980 or 1995 databooks, nor on the website, but I can find references to it on the web. It does sound like just a fixed output regulator.GLORY_MORRIS said:
We'll need that Schematic.
Here's a link to the datasheet:
http://www.alldatasheet.com/datasheet-pdf/pdf/8857/NSC/LM342.html
Here's a blow-up of the PSU:
http://www.alldatasheet.com/datasheet-pdf/pdf/8857/NSC/LM342.html
Here's a blow-up of the PSU:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
That's what dmills suggested it was. 2 independent +ve supplies, connected in series.
Yup, 7818 should work. LM340 (18V version) too.
Yup, 7818 should work. LM340 (18V version) too.
Thanks fellas. I'll get right on it!
If you feel like it, I'd love an explanation on why this works.
If you feel like it, I'd love an explanation on why this works.
The ps is using a virtual ground. The ground from the output is not connected to the earth ground. If you measure the +18v output with reference to an earth ground it will actuly measure +36 volts. If you do the same from your ground point with reference to earth you should measure +18 volts. and from The -18v to earth will actuly measure 0 volts.
Picture a string of 9V batteries all wired in series. say 10 of them but it could be any number more then 4.
Pick a spot in the center of the string or close to it. Now go up 2 batteries and down 2 batteries. the center point is your common and up two batteries is your +18V and down two batteries from that center point is your -18V
It doesn't matter where in that string you pick the center point as long as it is at least 2 batteries away from the end point.
you will always have +18V two batts up and -18V 2 batts down from that centerpoint.
The power supply you have shown works the exact same way only is regulated. The potential of voltage is always the same at the 3 points in relation to each other.
Pick a spot in the center of the string or close to it. Now go up 2 batteries and down 2 batteries. the center point is your common and up two batteries is your +18V and down two batteries from that center point is your -18V
It doesn't matter where in that string you pick the center point as long as it is at least 2 batteries away from the end point.
you will always have +18V two batts up and -18V 2 batts down from that centerpoint.
The power supply you have shown works the exact same way only is regulated. The potential of voltage is always the same at the 3 points in relation to each other.
I have definitely seen a app note somewhere with this configuration. I think it dates to a period when negative regulators were non-existent or at least expensive / hard to find. It's not so odd looking if you just keep in mind that voltages are relative. The part I'm unclear on is whether common in the schematic has to float relative to the mains earth potential.
Actually, it doesn't seem that odd now. Maybe in the context of modern designs. Anyway, the 7818 did the trick and this White EQ is up and running again.
Thanks for the explanation and the help!
-Morris
Thanks for the explanation and the help!
-Morris
No, it won't (necessarily) because earth ground isn't connected to the output side of the circuit.ph_christensen said:
Put it this way, you can bond the output common to earth without any problem (if that fits your application).
This works because everything between the transformers and the input side of the regulators is floating.
The only problem with this architecture is a possible slight increase in noise due to capacitive coupling via the transformer windings in the absence of an electrostatic screen, but in practise it seems to work fine.
I actually quite like a variation on this trick for larger power supplies as it can be used to produce a supply where the cases of the pass transistors are at ground potential, which simplifies mounting and improves thermal resistance numbers.
Consider a positive output supply using say a '3055 where the regulator is in the negative return, now the collector (can) is at case ground and just needs a little heatisink paste (no mica washers or the like).
Regards, Dan.
This works because everything between the transformers and the input side of the regulators is floating.
The only problem with this architecture is a possible slight increase in noise due to capacitive coupling via the transformer windings in the absence of an electrostatic screen, but in practise it seems to work fine.
I actually quite like a variation on this trick for larger power supplies as it can be used to produce a supply where the cases of the pass transistors are at ground potential, which simplifies mounting and improves thermal resistance numbers.
Consider a positive output supply using say a '3055 where the regulator is in the negative return, now the collector (can) is at case ground and just needs a little heatisink paste (no mica washers or the like).
Regards, Dan.
- Status
- Not open for further replies.