Nuuk said:
Celestion 7's (the original ones not MKII, if that matters) at the moment. They're 8ohm "nominal" is all I know. Tried looking for proper specs on the drivers, but they're unlabelled so.... I couldn't find anything. I am guessing they would work OK on an unregulated supply though...
If it wouldn't work, I wouldn't encourage you (or anybody else) to go ahead.
So yes, give it a try. Get the amp working and then try a pre amp. At least you will have something to listen to while you work on those regulators. Depending on the synery with your speakers, you may not bother with the regulators!
So yes, give it a try. Get the amp working and then try a pre amp. At least you will have something to listen to while you work on those regulators. Depending on the synery with your speakers, you may not bother with the regulators!
There is a problem with the schematic in this post.
Let's break it up into parts so it's easier to see what's going on. Look at the positive rail regulator circuits first. The regulators share the same input voltage and return nodes but their output voltages may be independent of each other. This is important because the positive linear regulators have an NPN pass transistor inside that is used to regulate the voltage at the output.
Now, look at the negative rail regulator circuits. They also share the same input voltage and return nodes but they also share the same output node because their output nodes are connected to return nodes of the positive rail regulators. Any difference in output voltage between the two negative rail regulator circuits will cause currents to flow and make the lower side regulators dissipate a lot of heat (what you saw markie).
To make this split rail positive regulator circuit work you need 4 bridge rectifiers (2 on each secondary) to isolate the "virtual" grounds of the two regulator circuits.
Refer to the attached schematic if you prefer a visual explanation.
Cliff notes: Notice how the inputs, returns, and outputs of both negative rail regulators are tied together. That isn't kosher.
Let's break it up into parts so it's easier to see what's going on. Look at the positive rail regulator circuits first. The regulators share the same input voltage and return nodes but their output voltages may be independent of each other. This is important because the positive linear regulators have an NPN pass transistor inside that is used to regulate the voltage at the output.
Now, look at the negative rail regulator circuits. They also share the same input voltage and return nodes but they also share the same output node because their output nodes are connected to return nodes of the positive rail regulators. Any difference in output voltage between the two negative rail regulator circuits will cause currents to flow and make the lower side regulators dissipate a lot of heat (what you saw markie).
To make this split rail positive regulator circuit work you need 4 bridge rectifiers (2 on each secondary) to isolate the "virtual" grounds of the two regulator circuits.
Refer to the attached schematic if you prefer a visual explanation.
Cliff notes: Notice how the inputs, returns, and outputs of both negative rail regulators are tied together. That isn't kosher.
Attachments
Nuuk said:If it wouldn't work, I wouldn't encourage you (or anybody else) to go ahead.
So yes, give it a try. Get the amp working and then try a pre amp. At least you will have something to listen to while you work on those regulators. Depending on the synery with your speakers, you may not bother with the regulators!
Ok cool, cheers. I'm back listening on the old Cambridge Audio A5 while I mess about. It certainly seems to sound quite different...
I think I will move the preamp in to a separate case after all, after a bit of thought. Seems like a better / easier idea really, as I can complete the power amp and even use it while building the preamp.
BWRX said:
BWRX, good question. No, as it goes there wasn't. I was too scared of blowing up opamps given the simulation results. Should I test with 'load resistors', one between +ve and ground and one between -ve and ground presumably? I simulated this and it was, unfortunately, no different.
BWRX said:
BWRX, thank you so much for looking at things and taking the time to debug and explain this for me! It seems like you can see what I see, but actually properly understand it and explain it a lot better than myself.
This certainly explains why one channel was getting so much hotter.
Presumably then, if I were to use three regulators (edit: three regulated split supplies i mean...) off of one transformer, (one per channel for the power amp, and one for the preamp) I would just as easily be able to achieve this by using three bridge rectifiers on each secondary, right? If so, you've just made my day! I was totally bummed out that I couldn't get it to work properly!
BWRX said:
Truly fantastic
I can't thank you enough.I will see if I have enough parts to try this out tomorrow.
If I understand correctly, in doing this, I presume that because the virtual grounds after the regulators are... well.... virtual, and 'floating' (I think that's correct?) they shouldn't be connected together to safety earth? This would put me right back where I started I think and defeat the whole purpose of isolating them, right?
markiemrboo said:
Since both grounds are now floating they could be connected together without any problems. I wouldn't directly connect them together to each other or to safety ground. There are other recommended ways of doing it with diodes, a resistor, and a cap. I believe Rod Elliot has an explanation of it on his ESP web pages.
If the source you will be using is single ended and has a common ground for both channels you will in fact be connecting the grounds of both power supplies through the source when all is said and done.
This circuit is deceptively simple but you have to know what's going on inside the black boxes (the regulator ICs) in order to see what the issue is. I admit I had to let it stew in my brain for a little while before realizing what the problem was.
BWRX said:
Yeah, I've made a 'disconnecting network' as on his site. I was just wondering if I could still connect the "0v" outputs of the supplies to that still. Seems I can then, but I think I will have to sleep on that one
I have only just learnt the basics of 'the transistor', bias and all that is still a tad over my head, but I managed to make a working fan controller for my computer and actually understand it. So... I probably didn't really stand much of a chance if you need to know what's going on inside the regs 
Thanks again though. I am gonna get some sleep now. Good night all.
Nuuk said:Brian, thanks for sorting out this issue and explaining it so well! I have now edited the options diagram on Decibel Dungeon, and added a note (that includes a link to your post).
Just got another bridge rectifier to do this today. It still doesn't seem to work as expected. I have tried with four bridge rectifiers initially, two for one regulated power amp supply and the other two for the regulated preamp supply.
This time, the power amp negative rail is at -12.8v and positive at the expected 28v. The preamp supply is correct at +-13.8v. Rather than the power amp supply being OK and the preamp negative rail being that of the power amp negative rail.
When I disconnect the preamp supply the power amp supply negative rail measures the expected -28v.
Thoughts?
To clarify, you have four positive regulators wired up as shown in this schematic and RAIL1 is supposed to be +/-28V while RAIL2 is supposed to be +/-14V, but RAIL1 is +28V and -12.8V and RAIL2 is +/-14V. Correct?
I thought isolating the input voltages by giving each regulator its own bridge rectifier would fix the problem but apparently there is still interaction between the negative rail regulators. The problem still lies with the negative rail regulators trying to regulate two different voltages while sharing a non galvanically isolated input.
I thought isolating the input voltages by giving each regulator its own bridge rectifier would fix the problem but apparently there is still interaction between the negative rail regulators. The problem still lies with the negative rail regulators trying to regulate two different voltages while sharing a non galvanically isolated input.
Ok. This requires some more brain power. I'll think it over some more tonight after work.
One thing to try would involve putting a resistor in between the bridge rectifier and the input of the negative rail preamp regulator circuit. The value of the resistor would depend on the current draw of your preamp but a good value would probably be in the neighborhood of 400-1kohm with at least a 1/4W power rating.
One thing to try would involve putting a resistor in between the bridge rectifier and the input of the negative rail preamp regulator circuit. The value of the resistor would depend on the current draw of your preamp but a good value would probably be in the neighborhood of 400-1kohm with at least a 1/4W power rating.
BWRX said:
Thanks very much for the continued help, really appreciate it!
Thinking about it, it does seem slightly different than before. The negative rail isn't identical to the negative rail of the other supply, it's about 1v lower!
e.g (just to clarify)
before, I had:
+28v, -28v
+14v, -28v
now I have:
+28v, -12.8v
+14v, -14v
Look forward to hearing from you again soon, and enjoy the rest of your day at work
Just thought I would post with some info, which may or may not help.
I figured exact voltages would help. So rather than try and remember them, I wrote down the outputs.
It was actually:
Power amp: +28.5v, -12.6v
Pre amp: +13.6v, -13.6v
There's a mysterious 1v difference here.
Also, I measured the output from the caps for the power amp supply, still with the regulator (+28.5v and -12.6v) connected, and I am still actually seeing an ordinary / normal +36v from both.
Just as a reminder, when I disconnect the preamp regulator the power amp supply works as normal (+-28.5v).
The outputs are *not* connected to any kind of load at all, and I am not taking the 0v outputs to safety earth. Quite frankly, until I am getting the correct output voltages I am a bit scared to! Perhaps I should be? But I think it should be measuring correct without doing this anyway..?
Maybe something here is significant somehow, I don't really know!
I figured exact voltages would help. So rather than try and remember them, I wrote down the outputs.
It was actually:
Power amp: +28.5v, -12.6v
Pre amp: +13.6v, -13.6v
There's a mysterious 1v difference here.
Also, I measured the output from the caps for the power amp supply, still with the regulator (+28.5v and -12.6v) connected, and I am still actually seeing an ordinary / normal +36v from both.
Just as a reminder, when I disconnect the preamp regulator the power amp supply works as normal (+-28.5v).
The outputs are *not* connected to any kind of load at all, and I am not taking the 0v outputs to safety earth. Quite frankly, until I am getting the correct output voltages I am a bit scared to! Perhaps I should be? But I think it should be measuring correct without doing this anyway..?
Maybe something here is significant somehow, I don't really know!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.