advertisement


regulator booster

mikesnowdon said:
From the Flea manual:



So to implement that into the superreg we have this:

superegoutputfilter.png


Im going to try it out as its easy enough to add it, although space is a tight.
Click to expand...

Mike - as Dan says, you really cannot use this simple LPF on the output of a reg. Think about Ohms law - once the reg is supplying just a few mA - that resistor will drop ALL the available voltage - leaving you with nothing at the output. It's use in the Flea was to filter a voltage reference - a different thing altogether from a circuit designed to supply CURRENT. You could in theory use a low value resistor and up the capacitor value to match - but it would then be huge - and the voltage output will always vary as the current demand changes! (ohms law again) - kind of negating the whole point of a VOLTAGE REGULATOR. Sorry, but the whole idea is not a particularly elegant thing to do on the output of a low voltage reg.

I would stick to just the vbe before the reg.
 
I have a lot of BC548s. Having looked at the datasheet it seems that these have the same noise figures as the 547, but have a lower voltage rating, which if being used at 5 or 12v should not be a problem?
 
Puffin said:
I have a lot of BC548s. Having looked at the datasheet it seems that these have the same noise figures as the 547, but have a lower voltage rating, which if being used at 5 or 12v should not be a problem?
Click to expand...

They're gonna see the full raw DC voltage of 21v as they sit in front of the 7805. Check if they can cope with that. If so then they should be fine I think.
 
Hi all.

Sorry about last nights confusion :D I'm still quite embarrassed about that.

Earlier Teddy suggested using the VBE on the output would be better as the WM8740 DAC's in my CA640Cv2 have quite poor noise rejection. My only consideration on this was that I thought the DAC's needed to be on the same supply voltage as the master clock = 5V. However, Matt Farley managed to upgrade the DAC's in his 640 to WM8741 which require 3.3v on the digital supply rails. He simply changed the regs to 3.3v items, and it worked fine. So my considerations about needing 5v for all clock/dac rails are obviously wrong. Now if using the VBE on the output of my 7805 regs I'll be getting around 4v or thereabouts. The WM8740 is fine on 3.3 to 5v supply so it should work fine, but....

In the datasheet there is a note in big letters:

"DVDD must be equal to of less than the AVDD supply"

So, with the VBE on the output am I guaranteed to see the same voltage drop on all 4 regs, or will there likely be some variation? If not how much variation is likely to cause potential problem's?

Another thing. Below is a circuit for having the VBE on the reg output. Is it correct and do I really need the input R/C filter? It would be handy if I dont as the reg can be made more compact then.

vbeversion2.png


Thanks in advance :)
 
Hi Mike, the new circuit has a couple of problems. First, do you understand why your circuit in post 101 won't work? If so, apply the same reasoning to your new circuit.
 
Hey Dan.

I got that earlier. I just assumed that when Teddy sugessted having the VBE on the reg output (as he has done himself), that it was a simple case of swapping the circuit across from the input to the output. I assumed this would require the 10k resistor for the VBE to function properly.

How would you suggest it be done?

EDIT: Regarding the input R/C filter. This could be an issue for similar reasons as 'herewegoagain' refers to in a previous post. Now, my understnading is that this res will see alot more voltage (and available current) so may not be as much of a problem as having it on the output. However, with the VBE sweeping up all the output noise its probably just not necessary. If this is the case why not just say so in the first place. Im just trying to understand.
 
Hi Mike,

Are you familiar with ohms law? V=IR and it's derivatives? If so you can use it to calculate how much voltage will be dropped across a resistor. You know the resistor value. Think about how much current the load will require .... an opamp might need 5mA, a DAC chip maybe 100mA. This info will be in their datasheets.

A Vbe across the output of the regulator will work, but there are some issues to work out. First how much current will be dropped by the Vbe? Obviously a DAC chip needs a carefully defined voltage, so this is important. Also, a Vbe will have much higher impedance than a regulator so will be less good at regulating a fluctuating load. I wouldn't use one except for loads requiring a constant current.

Finally, the load current will go through both the regulator and the BC547C transistor. The reg is specified for 1.5A, but what about the BC?

I admire your willingness to learn and experiment, but I'm not sure how much knowledge you have, so I'm not sure how to pitch my answers. If anything that I've said doesn't make sense, ask! I'm not an expert - many years ago I did a BTEC in electronics, but I'm not half as clued up as guys like Piglets Dad or Greg Ball. Software is my day job these days. Anyway, I'm happy to pass on whatever I remember.
 
Cheers Dan.

Much earlier in the thread Teddy and Martin (I think) discussed this DAC and the merits of having the VBE before or after the reg. I think it was Teddys view point that any impedance issues caused by having the VBE on the output would be outweighed by the improved noise rejection. I think this conclusion was made based on the relatively poor PSSR of the WM8740 dac chip.

To be honest what they and others were discussing went over my head somewhat so I opted for the simplest (from my viewpoint) option of using the VBE in the input side of the reg. Having successfully done this I'm keen to learn if it can be further improved while maintaining a nice easy to assemble and compact stripboard approach, hence my questions on R/C filtering.

Maybe you could go over the whole thread and interpret what the more technically mined members were discussing, and relay it to me in a more newbie friendly terms?

PS: Need to look up ohms law....


PPS:I really do appreciate the help I get from those who are patient enough with me to offer it. I know I can be a pain sometimes and probably ask too many questions, getting all in a spin over simple things. Im quite keen to learn more. I should have gone to college/uni when I was young enough and free from familial responsibilities but alas there were temptations that led me astray. Now that Im old(er) and more boring I can pursue my interest in electronics through hifi bodging :D

To all of you who have leant me a hand since I joined this site, thank you :)
 
mikesnowdon said:
PPS:I really do appreciate the help I get from those who are patient enough with me to offer it. I know I can be a pain sometimes and probably ask too many questions, getting all in a spin over simple things. Im quite keen to learn more. I should have gone to college/uni when I was young enough and free from familial responsibilities but alas there were temptations that led me astray. Now that Im old(er) and more boring I can pursue my interest in electronics through hifi bodging :D

Click to expand...


Blimey I could have written that myself.
 
Mike, the regulator with Vbe on the output will work for opamps. Try it to see if you like it better than the configuration you have now. As for the DAC, is it the 8740 or 8741?
 
Hey Dan

Glad I gave you a laugh :D Though it was a genuine ask for a little help though ;)

I already have some different regs on the opamps (S-power). Im using the VBE'd 7805 diy regs on the DAC's which are WM8740. I have some concerns on using the VBE on the output which I expressed in a previous post....

Whats your take on this mate?

myself..... said:
Earlier Teddy suggested using the VBE on the output would be better as the WM8740 DAC's in my CA640Cv2 have quite poor noise rejection. My only consideration on this was that I thought the DAC's needed to be on the same supply voltage as the master clock = 5V. However, Matt Farley managed to upgrade the DAC's in his 640 to WM8741 which require 3.3v on the digital supply rails. He simply changed the regs to 3.3v items, and it worked fine. So my considerations about needing 5v for all clock/dac rails are obviously wrong. Now if using the VBE on the output of my 7805 regs I'll be getting around 4v or thereabouts (due to the ca.1V dropout across the VBE). The WM8740 is fine on 3.3 to 5v supply so it should work fine, but....

In the datasheet there is a note in big letters:

"DVDD must be equal to or less than the AVDD supply"

So, with the VBE on the output am I guaranteed to see the same voltage drop on all 4 regs, or will there likely be some variation? If not how much variation is likely to cause potential problem's?
Click to expand...
 
Mike. They are all positive regs. I left it running for a few hours before critical listening. I have to say I think it is an improvement. Low level detail seems better.

Good bang for the buck.

Rob.
 
Puffin said:
Mike. They are all positive regs. I left it running for a few hours before critical listening. I have to say I think it is an improvement. Low level detail seems better.

Good bang for the buck.

Rob.
Click to expand...

Nice one :)

How did you do the one that sit's on the big heatsink on the left hand side? Did you use a power transistor? If so is that power transistor mounted on the heatsink instead of the regulator? (cant quite tell from the pic)
 
You must log in or register to reply here.


advertisement


Back
Top