Regulated B+ Solutions?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
BluegrassDan said:
The thread has several revisions. Is Rev 4 the final (and merged with the earlier schematics showing the bridge rectifier and caps)?


That message has the definite version, yes. It's not tied to a bridge rectifier which is why I didn't draw it, but yes just connect the dots to the earlier ones. Change ZD1 (or chain several in series) to achieve desired voltage, and drop it further 5-10VDC with the VR1 trimmer to get perfect noise floor.
 
Ian,  is there any secret to selecting the resistor values for the LR8 regulator circuit?
I'm experimenting with making a regulated supply for a valve mic 105 volts, and have a CRCRC pre smoothing
before the LR8 and then 10uF output cap. Current draw att output is about 4mA including LR8 setting resistors.

Unfortunately the LR8 output has more noise than its input!  In is about 2mV of noise on an unstable 137volts dc (i.e. it bounces about 10mV). Output is a stable 105 volts but noise is now at least 5mV and very spikey in nature!

I had a zener/transistor based regulator before, and that was clean. Just thought I'd try the LR8 as I could then adjust the voltage exactly.
 
Walrus said:
Ian,  is there any secret to selecting the resistor values for the LR8 regulator circuit?
I'm experimenting with making a regulated supply for a valve mic 105 volts, and have a CRCRC pre smoothing
before the LR8 and then 10uF output cap. Current draw att output is about 4mA including LR8 setting resistors.

Unfortunately the LR8 output has more noise than its input!  In is about 2mV of noise on an unstable 137volts dc (i.e. it bounces about 10mV). Output is a stable 105 volts but noise is now at least 5mV and very spikey in nature!

I had a zener/transistor based regulator before, and that was clean. Just thought I'd try the LR8 as I could then adjust the voltage exactly.

Hi Kevin, I don't know of any secrets in resistor value selection. However, for noise reduction, the example circuit given in The Art of Electronics recommends adding a capacitor from ADJ to ground (and a protection diode from ADJ to OUT). They do not recommend a value for the capacitor but the impedance will be quite high so a few uF should be enough..
Cheers

Ian
 
Thanks Ian, I'll try a cap on there tomorrow.
It's just wierd how it is introducing noise that wasn't there before!
 
Walrus said:
Thanks Ian, I'll try a cap on there tomorrow.
It's just wierd how it is introducing noise that wasn't there before!
What I think is happening is that any noise picked up on the ADJ pin is amplified by the error amp and appears at the output.

Many regulator chips are not designed to give the sort of very low noise we like to have on audio power rails so sometimes they need a helping hand.

Cheers

IAn

Cheers

Ian
 
Well, putting a 2.2uF cap on the adjust terminal improved things, but it still had noise on the output.
So I went back to a standard Zener Transistor setup and that was much better.
All this is being done using a scope to monitor the noise, so I'll go with the Zener version on the basis that if it looks better,
it should be better....  :)
 
Yes, less noise and it was a lot smoother looking too.
But the Zener/Transistor cct better.

It's interesting that on the LR8 data sheet, they show it dropping  over 200 volts (can't remember exact number) down to 5 volts.
Maybe it's not designed to be really low noise??
 
Walrus said:
Yes, less noise and it was a lot smoother looking too.
But the Zener/Transistor cct better.

It's interesting that on the LR8 data sheet, they show it dropping  over 200 volts (can't remember exact number) down to 5 volts.
Maybe it's not designed to be really low noise??

That could be. The data sheet talks about it being used for industrial controls, battery chargers and as a start up supply in a SMPS none of which require low noise..

Cheers

Ian
 
buildafriend said:
The Orange 86 Zener TIP50 power supply schematic has worked great for me

Are TIP50/zeners used as cap multipliers? We work with very different currents, so there is  no same solution for everyone.
I remeber having ripple problems with LR8 no matter how it was used. That was 1ch of 250V at 10-20mA max.
 
buildafriend said:
The Orange 86 Zener TIP50 power supply schematic has worked great for me
Can you post a schematic of that? Google brings up all kinds of stuff. Not sure which answer is the one you refer to.

Cheers

Ian
 
Just finished building Kingston's MOSFET B+ design. The proof is in the pudding.

Preamp at full gain. 150 Ohm input load resistance. Duplicate external PSUs, except one is the MOSFET regulator and the other a traditional RCRCRC. Heaters are elevated identically in both designs.

Analyzers are frozen for 60 seconds. 3dB/octave tilt. Not scientific by any means, just an initial A/B.

Unless someone can convince me otherwise, I gotta say the MOSFET seems to be the winner so far.

Pic attached.

 

Attachments

  • DBTube,-MOSFET-vs-Traditional.jpg
    DBTube,-MOSFET-vs-Traditional.jpg
    364 KB · Views: 74
Interesting comparison. It seems to have eliminated the very low frequency mains voltage variations that the CRCRCRC circuit cannot. Your timing is excellent because I am just about to try a CRCRCRC followed by a MOSFET stabiliser for a 200mA plus HT mixer supply.

Cheers

Ian
 
ruffrecords said:
Can you post a schematic of that? Google brings up all kinds of stuff. Not sure which answer is the one you refer to.

It's a clumsy voltage stabilizer with the high voltage zener reference directly at a pass transistor base. It dumps all the zener noise to B+ line but works fine if there's an additional RC stage after it.

I have used it plenty before this more elegant MOSFET circuit that we have crowd sourced here. Example PSU here with better suited transistor than TIP50: https://groupdiy.com/index.php?topic=41509.0
 
Kingston said:
It's a clumsy voltage stabilizer with the high voltage zener reference directly at a pass transistor base. It dumps all the zener noise to B+ line but works fine if there's an additional RC stage after it.

OK, got it. MOSFET with noise reduced zeners is definitely a better bet.

Cheers

Ian
 
There are some great and cheap pcbs on eby for this ..  search for 'kubota regulator' - they have a schematic which is very straight forward but for the small amount of increased complexity,  well outperforms my previous favourite - the 'Orange 86 hv psu'  :)

I've built and used several  - they are solid, even when used as a 'bench hv' where there is frequently much 'fussing about' connecting/disconnecting, power up/down cycles  ... 

It has a lot flexibilty  without being too fancy  ..  and remains easy to repair if anything does go wrong.

I've been using mine for reg outputs of+375V at 250mA with no probs at all - I fit 2 of these modules in a 1RU rack with 'transformers and caps  on top'  ..  no issues with heat or anything else.

Regulation is ridiculously good with the high-speed-discrete-error amp of the 'kubota' module - it uses a constant current zener adjustable voltage reference and a trimmer/potentiometer to adjust for the desired voltage. 

From memory,  it gave me a hum floor better than  -120 dBu in REW    ...  meaning you can approach 130dB of freedom from hum even in a decent powered speaker amp context  :)

 

Latest posts

Back
Top