Super Regulator

Should be no problem for the performance but depends on your dissipation, how you heatsink them.
What's your expected dissipation? Input/output difference x output current?

Jan

Thanks a lot for the quick response (and your amazing work)! Much appreciated.

Input is 15V, output is 5V at 29mA, so 290mW (per side).

I have four of them for a DAC powered by the AK4499EQ.

At that level of dissipation, any recommendation regarding the heatsink? Is it even needed? I have very, very tight space constraints to deal with.
 
Last edited:
Thanks a lot for the quick response (and your amazing work)! Much appreciated.

Input is 15V, output is 5V at 29mA, so 290mW (per side).

I have four of them for a DAC powered by the AK4499EQ.

At that level of dissipation, any recommendation regarding the heatsink? Is it even needed? I have very, very tight space constraints to deal with.

Pics of your 4499EQ DAC?
 
Pics of your 4499EQ DAC?

Not yet, but it's an open source project that you can follow on this thread (you might want to start with this post).

I must warn you though: it's a bit crazy...

I hope to have the PCB layouts for all boards (PSU, DAC, XLR) ready within a month or so. The whole DAC including these three boards will be very small (3" × 1.5" × 1.75"), which is why I want to use surface-mounted components only.

I am working against the AKM evaluation board that used to be available on DigiKey. The next batch is expected to be available again in November. They use a custom version of the super regulator based on the AD817A/AD (four of them). You can get the manual for the evaluation board with the schematics for the regulator by contacting AKM's sales department.
 
Last edited:
AX tech editor
Joined 2002
Paid Member
Thanks a lot for the quick response (and your amazing work)! Much appreciated.

Input is 15V, output is 5V at 29mA, so 290mW (per side).

I have four of them for a DAC powered by the AK4499EQ.

At that level of dissipation, any recommendation regarding the heatsink? Is it even needed? I have very, very tight space constraints to deal with.

You may not need heatsinking, but if you are using the Superreg board, you can replace the foreseen heatsink with a small piece of copper sheet, and solder the pass device to it. That's probably simplest. Just for peace of mind.

Jan
 
Not yet, but it's an open source project that you can follow on this thread (you might want to start with this post).

I must warn you though: it's a bit crazy...

I hope to have the PCB layouts for all boards (PSU, DAC, XLR) ready within a month or so. The whole DAC including these three boards will be very small (3" × 1.5" × 1.75"), which is why I want to use surface-mounted components only.

I am working against the AKM evaluation board that used to be available on DigiKey. The next batch is expected to be available again in November. They use a custom version of the super regulator based on the AD817A/AD (four of them). You can get the manual for the evaluation board with the schematics for the regulator by contacting AKM's sales department.
Thanks for the info. Subscribed.
 
An inexpensive way of shedding Watts (and asking the regulator to do less work) is to just use a resistor in the input path. As the SR is not capable of delivering a lot of current, you're not creating a lot of heat in the pass transistor.

The pass transistor was horizontally mounted in the original Old Colony PCB's -- (a lovely design btw Mr. Didden!) I think that the heat sink I am using is an Aavid Thermalloy5070 series.
 
An inexpensive way of shedding Watts (and asking the regulator to do less work) is to just use a resistor in the input path. As the SR is not capable of delivering a lot of current, you're not creating a lot of heat in the pass transistor.

The pass transistor was horizontally mounted in the original Old Colony PCB's -- (a lovely design btw Mr. Didden!) I think that the heat sink I am using is an Aavid Thermalloy5070 series.

Great idea! Thanks a lot Jack.
 
I think I bit off more then I can chew. My electronics class final project (The BA-3 as preamp build guide) Recommended this power regulator.

I'm the worst student in math in ELEC 102a class. I read the first 10 pages and last ten pages and a few in the middle of this 173 page thread. Is there a exact page of other thread or resource to build a 24V version. The first few pages mentioned higher voltage opamps i understand but start getting lost. I do realize I should pick up the cap voltage to 30v - 35V. and possible other diodes.

I'm ok at cut and paste or copy and then mod after I build my first one.
Need training wheels and hand holding on this Super Regulator
BA-3 to pair with my ACA mono blocks I built last year as my ELEC 101 class project. I'm real good at making magic smoke in every one of my projects, that I get A+ grade
 
With my lack of knowledge but reading through the post I think I understand A few opamps can be used for 24V. The AD825 and AD797, with favoritism going towards the AD825. When I put this part number into Mouser it gives me 20 options. My experience level is limited to slightly higher voltage capacitors are OK slightly bigger resistors are OK and bigger transformers can be used. But opamps,MOSFET, Jfets are still way over my head with all their minute specifications differences.
I know I need to 12 V zener diode. At Mouser I searched 12 V,Zener diode, through hole. I got 187 results. I did not know what wattage to pick but I noticed the results options dropped dramatically if I just picked one wattage at a time.
I also picked up through my reading the post some people had questions at what locations the diode’s should be changed. D7 I got a little lost trying to understand Jan Didden ( customizing the output voltage of the DIYaudiostore superreg project).

As for position D5 referred to as the LM329 reference diode. Do I understand I need to change the value or part number of this item to?.

When I finish this project and it works beautifully I want to add a post to this thread of every part and part number that are used to make a 24 V regulator so the next Nubee doesn’t have to spend every night after work for two weeks wondering what parts to select. I have not read my electronics class homework or turned in any of my sheets just researching these parts for superreg and parts for the BA-3.

Putting in my order tonight at Mouser for my parts for my BA-3 at least that part of my class project I can finish but the superregis killing me on the amount of time just trying to select parts.
 
AX tech editor
Joined 2002
Paid Member
OK, one more try.

Step 1: realize that in operation, both the opamp inputs are at the same voltage. The +input is at the 6.9V of the reference. So your task is to dimension the two resistors feeding the -input from Vout such that -vin also gets 6.9V.

So you know that there is 24V-6.9V= 17.1V across the top R, and 6.9V across the bottom R. So those resistors must be in that ratio: 6.9 to 17.1. Which is 2.48, so the top R is 2.48* bottom R.

A quick one would be to make the bottom R 6.9k and the top R 17.1k, but those are non-standard values.

You could take 4.7k for the bottom R, a standard value, and then the top R would be 4.7k*2.48=11.6k and 12k is a standard value, close enough.

Step 2: the zener diode at the opamp output should be around half Vout, so a 12V zener is fine. Not critical; if you have 10V or 15V in your parts box, that's fine too.

Step 3: select opamp. Since the opamp works between Vout and ground, it can use its total supply capability. For a +/-15V opamp that is 30V and that would be fine here. An AD825 is great; an AD797 also in this respect but the '797 is less stable so only recommended if you know what you are doing. Walt and I never used it, too much hassle.

Step 4: make sure all caps can handle the Vin and Vout + some safety.

Done.

Jan
 
I think AD797 is accepted as the best that has been tried (though there are a few possible better candidates): noise=0.9nV/sqrtHz, Iout=50mA. Only it has high tendency for oscillation: bandwidth=110M, so requires extra care in building.

I guess I could select both opamps and see which one I have the easiest time with my experience level and knowledge is not high enough to debug a problem with an op amp.

This being a class project and time is ticking down to the end of class if I could have this up and running by November. If you can possibly recommend one more preferred op amp I can purchase that too because just the shipping from Mouser for one part is $9.79 so I’m trying to purchase as many of the parts that I need all at once.
By looking at my bucket list of components to build I’m very sure I will be rebuilding this superreg more than one time .
 
AX tech editor
Joined 2002
Paid Member
Hi,

I am a green DIYer here and I want to build this board to output 3.3v on one half. From My understanding, if I change Volt Ref to 2.5V, and use opa162 as Jung originally wrote, I should change R14 to 1.6kOhm, and R13 to 5kOhm.

Correct? Thanks!

I don't have the schematic in front of me, but assuming that R13 and R14 are the feedback resistors from output to -input, with 1.6k being the bottom one. In that case, I think you are wrong.

Remember, remember, remember: in operation, both opamp inputs MUST be at the same voltage (within a few mV).

With 2.5V ref on +input, the lower R also has 2.5V to the -input, so the top R must have 3.3-2.5=0.8V.
With 2.5V across 1.6k, Rtop must of course be 0.8/2.5*1.6k=512 ohms nominal.

Jan
 
I don't have the schematic in front of me, but assuming that R13 and R14 are the feedback resistors from output to -input, with 1.6k being the bottom one. In that case, I think you are wrong.

Remember, remember, remember: in operation, both opamp inputs MUST be at the same voltage (within a few mV).

With 2.5V ref on +input, the lower R also has 2.5V to the -input, so the top R must have 3.3-2.5=0.8V.
With 2.5V across 1.6k, Rtop must of course be 0.8/2.5*1.6k=512 ohms nominal.

Jan

Hi Jan,

Thanks for quick reply. But by your schematic, R13 is the bottom one in parallel with 120uf cap.
 
AX tech editor
Joined 2002
Paid Member
Hi Jan,

Thanks for quick reply. But by your schematic, R13 is the bottom one in parallel with 120uf cap.

There's two resistor on top of each other, one end to Vout, the other end to gnd, the mid to -input opamp. Those are the ones that set the output voltage by 'multiplying up' the ref voltage at the +input of the opamp.

On the schematic in the diyaudio store, R7 (pos) and R14 (neg) are the 'bottom' ones that are connected to gnd, just as the reference. The 'top' is R7 (pos) and R13 (neg), these are connected to Vout.

Agreed?

Edit: 'bottom' I mean at gnd, not physically, sorry. I think in voltages ;-)

There is indeed a cap // R7, R13 to make the voltage division for AC equal to one, for better regulation.

Jan