Power Supply Voltage Out Mod - going from +-15V to +-18V

[Nen'O]

Hi,
Hope someone here with experience can advise if this mod is doable.
It's Yamaha PW8 linear power supply I want to mod.
Service manual available online

https://audio-circuit.dk/downloads/yamaha10/Yamaha-PW8-psu-sm.pdf
From what I've been told the regulators on PW8 are already a discrete design, the best way to get higher voltage is to replace zener diode references that are in there.
Changing ZD101 and ZD102 from 5.6V to 6.8V will get out close to 18V.

Anything else I should take a look at?

Cheers
Neno

 

[radardoug]

Why? If you are doing it to get more audio headroom, do the maths on what the increase will give you. Not worth doing for headroom improvement, and puts everything under more stress.

 

[user 130681]

Replace R131 and R132 with a 1k pre set.
You will have a range from the maximum that the DC can supply to about what you have at present.
Far simpler than changing zener values.
Bear in mind the op amps are not happy with more than +-16volts!

 

[saint gillis]

Replace R131 and R132 with a 1k pre set.
You will have a range from the maximum that the DC can supply to about what you have at present.
Far simpler than changing zener values.
Bear in mind the op amps are not happy with more than +-16volts!

Douglas Self recommends +/-17V in his books :
" Soundcraft ran all the op-amps in their mixing consoles at 17 V for at least two decades, and op-amp failures were almost unknown. This recommendation assumes that the op-amps concerned have a maximum supply voltage rating of 18 V, which is the case for the Texas TL072, the new LM4562, and many other types. "

 

[Nen'O]

Hi all,
Thanks for advices.
This PSU is powering A&H ZED R16. Ive already modded this board and took out stock SMPS and replaced it with connector for external linear PSU.
Everything works great. I'm actually surprised how good PSU is Yamaha PW8.
You are right , the mod is ment to increase headroom.
Actually I'm in touch with Mike from A&H who designed this mixer. He confirmed that op-amps will be ok up to +-18V at least. ZED R16 shares the same circuits with 24 channels version A&H GS R24 which is powered with +-17V.

Op-amps in these boards are 4850 for preamps, summing and buffering and good old TL072 for eq's.

This is exact instructions I got from some old engineering cat.

I think the easiest way to do it is by sticking a diode in series with the zener.

Just remember to put them either anode to anode or cathode to cathode. That is because zeners operate in the reverse direction but the diodes need to operate in the forward direction.
This will basically add a diode drop (0.7V) to the zener voltage, which is 5.6V for the 15V supply rails.


Thanks

 

[Nen'O]

Hi Jon,
Can you please explain why Op-amp is not happy with more than+-16V.

Thanks

 

[JohnRoberts]

MTBF (mean time between failure) doubles for every 10'C higher ambient temperature. Power dissipation increases roughly linearly with increased voltage. Headroom in dB only increases slightly from several volts of higher rails.

Prudent designers figured this out decades ago...

JR
[edit- typical IC processes have 36v maximum tolerance setting rail limits at +/-18v max, it is not prudent to stray too close to the edges./edit]

 

[user 130681]

Hi Jon,
Can you please explain why Op-amp is not happy with more than+-16V.

Thanks

Look at the data sheets for different op amps. Most were rated at +-15 maximum and having no clue what you were supplying, erring on the side of caution can save an awful lot of fault finding time when your 4558s have all gone bang or extremely noisy due to a voltage overload.
Most newer op amps are happy with +-18volts but no more.
Now I know you are using modern A&H equipment, it will probably be OK. (The boards are protected internally with 100R 1/2W Fire Proof resistors in case of over voltage, as you will know).
It is always best to avoid confusion by stating the purpose as well as the device in question, I find.

 

[Tubetec]

I'm looking at a +/- 14 volt supply from 20 v lithium cells ,
Battery is 20.5v fully charged , 15 volts discharged , regulator starts to drop out at around .75 volts , So it seems a voltage of around +/-14.25 v on the op amps would be right , I'll use a relay board to detect when the battery reaches 15 volts and cuts out to prevent over discharge .

 

[radardoug]

Hi all,
Thanks for advices.
This PSU is powering A&H ZED R16. Ive already modded this board and took out stock SMPS and replaced it with connector for external linear PSU.
Everything works great. I'm actually surprised how good PSU is Yamaha PW8.
You are right , the mod is ment to increase headroom.
Actually I'm in touch with Mike from A&H who designed this mixer. He confirmed that op-amps will be ok up to +-18V at least. ZED R16 shares the same circuits with 24 channels version A&H GS R24 which is powered with +-17V.

Op-amps in these boards are 4850 for preamps, summing and buffering and good old TL072 for eq's.

This is exact instructions I got from some old engineering cat.

I think the easiest way to do it is by sticking a diode in series with the zener.

Just remember to put them either anode to anode or cathode to cathode. That is because zeners operate in the reverse direction but the diodes need to operate in the forward direction.
This will basically add a diode drop (0.7V) to the zener voltage, which is 5.6V for the 15V supply rails.


Thanks

I guess you couldn't be bothered doing the calculation, so here it is for you.
Do you really think you are going to be able to hear the difference?

 

[Nen'O]

Look at the data sheets for different op amps. Most were rated at +-15 maximum and having no clue what you were supplying, erring on the side of caution can save an awful lot of fault finding time when your 4558s have all gone bang or extremely noisy due to a voltage overload.
Most newer op amps are happy with +-18volts but no more.
Now I know you are using modern A&H equipment, it will probably be OK. (The boards are protected internally with 100R 1/2W Fire Proof resistors in case of over voltage, as you will know).
It is always best to avoid confusion by stating the purpose as well as the device in question, I find.

Thanks Jon,

My knowledge of el. circuits is very basic.
The mods I'm doing to my A&H Zed R16 are way to dip into world of analogue circuit design, possibly improve already great sounding mixer and try not to burn down the house.

It's really useful info you shared.


Cheers

 

[Nen'O]

I guess you couldn't be bothered doing the calculation, so here it is for you.
Do you really think you are going to be able to hear the difference?

Hi Doug,

I wasn't even aware there is such a thing as analogue circuit headroom calculator but i did came across this post here before I came to an idea to step up PS

https://groupdiy.com/threads/headro...om-isnt-always-a-good-thing.73979/post-936701
"Lack of headroom as a sonic choice is useless if the topology doesn't sound good when out of juice. Different question. Headroom itself has little to do with the Neve versus the SSL. The same is true of a Mackie versus an (older) Midas.

Worth considering this theoretical (16 versus 24 doesn't buy much):

9V (or +/-4.5V) = +12.27 dBu
12V (or +/-6V) = +14.77 dBu
24V (or +/-12V) = +20.79 dBu
30V (or +/-15V) = +22.73 dBu
32V (or +/-16V) = +23.29 dBu
34V (or +/-17V) = +23.82 dBu
36V (or +/-18V) = +24.31 dBu
40V (or +/-20V) = +25.23 dBu
44V (or +/-22V) = +26.06 dBu
48V (or +/-24V) = +26.81 dBu
60V (or +/-30V) = +28.75 dBu"

Going from SMPS to linear supply was totally worth mod, i can tell the difference. PW8 is 300W and made to supply much larger mixers and it's total overkill for 50w ZED R16 needs, but there is some magic going on with oversized power supply.

This guy explained it much better



If going from +-15V to +-17V is an easy and not risky mod I'm in. From what I understood the mixer is design to be powered with +-17V.
Increasing headroom by 1.5 db will be welcome in many situation especially when working with acoustic ensembles where every db counts, now I'm not sure if +-17 is going to improve headroom of ZED R16 A/D-D/A converters, can only hope it will.

Cheers

 

[Khron]

If anything, you might then need to slightly alter the attenuation before the ADC inputs, since it would be safe to assume those have been designed to clip around the same time as the opamps on the analog side.

What use is more analog headroom if you might end up clipping the ADC inputs?

This just once again brings up the "why?" question for whatever mods are desired.

 

[Nen'O]

Hey Khron,
Glad to see you here. Learned a lot from reading your blog.
Is your GTR amp coming out anytime soon?

Messing with circuits inside ADC is definitely no,no for me.
Good news is that this board has been very well calibrated, the analogue circuit will clip much sooner before ADC get overloaded.
Stock specs are actually not that bad at all... considering that A&H doesn't boost numbers for marketing purpose like some do...

Headroom

Analogue Headroom from nominal (0Vu) 21dB
Digital converter headroom from nominal analogue (0Vu) 16dB

Digital Performance

Analogue to Digital conversion 24bit 114dB dynamic range (A wtd)
Digital to Analogue conversion 24bit 118dB dynamic range (A wtd)

 

[abbey road d enfer]

This guy explained it much better

3 comments:

• This guy has someting to sell
• Don't disregard cognitive bias
• My personal opinion is that if the circuit shows different response to different PSU that deliver the same voltage, something is wrong in its design

 

[Tubetec]

It would seem from the figures going up much beyond 16v makes only a small difference ,
would it be better to concentrate on making the supply quieter instead of trying to increase head room ?

 

[Khron]

Hey Khron,
Glad to see you here. Learned a lot from reading your blog.
Is your GTR amp coming out anytime soon?

Messing with circuits inside ADC is definitely no,no for me.
Good news is that this board has been very well calibrated, the analogue circuit will clip much sooner before ADC get overloaded.
Stock specs are actually not that bad at all... considering that A&H doesn't boost numbers for marketing purpose like some do...

Headroom

Analogue Headroom from nominal (0Vu) 21dB
Digital converter headroom from nominal analogue (0Vu) 16dB

Digital Performance

Analogue to Digital conversion 24bit 114dB dynamic range (A wtd)
Digital to Analogue conversion 24bit 118dB dynamic range (A wtd)

I couldn't have known that such details were even part of the spec sheets, but that's a nice touch.

And the amps have been on the back burner for quite some time now - life keeps getting in the way, and all that

 

[Nen'O]

It would seem from the figures going up much beyond 16v makes only a small difference ,
would it be better to concentrate on making the supply quieter instead of trying to increase head room ?

Hi,
PW8 is surprisingly quiet and discrete design, even caps inside are good quality. Not sure if there is much left to do filter out noise and make it quiter.
I really didn't expect it will be that good PSU but this company (Yamaha) once again manage to wooow me.

 

[abbey road d enfer]

Hi Doug,

I wasn't even aware there is such a thing as analogue circuit headroom calculator but i did came across this post here before I came to an idea to step up PS

9V (or +/-4.5V) = +12.27 dBu
12V (or +/-6V) = +14.77 dBu
24V (or +/-12V) = +20.79 dBu
30V (or +/-15V) = +22.73 dBu
32V (or +/-16V) = +23.29 dBu
34V (or +/-17V) = +23.82 dBu
36V (or +/-18V) = +24.31 dBu
40V (or +/-20V) = +25.23 dBu
44V (or +/-22V) = +26.06 dBu
48V (or +/-24V) = +26.81 dBu
60V (or +/-30V) = +28.75 dBu"

It should be noted that these figures are oversimplified.
Actually, with a 9V (+/-4.5V) supply, the possible output voltage is not the 3.18Vrms that +12.27dBu suggests. Most opamps (discrete or monolithic) cannnot deliver rail-to-rail output*. They have a residual loss of about 1V each way, so the 9V swing becomes 7, resulting in a max output of +10dBu, or a loss of 2.2dB.
Of course, at the other end, with 36V rail max output is arespectable +23.8dBu, a loss of only about 0.5dB.

*Indeed tehre are rail-to-rail opamps, but very few can operate with the typical +/-15-18V commonly used in audio.

 

[Tubetec]

By the looks of the PW8 is an SMPS with linear regulators after , its probably very carefully designed to be noise free , your standard off the shelf switcher will have hundreds of mV of noise .