Jump to content
IGNORED

Buffalo GS20xx - Bypass 12V stage via 3 external PSs


Recommended Posts

5 hours ago, RickyV said:

So I Frankensteined this today and it works,😀.
 

2EC473DB-A603-48B8-B63D-4B76750E5754.thumb.jpeg.502460c43063ee215d3f7a43edff4b8c.jpeg

 

So I had to fall back to a smps to deliver the power, lower right corner.

From 7.5V I dropped it to 3.6V, that heatsink gets 60 degrees C. From this 3.6V I fed a 1A 3.3V module and a 2A 1.5V module. I used 2A module because dissipated heat/ power. 
The 1V, 5A was still a problem I started out with 4V on the input but that the regulators got about 80 degrees C, too hot. Than I tried it one diode, dropping 0.8V that dropped the temperature to 60 degrees C, still to hot. Than two diodes dropping 1.6V so now the 1V/5A regulator has 2V on its input and is about 50 degrees C now.

The heatsink’s from the first regulator and the diodes get about 60 degrees C. Even the smps get 43 degrees C.

So now that everything is working I will put it, temporary, in my system upstream of my Etherregen, connected by fiber.

So now I have to figure out what I need for a power supply. 
7.5Vx 5A= 37.5W    37.5W/3.6V= 10.4A at 3.6V ???

 

 

Hi Ricky

Could you please give more details about the stacked modules you are using to get the 1.5v and 1v supplies?

Are they 1 amp LDOVR modules? I don't get how you actually stacked them like that.

Link to comment
28 minutes ago, Soul Analogue said:

Ok guys

 

today i found out why i could not power the gs2016a with my variable lab psu.... i tried to load test it.. and it could only supply stable volt at 1.6a at 1v.... 

so i have built my own 1v discrete regulator (i dont like monolitic regulators🙄sorry) and now successfully power it 

and i managed to measure the currents levels of these rails. Power up with two copper ethernet connections, max current Results as follows:

1v 2.5a

1.5v 0.2a

3.3. 0.7a


6 hours of work, problem solving and measurements....hope these figures help

2 Gold Stars for effort!

Link to comment
21 minutes ago, Soul Analogue said:

Ok guys

 

today i found out why i could not power the gs2016a with my variable lab psu.... i tried to load test it.. and it could only supply stable volt at 1.6a at 1v.... 

so i have built my own 1v discrete regulator (i dont like monolitic regulators🙄sorry) and now successfully power it 

and i managed to measure the currents levels of these rails. Power up with two copper ethernet connections, max current Results as follows:

1v 2.5a

1.5v 0.2a

3.3. 0.7a


6 hours of work, problem solving and measurements....hope these figures help


Thanks that helps a lot. 
I think for the 1V we need a regulator that can dissipate a lot of heat so my 5A ldovr stack my not be the best solution in this spot. I also Have a couple of 5A newclassd regulators on hand that need a 5v voltage drop so then I can supply all regulators with 6v. The only problem with the newclassd regulator is that it’s minimal voltage is 1.8V so I still need a diode after it dropping 0.8 v.


02D96D59-B411-4894-9005-3CA7634DDB50.thumb.jpeg.0d1d625d0c4d54dc85404eda261c6eee.jpeg

 

oh and btw I had a listen last night with the buffalo in place and sounds really good.

Meitner ma1 v2 dac,  Sovereign preamp and power amp,

DIY speakers, scan speak illuminator.

Raal Requisite VM-1a -> SR-1a with Accurate Sound convolution.

Under development:

NUC7i7dnbe, Euphony Stylus, Qobuz.

Modded Buffalo-fiber-EtherRegen, DC3- Isoregen, Lush^2

Link to comment
6 minutes ago, Soul Analogue said:

Oops.. tested again with power cycle:

the 1v max out at 3.1a ... not able to add sfp+ connections yet... the current drawn maybe higher...


And I think when it’s boots up there are much higher current peaks.

Meitner ma1 v2 dac,  Sovereign preamp and power amp,

DIY speakers, scan speak illuminator.

Raal Requisite VM-1a -> SR-1a with Accurate Sound convolution.

Under development:

NUC7i7dnbe, Euphony Stylus, Qobuz.

Modded Buffalo-fiber-EtherRegen, DC3- Isoregen, Lush^2

Link to comment
10 minutes ago, RickyV said:


Thanks that helps a lot. 
I think for the 1V we need a regulator that can dissipate a lot of heat so my 5A ldovr stack my not be the best solution in this spot. I also Have a couple of 5A newclassd regulators on hand that need a 5v voltage drop so then I can supply all regulators with 6v. The only problem with the newclassd regulator is that it’s minimal voltage is 1.8V so I still need a diode after it dropping 0.8 v.


02D96D59-B411-4894-9005-3CA7634DDB50.thumb.jpeg.0d1d625d0c4d54dc85404eda261c6eee.jpeg

 

oh and btw I had a listen last night with the buffalo in place and sounds really good.

Haha.. your heatsink looks a bit too small

This is mine.. dropping from 8.8v

 

Building a 1v supply is quite a challenge... i used to have newclassd reg... but it could not get stable at 1.2v... the designer had no clue what can be done

so i resort to build everything by my own 🙄 this hobby is testing my endurance and patience..

 

using diode in series at the output of regulation should not be an ideal method... the diode can have different forward voltages at different current levels... making the voltage at the load not stable.... 

7BD1F7AA-40F0-4939-9E6E-1FFF69B8EC7B.jpeg

Builder of Linear Power Supplies

Link to comment
17 hours ago, tims said:

Hi Ricky

Could you please give more details about the stacked modules you are using to get the 1.5v and 1v supplies?

Are they 1 amp LDOVR modules? I don't get how you actually stacked them like that.


Here is how to do that. You have to connect all the inputs and all the outputs of all the modules. Plus the set points of all the modules, last picture red arrow. 

18D707A7-0E35-461D-896D-D374F66C6FF0.thumb.jpeg.d4cf3bd118530c9edd17efd919ef896a.jpeg
 

126E1D56-F1B4-4740-A1D7-4E735364B4F5.thumb.jpeg.b47d8764c4fbf08388764a5945fbfa0e.jpeg


You also have to put a drop of tin on the ENable pads where it says on off.

 

6505834C-8A3F-40C7-8AF1-0035DF6FA9B8.thumb.jpeg.71ba7052f64ac1bf7b3ce626f2a6a589.jpeg

 

Meitner ma1 v2 dac,  Sovereign preamp and power amp,

DIY speakers, scan speak illuminator.

Raal Requisite VM-1a -> SR-1a with Accurate Sound convolution.

Under development:

NUC7i7dnbe, Euphony Stylus, Qobuz.

Modded Buffalo-fiber-EtherRegen, DC3- Isoregen, Lush^2

Link to comment
11 minutes ago, Soul Analogue said:

using diode in series at the output of regulation should not be an ideal method... the diode can have different forward voltages at different current levels... making the voltage at the load not stable.... 

 

Thanks that is bad news that 1V is a real pain in the butt.

Meitner ma1 v2 dac,  Sovereign preamp and power amp,

DIY speakers, scan speak illuminator.

Raal Requisite VM-1a -> SR-1a with Accurate Sound convolution.

Under development:

NUC7i7dnbe, Euphony Stylus, Qobuz.

Modded Buffalo-fiber-EtherRegen, DC3- Isoregen, Lush^2

Link to comment

Great work Soul Analogue !

 

I'll do it shortly on my GS2024.

Hope that "bigger switch" does not mean "bigger amps" :-)

Interesting your remark regarding the higher current / SFP. I need 1...

 

Regarding power consumption of SFP ports, the rough figure I managed to get on my modded TP-Link SG5412F (3 external PS : 3.3V / 1.8V / 1VDC) : 1 more SPF activated => overall consumption up by 1W, but most of it on the 3.3V, not on the 1V. I'll need to check that point again.

Rgds

2.1 basic stuff => 2 mains are Dynaudio Core59 + sub Dynaudio 18s

Actives / digital AES in / active correction on PC side

Passive daddy setup is dead

Link to comment
7 hours ago, Soul Analogue said:

Haha.. your heatsink looks a bit too small

This is mine.. dropping from 8.8v

 

Building a 1v supply is quite a challenge... i used to have newclassd reg... but it could not get stable at 1.2v... the designer had no clue what can be done

so i resort to build everything by my own 🙄 this hobby is testing my endurance and patience..

 

using diode in series at the output of regulation should not be an ideal method... the diode can have different forward voltages at different current levels... making the voltage at the load not stable.... 

7BD1F7AA-40F0-4939-9E6E-1FFF69B8EC7B.jpeg

 

Hi, which voltage regulator do you use? Most have a minimum output of 1.2v...

Link to comment
21 minutes ago, basillus said:

 

Hi, which voltage regulator do you use? Most have a minimum output of 1.2v...

It is a discrete regulator designed by myself... 

mosfet usually have a Vgs of 1.2v ... that makes them only respond to voltage when it reaches 1.2v... and hence the minimum output...

it was a bit tricky to overcome this Vgs boundary

Builder of Linear Power Supplies

Link to comment

May not be of interest but here are a couple of LDO regs that offer <1v output voltage and 5A output:

"The LT3070 from Analog Devices has 5-A output with 0.15-V maximum dropout. Maximum input is 3 V; output voltage is adjustable from 0.8 to 1.8 V."

and

"The MIC69502 from Microchip Technology has 5-A output with 0.5-V maximum dropout. Maximum input is 5.5 V; output voltage is adjustable from 0.5 to 5.5 V"

from

https://www.electronicdesign.com/power-management/power-supply/article/21808454/5-ldo-regulators-with-high-output-current

Link to comment
11 minutes ago, tims said:

May not be of interest but here are a couple of LDO regs that offer <1v output voltage and 5A output:

"The LT3070 from Analog Devices has 5-A output with 0.15-V maximum dropout. Maximum input is 3 V; output voltage is adjustable from 0.8 to 1.8 V."

and

"The MIC69502 from Microchip Technology has 5-A output with 0.5-V maximum dropout. Maximum input is 5.5 V; output voltage is adjustable from 0.5 to 5.5 V"

from

https://www.electronicdesign.com/power-management/power-supply/article/21808454/5-ldo-regulators-with-high-output-current

This is a good news! Makes many lives easier😉

Builder of Linear Power Supplies

Link to comment

hi,

just an idea regarding the tricky 1VDC : we can "not to deal with it". (I'm quick a noob in electronic vs you, so be kind 🙂 )

 

FYI, see this post & the next 2 ("phile" overthere it's me, "tgb" here) : http://forum-hifi.fr/thread-16891-post-381881.html#pid381881

 

As I replied in the post in the link above & following the feedback of Audiodemat : the key voltages are 3.3V (first) & 1.5V (2nd).

The 1V does not seems to act on SQ (assumption to be confirmed... maybe completly wrong...)

Based on that =>

1. as the amps @ 3.3V & 1.5V are rather small, they are easy to manage vs heat => we can look further & plan a nice & good PS for them

2. as the amps @ 1V are high & lead to trouble vs heat => no brainer because the 1V may not play a role in SQ => let's choose an easy & cheap way to deal with it.

 

This is the cheap & easy way I use :

I use a lot the Drok buck converters (for inst. when I tested the Buff) : very efficient stuff, good price, not noisy.

This one is a good one => https://www.droking.com/dc-power-supply/dc-buck/Adjustable-High-Power-Adapter-Buck-Voltage-Regulator-DC8-40V-to-1.25-36V-8A-100W-Converter-Laptop-Charger

As a power DC source : any USB charger. I use this one for testing when I need 5VDC => https://www.amazon.fr/RAVPower-Chargeur-Universel-Technologie-110-240V/dp/B00PRTBCYW?th=1

1 port can deliver 2A => use 2 ports and you get 4A (may need to add a resistor or cap in between if trouble but it may be as simple as that).

Then : as this good Drok (that can handle heavy currents) can not go lower than 1.25VDC => simply add some power resistors downstream (like this one I use and works well vs heat => https://www.conrad.fr/p/weltron-resistance-de-puissance-033-sortie-axiale-11-w-10-1-pcs-411710 ) 

 

Finally, the "cheap & easy setup" to manage the 1V would be :

USB charger (15eur) => 5VDC output => 2 Droks (8eur each), 2 Droks to split the need vs heat transfer, maybe need to add a resistor between the 2 Droks to help vs voltage regulation => some power resistors (6eur max) => to finally get the 1VDC.

 

Tip : b y adding the resistors after the regulation steps using the Drok modules you can guess the final voltage, so you simply have to use the trimer on the last Drok to finetune & get the final 1VDC.

 

My point is only about money efficiency. Dealing with 1.5 & 3.3V with noce Mpaudio or any LT3045 regs willl be easy vs heat

Dealing with 1V at 3-5A add cost, and if the SQ is not much affected by the 1V  : better to put $$ where it is relevant.

Just an idea & brainstorming 🙂

Rgds

2.1 basic stuff => 2 mains are Dynaudio Core59 + sub Dynaudio 18s

Actives / digital AES in / active correction on PC side

Passive daddy setup is dead

Link to comment
13 hours ago, tgb said:

).

Then : as this good Drok (that can handle heavy currents) can not go lower than 1.25VDC => simply add some power resistors downstream (like this one I use and works well vs heat => https://www.conrad.fr/p/weltron-resistance-de-puissance-033-sortie-axiale-11-w-10-1-pcs-411710 ) 

 

Finally, the "cheap & easy setup" to manage the 1V would be :

USB charger (15eur) => 5VDC output => 2 Droks (8eur each), 2 Droks to split the need vs heat transfer, maybe need to add a resistor between the 2 Droks to help vs voltage regulation => some power resistors (6eur max) => to finally get the 1VDC.

 

Tip : b y adding the resistors after the regulation steps using the Drok modules you can guess the final voltage, so you simply have to use the trimer on the last Drok to finetune & get the final 1VDC.

 

a dropping resistor after the regulated voltage can only be used if the current drawn by the load is constant

however the load at the 1v at the buffalo is constantly changing

so the voltage dropped by the resistor will be constantly changing, according to Ohms law..and hence the changing volt at the load... 

i believe it would lead to instability of the switch... even if it happens to work, the sound resulted would be worse than the onboard smsp regulator

Builder of Linear Power Supplies

Link to comment

Hi,

quick feedback after long eve... 6hours... and finally half success (although I'm still not sure it will work :-) ).

I took the selfs off.

But then, I could not manage to boot the switch cause the 1V. I used a charger 5VDC, not enough amps cause limited to 2A ; I tried to use 2 ports but still not enough amps available.

Finally, I found in the bin the 1V self ... I roughly soldiered it back => reuse the 12V input to get the 1V like beofre (only 1V cause 3.3V & 1.5V are ok using external PS) => and the switch boots (nothing plugged) !

It's too late, I'll see tomorrow if it can really work.

 

The figures I got, without any RJ45 or SFP on the GS2024 :

1.5V : 0.32A, stable

3.3V : up & down 0.4 <> 1.4A

1V : ? amps. Based on the 12V input, I have up-down 0.26-0.3A at 12VDC.

 

I admit I tried to go back to the 1V powered thru the 12V because using an 5VDC external PS, to go down to 1V => the heat sink of the buck was too hot (>50°C). I felt not secure with that.

I'm a noob in electronics, so, managing 4A at 1V, I'm not used to :-)

 

I hope that tomorrow I can get some music out of this tricky switch ! :-)

 

Very strange this switch compared to my Tp-link SG5412F :

on the tp-link, all currents are stable => 0.92A for 3.3VDC, 0.77A for 1.8VDC & only 0.93A for 1VDC (very far from the 3-4-5?amps on the Buff at 1V).

2.1 basic stuff => 2 mains are Dynaudio Core59 + sub Dynaudio 18s

Actives / digital AES in / active correction on PC side

Passive daddy setup is dead

Link to comment

Hi tbh

 

Thanks for your continued work with the switch'es power supply.  I've been following your posts in forum-hifi.fr (via google translate) and read that the 1v supply works with 5v (on the 12v input) with the other voltages (3.3v and 1.5v) being fed separately. 

However, without the separate supplies for 3.3v and 1.5v,  do you think a linear 5v PS (feeding 3.3, 1.5 and 1v) would sucessfully boot the switch?

If so, I guess the 5v supply would need to output at least 4 or 5 amps?

I know that this single linear 5v supply supply would not be as good separate supplies for 3.3 and 1.5v but it would have to be better than the 12v switching supply that comes with the switch.

Link to comment

Hi,

sorry i'm lazy to translate rigth now, please use Google Translate, please find there the last mod => http://forum-hifi.fr/thread-16891-post-382981.html#pid382981 (this post and the next ones).

Just ask if you need more info, right here.

 

To make it short :

I'm not expert, thus impossible for me to feed 1VDC at 4A minimum. So, the 1VDC is fed thru the input connector with 5VDC. As the 2 selfs on 3.3 & 1.5VDC are out => all the current goes to the 1VDC circuit.

As I wrote 1hours ago or so on the forum-hifi :

current on 3.3VDC is quite high : moving from 0.7 to 1.5A. I even lost the Buff htis night so I thnik it needed a pick above the 2A that the PS can deliver. Thus a 2.5A PS for the 3.3V seems to be a mnimum.

All tests are done with only 2 ports in use : 1 SFP (to other switch) & 1 RJ45 (to the streamer).

 

Interesting stuff. Final setup is not the one I expected but the high current at 1VDC is not for nood in electronics like me :-)

 

@Tims :

powering the Buff with 5VDC instead of 12VDC should work, no problem. The issue is simply the amps : let's say thatyou need a PS able to deliver 3A minimum at 5VDC ; at 2A the PS will be close to its maximum output at steady state, so not secure.

Exact, as you say & as I learn yeateday on the forum-hifi from patatras : powering with 5VDC instead of 12VDC leads to better SQ.

Rgds

2.1 basic stuff => 2 mains are Dynaudio Core59 + sub Dynaudio 18s

Actives / digital AES in / active correction on PC side

Passive daddy setup is dead

Link to comment

@tims I correct my previous post : 5VDC at input instead of 12VDC => issues ahead.

Honestly, if you plan to try any mod on this Buff => wait for comments (I'll post my latest mods here for instance), that states that, over a few days, the mod is fine & the Buff runs OK at steady state.

 

Although you're an expert in electronics or you can short-circuit a Buff & you don't care because your retailer have plenty of them. I'm fine, try & bring some info :-)

If you're a noob like me : wait & see :-)

if you're a noob & like this tricky challenges : go for it :-)

 

News from the weekend : this Buff is a trap for DIYer, I think...

all weekend : it boots > Buff back to hifi setup > music streaming > the Buff loose the network > can't make it boot again > Buff out of hifi setup to desk to try another way to power it > back to hifi setup > boot > music streaming > the buff loose the network > can't... > back to desk > etc...

x times...

music has been in the air for 1h, but... don't know if this time it will last all night, or stop within 30min or 1h

 

If I wasn't a cool & calm guy, I could a play freebies withe the Buff tis afternoon & beat the longest throw :-)

Rgds

2.1 basic stuff => 2 mains are Dynaudio Core59 + sub Dynaudio 18s

Actives / digital AES in / active correction on PC side

Passive daddy setup is dead

Link to comment
On 8/1/2020 at 11:08 AM, tgb said:

Hi,

sorry i'm lazy to translate rigth now, please use Google Translate, please find there the last mod => http://forum-hifi.fr/thread-16891-post-382981.html#pid382981 (this post and the next ones).

Just ask if you need more info, right here.

 

To make it short :

I'm not expert, thus impossible for me to feed 1VDC at 4A minimum. So, the 1VDC is fed thru the input connector with 5VDC. As the 2 selfs on 3.3 & 1.5VDC are out => all the current goes to the 1VDC circuit.

As I wrote 1hours ago or so on the forum-hifi :

current on 3.3VDC is quite high : moving from 0.7 to 1.5A. I even lost the Buff htis night so I thnik it needed a pick above the 2A that the PS can deliver. Thus a 2.5A PS for the 3.3V seems to be a mnimum.

All tests are done with only 2 ports in use : 1 SFP (to other switch) & 1 RJ45 (to the streamer).

 

Interesting stuff. Final setup is not the one I expected but the high current at 1VDC is not for nood in electronics like me 🙂

 

@Tims :

powering the Buff with 5VDC instead of 12VDC should work, no problem. The issue is simply the amps : let's say thatyou need a PS able to deliver 3A minimum at 5VDC ; at 2A the PS will be close to its maximum output at steady state, so not secure.

Exact, as you say & as I learn yeateday on the forum-hifi from patatras : powering with 5VDC instead of 12VDC leads to better SQ.

Rgds


I have the 3.3v  running from a 1A lt3045 without it getting hot.

Meitner ma1 v2 dac,  Sovereign preamp and power amp,

DIY speakers, scan speak illuminator.

Raal Requisite VM-1a -> SR-1a with Accurate Sound convolution.

Under development:

NUC7i7dnbe, Euphony Stylus, Qobuz.

Modded Buffalo-fiber-EtherRegen, DC3- Isoregen, Lush^2

Link to comment
11 minutes ago, RickyV said:


I have the 3.3v  running from a 1A lt3045 without it getting hot.

oh oh !...

I think I screwed up somewhere...  🙂

2.1 basic stuff => 2 mains are Dynaudio Core59 + sub Dynaudio 18s

Actives / digital AES in / active correction on PC side

Passive daddy setup is dead

Link to comment
  • 4 weeks later...

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...