• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Swapping class AB amp linear supply with a SMPS - incl measurements

maty

Major Contributor
Joined
Dec 12, 2017
Messages
4,596
Likes
3,160
Location
Tarragona (Spain)
...Hypex SMPS installed and E1DA ADC: 5W/6.3V 8R


MicroAudio- noise-orchardaudio.png
 

fpitas

Master Contributor
Forum Donor
Joined
Jul 7, 2022
Messages
9,885
Likes
14,191
Location
Northern Virginia, USA

maty

Major Contributor
Joined
Dec 12, 2017
Messages
4,596
Likes
3,160
Location
Tarragona (Spain)
If you have to buy one, better one from MicroAudio.
 

Piere

Active Member
Joined
Feb 24, 2022
Messages
195
Likes
190
Have looked at it. Apart from pure Class-A amps I don't see what a PFC is doing in an audio PSU... :confused: But at least you can hook up as many uF on it as you wish!
 

Doodski

Grand Contributor
Forum Donor
Joined
Dec 9, 2019
Messages
20,748
Likes
20,759
Location
Canada
Due to the still large heatsinks it appears to me they still use their old Class-A/B Mosfet architecture. And why not? It was a landmark in the '80-ties and '90-ties. And sound wise to me it can still cope with most of the top amps today! Here a pin-up of the 3370 I restored recently:
I would take a Perreaux if one was offered to me. Nice gear. I remember something about a magazine test of a very large model and they used quarter dollar coins to sorta-weld another coin with it.
the feedback loop was modified and the poor thing cried (oscillated) out loud. Luckily all mosfets survived! You can still see the burned PCB under the zobel resistors at the top left. And if you ever touched the smell of burned epoxy PCB... The thing went firstly 2 weeks in quarantine for it.
I've seen maybe 2 or 3 amps that oscillated and they all required me sticking my fingers in the circuitry and touching it. :D This one is a beauty and it is fortunate that it managed to survive. Yes, I've seen some burned stuff. One was a SoundStream power amp for home use and it was wet inside from the electrolytic of the main smoothing caps that where exploded, there was fiber all over inside from exploded caps and the amp PCBs had burn marks too. I boxed it up and shipped it to the manufacturer for the customer and for in-warranty engineering evaluation. Too many toxins in that unit for me to clean it up... :D
 

maty

Major Contributor
Joined
Dec 12, 2017
Messages
4,596
Likes
3,160
Location
Tarragona (Spain)
This Class D amplifier has a very good PSRR -> no problema with Hypex, MicroAudio or...
 
OP
McFly

McFly

Addicted to Fun and Learning
Joined
Mar 12, 2019
Messages
903
Likes
1,861
Location
NZ
Have looked at it. Apart from pure Class-A amps I don't see what a PFC is doing in an audio PSU... :confused: But at least you can hook up as many uF on it as you wish!
Really? saving energy perhaps? The power factor I get with the hypex is like ~0.6, like I said earlier when I was pumping 400W x2 I got ~1300W at the wall, but 9.5amps at the wall -The power companies turbines are actually doing more like 2280W (technically ~2280VA) of work - That's nearly the maximum the socket is allowed to do here - 10A @ 240V 2400W/2400VA.
 
OP
McFly

McFly

Addicted to Fun and Learning
Joined
Mar 12, 2019
Messages
903
Likes
1,861
Location
NZ
Due to the still large heatsinks it appears to me they still use their old Class-A/B Mosfet architecture. And why not? It was a landmark in the '80-ties and '90-ties. And sound wise to me it can still cope with most of the top amps today! Here a pin-up of the 3370 I restored recently:

View attachment 229321

It was modified almost to death by some Audio Magician. When It arrived at my desk very large 4,7 uF 1000V Mundorf Silver foil caps, that merely belong to an audiophile speaker x-over, were strapped to the back as input capacitors. Very good hum catchers I know now! Also the feedback loop was modified and the poor thing cried (oscillated) out loud. Luckily all mosfets survived! You can still see the burned PCB under the zobel resistors at the top left. And if you ever touched the smell of burned epoxy PCB... The thing went firstly 2 weeks in quarantine for it. Top right the Hypex soft start.

After fondly restoration, the thing is now happily in service again. Here how it performs at 300W into 8 ohms:

View attachment 229323

Mains hum looks a bit misleading dramatic here but it is about -108dB and distortion quite low. It's an order lower than specified in the manual.
Very nice! perhaps they were at the limit of their measurement gear back then and they were actually a bit better than the specs. Probably not uncommon for the time with several manufacturers.
 

Piere

Active Member
Joined
Feb 24, 2022
Messages
195
Likes
190
PFC is only beneficial above a certain continuous power level. That holds only for class-A amps and to a certain degree for PA amps. But not for domestic class-A/B amps idling at 10W - 20W. With the MicoAudio SMPS it serves another trick: The PFC provides a stabilised +400V rail and then that is converted down brute force to the required + and - rail voltages without any regulation. That PFC is merely acting as a pre-psu.
 
Last edited:
OP
McFly

McFly

Addicted to Fun and Learning
Joined
Mar 12, 2019
Messages
903
Likes
1,861
Location
NZ
Good point - I forgot peak vs continuous :facepalm:. However there's still no denying any device appliance with 0.6 power factor throughout its entire operating load is less efficient than the same device with 0.95, at least from the power retailers p.o.v. afaik, we only pay for Watts, not Volt Amps, so why would we care. Here they fine (charge more p/a) large installations ~500A+ for not using PFC at the main boards. (probably in Aus too, @restorer-john ?)
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,280
Location
Gold Coast, Queensland, Australia
probably in Aus too, @restorer-john ?

I honestly don't know what is done in regard to PFC in Australia and what is allowed/not allowed.

 

Head_Unit

Major Contributor
Forum Donor
Joined
Aug 27, 2018
Messages
1,341
Likes
688
Tell you also what you don't see on their SMPS - massive output caps.
Yeah, to second what @fluid said I once attended an AES talk in the early days of Class D. I forget who the engineer was but to some questions in the audience he said "the math says this is how much capacitor you need, more is not useful" due to the high frequency recharge.

I gotta say this thread is SO interesting, upending "commonsense" assumptions that high frequency supplies would radiate more than old faithful types.
 

Piere

Active Member
Joined
Feb 24, 2022
Messages
195
Likes
190
If you look at the SMPS especially designed for audio amps like those from Hypex, you will see that the energy storage of massive output capacitors is moved to the primary side! But that storage is still there. That way with E= 1/2 x C x U^2 you can do with a lot smaller capacitors for the same amount of energy storage. Now the peak currents has to be switched by the converter! With a proper design that is not a problem as Hypex proves! The output capacitors can be quite small now, but still they need to carry heavy rms currents. That is a main concern regarding reliability.
 

Sokel

Master Contributor
Joined
Sep 8, 2021
Messages
5,840
Likes
5,775
The 20.000 μF that the SMPS has now is not small amount of capacity but is certainly a lot less than the 100.000 μF my previous amp had and I'm sure is exactly what is needed.
The only thing I regret is that I never measured the energy consumption in the old house and I have nothing to compare now.
 
OP
McFly

McFly

Addicted to Fun and Learning
Joined
Mar 12, 2019
Messages
903
Likes
1,861
Location
NZ
On the rather contrary to my thread, I picked up a lowly, beat-up-ass, perreaux pmf1550 last week. MOSFET output class AB 150wpc - Dated 1986. Looks like it had done flat/garage party duty at full noise most of its life. It had massive distortion and and oscillation in one channel. Has signs of being hot all through it.

I replaced the zobel capacitors and a couple driver transistors that really low gain.

Imagine my surprise (after all my shade thrown at transformers) when I checked distortion and it measures better than any class AB-linear-power-supplied amp measured here at ASR? You certainly can get linear power supplies to be quiet.

right channel, 5W, 4 ohms 1khz. (left channel was -94db) - might even have been better if it wasn't for the measurement artifact I get at 16khz. (shows up in all my testing - think its inherent in my laptop)

pmf1550 5w 8r a.jpg



Thats 95dB SINAD.
 

Piere

Active Member
Joined
Feb 24, 2022
Messages
195
Likes
190
Nice! I have revised several 3370's and 3400's (basically the same amps). Most had also degraded input transistors, causing excessive noise, oscillating and DC offset. Replacing them also with new (matched/paired) MPSA42's helps also a lot! I usually replace the standard bridge rectifier with a Schottky one (IXYS from Mouser) and bypass each diode with a 100nF PP cap. This helped to silence mains noise at upper high's further. Next point: The feedback lower resistor is AC grounded through a electrolytic cap bypassed with a MKT cap. I replaced them with a Panasonic bipolar electrolytic and a Panasonic PP. Appearently it improved resolution, but ok, that's highly subjective :D At the end measured distortion was 10-20 dB lower than stated in the manual.:D
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,497
Location
Monument, CO
Very nice, though I don't think it beats Benchmark. Back when I worked at an audio store, as Perreaux delivered their first shipment, they took one amp out of the box and dropped it from the loading dock to the ground (~5'). Took it in, hooked it up, worked fine, and met all specs on the bench. I still have a fond spot for them... :)
 

Piere

Active Member
Joined
Feb 24, 2022
Messages
195
Likes
190
No regarding SINAD it will not beat Benchmark me think. But hooked up to a good source and resolving loudspeakers it still can cope with it today IMHO. And presents you a pleasant soundstage that gives to the music the room it deserves. Few weeks ago I listened to a 3370 in such a set-up and presented me that "holographic" image you seldom hear. SINAD measured at a standard resistive load and at steady level is just a limited qualifiying parameter IMHO, although for sure a valuable one.
 
Top Bottom