DIY Audio Projects Forum

HIFI power amp discussions
Page 1 of 17

Author:  Yungman [ 08 May 2018, 22:43 ]
Post subject:  HIFI power amp discussions

I want to start a new thread of my own to continue discussion of SS power amp. I don't want to further hijack the other thread. I just continue from this thread:

This is an open discussion, any idea or opinion are welcome, just let the discussion led wherever it goes.



Author:  Yungman [ 08 May 2018, 23:11 ]
Post subject:  Re: HIFI power amp discussions

Hi Kochiya

I have a theory about why low THD give clearer sound, better separation, perceive quieter, clear and all. Don't laugh, I just throw this out what is in my mind from engineering point of view.

I am looking at this in the point of view of intermod IM. We all know one important test is two tone test when we look at the f1+/-f2, 2f1+/-f2, f1+/-2f2 etc etc.

Think of music/tv/movie, there are a lot of frequencies going on at the same time. For an amp with very low THD, the harmonics are very low, so the amplitude of the IMs are very low. Compare to an amp with higher THD, imaging at any given time, there are 100 frequencies being amplified. Each produce a few harmonics. The frequencies and harmonics start mixing together, the harmonics start mixing with the fundamentals and their harmonics. You literally create a noise floor correlating with the original sound. Things get muddier, it becomes a constant background.

Those IM being correlated with the original signals smear and blur the original sound. You loss transparency, separation. the noise floor ruin the clarity.

We are talking about at least mid fi to hifi, all the treble, bass should already be good enough, what make a difference to me from an ordinary system to a good system is the sense of openness, 3D, transparency. The feel of "biggness" from the lack of words. The feeling of very quiet and the detail stands out.

As I said, I yet to design and build a good tube amp to confirm or invalidate my believes, I built a SE tube amp, that was a disaster. I think it's because of my speakers. SE amp needs high efficiency speakers like Altec, Klipsch type of speaker. I never heard an Altec, I don't like Klipsch. So I decided to do a push pull tube amp. I have a rough design already, I am just having too many ideas, this will be 3rd in line and won't get to it at least till next year.

My next amp is error correction design on the OPS to eliminate crossover distortion. I like my first two amps, problem is it burn too much power. I run 1A to 1.1A per channel, it gets very hot really fast. I am running +/-33V supply, that gives me only 100W @4ohm or 50W @8ohm. I want to try to get the same quality but either getting higher power with the given chassis or a smaller chassis with less output transistor pairs. I just got the add on pcb back to attach to the existing OPS. I think this will take up a lot of time on this one. So everything else has to wait. Here are some pictures of my amp during assembling. You can see it's not a small chassis by any measure, it's only 100W @4ohm. It's getting pretty toasty.

Finished amp modules 1.jpg

ADD IPS filter 2.JPG

Combine front back and heatsinks.JPG

Author:  ILoveHiFi [ 09 May 2018, 01:26 ]
Post subject:  Re: HIFI power amp discussions

I think your very right about your theory on why lower THD sounds better.

Perosnally I'm not too keen on a valve amp, I've designed many valve pres and it does the job getting me the sound.
I own a sound luster and the sound is just........not up to standards.

Your amp is nice looking nice large heat sink
I know lots of transistors do make a diffrence in sound

Author:  Yungman [ 09 May 2018, 04:41 ]
Post subject:  Re: HIFI power amp discussions

Nice. You think valve preamp make a difference? One of my project is to design a hybrid power amp with tube IPS, this is my project after the error correction amp. Currently my preamp is just my simple design with an opamp buffering the volume pot to drive the power amp. I bought the remote volume control and input select from ebay and incorporate into my preamp.

How do you describe the tube preamp sound?

I like your little scope, I have the old CRT Tektronix 2465A scope. I needed it for my contract that worked on high speed circuits, so I need a 350MHz scope. It's big. My work space is small, real estate is of premium, I sure can use a small one like yours. But my scope is working great, I am too cheap to spend money to buy another one.

Author:  ILoveHiFi [ 09 May 2018, 05:37 ]
Post subject:  Re: HIFI power amp discussions

Well I'd say with standard or mid end valve pre amps, I'd rather go direct input. The distortion is too high to be sounding anygood. You can buy for low cost like martnz 7 from china sound is poor.

In picture shown its a valve only pre amp, fairly low distortion with plenty of valve sound.
Not in picture which I normally listen to is a valve pre amp hybrid, 250v regulated while having 20 something transitors per channel to get high gain and low distotrtion. All directly couppled.
Main goal was to build a extremely low distorion valve pre amp to see weither or not valve sound is distortion. It turned out that valve sound is not distorion.
I am getting plently of valve sound and warmth, while having all details and staying true to music, a little valvy twist that makes music good.

My signal generator is some chinese digital generator and the scope is 50mhz again some unkown budget chinese brand, budget but has FFT. Both had worked reliably so I am happy with it.

I've designed many valve pre amps from scratch I think total of 6+ pre amps.
Also very exprienced with designing zero feedback or single ended non op amp type pre amps using transitors running on 150-200v I've built 4 of these.

Amp schematics done 10+ already. Started off building things everyone done, modified things one by one per time. Ended up learning and comming up with entierly new schematics.
Actually started off with desinging just simple basic transitor stages put in series to build a few amps, got confident then went to discrete op amp type amplifiers.

I've also looked at online turorials on how things were done on the way I go, especailly op amp tutorials, without these my DIY audio world would be colourless today.

Author:  Yungman [ 09 May 2018, 19:57 ]
Post subject:  Re: HIFI power amp discussions

You use over 20 transistors per channel together with the tube for a preamp? Do you mind sharing your schematic on the hybrid preamp? I really have not look into preamp much. I designed one, but mainly the switching matrix to mux 2channels and the video from 3 different sources. The video is just mux, no electronics involved, just relays control by the remote module I bought on ebay. The muxed two audio channels go to the motorized volume pot at the front panel. I just use an OPA604 to buffer the volume pot to drive the coax to the power amp. Very simple. Here are pictures of my preamp

Preamp 1.JPG

Preamp 2.JPG

You can see the preamp pcb is very small, just a buffer with like +1dB at below 500Hz to give it a little warmth.

Ha ha, you build your preamps in a baking pan? My big boss will never let me get away with a baking pan in the living room. The chassis is most the cost of the DIY project!!!

All my effort are on power amp as I can get opamp for preamp below 0.001%THD. I never copy someone else circuit. To be honest, I have not seen one that I want to copy yet. Most of the circuits of power amps are quite similar. Differential IPS, VAS and OPS. Most of the commercial amps are simple 2EF OPS. The famous Krell KSA design is actually very primitive and it's not even that good in my book. The difference is they run in class with over 10 pairs of output transistors. Design wise, it's really backward. So is Mark Levingson.

My design is Blameless front end but with current mirror load on top. Then drive a darlington VAS. I design 3EF output stage similar to the Diamon driver described in Cordell's book (p193), but I did not do the complete diamond. I did the first pcb run with diamond, I found it is not stable because of signal feedback. I just connect the collector of the pre-driver stage to the rail to make I called a complementary 3EF ( PNP-NPN-NPN or the top and NPN-PNP-PNP on the bottom).

I have the complementary pre-driver and driver transistor on a separate heatsink closely tracking each other to compensate the temp coef of Vbe. The bias spreader bolted on top of one of the big transistor on the main heatsink. I got very good result, the idle current drill no more than 10% from first power on the normal temperature ( I would say about 60deg C, it's warm). I can post the schematic if you are interested.

Author:  ILoveHiFi [ 09 May 2018, 21:42 ]
Post subject:  Re: HIFI power amp discussions

I'm poor so a baking pan does the job in keeping my hifi projects in one pice. Not excatly safe but I never had any fatail accidents.
I'm a uni dude and my parrents dosen't want to spend money on me.

I'm sorry I have to say, I can't share my amplifiers because they sound good and I want to hopefully be able to sell them in the future and make money hopefully.

Baking pans allow me to make many projects at virtually no or very little cost, its not very safe but not fouls or children live in my house so thats ok. This allowed me to test more schematics and build more amps for experience.

If you wanna have a go at high voltage hybird pre amps, MPSA42 and 92 from aliexpress are the things I use, I've brought 200 pairs from two diffrent sellers and tested all of them to be able to operate on 250V and 400mW without breaking down, which means real deal.
I sometimes use TO-220 guys for pre output, I got them on rs components because aliexpress lacks high voltage audio to-220 and the chances of getting fake..... ... 0.0.E2ez1v ... 0.0.E2ez1v

Author:  Yungman [ 09 May 2018, 21:54 ]
Post subject:  Re: HIFI power amp discussions

I see, I guess you are still very young. I already retired. I am not interested in forming my own business anymore. To me, it's just fun. I had my 30 years of EE already, if I want to start something, I'll stay working. More money in my job than trying to sell amplifiers.

Are you working? What do you do?

I understand you don't want to share. Even when I have a good idea, I would try patent it even if I don't have plan to market it. I obtain a US patent in guitar noise cancelling 4 years ago and I did not really market it. Just nice to have one more plaque hanging on the wall.

Are you into power amps? It's a lot more challenging to me. Lots of mechanical work on top of electrical. Cooling, transistor matching ( critical), power supply, grounding layout. So much more than just the amplifier circuit.

Author:  ILoveHiFi [ 09 May 2018, 23:46 ]
Post subject:  Re: HIFI power amp discussions

I was into power amps however neighbours keep me away from it.

To me power amps is same as a head amp, just more output transitors and more money along with using the same gain stage.
I'd say power amps are more time consuming thats another reason why I try not to build them
My head amps have to drive 8 ohm headphones so it can also drive speakers.

It takes me arround a day to build a new schematic from scratch get it tested and fine tuned if everything turns out allright first time. This means a mono channel either head amp or pre amp is built for testing purposes

Author:  Yungman [ 10 May 2018, 01:22 ]
Post subject:  Re: HIFI power amp discussions

Not really, I don't think you can compare a headphone amp to a power amp. Headphone amp is low power, that make it much easier to design. Transistor selections are much easier, taming is a lot easier. Particular if you use over 4 pairs of output transistors, you have to worry about interaction between transistors. Parasitic capacitance is a whole lot higher for high current circuits.

IPS and VAS is the easy part, the OPS is everything. All the subtleties in the design. I used to think how hard can an audio amp be having designed much more advanced state of the arts RF, ultra low noise, even designing analog ICs. I can assure you there are a lot that goes into power amps. There is no comparison. of small signal vs power output stage.

All the difficult stability problems I faced are from the OPS, none in the IPS and VAS. Distortion mainly from OPS, not IPS and VAS.

For low power amps, 1 pair of TO220 as output transistor running in class A, and you are good. Easy to find transistor over 100MHz, Try the high power amp, it's a different world. Just choosing the transistors takes a lot of effort.

Actually, a lot of people prefer passive preamp, just a volume pot driving the power amp. That's it!!! Problem is to avoid capacitance of coax loading down the volume pot, the volume pot needs to be in the power amp. I am actually trying to move things around in the power amp to fit the extra circuit in order to get a total passive preamp. That actually is a challenge as my power amp is full.

Page 1 of 17 All times are UTC - 6 hours [ DST ]
Powered by phpBB® Forum Software © phpBB Group