DIY Audio Projects Forum
http://diyaudioprojects.com/Forum/

HIFI power amp discussions
http://diyaudioprojects.com/Forum/viewtopic.php?f=10&t=6215
Page 11 of 16

Author:  Yungman [ 24 Sep 2018, 02:11 ]
Post subject:  Re: HIFI power amp discussions

ILoveHiFi wrote:
Cause current mirror load on input stage greatly increases your openloop gain thats why the distortion is lower and making yours sound better than with out.

No current mirror load gain is less than 100, mostly less than 50.
Current mirror load and darington buffed vas easily go past 200x gain on input stage.


I have plenty of gain with just two stages as the THD spec confirmed it already. Also, loop gain is NOT the only thing determining THD. the THD of each stage matters. You can have lower loop gain with more linear stage and get better result.

What kind of speakers you have? If you get into hifi amp design, you want to have a very good pair of speakers. Or else you won't be able to tell. I am thinking about buying a better pair, I feel I am pushing beyond my current pair of speakers. It's going to cost an arm and a leg even used. I have a pair of JM Lab Focal Spectral 913.1. It was retail at $5500 in 1998. I bought it as demo in the showroom. I am thinking about buying a pair of JM LAB Utopia used. Cheapest on ebay is $5500!!! That's going to hurt.

Author:  ILoveHiFi [ 24 Sep 2018, 02:17 ]
Post subject:  Re: HIFI power amp discussions

Nothing too fancy, just quality cheapo ones like slik dome tweeters and full range 8 inch crossed over at higher frequencies.
15inch is quite fancy speaker box maze one with high sensitivity and high frequency response 5khz crosed over well.
Very flat and true frequency response I can hear because cross over system had been well done with erse pulse x caps and high premiativity ferrite rings so less turns required for inductor less resistance added.

Author:  ILoveHiFi [ 24 Sep 2018, 03:11 ]
Post subject:  Re: HIFI power amp discussions

As mentioned before I listen to headphones mainly because dad rents the other half of the house out, getting complaints cause I like loud all day.

I did listen to expensive modern headphones didn't like it quality was poor.
I like old paper cone headphones like 3 inch ones. I have jokumi te-1025, prinzsound 909vs and hp880 pro signal.

HP-880 is 3 inch cheapo you can buy from element 14 but my favourite, best vocals slightly lacking on highs but overall best sound. Orignal wires are cheap and major draw back to sound quality we replaced with awg 16for gnd and awg 18 for L R wire, running from left to right wires are awg20 which is max supported size without any physical mods.

We have zobel network on both 909vs and hp880 since it flatens frequency resonse increases mid and high freueqncy sound quality and extends frequrnecy response slightly. We used erse pulse x caps for zobel.

909vs is extended on both high and lows good frequency response also larger 3.5inch speaker with tripple fold paper conesides compared to hp880 only doubble fold sides and 3inch.
909vs has poor midrange and vocals are more harsh and rough, hp880 is warm realstic low distortion and good. 909vs also got oversied wire mods since original wire broke.

te-1025 uses 3inch paper cone but plastic coneside less extended on base but highs are as extended as 909vs. te-1025 is like downgraded version of 909vs in all areas

Author:  Yungman [ 29 Sep 2018, 15:09 ]
Post subject:  Re: HIFI power amp discussions

I have an interesting observation. I am bringing up a new amp pcb, on one, did to tighten the screws that mount the transistors onto the heat sink, I forgot. The current went up high. After that, I notice one transistor has Vbe about 20mV lower than the others, It is hocking current. I had to change it out.

I match transistors by the set, I usually get matching to within less than 2mV between all 9 transistor on each side. I change that transistor to a new one ( I have a spare for each set), everything is ok.

Then last night, I accidentally shorted the emitter of one of the big transistor to rail voltage. It cause a large spark. I measure the Vbe of all the transistor. The one I shorted has about 3mV lower than the rest and is drawing more current.

Two seems to form a patent. Just an interesting observation.

Author:  ILoveHiFi [ 29 Sep 2018, 21:36 ]
Post subject:  Re: HIFI power amp discussions

That is interesting to see vbe lowering.
I think I can get a patent for my designs because its unique and I don't think anyone has done it.

Author:  Yungman [ 30 Sep 2018, 01:30 ]
Post subject:  Re: HIFI power amp discussions

I wrote the wrong word, I meant two is a pattern that when the transistor is stressed, seems like the Vbe gets lower. The loose screw might have cook the transistor really bad. I don't think the one arc damage the transistor too much, so it's only like 2 or 3mV difference. But I am going to change it out.

There's no patent to talk about on this!!!

Be careful, circuit by itself is very hard to patent and enforce the patent, it has to be a new concept, not just a wisely design circuit. Or else, I would have a lot of patents!!! Here is a patent I wrote and granted https://patents.google.com/patent/US8704074B1/en?assignee=yungman+alan+liu&oq=yungman+alan+liu

When I first submitted, they rejected as they don't care about the circuit, there is another circuit that use voltage summing to cancel the noise. I have to rewrote the concept that my idea is injecting the noise cancelling current into the pickup, the reactance of the pickup transform the current into the noise cancelling voltage to cancel the noise. Basically the concept is a Transconductance amplifier. That I got the patent. Also, even if the battery goes dead in the middle of the performance, the guitar still can function, only there is no noise cancellation. This is something the patent can be granted.

It's the concept, not the circuit. It's like a circuit that can do the same job with 4 transistors where other people have to have 6 transistors, is going to be very hard to get a patent.

Author:  Yungman [ 30 Sep 2018, 15:03 ]
Post subject:  Re: HIFI power amp discussions

I don't mean to discourage you on your patent. You just need to do a lot of research, not just existing product on the market. You need to go on Google Patents to do search. It's time consuming, you have to find the right place where your circuit belongs, then look through any patent that sounds like yours. You'll be surprise so many patents that did not make it to real product. Like my patent, I never even go out of my way to market it as there's no money in the field, and I have to do a lot of marketing, go to conventions, spending thousands of dollars to get a booth in the convention to do demo.......I am just too old for that. I just want to get the patent so I can hang on the wall!!! Just like no matter what I invent in hifi and produce great amps, I have no intention of marketing them. Market too small, too much marketing and sells. All the money is earned by big companies like Pioneer, Onkyo, Yamaha that produce crappy amps for cheap. I seriously doubt Nelson Pass is making boat load of money. Ask how many people even know the names of Krell, Mark Levinson, Pass Lab etc. I bet most don't even know these names.

Author:  ILoveHiFi [ 06 Oct 2018, 16:01 ]
Post subject:  Re: HIFI power amp discussions

Thanks for the tips

http://www.onsemi.com/pub/Collateral/KSD882-D.pdf
I brought some KSD882 and KSD772 latley geniune from rs components, only $36.5nzd. Fairly high frequency response excess of 60MHz when biased with enoug current going to be perfect for head amps. Unfortunatley the better ones with higher frequency response are not avalible so I didn't buy any.

FT of 50MHz should be enough so the bottle necking of higher frequency response is limited by op amp schematic not transistors

Author:  ILoveHiFi [ 06 Oct 2018, 17:00 ]
Post subject:  Re: HIFI power amp discussions

36.5nzd for 60 pairs

Author:  Yungman [ 07 Oct 2018, 01:44 ]
Post subject:  Re: HIFI power amp discussions

Check out these:

https://www.digikey.com/product-detail/en/on-semiconductor/KSC3503DS/KSC3503DS-ND/1049763
https://www.digikey.com/product-detail/en/on-semiconductor/KSA1381ESTU/KSA1381ESTU-ND/1050068

These are the fastest pre-drivers I can find so far. 150MHz and 300V. You need the high voltage for higher power amps. Notice the Cob is very low.

Fast transistors are not for high speed, that's only the by product. More important is to kick the poles high enough to get it out of the stability picture.

Check out the datasheet of the transistors I use, I spent a lot of time choosing them. For the ft, voltage, Cob and beta. If you can find better ones, let me know. The new amp I am working, I get a little over 40V/uS slew rate of 40V swing driving 4ohm.

Page 11 of 16 All times are UTC - 6 hours [ DST ]
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/