JLH1969 MOSFET Class A/AB amp

LAOQIANG

能否将ocl版本图上一下。

xiaolu36

偶做了一个声道，在原来的1969上修改，用43Vde电源，上电后中点基本不用调整，电流在200mA左右，不知道功率多大。。等高手帮偶算算

石城

530没听过。要是和540一个样就惨了。

石城

谁能标准地反译一下楼主的文章吗

millwood

用43Vde电源，上电后中点基本不用调整，电流在200mA左右，不知道功率多大。

it is about (43/2-5)^2/8=...

要是和540一个样就惨了。

you must have done something wrong.

xiaolu36

原帖由 石城 于 2009-3-9 17:27 发表
530没听过。要是和540一个样就惨了。

540有啥问题？

xiaolu36

请问millwood ，并管能扩大功率吗？方便的话请打拼音字母。。哈哈，谢谢了！

millwood

原帖由 xiaolu36 于 2009-3-10 09:16 发表
请问millwood ，并管能扩大功率吗？方便的话请打拼音字母。。哈哈，谢谢了！

power dissipation in the output devices is never a problem with this design, as long as the output stage works in Class B/AB.

there is no theoretical limit to how much power you can get out of this design. But practically, 60v is about as high as you can with a MOSFET output stage. that works out to be about (60/2-5)^2/(2*8)=40w rms into an 8ohm load.

at that point, each output device is dissipating about 20 - 30w rms so one to247 or to3p device will more than suffice.

millwood

the limiting factor is the dissipation in the bootstrap network. the power dissipation in the bootstrap network is about Id*Vcc/2, where Id is the idle current for the driver, and is typically 20ma. so at Vcc=38v, the power dissipation is 20ma*38v/2=.38w, so a good 0.5w trimmer will do.

since it is very hard to find a trimmer bigger than 1w, it quickly becomes an impossibility to use higher voltage without burning up the trimmer.

now, if you really want to build a super high voltage JLH, you can replace the bootstrap network with a constant current source. However, in that case, you will need to work on temperature compensation and stability.

xiaolu36

请高手翻译。。。

xiaolu36

有人对irf530提出非议。。。看下面文字

6．使用IRF9530和IRF530场效应管做功率输出管是否合理？

V300所谓的“分立元件电路”是根本没有达到预期效果的，音质实际上还不如集成功放。功率管选型也欠妥，IRF530/9530这类MOS管实际上是开关管而非音频管，仅适合工作在开关状态，根本不适合音频应用。若非要作音频应用时，必须在极高的静态电流下才能获得可接受的音质，也就是说必须工作在纯甲类状态下，但看看V300的散热系统就知道这是不可能的。我拆机时候顺手测量了V300功放板的射级电阻，使用的是3位半的数字表，测量的阻值为0.2欧姆，短路功放输入端，在这个电阻上测量的电压为0.01伏特。换而言之，V300功放的静态电流仅仅只有0.05安培！这样的偏置状况，如何能够奢望IRF530/9530能够出好声？

另外，对于音频系统来说，任何控制电动扬声器的功率放大器，都在追求最小的输出阻抗。V300的设计却反其道而行之，竟然选用了内阻达2欧姆左右的场效应管，不说这样的内阻在推动4欧姆的扬声器时候要白白浪费多少输出功率，这也使得本来就是中功率工业用管的IRF9530/IRF530驱动扬声器更加的无力，几乎处于失控的边缘。

另外偏流调整管竟然没有贴在散热片上，会不会造成稳定性不好？要知道工业场效应管也是正温度系数的！而且在其中一套产品中我们发现末级4个功率管都不一样，其中有一个是IRF9630（注意：实际应该是IRF9530），还有一个IRF530外形也不一样，可见批次完全不同。三级管批次不同，特性就会有变化，何况三极管还要配对使用最好，就更不用说型号不一样了，这可是影响音质的大忌。

oldboy

Dear Millwood,
I just found this thread and it is so long to fiinish !

Allow me to have some questions since I havn't go through it !

Instantly, it make me think about :

1) I have a lot of K135, K405 in my drawer that bought 18 years ago. Can I use it in 1969 ?
What componts/values should be changed ?

2) If just compare between using 2N3055 & MJ 15003, will it make a different ? I have look at their data sheet. Both their FT=2Mhz

3) What is your comments if use PNP power transiitor ?

Thank you for help for the LAZY MAN !

millwood

oldboy:

1) I never tried those transistors myself. from a quick search, 2sk135 looks like a lateral-MOSFET. Those are great MOSFETs, with much lower Vgs(threshold) than the IRF types. so I would say that the MOSFETs you have are too good for the JLH1969M. Having said that, I don't see a reason that they wouldn't work. You may need to play with the amp a little bit to account for their lower Vgs.

2) MJ15003 should work. They have built-in bias resistors so they behave actually much like MOSFETs than BJTs in this amp. there are some discussions on that lately so you may want to search for those posts.

3) PNP power transistors tend to have better beta linearity so everything else being equal, they should perform better in terms of distortion / open loop frequency responses. However, they also tend to have smaller safe operating area so they are not as good for high power applications. Again, check out some recent discussions.

millwood

IRF530/9530这类MOS管实际上是开关管而非音频管，仅适合工作在开关状态，根本不适合音频应用。

just because something is designed for A doesn't mean they are not good for something else.

MOSFETs, surprisingly, were originally designed for audio applications - being a voltage device makes them easier to drive.

today, most MOSFETs work as switchers. does that mean that MOSFETs should be used for switching?

V300功放的静态电流仅仅只有0.05安培！这样的偏置状况，如何能够奢望IRF530/9530能够出好声？

there is no correlation between high idle current and good sound quality, once the transistors start to conduct.

50ma is a respectable level of idle current for most applications.

oldboy

原帖由 millwood 于 2009-3-19 02:19 发表
oldboy:

1) I never tried those transistors myself. from a quick search, 2sk135 looks like a lateral-MOSFET. Those are great MOSFETs, with much lower Vgs(threshold) than the IRF types. so I wou ...

It is too impatient to wait, where I can find the latest ver. of JLH 1969M ?

Thanks !

millwood

oldboy: they are all roughly the same and anyone of them will be fine.

please let us know how it goes. thanks.

oldboy

原帖由 millwood 于 2009-3-19 04:52 发表
oldboy: they are all roughly the same and anyone of them will be fine.

please let us know how it goes. thanks.

Is it just direct replace the 2N3055 by K405 and only add the 220 gate resistor ?

Thanks !

millwood

oldboy:

1) you should have a gate stopper. However, the l-mosfet has much lower Ciss (600pf according to the datasheet, vs. typically >2200pf for v-mosfet), you have broad room here. I would use 110ohm - 470ohm gate stoppers here.

2) the l-mosfet also have negative temp coefficient at much lower Ids. so you don't have to have the 0.2ohm Re resistors - though having them helps.

3) you will certainly play with the bootstrap network to account for lower Vgs of the l-mosfet (~1v vs. 4v for v-mosfet). so the two resistors will need to drop 3v more at the given current.

millwood

oldboy:

I am looking at the schematic in post #4. for it to work with a l-mosfet, you need to make the following changes to the bootstrap network:

1) R4: with v-mosfet, the current going through R4 is about 4v/220ohm=20ma. to maintain the same current for l-mosfet (Vgs=1v), R4 needs to be 1v/20ma=47ohm.

2) R7/R8 needs to drop 32/2-1v=15v, at 20ma. so R7+R8=15v/20ma=750ohm. R7=R4=47ohm and use a 1k trimmer for R8.

oldboy

Dear Millwood,
Thanks for your help !
I think I should finish reading the Thread ASAP.
My plan is :

Becuae of I have already bought the PCB below :

http://bbs.hifidiy.net/viewthrea ... ;highlight=%2Bxr000

I will solder it but with PNP MJ15004, I have so many in my drawer too.
Use this one as reference to compare i) 1969M BJT ii) 1969M IFR510 iii) Pass F5

My Pass F5 is in half way, have you tried it ?