给初学者的礼物之一：和我一起玩简单好听的P3A功放

筑明

    很久以前就见过这个线路，认为比较简单就没有起心制作，论坛的99990000兄弟的制作贴及推荐让我颇感兴趣，上diyaudio上一翻，爬了几百楼，几乎都是正面的评价，后续对这个线路的改进也很多。
    加入论坛的第二十个年头，在论坛中学到了很多知识和技巧，也看到很多初学者都只做集成块放大器，其实分立元件的功放更有可玩性，所以我找了几个简单好听的线路一一制作调试，方便初学者借鉴参考，或者照猫画虎也会有好的结果。
   
   原著作者为diyaudio的大神：Rod Elliott
   链接：https://sound-au.com/project3a.htm

60-100W Hi-Fi Power Amplifier
© September 2000, Rod Elliott (ESP)
Last Updated Jan 2021
PCB  Please Note:  PCBs are available for the latest revision of this project.  Click the image for details.
Introduction
Update - 25 June 2009 - Although the last update highly recommended the latest OnSemi power and driver transistors, they remain hard to get in most countries.  As a result, the recommended power transistors are now the much more readily available MJL21193/4.  While in theory these are not quite as good as the latest versions, they are still excellent devices.  It is extremely doubtful that anyone would ever pick any difference with test instruments, and there will be no change that is audible.

Much the same applies to the driver transistors.  Although the BD139/140 devices are not considered to be the 'finest' audio transistors available, they work very well indeed, and have been used in all of the P3A amps I've built for my own use or as part of a system.  Again, it is highly unlikely that there will be any meaningful measurement that will show these transistors to be inferior to 'audiophile' parts.

24 Jul 2003. - OnSemi released a new range of transistors, designed specifically for audio applications.  These new transistors have been tested in the P3A, and give excellent results.  As a result, all previous recommendations for output transistors are superseded, and the new transistors should be used ... if you can get them.  Several years after release, the new devices may still not be readily available.

The output devices are MJL4281A (NPN) and MJL4302A (PNP), and feature high bandwidth, excellent SOA (safe operating area), high linearity and high gain, however they are not inexpensive.  Another output transistor option is the 200W NJW3281 (NPN) and NJW1302 (PNP).  These are more economical, but still have excellent specifications.  Driver transistors are MJE15034 (NPN) and MJE15035 (PNP).  All devices are rated for at least 250V, with the power transistors having a 230W dissipation and the drivers are 50W.

For a budget system, you can use TIP35C (NPN) and TIP36C (PNP) output transistors.  If you can get the 'full-pack' TO-247 case versions they can be mounted under the board in the same way as the MJL devices.  These are limited to 125W dissipation (25°C case temperature), but despite that apparent limitation they can still drive a 4 ohm load from ±35V supplies.  In theory the peak dissipation may be exceeded, but these are extremely rugged transistors and handle abuse with ease.  Don't push your luck though - the maximum unloaded supply voltage is ±35V!

note        Note that there is a major reason that P3A is different from most amp projects you will see on the Net - it uses complementary feedback pairs (aka Sziklai pairs) for the output stage, and quiescent current is controlled by the driver transistors.  If the bias servo is mounted on the heatsink, it will provide over-compensation and crossover distortion will result.
The basis for this amplifier was originally published as Project 03, and although the base design is almost 30 years old, as an amplifier it remains 'state of the art' - this is an extremely good amplifier.  It is simple to build, uses commonly available parts and is stable and reliable.  The design featured is a full update on the original project, and although it has many similarities, is really a new design.

This new amp (like the original) is based on an amp I originally designed many years ago, of which hundreds were built.  Most were operated as small PA or instrument amps, but many also found their way into home hi-fi systems.  The amp is perfectly capable of driving 4 Ohms, provided the supply voltage is maintained at no more than ±35V.

This amplifier, although very simple, is capable of superb performance.  This is not an amp to be under estimated, as the sound is very good indeed, and this is due (in part, at least) to the inherent simplicity of the design.  The amp is exceptionally quiet, and is reasonably tolerant of difficult loads.  It is an ideal amplifier for biamped systems, and may be operated in bridge mode (BTL) if you use the recommended output transistors (which have the necessary power ratings).

The design has had the benefit of many, many years of consistent use, and this version is the best of all - the refinements ensure minimum 'switch-on' or 'switch-off' noise, and the availability of really good output devices has improved on a known and very stable design.

I have heard nothing but praise from those who have built this amplifier - all feedback I have received has been very positive indeed.  It is highly recommended, based on the reports from countless people who have built it.  Considering that the design has been available for 20 years with almost no changes (other than output transistor changes due to availability), it's safe to say that it has stood the test of time.  It makes no pretense at being 'better' than anything else, but the results of several thousand constructors is testament to its inherent reliability and sound quality.

60-100W Hi-Fi 功率放大器
© 2000 年 9 月，罗德-艾略特（ESP）
最后更新于 2021 年 1 月
印刷电路板 请注意：本项目的最新修订版可提供印刷电路板。 点击图片了解详情。
简介
更新 - 2009 年 6 月 25 日 - 虽然上次更新强烈推荐使用最新的 OnSemi 功率和驱动晶体管，但在大多数国家仍然很难买到。 因此，现在推荐的功率晶体管是更容易买到的 MJL21193/4。 虽然从理论上讲，这些晶体管的性能不如最新版本，但它们仍然是非常出色的器件。 我们非常怀疑是否有人能用测试仪器检测出它们有什么不同，而且也听不出有什么变化。

驱动晶体管也是如此。 虽然 BD139/140 器件并不被认为是目前 "最好的 "音频晶体管，但它们的工作性能确实非常好，我自己制作的所有 P3A 放大器都使用了它们，或将其作为系统的一部分。 同样，任何有意义的测量都不太可能表明这些晶体管不如 "发烧级 "部件。

2003 年 7 月 24 日 - OnSemi 公司发布了一系列专为音频应用设计的新型晶体管。 这些新晶体管已在 P3A 中进行了测试，结果非常出色。 因此，以前关于输出晶体管的所有建议都被取代，应该使用新型晶体管......如果你能买到的话。 在发布数年后，新器件可能仍然不易买到。

输出器件为 MJL4281A（NPN）和 MJL4302A（PNP），具有高带宽、出色的 SOA（安全工作区）、高线性度和高增益等特点，但价格并不便宜。 另一种输出晶体管是 200W NJW3281（NPN）和 NJW1302（PNP）。 这些晶体管更为经济，但仍具有出色的规格。 驱动晶体管是 MJE15034（NPN）和 MJE15035（PNP）。 所有器件的额定电压至少为 250V，功率晶体管的耗散功率为 230W，驱动器的耗散功率为 50W。

对于经济型系统，可以使用 TIP35C（NPN）和 TIP36C（PNP）输出晶体管。 如果能买到 "全封装 "TO-247 外壳版本的晶体管，它们就能以与 MJL 器件相同的方式安装在电路板下。 这些器件的耗散功率被限制在 125W（25°C 外壳温度），但尽管有明显的限制，它们仍然可以在 ±35V 电源下驱动 4 欧姆的负载。 理论上，可能会超过峰值耗散，但这些晶体管非常坚固耐用，可以轻松应对滥用。 不过，千万不要铤而走险--最大空载电源电压为 ±35V！

注意 注意 P3A 与您在网上看到的大多数放大器项目不同的一个主要原因是，它的输出级使用互补反馈对（又称 Sziklai 对），静态电流由驱动晶体管控制。 如果偏置伺服器安装在散热片上，就会产生过度补偿，导致分频失真。
这款放大器的基础设计最初是作为项目 03 发布的，虽然基础设计已有近 30 年的历史，但作为一款放大器，它仍然是 "最先进的"--这是一款非常出色的放大器。 它制作简单，使用的零件很常见，而且稳定可靠。 这次介绍的设计是对原始项目的全面更新，虽然有许多相似之处，但实际上是一种全新的设计。

这款新音箱（与原版一样）基于我多年前设计的一款音箱，当时制造了数百台。 其中大多数被用作小型扩音机或乐器放大器，但也有许多被用于家庭高保真音响系统。 只要电源电压不超过 ±35V，该放大器完全能够驱动 4 欧姆。

这款功放虽然非常简单，但却具有卓越的性能。 这款功放的音质确实非常好，这（至少部分）归功于其固有的简单设计。 该放大器异常安静，对困难负载的承受能力也相当强。 它是双功放系统的理想放大器，如果使用推荐的输出晶体管（具有必要的额定功率），还可以在桥接模式（BTL）下工作。

多年来，该设计一直在不断改进，现在的版本是所有版本中最好的--这些改进确保将 "接通 "或 "断开 "噪声降至最低，而真正优秀的输出设备的出现又改进了已知的、非常稳定的设计。

我从制造过这款放大器的人那里听到的都是赞美之词--我收到的所有反馈都非常积极。 根据无数制造过它的人的报告，我们强烈推荐这款放大器。 考虑到这种设计已经问世 20 年，几乎没有发生过任何变化（除了输出晶体管因可用性而发生变化），可以说它经受住了时间的考验。 它没有任何 "比其他产品更好 "的噱头，但它的结果却让人信服。

The photo above shows how the board mounts to the heatsink and clamps the output transistors beneath the PCB.  This is the latest Revision-C version of the PCB, but the general arrangement hasn't changed very much over the years.  It's always been possible to cut the topmost section of the PCB so the output transistors can be mounted vertically, and the board can also be cut in half so each amp can be on its own heatsink or even in a separate enclosure.  You may notice that the photo shows 0.22 ohm emitter resistors and the schematic shows 0.33 ohms.  The higher value is recommended as it improves bias stability.

For those who want to build a Class-A power amp, see Project 3B.  It is virtually identical to the design shown here, but needs a far more robust power supply to accommodate the high quiescent current.  Although most people seem to think that Class-A is 'better' than Class-AB, it's not really the case.  P3B has performance that's really no better than the P3A, but has far less power and develops a great deal more heat.
上图显示了电路板如何安装到散热片上，并将输出晶体管夹在 PCB 下面。 这是印刷电路板的最新修订版-C，但多年来总体布置没有太大变化。 一直以来，我们都可以切割 PCB 的最顶端部分，以便垂直安装输出晶体管，也可以将电路板切成两半，这样每个放大器都可以安装在自己的散热片上，甚至可以安装在单独的机箱中。 您可能会注意到，照片上显示的发射极电阻为 0.22 欧姆，而原理图上显示的电阻为 0.33 欧姆。 建议使用较大的电阻值，因为它能提高偏置稳定性。

Description
Note that like the original, there is (still) no output short circuit protection, so if speaker leads are shorted while the amp is working with a signal, there is a very real risk of the transistors being destroyed.  I suggest and recommend the use of Speakon connectors both at the amplifier and speaker ends.  The specifications are very similar to those of the original project, but the use of a current sink in the differential pair input stage means that there is virtually no thump at turn on or off.
说明
请注意，与原版一样，该产品（仍然）没有输出短路保护功能，因此如果在放大器工作时扬声器引线被短路，晶体管很有可能被毁坏。 我建议并推荐在放大器和扬声器两端使用 Speakon 接头。 产品规格与最初的项目非常相似，但在差分对输入级中使用了一个电流沉，这意味着在开启或关闭时几乎不会产生砰砰声。
I have also added the ability to adjust the quiescent current, and with the transistors specified the amp will provide 100W into 8 ohms, at a maximum supply voltage of ±42V.  This supply is easily obtained from a 30-0-30V transformer.  Consider that increasing the supply voltage from 35V to 42V represents an output power increase of only about 1.6dB, but the potential for output transistor damage is almost doubled.  IMO, it's just not worth it.Description
Note that like the original, there is (still) no output short circuit protection, so if speaker leads are shorted while the amp is working with a signal, there is a very real risk of the transistors being destroyed.  I suggest and recommend the use of Speakon connectors both at the amplifier and speaker ends.  The specifications are very similar to those of the original project, but the use of a current sink in the differential pair input stage means that there is virtually no thump at turn on or off.

我还增加了调节静态电流的功能，使用指定的晶体管，放大器可在最大电源电压为 ±42V 时为 8 欧姆提供 100W 功率。 这个电源可以通过 30-0-30V 变压器轻松获得。 考虑到将电源电压从 35V 提高到 42V，输出功率只增加了约 1.6dB，但输出晶体管损坏的可能性却几乎增加了一倍。 在我看来，这根本不值得。


As can be seen, it is not a complex amp, but the performance is excellent.

For use into 4 ohms (including bridging into 8 ohm loads), do not exceed ±35V (from a 25-0-25V transformer).  Most applications will be satisfied with the lower voltage, and the reliability of the amp is assured with almost any load.  In bridge mode, this amp will happily produce 200W into 8 ohms, and will do so reliably even for continuous high power levels.  Never attempt to operate the amp in bridge mode into 4 ohms, as this represents an equivalent load to each amp of 2 ohms.  The amp was not designed to handle this, and will fail.  ±42V is the absolute maximum voltage, and should only be used where 4 ohm loads will never be applied.

D1 is a green LED, and should be a standard type.  Don't use a high brightness LED, or change the colour.  This is not for appearance (although the green LED looks pretty neat on the board), but for the voltage drop - different coloured LEDs have a slightly different voltage drop.  The aim is to have a voltage across the LED of around 1.9-2V.  This may seem to be on the low side for typical green LEDs, as they are normally rated at 2-2.2V (although some are much higher and cannot be used).  However, a nominal 2.2V LED will have the right voltage across it at low current - only 1.6mA is provided by R8 with a ±35V supply.

VR1 is used to set the quiescent current, and normally this will be about 50-100mA.  The amp will work happily at lower current, but the distortion starts to be noticeable (on a distortion meter monitored by an oscilloscope) at less than around 20mA (the recommended minimum quiescent current).  The Class-A driver (Q4) has a constant current load by virtue of the bootstrap circuit R9, R10 and C5.  Stability is determined by C4, and the value of this cap should not be reduced.  With fast output transistors such as those specified, power bandwidth exceeds 30kHz.

With the suggested and recommended 35V supplies, Q4 and the output drivers (Q5 and Q6) will normally not require a heatsink.  With 4 ohm loads, you may find that a heatsink for Q5 and Q6 is needed, but my experience is that these transistors should not get hot under most operating conditions.

If using the amp at ±42V, a small heatsink should be used for Q4, as the dissipation will be quite a bit higher and the device will get very warm.

Although I have shown MJL4281A and MJL4302A output transistors, these have been available for over 6 years and are still hard to get.  The recommended alternatives are MJL21193 and MJL21194, or NJW3281 (NPN) and NJW1302 (PNP).

Note: It is no longer possible to recommend any Toshiba devices, since they are the most commonly faked transistors of all.  The 2SA1302 and 2SC3281 are now obsolete, and if you do find them, they are almost certainly counterfeit, since Toshiba has not made these devices since around 1999~2000.

Before applying power, make sure that VR1 is set to maximum resistance to get minimum quiescent current.  This is very important, as if set to minimum resistance, the quiescent current will be very high indeed (almost certainly enough to blow the output transistors!).
可以看出，该放大器并不复杂，但性能卓越。

用于 4 欧姆（包括桥接 8 欧姆负载）时，电压不要超过 ±35V（来自 25-0-25V 变压器）。 大多数应用都能满足较低电压的要求，而且几乎任何负载都能保证放大器的可靠性。 在桥接模式下，该放大器可以在 8 欧姆范围内产生 200 瓦的功率，即使在连续大功率情况下也能可靠地工作。 切勿尝试在桥接模式下以 4 欧姆的功率运行放大器，因为这意味着每个放大器的等效负载为 2 欧姆。 放大器并非为处理这种情况而设计，因此会出现故障。 ±42V 是绝对最大电压，只能用于绝对不会使用 4 欧姆负载的场合。

D1 是绿色 LED，应为标准类型。 不要使用高亮度 LED 或改变颜色。 这不是为了外观（虽然绿色 LED 在电路板上看起来很整洁），而是为了压降--不同颜色的 LED 的压降略有不同。 目的是使 LED 两端的电压在 1.9-2V 左右。 这对于典型的绿色 LED 来说似乎偏低，因为它们的额定电压通常为 2-2.2V（尽管有些 LED 的电压更高，但不能使用）。 不过，标称电压为 2.2V 的 LED 在低电流时两端就会有合适的电压 - 在 ±35V 电源下，R8 只提供 1.6mA 电流。

VR1 用于设置静态电流，通常约为 50-100mA。 放大器可在较低电流下正常工作，但在低于 20mA 时（建议的最小静态电流），失真开始明显（通过示波器监测失真表）。 通过自举电路 R9、R10 和 C5，A 类驱动器（Q4）具有恒定电流负载。 稳定性由 C4 决定，该电容的值不应减小。 使用指定的快速输出晶体管时，功率带宽超过 30kHz。

使用建议和推荐的 35V 电源时，Q4 和输出驱动器（Q5 和 Q6）通常不需要散热器。 使用 4 欧姆负载时，您可能会发现 Q5 和 Q6 需要散热片，但根据我的经验，这些晶体管在大多数工作条件下都不会发热。

如果在 ±42V 电压下使用放大器，则应为 Q4 使用小型散热片，因为耗散量会高很多，器件会变得很热。

虽然我已经展示了 MJL4281A 和 MJL4302A 输出晶体管，但这些晶体管已经上市 6 年多，仍然很难买到。 推荐的替代品是 MJL21193 和 MJL21194，或 NJW3281（NPN）和 NJW1302（PNP）。

注：由于东芝器件是最常见的伪造晶体管，因此不再推荐任何东芝器件。 2SA1302 和 2SC3281 现在已经过时，如果你找到它们，几乎肯定是假冒的，因为东芝从 1999~2000 年左右就不再生产这些器件了。

通电前，确保将 VR1 设置为最大电阻，以获得最小静态电流。 这一点非常重要，因为如果设置为最小电阻，静态电流将非常大（几乎肯定足以烧毁输出晶体管！）。
Construction
Since I have boards available for this amp, I obviously suggest that these be used, as it makes construction much easier, and ensures that the performance specifications will be met.  Note that the layout of any power amplifier is quite critical, and great pains were taken to minimise problem areas - if you make your own PCB, it is unlikely that you will be able to match the published specifications.

All resistors should be 1/4W or 1/2W 1% metal film for lowest noise, with the exception of R9, R10 and R15 which should be 1/2W types, and R13, R14 must be 5W wirewound.

The bootstrap capacitor (C5) needs to be rated for at least 25V (preferably 35V), but the other electrolytics can be any voltage you have available.  The trimpot (VR1) should ideally be a multiturn, but an ordinary single turn pot can be used (but is not recommended).  Setting the current will be a little more difficult with a single turn pot, and they are not as reliable.

A pair of these amps will be quite happy with a 0.5°C/W heatsink for normal hi-fi use if the quiescent current is maintained at the minimum recommended of 20mA.  You will probably be able to get away with a smaller heatsink if the supply voltage is reduced to ±30V, but you have to ask yourself if it's worth it.  For higher quiescent current or if you expect to push the amp, use a larger heatsink.  Consider using a fan if you are going to push the amp hard.  Remember - there is no such thing as a heatsink that is too big.
构造
既然我有这款功放的电路板，我当然建议使用这些电路板，因为这样会使构造更加简单，并确保达到性能指标。 请注意，任何功率放大器的布局都相当关键，我们会尽最大努力减少有问题的地方--如果您自己制作电路板，就不太可能达到公布的规格。
所有电阻都应为 1/4W 或 1/2W 1%金属膜电阻，以降低噪音，但 R9、R10 和 R15 应为 1/2W 类型，R13 和 R14 必须为 5W 绕线电阻。
Basic Specifications
The following shows the basic measurement results ...
基本规格
下面显示的是基本测量结果 ...


Notes

The frequency response is dependent on the value for the input and feedback capacitors, and the above is typical of that when the specified values are used.  The high frequency response is fixed by C4, and this should not be changed.
Operation into 4 ohm loads is not recommended with the 42V supplies.  Peak dissipation will exceed 110W in each output transistor, leaving no safety margin with typical inductive loads.  All supply voltages are nominal, at no load - your transformer may not be capable of maintaining regulation, so power may be slightly less than shown.
This figure is typical, and is dependent on the regulation of the power supply (as are 1 and 2, above).  Worst case power with 8 ohm loads is about 50W, but the supply will be seriously inadequate if the power falls that far.
This is an extremely pessimistic test, because the bandwidth extends well above and below anything that is audible.  The response of my meter extends from around 3Hz to well over 100kHz, so the measured noise is much greater than would be the case with any weighting network.
备注：
1、频率响应取决于输入电容和反馈电容的值，以上是使用指定值时的典型频率响应。 高频响应由 C4 固定，不得更改。
2、不建议使用 42V 电源在 4 欧姆负载下工作。 每个输出晶体管的峰值耗散将超过 110W，在典型电感负载下没有安全余量。 所有电源电压均为空载时的标称电压 - 您的变压器可能无法维持稳压，因此功率可能略低于所示值。
3、此数据为典型值，取决于电源的调节能力（如上文的 1 和 2）。 8 欧姆负载的最差功率约为 50 瓦，但如果功率下降到这个程度，电源将严重不足。
4、这是一个极其悲观的测试，因为带宽远远超出和低于任何可听的范围。 我的测量仪的响应频率从 3Hz 左右一直延伸到 100kHz 以上，因此测得的噪声比任何加权网络都要大得多。
Four of these amps in a biamped arrangement will give you prodigious SPL, and is similar to the arrangement I am using.  Coupled with a Linkwitz-Riley crossover, the amplifiers can be mounted in the back of the speaker box, so only signal and power are needed for a complete system that will leave most commercial offerings for dead.
四个这样的功放组成双功放排列，可以提供极高的声压级，与我使用的排列方式类似。 与 Linkwitz-Riley 分频器配合使用，功放可以安装在扬声器箱的背面，因此只需要信号和电源就能构成一个完整的系统，让大多数商业产品望尘莫及。
Powering Up
Make sure that the amp board is mounted to a heatsink before applying power.  Operation without a heatsink is possible, but only if you know exactly what you are doing, run the amp from a lower than normal supply voltage, maintain zero quiescent current and do not connect a load.  Any attempt to run the amp 'normally' without a heatsink may result in almost instantaneous failure of output transistors and in some cases driver transistors as well.

If you do not have a dual output bench power supply - before power is first applied, temporarily install 22 Ohm 5 W wirewound 'safety' resistors in place of the fuses.  Do not connect the load at this time! When power is applied (typically ±35V), check that the DC voltage at the output is less than 1V, and measure each supply rail.  They may be slightly different, but both should be no less than about 20V.  If widely different from the above, check all transistors for heating - if any device is hot, turn off the power immediately, then correct the mistake.

If you do have a suitable bench supply, the initial test is much easier! Slowly advance the voltage until you have about ±20V, watching the supply current.  If current suddenly starts to climb rapidly, and voltage stops increasing then something is wrong, otherwise, continue with testing.  Note: as the supply voltage is increased, the output voltage will decrease - down to about -2V, then quickly return to near 0V.  This is normal.

Once all appears to be well, connect a speaker load and signal source (still with the safety resistors installed), and check that suitable noises (such as music or tone) issue forth - keep the volume low, or the amp will distort badly with the resistors still there if you try to get too much power out of it.

If the amp has passed these tests, remove the safety resistors and re-install the fuses.  Disconnect the speaker load, and turn the amp back on.  Verify that the DC voltage at the speaker terminal does not exceed 100mV, and perform another 'heat test' on all transistors and resistors.

When you are satisfied that all is well, set the bias current.  Connect a multimeter between the collectors of Q7 and Q8 - you are measuring the voltage drop across the two 0.33 ohm resistors.  The most desirable quiescent current is 75mA, so the voltage you measure across the resistors should be set to 50mV ±5mV.  The setting is not overly critical, but at lower currents, there is less dissipation in the output transistors.  Current is approximately 1.5mA / mV, so 50mV will represent 75mA quiescent current.

After the current is set, allow the amp to warm up (which it will), and readjust the bias when the temperature stabilises.  This may need to be re-checked a couple of times, as the temperature and quiescent current are slightly interdependent.  When you are happy with the bias setting, seal the trimpot with a dab of nail polish.
接通电源

通电前，确保放大器电路板安装在散热片上。 不使用散热片也可以运行放大器，但前提是您必须清楚自己在做什么，使用低于正常的电源电压运行放大器，保持零静态电流，并且不连接负载。 任何不使用散热片而 "正常 "运行放大器的尝试都可能导致输出晶体管瞬间失效，在某些情况下还会导致驱动晶体管失效。

如果没有双输出台式电源，则在首次接通电源之前，临时安装 22 欧姆 5 瓦绕线 "安全 "电阻器以代替保险丝。 此时不要连接负载！通电后（通常为 ±35V），检查输出端直流电压是否低于 1V，并测量每个电源轨。 它们可能略有不同，但都应不低于 20V 左右。 如果与上述值相差很大，请检查所有晶体管是否发热 - 如果任何器件发热，请立即关闭电源，然后纠正错误。

如果有合适的台式电源，初始测试就容易多了！缓慢升高电压，直至达到 ±20V 左右，同时注意电源电流。 如果电流突然开始快速上升，而电压停止上升，则说明出了问题，否则继续测试。 注意：随着电源电压的升高，输出电压会降低-2V 左右，然后迅速恢复到接近 0V。 这是正常现象。

一切正常后，连接扬声器负载和信号源（仍安装安全电阻），并检查是否有合适的声音（如音乐或音调）发出 - 保持低音量，否则，如果您试图从放大器获得过多的功率，在电阻仍然存在的情况下，放大器将严重失真。

如果放大器通过了这些测试，则移除安全电阻并重新安装保险丝。 断开扬声器负载，重新打开放大器。 确认扬声器终端的直流电压不超过 100mV，并对所有晶体管和电阻器再进行一次 "热测试"。

确认一切正常后，设置偏置电流。 将万用表连接到 Q7 和 Q8 的集电极之间，测量两个 0.33 欧姆电阻上的压降。 最理想的静态电流为 75mA，因此测量电阻上的电压应设置为 50mV ±5mV。 这一设置并不过分重要，但在较低电流时，输出晶体管中的耗散较少。 电流约为 1.5mA / mV，因此 50mV 代表 75mA 静态电流。

电流设定后，让放大器预热（会预热），待温度稳定后重新调整偏置。 这可能需要重新检查几次，因为温度和静态电流略有相互影响。 当您对偏置设置感到满意时，用指甲油涂抹一下，将微调器密封起来。
If the temperature continues to increase, the heatsink is too small.  This condition will (not might - will) lead to the destruction of the amp.  Remove power, and get a bigger heatsink before continuing.  Note also that although the power transistors are mounted to the board, never operate the amp without a heatsink - even for testing, even for a short period.  The output transistors will overheat and will be damaged.
如果温度继续升高，说明散热片太小。 这种情况将（不是可能--将）导致放大器损坏。 卸下电源，换一个更大的散热片后再继续。 还要注意的是，虽然功率晶体管安装在电路板上，但切勿在没有散热片的情况下操作放大器，即使是测试，即使是短时间的操作。 输出晶体管会过热并损坏。
完成所有测试后，关闭电源，重新连接扬声器和音乐源。

Power Supply
Before describing a power supply, I must issue this ...
电源
在介绍一款电源之前，我必须先说一句...
WARNING:   Mains wiring must be done using mains rated cable, which should be separated from all DC and signal wiring.  All mains connections must be protected using heatshrink tubing to prevent accidental contact.  Mains wiring must be performed by qualified persons - Do not attempt the power supply unless suitably qualified.  Faulty or incorrect mains wiring may result in death or serious injury.
警告：主电源接线必须使用主电源额定电缆，并应与所有直流和信号接线分开。 必须使用热缩管保护所有电源连接，以防意外接触。 主电源接线必须由合格人员进行 - 除非具备相应资格，否则不得尝试供电。 错误或不正确的主接线可能导致死亡或严重伤害。
A simple supply using a 25-0-25 transformer will give a peak power of about 75W into 8 ohms, or 60W or so continuous.  This is influenced by a great many things, such as the regulation of the transformer, amount of capacitance, etc.  For a pair of amps, a 300VA transformer will be enough.  The 4,700µF caps shown should be considered the minimum, and in general I suggest that you use two in parallel for each supply (providing 9,400µF).  Feel free to increase the capacitance, but anything above 15,000µF or so brings the law of diminishing returns down upon you.  The performance gain is simply not worth the extra investment.
使用 25-0-25 变压器的简单电源在 8 欧姆时的峰值功率约为 75 瓦，连续功率约为 60 瓦。 这受到很多因素的影响，如变压器的调节、电容大小等。 对于一对放大器来说，300VA 的变压器就足够了。 所示的 4,700µF 电容应视为最小值，一般来说，我建议每个电源并联使用两个电容（提供 9,400µF 的电容）。 您可以随意增加电容，但任何超过 15,000µF 左右的电容都会带来收益递减规律。 性能提升根本不值得额外投资。

For the standard power supply as noted above I suggest a 300VA transformer.  In 230/240V countries, use a 3A fuse or the value suggested by the transformer manufacturer.  For 115V countries, the fuse should either be 6A or as advised by the manufacturer, and in all cases a slow blow fuse is required because of the inrush current of the transformer and capacitors.  The fuse rating may need to be increased slightly if you use more than the suggested capacitance, C3 is an X2 mains rated capacitor.  When placed in parallel with the transformer secondary it reduces RF interference (conducted emissions) by a useful amount.  It's not essential, but is recommended.

The supply voltage can be expected to be higher than that quoted at no load, and lower at full load.  This is entirely normal, and is due to the regulation of the transformer.  In most cases, it will not be possible to obtain the rated power if the transformer is not adequately rated.

The bridge rectifier should be a 35A type, and filter capacitors must be rated at a minimum of 50V for ±35V supplies.  Wiring needs to be heavy gauge, and the DC must be taken from the capacitors - not from the bridge rectifier.
   对于上述标准电源，我建议使用 300VA 变压器。 在 230/240V 国家，使用 3A 保险丝或变压器制造商建议的值。 对于 115V 国家，保险丝应为 6A 或制造商建议的值，由于变压器和电容器会产生浪涌电流，因此在任何情况下都需要使用慢熔型保险丝。 如果使用的电容（C3 为 X2 市电额定电容）超过建议值，保险丝额定值可能需要略微提高。 当它与变压器次级并联时，可以有效减少射频干扰（传导辐射）。 这不是必须的，但建议使用。
在空载时，电源电压可能会高于设计值，而在满载时会低于设计值。 这完全正常，是变压器调节的结果。 在大多数情况下，如果变压器的额定功率不足，则无法获得额定功率。
桥式整流器应为 35A 型，对于 ±35V 电源，滤波电容器的额定电压必须至少为 50V。 接线必须采用重型线规，直流电必须取自电容器，而不是桥式整流器。

筑明

原著对制作写得很详细，我先仿真，这个线路失真很低，失真中二次谐波比三次谐波高。

P3A - tip35.rar (279.99 KB, 下载次数: 349)

2023-10-7 11:54 上传

点击文件名下载附件 P3A-BOM.rar (7.99 KB, 下载次数: 247)

2023-10-7 12:03 上传

点击文件名下载附件

特别提醒，C3如果用极性电解，我仿真图上是接反了，应该是正极接地。

第50楼有PCB文件下载，板子上的电容C3也画反了，经验主义害死人啊。

筑明

给初学者的电路图解说

R1C2R2构成低通高切电路，调整C2可以微调高音；
Q1Q2是经典差分放大线路，Q3 D1 R7构成恒流电路，恒流值为（D1的压降-0.58V）/R7=2.2mA，R8给LED供电，电阻的大小可以微调LED的压降，同时也会影响差分管的电流；
C4是米勒电容；
Q4是VAS管，负责电压放大，其工作电流为（35V-1/2Q9压降）/（R9+R10）=5.23mA，Q9压降=2个pn结电压=1.2V
R9、R10、C5构成自举电路；
Q9、R16、VR1组成温度补偿电路，VR1减小则大管电流增大，这个电路中Q9必须和Q5热耦合在一起，不用装在大管的散热器上，我的PCB设计的是把Q9、Q5、        Q6装在同一块小铝板上，实际测试静态电流很稳，没有发现漂移。
Q5Q7和Q6Q8是倒置达林顿电流放大电路，简称CFP输出电路。

这个电路制作的关键点：
1、因为Q4的B、E极直接跨接到R6上，Q4的BE结的压降是固定的，约为0.58V（实测ST公司的BD140），那么Q1的电流为0.58V/560=1.04mA，前面计算了Q3组成的恒流电路的工作电流为2.2mA，Q2的工作电流=2.2-1.03=1.17mA，这时，Q1和Q2的电流不相等，这样差分电路的失真会变大，所以，精选LED的压降、R8值、供电电压值都会影响差分电路的平衡性，而差分电路的不平衡直接影响高音的失真度。
2、Q5Q7和Q6Q8的组合搭配也是关键点，R11R12的取值关系到达林顿组合管的线性，原创的搭配为BD139 /140+TIP35c/36c，或者BD139 /140+mje21193/21194，mje15031/32+mje21193/94，diyaudio论坛上推荐的是BD139 /140+C5200/A1943，我手头正好有这对管就选择了这个组合，其它组合还没来得及测试。

swsw4321

这个简单分立也不错
http://bbs.hifidiy.net/thread-527834-1-1.html

无语密码

适合初学者

microstar

很熟悉的单差分结构，电流放大级有点意思，学习一下

99990000

别看架构简单，初学者不一定能玩好CFP。

freefreemen

各位大神开始研究HIFI之后，功放的设计逐渐工业化了呀

cdzzl

没明白C3为啥正极接地？

筑明

先上PCB图，感兴趣的坛友可以自己画板或者等本帖完成后分享。


准备元件，最重要的是差分管，海外代购买了200对安森美的BC547C和557C


从柜子角落里面翻出来的箭猪牛



从低到高开始焊接，先焊接电阻，注意1W以上的电阻要架空，距离PCB要有一点点间隙


再焊接电容，我这里选了680uf的大电容，这里的退藕电容我留了18.5mm的直径，尽量装大一点的，注意耐压为50v及以上


散热片打孔、装推动管，我这里选用的是珍藏多年的4对管子，飞利浦的BD139和140，引脚是那种很软的铜脚，注意三极管与散热片之间要装绝缘垫。
铝散热片为50*30*1.5mm，无它，主要是前期做金刚狼时买了几片50mm的铝板，这次正好用上，如果没有4欧的负载，可以不装，我选择还是装上保险。


在推动管的背面装上温补管，我选的温补管是国产的2SD669C，贝塔值大于230

筑明

cdzzl 发表于 2023-10-7 15:24
没明白C3为啥正极接地？

Q1Q2是NPN管，基极为偏负的电压，你装好后测量就知道，如果选无极性电容可以无视这个。

NE5532

感谢分享，持续关注

microstar

灰色的飞利浦139/140可是不多见了

筑明

继续，装好三极管和电容，输入电容选的是松下4.7uf无极性，反馈电容选的是100uf25v尼康mouse无极性电容，没有考虑调音，主要是手头库存元件，先装响再说。

装大管，大管经测试仪测试筛选过。
上散热器，这里省略了金工钻孔攻丝工序的照片，主要是忘记拍照了。


侧面照


背面照，注意两个0.33欧电阻要选绕线电阻，增加高频稳定性。


通电，注意第一次通电前将可调电阻调到最大值，用万用表测试确认。
正负电源先串联两个30欧20W电阻再接到电路板上。



通电检查没有发热、冒烟的情况，检查输出电压是否低于20mV，一切正常后开始调整测试大管的静态电流，万用表表笔换成鳄鱼夹夹到0.33欧电阻上，万用表打到mV档上，通电测量，这时候应该显示0V，慢慢调整可调电阻，我PCB板上是逆时针为增加电流，调整到17mV左右，这时候大管的工作电流为50mA。
接上信号发生器和失真仪，1K 2.5W的失真：


20K2.5W的失真


20K方波


30K方波


50k方波

yygbs

cdzzl 发表于 2023-10-7 15:24
没明白C3为啥正极接地？

因为输出端是负电压

lommxsdiy

高质量的制作帖子，对初学者有很好的指导作用，论坛就需要这样的热心的坛友，持续关注！

yygbs

速度真快，这原理图我都没看懂，想玩，怕烧管

筑明

继续，大家可以看到，20K的失真偏大，原著设计是绿色的LED，我也是装上绿色的，怕电压不符合买的是绿发绿的LED，没有买那种白发绿的LED，在线测量这个LED的压降是2.78V！这时候Q1Q2的电流严重不平衡，影响了高频失真率，这时候板上Q4装的是2SA1381，取其Cob值小，高频失真会好一些，测量R6上的电压是0.615V，经计算，我要找一个1.82V左右的LED，搭了一个测试平台，把手头在立创买的的几百个台湾亿光红色LED逐一筛选，挑出来十多个1.82V左右的，装上后上AP再测试。

yygbs

筑明 发表于 2023-10-7 16:53
继续，大家可以看到，20K的失真偏大，原著设计是绿色的LED，我也是装上绿色的，怕电压不符合买的是绿发绿的 ...

在R7上并一组由电阻+可调电阻串联的组合就可以随意调电流了，要计算好组织，让R6的压降处在合适的区间，或者是让输出端电压在-1到-5mv即可。
这样就可以无视LED灯的压降了

筑明

yygbs 发表于 2023-10-7 17:17
在R7上并一组由电阻+可调电阻串联的组合就可以随意调电流了，要计算好组织，让R6的压降处在合适的区间， ...

您说得对，原设计也没有什么问题，关键是元器件的选用，我的P3A+已经设计好了，还没有来得及焊，不过对初学者而言越简单越好。