Quad 405 Clone Amplifier Kit. Part 1

[@kelvinjinxd]

The final output stage of Quad 405 has 0 quiescent current. Its nfb network is quite clever to hide crossover distortion completely. BTW, I bought the same board from eBay. I was surprised that it didn’t fit into the original Quad 405 case. The board is a little bit wider. I returned the board.

[@MichaelBeeny]

Yes, Kelvin you are correct. These clone boards are made to fit the clone case they make. Does make you wonder why they did this. Quad call the unusual biasing as current dumping. It's quite common these days to use feedback to lower crossover distortion. Not it's intended use, but it does work.

[@pedrojmorais]

I've got an homemade 405 mod3 with opa134 ampop and paper in oil Russian caps, mod3 changes the circuit for modern ampops, it's the best circuit design ever made for solid state devices, all audio books say so, the current dumping design is just brilliant, class A+2C very low output impedance, capable of driving 1 ohm loads, single ended voltage amplifier with two class c dumpers not related to the A gain of the amplifier, no influence from the load, neither power transistors fluctuating specs, with heat and load, on our A gain and overall stability, a Columbus egg design and master degree in electronics for Peter Walker, ( Krell needs 8 pairs of transistors to achieve close output performance of the 405 with just one pair.) Other amplifiers are adjusted for dc on factory, quad does that 10 times every second, that's why Jamiroquai travelled with one in is bags for mixing, the bass is just right. The original papers say not to test it with square waves, there are no such waves in real music, they say, the poor results are from dc feedback above 16 hertz, and the roll off beyond 20 kHz. I have it now for 30 years with my audio note M7 clone, and I've tested against the audio note Ongaku and it sounds very similar. My home sound is the best I've ever heard, only mágico speakers with dartzeel came close. I wouldn't replace it for nothing in the world. The sound is not old is just the best there is, but you'll need mod3 transistor version, 500 watts transformer 100 000 microfarad dc caps or more, Muse electrolitic caps and paper in oil caps and silver cables, OPA134 and polystyrene cap on Wheatstone bridge cap.

[@ryansupak3639]

Came because I'm deciding between these boards and the Dada boards (at a substantially higher cost), but stayed because I love listening to you speak.

[@MichaelBeeny]

You will probably pay more for Dada boards as they tend to use better parts. I thing some of the Ali/Ebay PCBs use fake transistors the TO3 output transistors almost are for sure.

[@ryansupak3639]

@@MichaelBeeny I ordered both a set of Dada boards, and a full 405 clone from Alibaba. Both are arriving soon -- and yes, my heart tells me the Dada boards will be the ones.

[@MichaelBeeny]

@@ryansupak3639 Keep in touch Ryan, let us know what you find.

[@izzzzzz6]

Should it not be mounted in a metal case before testing for quality?

[@MichaelBeeny]

Not really, it relatively low impedance, at worse it could, in theory pick up a little noise. If I cannot hear it, it really does not matter.

[@stevesmyth4982]

The Quad 606 has a similar circuit design, the biggest difference being that it has six output transistors. Looking at the schematic the power rails are 57.8 V for the +ve rail, and 53.4 V for the -ve rail so it would appear that the assymetrical clipping is a function of the circuit design and Quad has compensated by using assymetrical rail voltages in the 606.

[@MichaelBeeny]

It's been many years since I've looked at the original circuits of the Quad. I must try that, makes sense now. Almost impossible to get those voltages from a transformer but I can do it on the bench. That would also explain the slightly difference power draw from each rail.

[@redrobbosworkshop]

Nice clean boards and looks like as decent kit. One suggestion is to test the amp at the recommended rail voltage since this design uses a crude resister dropper and diode clamp to set the voltage for the input chip. IIRC this amp expects to see 55v rails so the input IC might be starved of voltage, and this IC is the input stage so it needs be working optimally.

[@MichaelBeeny]

Yes, you are quite right. I always start tests on a current limited supply to do the first power up. Full tests are always done with the correct voltages. One of the clone complete 405 amplifier does in fact use a 38 volt supply but only offers about 65 watts output. You've gotta watch them, any way to rip people off, they will try it.

[@nigelpearson6664]

The Naim. Nap250 was touted as a 303 replacement. I prefer the 303. If understood the 303 sets out to be better than the Quad valve design. It is massively better. It has almost an impossibly ideal distortion spectrum for an early transistor design. The protection circuit is simple and effective. The feedback loops are interesting. The power supply is ingenious. The amplifier output inverts. Invert the speakers if not using a 33 . I found Quad a bit smug. They were always nice to me. I miss those days. Thinking back they deserved to be proud.

[@MichaelBeeny]

I never liked the Quad 303, and still don't. Infact even Quad did not like it, they had to enter the transistor market as sales for the valve stuff was dwindling fast. It's made to go with the electrostatic. Quad was always amazed anyone would buy any other type of speaker. In fact I also was not fond of ANYTHING Niam made. I put them both in the same camp. The later amplifiers from the 405 got better but by today's standards, it's slow, un dynamic. Things have moved on. Everything has it's day.

[@Andrew-rc3vh]

There is probably a low frequency filter so it rolls off the bass. Look for an R & C which give about 20Hz pole and then increase the capacitance. It's probably designed like that on purpose due to bad bass distortion in old speakers.

[@Annies_E_P]

​ @Michael Beeny Hi Michsel. Quad spec's +/-50 rails - transformer secondary of ~2x35V. When the driver transistor broke its leg, it looked like they were recycled parts, as the tabs & legs looked none to clean - IMHO. I'm sure the kit suppliers are are above reproach tho...(!?)

[@DrBroncanuus]

brilliant video...I learnt a lot....so are McIntosh amps on a par with the Quad 303 ?

[@MichaelBeeny]

To be honest I don't have any experience of Mac amplifiers. Either technically or listening to them.

[@nigelpearson6664]

My Pro Audio friend said the 405 was deliberately truncated to give extreme reliability. The feed forward distortion cancellation doesn't need extended bandwidth. For 4 ohms 405 can be parallel bridged. Gale 401 speakers sounded dramatically better that way. They were notorious.

[@ianmckenzie8685]

The connection point on the power transistor heatsink is likely to be for safety grounding. To comply with many electrical codes, the metalwork to which the heatsinks are bolted needs to be connected to safety earth, unless it is made inaccessible or the amplifier power supply is certified as double insulated. Neither of which is guaranteed in a kitset amplifier!

[@Unker_Spunkanathan]

Are you looking for a larger chassis for your future projects? The chassis you have been using looks too congested. Maybe those full sized 430mm width types would provide more flexibility in mounting the parts

[@MichaelBeeny]

Yes, I really do need a bigger box. They cost so much these days. We now have to pay tax @ 15% and postage has doubled on big stuff. I've been thinking about using a 19 inch rack case.

[@bonobo9904]

The amplifier itself has faster response. The RC filter between the opamp and transistor limits slew rate of input signal making it lower than amplifier own slew rate. As you have the PCB at the test bench you can replace 1000 pF filter capacitor to 220 or 100 pf and take a look at the output square signal. I did it virtually using MicroCap and I saw better slew rate. But, regarding the sound quality, especially TIM level, it's better to have this filter as it was designed.

[@MichaelBeeny]

Thanks for the info, I might try that just to see what the results are thank you.

[@jonlaws4493]

Space the higher power resistors from the board as they get hot.

[@MichaelBeeny]

I did cover this point in the video Jon.

[@HillsWorkbench]

Wouldn't be a design I'd favor but it does have some interesting features, like built-in crowbar protection, which might be the cause of that early clipping. As you have overcurrent protection in the power supply, it would be interesting to see if it tests better with those two transistors removed (Tr5&6?). A crude thermal/bias circuit, just the old two diode approach, probably makes it just Class AB1, not 2 as it likely doesn't run much idle current.

[@MichaelBeeny]

I'm putting together a proper power supply with 50/0/50 DC. Lets see if it works better like that. 30/0/30 DC is borderline really but fine for first boot up. Nothing like a limited supply to prevent the smoke and possible destruction of everything.

[@CoolMusicToMyEars]

I used to sell & repair the Quad Equipment only thing really annoying was the laminations on the E Core transformer, from the waveforms it looks like you haven't loaded the output properly, use none inductive load, plate resistors from Vishay, the waveforms looks like a probe problem switch scope to DC input instead of AC then your frequency at the low end will be far better, trim scope probe for your scope I presume your using a x10 probe, the pickup will be because its not in it's case, use good quality phono leads, short leads, don't forget to make sure your output impedance is matches the 33 or 34 or 44 output impedance, best to scope both input & output so you can check for any Oscillator limitations, The power supplies should have a very low output impedance, Quad 405 power supply is quiet basic but very effective, I started off as professional audio sales & service engineer, then landed into instrumentation, the at the same time Aerospace Defence Electronics Test Qualification Verification Consultant done the rounds with the giant's in the industry, now retired, I had a Quad 405 & Quad 44 unit until they were stolen 😢, I can say the level of noise on the QUAD 405s versions 1 & 2 is low as long as you don't put the control unit on top of the Quad 405 power Amplifier, I've seen clones on quad 405 not sure what versions in a case for about £188, dont expect the build quality of the case to be anything like the solid Quad cases, at the End of the Day You Pay for What You Get, There's been a few companies that have modified the 405, that's for you to decide if you want to go that far 🤔 PS make sure even none electrolyte capacitors have a right way & wrong way around, outer foil markings should be marked with either a line or sometimes a dot, stand the resistors with ceramic stand off's

[@nemesissanity8132]

Unmatched or/and not genuine transistors

[@HillsWorkbench]

Probably the crowbar transistors have more gain than the original design and are cutting in early.

[@vit4min4s]

Great work MB, thanks! I can´t wait for the next project. L20SE?

[@allthegearnoidea6752]

Really enjoyed this video Great review - im thinking of building one of these kits. You almost made me cry my only Quad to date is a 303 and I bought it because I like the industrial design look as 95 percent of the time a amplifier is just furniture. never really been into hifi or listen too music but I am interrested in electronics and testing so I hope to maybe learn about amplifiers next year. I was hoping to build a Mullard 510 kit but not found anything yet. Are there any amplifier kits you would recommend. thanks regards Chris

[@MichaelBeeny]

I remember the Mullard 5/10 very well Chris. The first HiFi amp I build as a spotty youth. Not seen any kits for years now. I did build a valve amp a few years ago that was quite good. Take a look here. kzbin.info/www/bejne/aoipdJmCq7GNp9U. I'm doing listening tests on the Quad clone right now. First impression are sounds good. Time will tell.

[@essexheraldcom]

Nice video, thank you for posting. To be fair, the original 405 (on witch those Chinese boards are based) was not very good with anything below eight ohms. It also lacked in overall dynamics. The later versions, (405 Mk2) were much better in this respect. I would recommend altering the input sensitivity though. It was designed to match the output from a Quad 303, 500mv. It really needs to be set at about 1.5V to get the noise level down. It is possible with only a few minor mod’s to get a 405 Mk2 up to a very high sonic standard indeed. Thank you again for posting. Harry.

[@MichaelBeeny]

Good morning Harry. Yes you 100% correct. I've not finished with this board yet. I'm waiting for a couple of different clones to arrive. More on this later, with a few mods of my own. It is quite noisy but many modern amplifiers for sure so the gain mod is a must. Thanks for your input. Regards, Mike. PS If you think the 405 is lacking try the 303, not a good sounding amplifier at all!! Lots of crossover. Quite visible of a distortion analyser. I think you meant the 33 pre amp!

[@MichaelBeeny]

@Tim Ring Tim, it's not often say this but I agree 100% with you. The 33/303 do look good as you say but sound, shall we say, less than good. They were really made to work well with the electrostatic speaker that needs current limiting and no low bass.

[@Radfordperson]

@@MichaelBeeny Bought a second hand 303 many years ago, did not like the sound, gritty treble, after a good service it was quite different, very much better, still use that amp today. Main speakers are Rogers BBC LS3/6.

[@alfredopaglia4322]

anche io oggi ho assemblato le due schede identiche a queste ma senza il connettore GND al centro sheda..quindi ho usato quel buchetto dove c'è scritto GND non so se devo anche portarlo al centro stella per ora sta così e non ho strumenti per misurare ma l'ampli so comporta molto bene

[@giuseppebeppe_tv8257]

Good video Michael! Thanks! I've just assembled my two kits, but using all genuine transistors ( NXP, Cdil and ST brands) because I don.t trust the original chinese semiconductors provided ....( I think the two power transistors are fake....) . on one kit I have used two genuine BUW46 from ST, in the other kit have mounted two MJ15024 from RCA, the two ampli sound very good ! the LM301 are the two provided in the kit. till now i have tested them with + and - 27 volts from two regulated power supply.

[@MichaelBeeny]

Thanks for the info Giuseppe, much appreciated. I thought the output transistors must be fake as new ones from Farnell cost more than the kits. Keep in touch as you progress with the project.

[@mikemargo]

Definitely fake, as is the op amp not even up for debate

[@ianshepherd917]

Silicon grease is not heatsink compound. They are two separate products!

[@MichaelBeeny]

A slip of the tong, that's what happens when you have no script and shoot live. I'm sure most people realized that.

[@moodyga40]

good amp

[@Unker_Spunkanathan]

Surprised you bought this coz you didn't like the naim clone... I presume this will have similar 70s/80s brit sound They sell this complete with a clone quad chassis too... Think it is roundabout 1k yuan before shipping My 500 yuan nap140 clone plug n play amp arrived last weekend. Still fooling around with it. Regarding grounding the heat sink, it probably matter less if it is standalone and floated. BTW, do you normally float the DC ground or join it to the chassis Earth? As for the driver transistors getting hot, 1 option is to mount them onto the main heatsink away from the output transistors. Now, whether or not to keep, perhaps it would be better to conduct the performance measurements after listening to it for some time so that you do not already have preconceived idea how it would sound

[@MichaelBeeny]

I looked at the clone complete but some of them only use 40/0/40 DC with quite a small transformer. To get full rated power you need 50/0/50 DC I'm sure some people get coned with this version.

[@Crazy80ivan]

I think that the poor performance has something to do with fake transistors. Or that they are bad copies of the old ones. I read that the brand "ON" makes them. I've seen more video's on youtube, with different boards (different transistors), with the same problems. Replacing them did good. Try to obtain Motorola transistors. There is enough info on the web on this matter. Do the same with the opamp.

[@HillsWorkbench]

ON was Motorola, got split off.

[@robmacdonald7004]

Square wave performance at 10 kHz and 20 kHz is about as expected. A square wave is made up from sine waves at the fundamental frequency plus odd harmonics. Odd harmonics of 10 kHz are 30, 50 70 kHz. 20 kHz odd harmonics start at 60 kHz! Your amp is not designed to reproduce these frequencies, and you can't hear them anyway.

[@MichaelBeeny]

I agree with you Rob, but we need away to testing amplifiers and, in my experience, and amplifier that gives a good square wave, usually sounds better than one that does not. A fast rise time tends to give more detail. For example, the L12/2 has very good square waves, much better than the Quad and ALL valve amps. It has more attack and detail. So, it is an indicator of better sound, probably not the only one, however.

[@ford1546]

My absolute best amplifier kit of all is. I have L15D. L25D. L25 CLASS AB NAIM NAP250. TPA 3116. TDA7492. HYPEX UCD180. without a doubt the HYPEX UCD180 has absolutely the best sound quality of all CLASS D at a good distance from the others. And has better sound than the NAIM NAP250 has. then comes the best class ab which is L25 Class AB which has really good sound

[@joeneticchan9640]

What is your comments to compare L15D and Hypex UCD180? Thanks!

[@ford1546]

@@joeneticchan9640 Hypex ucd180 is a more expensive and better amplifier that has better sound especially in brighter areas. When compared to the L15d you hear well that the hypex ucd180 has better sound. L15D is at 150w in 4 ohms when I test it. l15d class d gets a little hot. L25D does not like 4 ohms!

[@ford1546]

@@joeneticchan9640 Almost all cheap popular class-d on ebay have poor sound especially in brighter areas.

[@joeneticchan9640]

Thanks for your quick reply! I owned L15DX2, a whole amplifier ordered from LJM, cost about USD200, 150W into 8ohm, about 300W into 4ohm I suppose, use it to drive Dynaudio Audience 42, sound pretty good, very dynamic and 3D sound, very high value amplifier in my experiences. But I am considering to upgrade it to Hypex or Purify class D amplifier if there has big gap between L15D and Hypex/Purify.

[@ford1546]

@@joeneticchan9640 L15D measures only 1x150w in 4 ohms with optimal power supply. If you want good sound, buy L25 class-ab or L12.2 Class-ab If you are looking for the best possible sound quality then the L15D is not for you. Not the amplifier with the worst sound but not the best either.

[@ford1546]

Quad 405 Stereo Power Amplifier (1978-84) This is an old design

[@nigelpearson6664]

303 did what a class A design would do except nearly class B ( very little bias requuired ). It must have been tedious to build. It's PSU divides people. Black speaker isn't at 0 volts. Even though made too cheaply Sinclair Z30 showed the way. An opamp layout. The biasing of the Z30 is not unlike the 303. The Sinclair Neoteric of a year before ( 1968 ) was like the Quad or Tobey and Dinsdale. The 405 was a not an amp I dream of. 303 given a load it likes can really sing. Most domestic speakers doubtless don't suit. Klipsche? My Magnepan SMGa were ideal. 4 ohms resistive. JLH Mosfet I have heard, it was better. Not unlike the output triple, Z30 in places. RCA isn't the origine design as is often said. UA741 partly. 741 is a mini power amp.

[@tinagladu3744]

dont always believe the numbers if a amp sounds good thats and old wives tail you should know that

[@ianf1008]

Certainly a nice video, but a couple of things I noticed: - Old school, plain metal tab power transistors (TO3, TO220, TO218 TOP3, T0P5 etc.) all used a common, little plastic insulating bush which now costs almost nothing - a dollar or two for a bag of literally hundreds of them. Most kit suppliers throw in a couple of extras for butterfingers like me. The opamp you pointed out as an improved type was actually the original LM301 type but as several engineers including Quad expert Keith Snook have all commented, the opamp in the 405 specifically, has little to do with audio. Of course, DIY builders like to fit the highest available grades of any component they can lay their hands on, whether relevant or not to any purpose. Quad 303? Yes, awful sound but the industrial design of the case is a timeless work of industrial art, as is the manufactured quality and finish of the times. Compared with modern products sporting chunky, chiselled and brutally raw aluminium block fascias, its old but still wonderful to look at.

[@iuliandumitrescu9441]

Clone better! Imput stage lme 49710, protection quad 405 out, change limiter (go to quad 520)ouput stage, (520 quad). Tr1 change to Wilson current mirror!!! My Quad clone is ouput dc is 0,3 mv!!!

[@MichaelBeeny]

Interesting Julian, was your clone a kit? from where? I have another kit from a different supplier, should be here soon I hope. I will build and test this one.

[@Theineluctable_SOME_CANT]

Dreadful amp...

[@MichaelBeeny]

I'm not sure I would go quite that far! but it's not my favorite for sure. Many MUCH better our there these days.

[@Theineluctable_SOME_CANT]

@@MichaelBeeny I never thought the Quad 405 was this bad. A 1 kHz square wave should be very clean...

[@EduardoCruz-ur4wq]

@@Theineluctable_SOME_CANT This is not a Quad 405 this is a cheap clone with fake power transistors, crap electrolytic capacitors, and the input ic is also a fake one, the correct voltage is not 30V +/- but 40V+/- .