First LM3886 Gainclone Build - Need Guidance/Suggestions

novice-akki

New Member
Joined
Aug 7, 2010
Messages
255
Points
0
Location
Delhi,NCR (India)
Hi All,

I got two pieces of old electronic junk from a cousin who's left for Canada for good.
First is an 80's Woodstock amp made here in Delhi, quite popular at one time and the other is an el cheapo cassette player - both in non-working condition.
However, chasis/cabinet of both are in pretty good shape, no rust, cracks, warp etc. I've already ripped them apart and salvaged a few items in process to use in my project - 2 channel amp

I've never had a 2 channel amp - have owned a mini component system earlier and use a Yamaha AVR, presently

Coming back to the build, from the salvaged parts I've an EI transformer with 18 0 18 secondary winding and a power supply. I read a lot of threads on gainclone builds and figured out the primary winding from secondary.

I connected the transformer to AC mains and am getting ~ 230 volts at primary and ~ 37 volts at secondary outer windings and when I measure from center to outer windings I get ~ 18 volts at both ends.

Next, as seen in fig. 4 the power supply PCB has 4 X 1N4002 diodes and in between these diodes are 2 X 1000uf 50V radial caps [BLUE], there's another 1000uf 25V radial cap [WHITE] next to a 1k resistor and there are 2 caps at the PCB input with color bands which I didn't recognize. I checked for continuity on PCB and it was ok. So, I soldered the secondary windings to PS pcb.

Now, on the PS pcb there are two outputs where I'm getting rectified DC voltage. At one o/p am getting ~ 52 volts DC and at the other o/p I'm getting ~ 26 volts DC - So, I believe that's OK as 18 volts AC after rectification and filtering should give 26 volts DC.

I get +26 DC volts and if I swap the DMM probes I get -26 volts DC from same pads on PS pcb - is this normal ? and same thing at the other o/p with +/- 52 volts DC

If the above sounds OK - then I'm planning to connect it to this dual mono kit I recently purchased.

Please confirm if it's OK to proceed - So, I may ask the next set of questions.

Help, guidance, suggestions earnestly sought.

Regards
Abhishek

Pics of the Woodstock amp I'm using in this project
 
Last edited:
Hi,

That's a good chassis for a DIY project. You mentioned, you have purchased that LM3886 kit. good choice for a start if you are a novice.
Do you plan to have Bass/treble balance control integrated, in that case you would need a preamp. lots of designs out there.
Except from the chassis, transformer, knobs and connectors nothing else is usable from this amp, forget the power supply and ALL the capacitors.

If i were in your place, i would have rebuilt/refurbished the amp.

Good luck on your project.

Cheers!!!!
Aniket
 
It has been a slow moving build but I've finally made some progress.
Received rectifier boards and snubber parts from audiosector a few days back.

So, I soldered the parts on rectifer pcb and also finished soldering the amp boards.
Got some cheap RCA and Speaker o/p hardware locally.

As suggested by Peter Daniel of audiosector, I've used a single bridge rectifier for both amp boards since I'm using a 18 0 18 CT transformer. After a couple of more email exchanges I managed to wire everything up and have it working quite easily in the end.

Happy to report that it works like a charm...listening to it as I type.

There's no thump sound when the amp is switched on/off or when source is connected or disconnected. There's no hiss/hum when music is playing. But, there's a faint hiss from speaker drivers which becomes audible at around 3" from them.

However all this may change when it goes in a proper metal cabinet.

Another thing I've noticed is on one channel dc offset is 66mV and on other 116 mV, how do I go about lowering it?

Here are a few pics
 
Last edited:
Sir,
From the photo, it seems that the positioning of the heatsink not the way it has got to be. If placed such that the fins are in vertical orientation it could have provided more effective cooling with better air circulation.
 
I've received some good quality parts from Diyaudiocart. Great service from the brothers.

These RCA and Banana sockets due to their better quality are different from the one's I've used before. Should I simply solder these or find suitable connectors ?

Also, to mount chips on heatsink shall I use self screws or nut bolts?
 
Last edited:
Thanks DAC.com !

I dont have tapping bits so instead I've used a 12 mm SS spacer between heatsink fins to fasten the LM3886 with a 3x15 mm screw and the contact between HS and chip is perfect.
RCA sockets will be soldered and female receptacles will be used for speaker banana sockets.

Since I'm not getting many replies here on this thread I may send a few more emails to diyaudiocart.com ;)
 
Everything now sits in a wooden enclosure.

PS: 18 0 18 CT transformer with single bridge rectifier for both amp boards,
There are 2 X 10,000 uf / 50 V caps with snubber parts on rectifier.

I get +/- 28 VDC on both boards without load and the voltages drop to +/- 26 VDC under load (6 ohms speakers).

The amp plays fine at 1/2 volume level for 30 mins after that if there's a song with a demanding passage one of the channels has a dip in volume for a sec and then its up again. This will continue till either I lower the volume or the passage finishes.

The amp is on the brighter side, the sound is almost fatiguing at 1/2 volume and absolutely harsh/fatiguing at 3/4 volume.

Also, how can I get tighter bass, more detail in midrange and clear highs.
 
The amp plays fine at 1/2 volume level for 30 mins after that if there's a song with a demanding passage one of the channels has a dip in volume for a sec and then its up again. This will continue till either I lower the volume or the passage finishes.

What is the VA rating of your transformer?
 
Not sure, it was salvaged from an old amp. Is that the weak link ?

If I remember correctly from the LM3886 datasheet, the transformer needed for 8 Ohm load is about 230 VA. For your 6 Ohm load, I don't recall exact rating.

If the transformer is unable to supply instantaneous peak power requirement as demanded by the nature of the song being played, the power rails will sag, leading to lowered volume as you had observed.
 
Not sure, it was salvaged from an old amp. Is that the weak link ?

Looks like, It is transformer plus power bank capacitors. I am building with 300VA (2x 25-0,6A) transformer for stereo. Each channel has 2x10000F power capacitor.

On DIY audio people say 150VA pc (if using toroidal transformer) or 120VA pc (if using R-Core transformer). Here in your pictures I can see its EI transformer which is highly inefficient.
 
If the transformer is unable to supply instantaneous peak power requirement as demanded by the nature of the song being played, the power rails will sag, leading to lowered volume as you had observed.

Looks like, It is transformer plus power bank capacitors.Here in your pictures I can see its EI transformer which is highly inefficient.

Thanks for confirming my doubt (weak transformer)

Is power rail sag = clipping ?

Also, per this graph I will need +/- 35 VDC for an ~ 65 W/Ch output since I will pair it with 6ohm speakers with 86 db sensitivity. Please confirm so that I start collecting parts and hopefully build my first power supply.
 
Last edited:
Thanks for confirming my doubt (weak transformer)

Is power rail sag = clipping ?

Also, per this graph I will need +/- 35 VDC for an ~ 65 W/Ch output since I will pair it with 6ohm speakers with 86 db sensitivity. Please confirm so that I start collecting parts and hopefully build my first power supply.

Go through this page to determine the transformer rating you need:

A Complete Guide to Design and Build a Hi-Fi LM3886 Amplifier - Circuit Basics

I believe clipping happens when the amplifier is overdriven beyond its linear operating range. Typical example: as one increases the gain of an amp, a sinusoidal wave being amplified will start developing a flat top and bottom once the amp is driven beyond its linear range. A severely clipped sine wave will start looking more and more like a square wave.

In case of power rail sag, the amp won't be able to reach its peak ever, so it will never clip. It simply doesn't have the energy to reach peak or near peak.
 

In short they are size to power ratio is bigger, and only advantage of noise coupling from mains which can be achieved by shielding in toroidal. Other leakage is magnetic flux and heat are more due to non ideal EI core joint. Core saturation in toroidal is avoided by proper turns ratio and core sizes.

I know you are pulling leg but I end my statement here.
 
OK, I've been using the amp for 2~3 hrs daily, no issues.

The amp is powered with a single 18 0 18 CT (center tap) trafo connected to single bridge rectifier with 4 MUR860, 2 X 10,000 uf /50V and snubber parts.

This is what I did with the wiring; there's a jumper wire between PG+ & PG- on rectifier and a jumper wire between output grounds of both amp boards.

Both these jumpers are connected with a thick wire.

From trafo CT there is a twisted triplet wire. One/single wire goes to each amp board and the third wire goes to the thick wire between the jumpers (mentioned above).

Can someone please confirm, if the caps and snubber parts on rectifier are in circuit ?

Also, should the earth tab from power entry module to be bolted/attached to chasis (even though its all wood).

There's a 3rd tap on trafo primary which is absolutely isolated from all other taps including trafo chasis. Should this tap also be grounded/bolted along with earth tab from power entry module ?

Many thanks in advance
 
Last edited:
You should provide some Photos of your wiring, just by describing the issue it is difficult to understand. It will be even better if you can provide a high resolution photo. If you have a google account then you can upload the Photos on Google Photos & Share the link in forum.

If you are using all wood cabinet then it is no use to attached the earth point in chassis.

Sadik
 
You should provide some Photos of your wiring, just by describing the issue it is difficult to understand. It will be even better if you can provide a high resolution photo.
Sadik

Sure, I will put up some images by this evening once I get back home, thanks !
 
Last edited:
If the casing is of wood and you are not tinkering inside while the amp is playing you don't need to connect the circuit ground to electric ground. You are using the default wiring as prescribed by Peter daniels and I am using the same without any problem.

You need to cnnect the central tap to the common ground in case of a star ground.
 
Last edited:
Check out our special offers on Stereo Package & Bundles for all budget types.
Back
Top