Amperes required by LM1875 Guitar Power Amp

[exztinct01]

accdg. to Rod Elliott "no heatsink is too big"

[phatt]

Toroidals require slow fuses because of their much higher inrush current , even if unconnected to any load, it takes some power to magnetize all that iron.

A fast fuse blows on turn on.

EI are lossier so in a way somewhat self protected.

@ phatt, after originally liking that Jaycar PCB, it looks nice, I noticed input and output are within millimeters of each other, WTF?  :duh and also that output Zobel (2r7 + 100nF) are printed but not fitted.

Chipamps often work without it ... until you connect speakers with a long cable that is  8|

Re the Zobel,
Those parts were moved so to be closer to the pin. I recall our dear friend Roly mentioning best to keep the Zobel as close to output as possible so I altered the PCB slightly.
(The big resistor standing on end covering the Jaycar name) :tu:

Yes I did notice the input/out issue but they were cheap.
I doubt this PCB design would win awards for the best layout, even with my DIY skill I would have done that layout in a different way and I don't like the narrow traces.
I've not yet fired these units up so I can't make judgement as to how well they work. ??

Phil.

[phatt]

accdg. to Rod Elliott "no heatsink is too big"

[exztinct01]

that one rod elliott's referring to that comes close to providing proper heatsink to their amps

from here http://sound.westhost.com/articles/guitar-amps.htm

"Of all the brands, there is one US maker who seems to generally get most things more or less right. There have been some spectacular blunders with early valve amps and some of the re-issues, and the continued use of completely unshielded pickups and wiring inside many of their guitars is a constant source of irritation. However, they do seem to enjoy comparatively better overall reliability than many of the others, but there will still be exceptions. Many of their transistor amps are borderline IMO, but don't often fail - mainly because modern power transistors are extremely rugged and regularly outperform their datasheet maximum ratings and published safe operating area.

[exztinct01]

Hey Phil, in your LM1875 layout, I see that you had a separate pcb for the chip. Besides the chip, what components are also located in that pcb?

[phatt]

Just the 5 jumper wires with a small cap across pin 3 and 5 which are the supply rails. :tu:
The power chip was originally just bolted to a small plate of Alloy which obviously worked until it blew up,, which is why I ended up with it. by chance I had a spare chip so I got it working again. This time with a much larger heat sink.  :dbtu:

If you read the whole ESP page you will note he mentions chip amps getting too Hot.
If you read the data sheet for LM3886 the fine print reveals that you have to limit supply to +/- 28VDC if you want to drive 4 Ohm Loads. Hum not so great after all

As to what Amp Brand ESP refers to,, well you will have to ask Rod. :tu:
Some of the early peavey stuff was good but later maybe not so good. Then again you could probably say the same about a lot of brand name gear now 

Because he mentions guitars with no shielding cable then I'd guess Fender would be the obvious brand as they are still wiring up Strats without coax cable which is Plain dumb. :loco
Phil.

[J M Fahey]

Because he mentions guitars with no shielding cable then I'd guess Fender would be the obvious brand as they are still wiring up Strats without coax cable which is Plain dumb. :loco
Phil.

What do you mean by "dumb" young Australian kid?

I have been playing my unshielded wiring 1956 Stratocaster plugged into this 1 knob (volume) all of 5W beauty since forever

and I have never ever had a problem although I play LOUD.
At least that's what Nurse Emilia at the Retirement Home always complains about.

Oh!!!! I see!!!! you mean that radio which is heard in the background?
Hey !!!! , a guy playing alone needs a little backing track to play along, doesn't he?
And if I set tone control to "dark" I almost make it disappear 

Shielded wiring!!!!!!!
You young kids are all spoiled kids !!!!!!!

I alway say: "spare the rod and spoil the child!!!!!! "  :trouble

[exztinct01]

[phatt]

Hey I'll take the "Young" as a compliment. :tu:

Yes now that I'm older and wiser I've dramatically reduced most of the Dumb stuff I do.

I read a good one the other day;
Intellect is God given,, Wisdom comes by learning from all your dumb mistakes. :lmao:
Phil.

[DrGonz78]

Intellect is God given,, Wisdom comes by learning from all your dumb mistakes. :lmao:
Phil.

If that is true then I thought I would be a wise genius by now...  :duh :lmao:

[exztinct01]

i wasted about 5 hours today trying to figure out (draw) how to make a layout for my lm1875 amp based on ESP Project 72. i looked at the national datasheet and found only small difference in the schem. Someone mentioned that the bypass caps and the zobel should be as close as possible to the chip but how far is not advisable?

Phil, I've seen your lm1875 layout where the chip is separated from other components of the power amp. Is it okay if the chip is a bit far from those bypass caps?
Also, which traces or wires should I avoid getting close to each other?

[phatt]

I'm assuming you are using some kind of software to design the pcb?
If So, ideally mount all the power amp parts on the same PCB and keep the Zobel close to output pin of chip, within 20~50mm.
Don't use that LM1875 Casino amp I posted as a good example as it's far from perfect layout wise.  But it does show that in spite of crap layout they do work.

Star grounding for common is not just for the pcb of power amp layout,,
(look at incurably_optimistic star ground layout)
You have to realize it starts at the rectifier supply,, Not at the common of the PCB.
Usually it's the Com1 Join at the first 2 main filter capacitors.
Sure if the Rectifier and Filter caps and Poweramp parts are all on one PCB then it's easy to make the Com 1 point but if supply is on a separate board or chassis mounted then it's easy to loose sight of the Common path.

Somewhere on ESP site Rod mentions this point and although it won't stop the circuit from working,, it makes it harder to track down hums and buzzes. I'll have a hunt for it later.

If your rectifier and Main filter caps are separate from power amp PCB then keep the wire connections as short as you can. Say no more than 150mm,, 6 inch long.

Oh yes try and keep input and output on other ends of the Pcb is good design.
That issue is a big problem with hiZ preamp circuits and I've walked into that trap when breadboarding circuits,,, but hey,,that is how us diy folks learn,, I come highly qualified,,,,,,,,from the School of Hard Knocks. :lmao:

Hope it helps,, Phil

[exztinct01]

Thanks Phil. BTW I'm using Fritzing for designing the PCB and I'm having a hard time designing coz I don't have the LM1875 part. And I also didn't know what components should be far from each other. But your post helped me a lot.
+1 

[phatt]

All good,,
Adding

Oh yeah,, another point regarding PCB design; Most of them are geared towards computer Mother board type circuits (low voltage) so track width and pad sizes are set small.
That is not what you want for diy home drilled and hand soldered pcb.
So find the "global parameters" and setup some wider tracks and larger pad sizes before you go joining the dots.

I use the lazer printer and hot iron technique for making my PCB's and found that tracks less than .5mm wide  gives me lotsa trouble. :grr
I use 1mm drills as .8mm drills break real easy so Pads less than 2mm leave only
.5mm of copper to solder. Larger components get 3mm pads where possible.
Phil.

[J M Fahey]

Thanks Phil. BTW I'm using Fritzing for designing the PCB and I'm having a hard time designing coz I don't have the LM1875 part. And I also didn't know what components should be far from each other. But your post helped me a lot.
+1 

Ouch!! that "free" Fritzing software is f***ing EXPENSIVE!! :loco

I was about to download and try it when I read how to use it after design: they use a proprietary file format only understood by them, must be sent to their own fabrication house, prices are crazy  :duh

They charge U$ 1 a sq cm, or ~U$6 a sq in, so my typical 100W board, 10x15 cm (4"x6") would cost U$150  :grr  :trouble

I make it for about $2  :lmao:

Check that you can "export" or save the final design as a printable *graphic*  file (.gif / .jpg / .bmp / etc.) so you can print it on paper (or transparent/drafting paper)  and use it as a real size guide to make your own PCB

Or use pedal maker oriented http://diy-fever.com/software/diylc/  which is simplest and fastest to learn.
It also designs perfboard and Veroboard, besides etched PCBs.
Also, since it is VERY popular, there is a arge free library of user contributed designs (in fact, you are encouraged to add yours after successfully tested) , I suggest you look at all designs to have ideas on how others, with similar background and problems, solved theirs.

Just one small detail: there are 3 versions, first one Windows only, other two try to be "universal" (Mac, Linux, etc.)  by running on clumsy Java code ... buggy and full of problems so stick to V1

https://github.com/bancika/diy-layout-creator/releases/download/REL_3_11_0_BETA/diy_layout.exe

http://diy-fever.com/wordpress/wp-content/uploads/2009/12/layouts-01-19-2010.zip

You don't need to have an actual LM1875 in your hand to design for it, software package *must*  have a part symbol called something like "TO220 5 leg" because many chips have same size and pinout: TDA2002/2003/2006/2030/2040/2050 , LM1875, etc.

If not, draw it yourself, chip datasheet usually shows case size, leg diameter and separation, etc.
http://www.ti.com/lit/ds/symlink/lm1875.pdf
check page 13 for "mechanical dimensions" .

They also suggest a couple generic PCB designs which you can copy or use as a guide, on pages 9 and 10 .

This is what you may end up with  :
https://www.youtube.com/watch?v=Fq56qW4gP98