debug my lil mosfet follower amp

[wavenator]

http://www.miguelovitch.dk/free%20schematic/15W_MOS_IRF350_follower.jpg
i've took half of that amp
i mean one side of it
its the same amp x 2

anyway
it is driving my speaker
but it is too quiet
i mean it sound like a ruby , even more quiet

well
i think thats a bias/voltage problem
i am using a 34 volt power supply with 500ma
those are the volt reads:
IRFP350(btw i used irfp and not irf)
drain: 34
sorce: 0.5
gate : 3.8

tip142:
collector: 0.5
emiter: 0
base: 0

well
where can be the problem
btw , for the 3.3k 2w resistor i used 2k+1k+330r in series
thx guys

[alanlan]

From what I can tell, your power supply ain't up to the job.  Each side of each channel will be running at something like 3A quiescent (6A in total).  500mA just isn't enough.

[wavenator]

ohhh i got ya
so the 2 watt resistor doesnt even give the supply to pass ?

i need a 34volt/3amp atleast?
sounds right

[alanlan]

Hang on, are you using the circuit as shown or just one MOSFET and one tranny?

[wavenator]

1 mos and 1 tranny
look
there are 2 amps bridged

i built 1 amp
i used "input 1" and "output 1"

[bool]

You did insert a dc-blocking capacitor between "out" and the speaker?

[wavenator]

You did insert a dc-blocking capacitor between "out" and the speaker?

dunno what you mean
i did excatly what you can see in the picture
did you mean C4? yes i did

[wavenator]

except the power supply
is there any other assumption?

[alanlan]

You did insert a dc-blocking capacitor between "out" and the speaker?

bool is correct.  I originally thought the diagram showed one half of the amplifier.  As it stands, there will be a significant DC current flowing through your speaker = very bad for your speaker coil and it won't work very well at all.  In a way, you are lucky that you tested it with a weedy power supply.

You need a DC blocking capacitor between OUT1 and your speaker (+end of cap to OUT1 and - end to speaker and then the other end of the speaker to 0V), but it has to be quite large to maintain bandwidth at the low end.  I would suggest at least 200uF 35V.  100uF may be OK - any less and you will start to lose bottom end which is OK if it sounds OK to you.

Your power supply isn't up to the job - needs to be at least 6A.  If you think about it, Q2 takes a constant 3A all the time, but if you have a large signal peak which pushes the source of Q1 near to the supply rail, then it has to supply an additional 25V ish / 8 ohms speaker impedance - roughly 3A or so.  This is a class A power amp, so it's not very efficient and requires a lot of current steady state current.

Do you happen to have the circuit for the power supply of the original power amplifier?

A few other things:
a) you will get a large pop on power up as the DC blocking cap required charges up - not very good design really - this is one good thing about the originally intended balanced/differential concept because both sides of the speaker should pretty much come up together at start up.
b) I assume you are driving this with a suitable pre-amplifer as the power amplifier does not provide any voltage gain and you want a maximum 25Vpp across the speaker - this suggests a pre-amplifer with +/15V rails itself.  Your guitar will NOT drive this by itself.
c) I assume your transistors are mounted on heat-sinks - Q2 will be dissipating 36W all by itself and Q1 will be a lot more than this.

I'm not trying to frighten you off, but there are some not inconsiderable challenges to designing/constructing power amps, and I'm no expert on these things - there are probably other issues to think about.

Personally, unless you *really* want a high power amp as it is currently, I would suggest looking at a lower power design - perhaps to fit within your 30V 500mA, it should be more like a 5W or 10W amp.

Good luck.

[brett]

Hi

i've took half of that amp
i mean one side of it

As bool said, you need a DC-blocking capacitor, because the source of the MOSFET  will be (DC volts) at about half of the supply voltage (17 volts).  That much DC is enough to really upset a speaker.  So insert a large capacitor (470 uF) between the output and the speaker.  (+ve to the MOSFET and -ve side to the speaker.  The other side of the speaker connects to ground (I assume that's what you've already done).  If you want to get fancy, do two extra things: (i) run a 0.1uF film cap in parallel with the 470uF output cap, and (ii) run a 10 ohm resistor and 0.22uF film cap in parallel with the speaker (this will "shunt" any output at high frequency (RF) signal to ground).

With that power supply, check that it doesn't get too hot.  Also, definately make C4 a decent sized cap instead of that pathetic 47uF.  A good rule of thumb is 2200uF per amp of supply.  So it should be at least 1000uF, or use 2 (redundancy and higher efficiency) ie. 2 x 470uF.
cheers

[wavenator]

ill try those things
any way
do you have any simple mosfet amp to build
even 5 watt

btw
this amp i built
how much power it has?

[petemoore]

how much power it has?
  Probably about the amount mine did, not much, just what it sounds like.
  1/2w or so I guess.
  If the input source voltage is higher you get of course higher output...up to 'the point, stick a booster 'preamp' on the front end and keep 'feeler on the heat dissipation.

[wavenator]

i mean how much power it has on full work
when it bridged it has 15
and now?

[alanlan]

In theory, it would be 1/4 of the bridged power so if your figures are correct that would mean 3.75W, but this seems a little low for an amplifier with a 3A quiescent current.

[wavenator]

dont you have any other scheme i can use
i want to build a simple mosfet amp

[brett]

Hi
I *think* there's a MOSFET follower in my gallery (follow |Layouts Gallery| at the top of the page. 
If you use the search function, my gallery is called Brett's Bin of Bogus Bits.) 

Also see http://www.diystompboxes.com/smfforum/index.php?topic=54700.0

cheers

[brett]

Hi
for a really hot design (that's a pun - it's great AND it needs a *big* heatsink), check out Figure 1 and the notes below it here:
http://sound.westhost.com/project83.htm
have fun
Brett Robinson

[wavenator]

thx for all the schemes
ill try some
dunno how big the heatsink should be but ill get some big one for both

[alanlan]

A tip - you need to consider °C/W.  That way you don't pay too much or too little for the heatsink required for the job.  Also consider heatsink compound or a gasket.

[wavenator]

גםdo you know how much c/w does it put?