bias in leach amp

[rookie amp builder]

yes its me again with yet another stupid question (at least i think so)

I have finally built one channel of the leach super amp, and dammit it works.

but i dont understand how to set the bias. I have read other peoples posts and some are setting it to 230 mA or so ???????. in the instructions im supposed to set the bias till my DC current draw reaches 100mA at the power entrance to the board (is this correct), but at this level i see 2 problems. the output transistors get ridiculously hot even when it is at idle. When looking at the output signal (no input or anything else connected but power) on the oscilloscope there is some sort of noise/distortion surrounding the flat line, if i turn it down to just below the threshold where the "noise" disappears the transistors run a lot cooler, but my draw drops to approx 30mA. what am i doing/not doing right

thank you once again for your help

[BobEllis]

Is the behavior the same when the input is shorted? It sounds like you are either getting some RF into the front end or you have an oscillation issue.

Pictures of your layout would be useful

[rookie amp builder]

Originally Posted by BobEllis

Is the behavior the same when the input is shorted? It sounds like you are either getting some RF into the front end or you have an oscillation issue.

Pictures of your layout would be useful

i havent tried shorting the input (im guessing you mean input signal to common ground) id love to send pics, but dont know how to put them up, or send them for that matter, is there an "attach" button i dont see on here

thanx!

[nigelwright7557]

Having designed my own amplifiers I use my own method for setting bias.

I apply a small sine wave and monitor the output.
I turn the bias right down to start with then increase it until crossover distortion disappears on the scope. This method means the setting is not too little you get crossover distortion but not so much the heat generated is wasted.

If the problem is oscillation you will see it on the output.

[BobEllis]

If you click on the "Go advanced" button there is an icon that looks like a couple mountains. That will insert a link to a picture you have uploaded to a website like photobucket or you can put in a link to your own website.

Down below in the additional options section is a button labeled Manage attachments.

Yes, connect the signal input to ground or connect it to a low impedance source as Nigel suggests. See if you can adjust the time base on your scope to find the frequency of the noise you see.

You may be seeing an oscillation due to lead length or dress. Another potential problem area is if the base stopper resistors are on the board and the leads to the bases of the output devices are more than an inch or so long. Better to mount them right on the base lead of the device.

[rookie amp builder]

dammit, you cant just attach a pic like everything else, i dont use photobucket, its been 2 1/2 years since i was on facebook lol

ok im going to go check out photobucket

[AndrewT]

manage attachments allows one to attach pics and diagrams to one's post.
If you upload your first 3 then you can repeat the process for a further 3 attachments.

Before you save them:
crop the pic to what you need us to see.
resize to as few pixels as allow us to see the details.
compress using an appropriate jpg, jpeg, png, compression system.

Then upload to DIYaudio.

Please do not use other servers. They are too slow and you eventually lose links.

[AndrewT]

measure the bias current across the output emitter resistor, all 4 should be within 10% of total range (20 to 22mV) .
Leach tells you the voltage/current to set for Re=0r33

The Iq measured at the supply rails is the total of all +ve (or -ve) output bias plus quiescent current of the voltage amp stage plus the driver and pre-driver bias currents, plus the output offset current (if any).

Short the input.
Leave the output open circuit.
Set bias and then check Iq (volts across 1r0) and output offset (mVolts) and output noise (mVac)

Check what happens when you switch ON and OFF. You don't want big pulses of current going to your speakers at ON/OFF

Connect your Source and recheck everything with the Source switched OFF and again with the Source switched ON.

[rookie amp builder]

please excuse my horrific setup, but it is my first time, im nervous "tee hee"

1. the bias pot turned down
2. bias approx half way
3. time on scope slowed right down with bias at same level

(im reading from channel 2 if you can read my settings. (i really dont know how the hell to use that thing, but i understand it for the most part)

[BobEllis]

You want to turn the time base the other direction to see the waveform - this is a HF oscillation. Start @.2us and work your way down to see an entire waveform if possible.

Are your base stoppers on the board or at the output devices? Those long leads are likely contributing to instability. The lead inductance and the input capacitance of the output devices form an LC resonator. Having the resistors at the base of the output devices lowers the Q of the circuit helping to prevent oscillation. Make the leads as short as possible. What devices are you using? The original devices were slower than many modern replacements. Faster devices are more likely to oscillate.

The only comment I have about your bench is clean up a bit to avoid accidentally contacting high voltage, whether that partially inserted mains plug on the left or the rail fuses on the right. We'd all love to have a bench that looks like a NASA clean room, but most look more like yours. Mine is often more cluttered.