Need some ideas for a power-supply problem

[dutch diy]

I'm currently building the 1996 John-Linsley-Hood amp (2 for a dual-mono setup).

I've build the regulated power supply from the article but ran into a problem. I've slightly modified the circuit to accomodate a fine tuning to + and - 22V.

The+22V half is ok. only the -22V half is not working ok.
With no load it works ok . With any load (from 470Ohm down to 2*12V20 W lightbulb) the voltage goes to 24.7V or more.
When the base of the power-transistor is not connected the output voltage is nearly constant (until the currentlimit of the 7915 is reached).

The problem probably is due to the totally different internal schematic of the voltage regulators (7815 and 7915).
I've tried so far Thomson and Motorola 7915's and this didn't make any difference.
The measured voltage on the 7915 from gnd=pin to out=pin (with a 470Ohm load from Vout to 0V) is 16,7 V
I've experimented with various NPN power transistors but the result always was an unstable -22V.

As i do not have access to Spice I could use some ideas.

See http:\\www.knoware.nl\users\cfw\amp for my amp site. I'll be posting a circuit diagram of the actual circuit after this.

 

[jam]

Regulators

Try using the LM317 and LM337.

Jam

 

[phil]

The 7815 and 7915 regulators are only rated at 1 amp. I thought this psu circuit used LM338K regulators which are rated at 5 amps. Ask Geoff Moss about it. He has the Class A site.

 

[dexter]

Dear Dutch DIY,

I also use the same psu as yours (the one in 96 article), no problem with mine except slight voltage difference between polarity to ground
.
Be careful.... 7915 and 7815 have a different pin out.
I suspect that was your problem, check your data book again
you also have to make sure that you got the boostrap transistor (ebc) rightly connected

However, do ask Geoff Moss, he's the real expert on JLH 10W (got an excellent web site build for the thing) and ....really kind and helpful

sorry if not much of a help..

kind regards.

 

[Geoff]

Dexter

Thanks for the complement!

Jos

As Dexter has suggested, check the connections to the 7915. The pins on the 7815 are input-ground-output, whereas those on the 7915 are ground-input-output.

If this does not reveal an incorrect connection, and assuming that all other wiring is correct and that you have no faulty components, try the following:

Reduce R4 from 1k5 to 1k0 and R3 from 470R to 220R. This will increase the current through the voltage divider so that the ground pin current of the 7915 will have less effect on the output voltage level. If this proves successful, I suggest you change R1 and R2 for the 7815 as well.

The National Semiconductor datasheet recommends a value of 1K0 for R1/R4 when the 7815/7915 is used in adjustable voltage circuit.

If you still have a problem after following the above, send me an email and I'll give the matter some further thought.

Geoff

 

[Jorge]

Hello,

Is the JLH - the old master - 96 circuit somewhere in the WEB?

TIA,

Jorge

 

[Geoff]

Jorge

It most certainly is! Have a look at my site:

http://www.gmweb.btinternet.co.uk/index.htm

Geoff

 

[dutch diy]

Thanks

Guys,

thanks for the suggestions.

I was already aware of the different pinlayout (smoked some 7915's before).

I'll go fo lowering resistor values first.

Dexter,

The difference in voltage to ground had me set up the small variable part in the power supply.

Phil,

the 7815/7915 are indeed rated 1A, that is why, in this design, there sits a booster transistor parallel, which should take the bulk of the current.

Jos

 

[dutch diy]

Problem solved

Guys,

problem has been solved thanks to a suggestion from Geoff Moss.

Fitting 100uF from Vout to 0V solved it.
Probably an oscillation problem.
On a completed system this would be suppressed by the capacitors on the amp pcb.

I'll close this thread for now.

Jos