advertisement


Quad 33 desoldering PSU

After a very enjoyable couple of hours listening to music this evening, unplugged the source & speakers and checked the settings again, now it was warmed up (though the heatsink was cool to the touch - 'room temperature' really). As expected, needed some slight adjustment tweaking.

Just a tad concerned about the left driver board though. Adjusting RV100 with the meter between pins 1 and 5, the 33.5V setting drifts around anywhere between 33.33V & 33.53V. It seems to be constantly shifting quite quickly around that area. In comparison the right channel is pretty much rock solid. Why the difference in performance there between channels?

Was wondering if this is normal, or if it might indicate a flakey component somewhere on the left channel? Was even wondering if it might be the new 4K7 pot? I thought it was only the quiescent current that was meant to drift a little?

John
 
Radfordman said:
V not mV I presume you mean.

If so, it's a small percentage, about 0.6%
Click to expand...

Doh! Well spotted - have edited my post.

Yes it's not much, but just curious why this channel does it and not the other which is pretty much rock solid. I'll have a check for any flux residue causing any problems tomorrow. I usually clean pcbs after I've soldered in new components, but the 303 pcb traces are so fragile I didn't fancy touching them.
 
I would be curious too, I can't remember any of mine doing that, a slow drift perhaps while the amp stabilized, possibly.

Perhaps a can of freezer to find possible faulty component(s). I expect that other ideas will be suggested.
 
I only got fairly long drifts on the voltage setting, i.e. go back 10 minutes later and it's shifted a little bit. The bias was a lot more sensitive. I found they both vary with the mains voltage a bit too, so I decided to do the final setup in the evening when I'd most likely be listening to it. I also found the adjustments rather slow acting, it seems any pot turn has almost a degree of momentum where the increase/decrease isn't instantaneous and kind of settles over a minute or two. With all the settings I tended to err a little on the side of caution and tried to set them so they didn't float up over the recommended setting, especially with the bias as the Dada recommendation is a little over Quad's anyway.
 
Our voltage drifts quite a bit here, but would have thought once past the regulator board it'd be stable regardless, which it is on the right channel.

Just had a look at the underside of the boards and there's a LOT of flux residue that just happens to be around RV100, which is one hell of a coincidence. So I'll carefully clean that all off tomorrow and see if its settled down. Was keen on avoiding using iso-p to remove flux as I don't want to risk it potentially softening the pcb trace glue. As you know just looking at those traces in the wrong way and they can lift! Will try and clean between traces only.

The flux from the original components was never cleared off by Quad. And soldering in the new trim pot in that area there was old flux residue mixed with the new from the Multi-core solder. Will get it all cleaned up tomorrow. Fingers crossed that'll solve this slight fluctuation...
 
Unfortunately I seem to have a definate problem with the left driver board :/
Switching it on this morning the right was still bang on 33.5V between pins 1 and 5.
On the left board I couldn't get it higher than 32.3V which is less than last night.
Quite a disappointment after the initial elation yesterday... ****!
I honestly don't know where to start. R110, R111, C104, RV100 (the latter two new items)? Some component that was fluctuating last night is now perhaps increasing in resistance to get the lower voltage reading via RV100?

Could C100 contribute to this? They're the only polarised caps (I thought it was a tant but might well be an axial electrolytic) I haven't replaced. Or C2 or C3, the original 2000uF caps?

quad%20303%20-%20stereo%20power%20amplifier%20-%20cct%20dia'.jpg
 
It's a well known technique. I used it years ago on old boards. Need a can of freezer spray with an extension tube, cool individual components to see if fault changes.

Not had it cause a problem.
 
Thanks Radfordman - looks like it's a well used & effective technique. After switch off last night I did do a very quick fingertip test on each of the transistors and board drivers, and nothing was obviously amiss, but it's something I'll bare in mind. We have a Maplin nearby and they do freezer spray there. Might grab some later, thanks.
 
Had another look at the circuit diagram (posted on the previous page), and using that as a guide measured the voltages 'underneath' R102 (22K) and 'above' R111 (2K2)' which are listed as 9V and 8.5V respectively in the diagram. I get 8.76V and 8.51V, so those appear to be pretty good. At these points the voltages are fine and, after a warm up, not fluctuating.

I'm obviously getting decent sound out of the unit, from last nights listening session, and that was well over 2 hours worth. So hoping it's something reasonably simple.

So logically that could point to R111 being flakey, as I cannot get higher than 33.2V even with the pot turned fully clockwise, or perhaps R110. I'll carefully desolder one end of each and see how they measure. Assuming at this point the new components fitted (C104 and RV100) are ok, but you never know... I ordered several of each fortunately.
 
Just measuring the component with a meter may not reveal that it is faulty. The freezer spray may. Heat can also be used, but be careful if you put soldering iron near components. Touch the iron on a component and you will of course ground it, possibly destroying it and associated components.

A hair drier is safer, but more difficult to direct accurately.
 
I may have to give that a go. No transistor casings are getting hot, compared with the good board, no resistors show any signs of distress or discolouration, and the sound quality last night was extremely good with no nasties at all.

Have desoldered one end each of R110 and R111 and both measure well within specs, from still warm from being desoldered, to having cooled down. I'll strike those off the list.

I have my suspicions about RV100. Thinking logically, it's a variable device, which in itself means it can easily cause fluctuations if not working correctly. Last night when I set voltages initially, all was fine. And the pot wipers on both channels were practically in identical positions after setting, between boards. After playing music for a couple of hours, then immediately checking voltages again, was when I noticed the 33.5V was fluctuating quite quickly within a very tight range.

Today I can't even adjust to 33.5V with the pot on full adjustment. It tops out at around 33.2, with no adjustment change through the last fifth of the pots travel.

As voltages are otherwise correct and perfectly stable around this area, I am pondering whether to change that pot.

quad%20303%20-%20stereo%20power%20amplifier%20-%20cct%20dia'.jpg
 
I would not have expected anything to get abnormally hot with the small issue that you have.

It may not be relevant, but have you checked to see if you have correct output transistors in there? Modern 2N3055's are not suitable and can make the amp oscillate.

The pot is a possibility, but a guess might be that there is a noisy transistor. The freezer spray would likely show this up. When the transistor is chilled, fluctuation may be less, or not at all.
 
Just checked R130 which feeds one side of RV101. Meant to be 150K but measures 164K, which is on the very edge of the 10% tolerance Quad quote for that resistor. Might explain why the voltage isn't able to be adjusted beyond 33.2V now via RV100.

Might replace that just to rule it out.
 
Radfordman said:
I would not have expected anything to get abnormally hot with the small issue that you have.

It may not be relevant, but have you checked to see if you have correct output transistors in there? Modern 2N3055's are not suitable and can make the amp oscillate.

The pot is a possibility, but a guess might be that there is a noisy transistor. The freezer spray would likely show this up. When the transistor is chilled, fluctuation may be less, or not at all.
Click to expand...

I'll definitely get some of that freezer spray over the weekend then, possibly tomorrow with a bit of luck. Will also check the output transistors, but as the amp hasn't been touched since it was bought I'm hoping they'll be untouched. Will try changing R130 first, just incase, as it is on the verge of being out of spec. Fingers crossed it isn't a transistor.
 
quad2fan said:
I'll definitely get some of that freezer spray over the weekend then, possibly tomorrow with a bit of luck. Will try changing R130 first, just incase, as it is on the verge on being out of spec. Fingers crossed it isn't a transistor.
Click to expand...

While you are out, why not get the transistors too, should not cost much. My only partly educated guess is that it is a transistor.

Do look at the output transistors though, just see if they are the 38494's
 
quad2fan said:
Just checked R130 which feeds one side of RV101. Meant to be 150K but measures 164K, which is on the very edge of the 10% tolerance Quad quote for that resistor. Might explain why the voltage isn't able to be adjusted beyond 33.2V now via RV100.

Might replace that just to rule it out.
Click to expand...

It's a possibility.
 
You must log in or register to reply here.


advertisement


Back
Top