I'm recapping my Quad 303, it has original 2200uf /100V caps. I have a four nice 5600uf/100V caps from hardly ever used 1000W PA amp I'm thinking of using. But I read too high capacitance can cause oscillations and some other problems in low end. I know 20-30% higher is ok but these caps are 100% more.
I've also seen some people using 10,000uf but think that's just way too much.
Anyone has some experience with Quad 303?
I've also seen some people using 10,000uf but think that's just way too much.
Anyone has some experience with Quad 303?
2200 to 5600 will be fine. They won't cause oscillation. Going to large on a cap causes the bridge rectifier to conduct harder but over a smaller angle per cycle of mains which increases effects due to copper losses in the mains transformer. 2200 to 5600 is no problem.
I agree with that. But it's only on initial startup before they charge. The Quad transformer looks big enough to handle some short term peaks (I think). I find this little more crucial when I build tube amps. Not only I usually need larger caps for the PS filtering but the tubes take a lot longer high current draw before they warm up. But I must say that some tube amps I've build had quiet under powered transformer so the initial load would be really high but I've never had transformer to burn up yet.
I just need to be careful with the Quad because I'm not too knowledgeable when it comes to SS amps.
I just need to be careful with the Quad because I'm not too knowledgeable when it comes to SS amps.
Keep in mind that 2 of the caps are filter caps -- making them larger is probably a good thing. The other 2 are the output caps, increasing them will decrease the low frequency cut-off, which may or may not be a good thing.
dave
dave
Not just on start up. You need to visualise the ripple voltage across the cap.
The bridge rectifier only conducts when the AC voltage from the transformer exceeds the ripple voltage on the cap. Its not supplying current all the time. When the AC voltage falls away each cycle the bridge drops out of conduction and the cap supplies the power. So a small reservoir cap causes high ripple but the bridge (and therefore the transformer) supply current for a 'long' time per cycle to replenish the charge.
Fit a massive cap and the DC voltage holds up between cycles (low ripple) but now the transformer and bridge have only a short time to put that charge back and so need to conduct heavily. That 'high current for less time' brings copper and iron losses in the transformer into play and also causes any charge related currents to develop higher volt drops in internal wiring and grounds.
So as always... its a compromise.
One might make the 2 caps a 2-stage dilter by putting them in series with a small R (but big wattage) between them.
dave
dave
That could work well actually, as long as the regulator had enough headroom. Even a very small resistance would cut the ripple fairly significantly. I don't know how well the active regulator holds up on the Quad.
Thanks Dave. That looks do-able. Even a 1 ohm would only lose a couple of volts or so at full output.
10,000uF 100v is the maximum that seems to be recommended for the reservoir caps.
6,800uF 63v is probably the optimum for the output caps.
The original caps are now well over 30 years old and need replacing anyway.
They were large caps for the time.
Remember to place the solder tags upward not down towards the PCB.
This was a QUAD mod due to leakage onto the boards.
Andy
6,800uF 63v is probably the optimum for the output caps.
The original caps are now well over 30 years old and need replacing anyway.
They were large caps for the time.
Remember to place the solder tags upward not down towards the PCB.
This was a QUAD mod due to leakage onto the boards.
Andy
Don't give me any modding ideas because I love to experiment and that's where things go bad for me 🙂
I'm just glad that the 5600uf will be fine in the Quad. It's not an amp I want to mess around with. I've not touched it for 25 years and once I change the caps I hope not to touch it for another 25. The preamp scares me a bit tho. Looks quiet complex inside, I'd hate for something to go wrong there. I was actually thinking of making basic preamp using single 6SN7 just in case the Quad preamp goes bad. Would have to be transformer-less preamp I think since it might be too hard or expensive to find good enough quality output trafo for the tube.
I'm just glad that the 5600uf will be fine in the Quad. It's not an amp I want to mess around with. I've not touched it for 25 years and once I change the caps I hope not to touch it for another 25. The preamp scares me a bit tho. Looks quiet complex inside, I'd hate for something to go wrong there. I was actually thinking of making basic preamp using single 6SN7 just in case the Quad preamp goes bad. Would have to be transformer-less preamp I think since it might be too hard or expensive to find good enough quality output trafo for the tube.
Ok, it's recaped now. Thanks for the advice. 3 out of the 4 original caps had a beer like head dried foam on top from leaking so they were pretty bad. I haven't listened to the Quad for a long time but the sound seems to be much tighter now, drums in particular are rally punchy. Maybe just my imagination tho.
Anyone knows what temperature the heat sing normally should run at idle or low volume? Seems to be quiet hot. I can't remember from the past past since it was behind cupboard for years. I can't find my temp meter now but just guessing by touch it's at about 40c or so.
Anyone knows what temperature the heat sing normally should run at idle or low volume? Seems to be quiet hot. I can't remember from the past past since it was behind cupboard for years. I can't find my temp meter now but just guessing by touch it's at about 40c or so.
Last edited:
That all sounds like a good result. 40C is pretty cool in the scheme of things. The quiescent bias current determines the power dissipation at idle in he output and regulator, so as long as that is set correctly it will be fine.
I didn't actually checked or set anything since it worked fine for years. I just changed the caps. I guess I should have checked things for drift but my good Voltmeter is dead and all I have right now is one of those $10 el-cheapo ones which I don't trust for any serious work.
Have been looking for a new meter for around $50 but it's so hard these days with so much rubbish from China. Every time I think I've found one, I do some deep research only to discover it has some major flaws like continuity not being instant etc. Can't judge by looks these days, they make them nice on the outside but junk on the inside.
Have been looking for a new meter for around $50 but it's so hard these days with so much rubbish from China. Every time I think I've found one, I do some deep research only to discover it has some major flaws like continuity not being instant etc. Can't judge by looks these days, they make them nice on the outside but junk on the inside.
- Status
- Not open for further replies.