Batteryman, I am no electronics expert and as you have seen in the previous comments there are some strong opinions about doing it, how its a waste of money. My understanding is that it's to clamp down the zero voltage. Please see the attached snippet from Quad 306 service manual, page 10. Now read this in conjuction with the comments from LesW in the attached photos.
Attachments
My guess is that positive and negative rails are working much better in conjuction with zero voltage, which in return is keeping the temperatures under control. All I can assure is that it works fine and nothing blows up by putting diodes there.
Yes, but this doesn't explain your claim about reduced dissipation.
It's not unusual to put a diode across the suppiies to prevent reverse biasing the caps.
It's not unusual to put a diode across the suppiies to prevent reverse biasing the caps.
I have absolutely no clue, why. Could be that as the power transistors are dealing with less of the voltage swings they run cooler. Or it could be that as I have changed almost everything, it's helping to run more efficiently. Cleaner and more stable powersupply equating to much more efficient operation. I am afraid to take these out now to double check as they were pain to solder in with the insulators.
Hi Ian,
Since you like the "singleton" input, you might enjoy this thread. I admire it greatly.
Best,
Steve
https://www.diyaudio.com/community/...-class-b-retro-amp.385080/page-6#post-7244225
Many thanks for the link, Steve. That is one very interesting reworking of some great, if superseded or simply forgotten design elements. There are still some other great sounding designs that use the bootstrapped capacitor VAS and some that use the CFP output stage but I've never seen some many notable elements together as in LKA's design which I would have totally missed without your suggestion. I think I understand why you admire it - awesome!
Thanks for posting up so much info, like you I have a 306 here that I have DIY'd a few new parts like the caps but you have gone much further than me.
My 306 goes through a Quad 34 pre amp which I have also been modding or should I say upgrading old components, sound is much better but looking back both amps were tired.
Problem I have now is a slight crackle in the right channel and a small pop in between tracks or when turning on/off, currently using a CD player (Marantz CD63 ki sig - next mod job!) I tried without the 34 and it still does it so its the 306, the left channel was terribly crackled and the culprit was the legs on the transistor near the transformer (T8) had broke so I temporarily soldered them which solved the issue, the others look fine even under magnification but I have decided to change all four as I read these can be an issue after years of use and heat.
I will really look into your mods and upgrades and probably apply to my amp as you seem really happy with the results, could I ask - what supplier did you use?
I too have done the capacitor and sensitivity reduction and they have made a big improvement in detail and clarity.
I haven't changed the 120k resistors but I did put 100n caps across the main reservoir caps, which were changed to 10,000uf.
I also put 0.33uf across C7 (100uf) and changed the 560r resistors on the input to Tantalum low noise types.
I bought the parts from hificollective except the 10,00uf which are from RS.
I haven't changed the 120k resistors but I did put 100n caps across the main reservoir caps, which were changed to 10,000uf.
I also put 0.33uf across C7 (100uf) and changed the 560r resistors on the input to Tantalum low noise types.
I bought the parts from hificollective except the 10,00uf which are from RS.
The change of 120k to 62k for R6 is seen implemented by Quad in the 909 amplifier. It ensures the base of TR1 a PNP is sufficiently negative to turn on, and yes 120k can cause vagueness where the input seems to not track with quiet music passages. Quad were considering the use of R8 which similarly would have made the base negative from the 6V8 voltage rail.
A further nice group of mods can replace zeners with the outstanding LM329 devices. TR2 then can be biased properly allowing for sufficient current for the J503, rather than excess current via the former zener at D3 and its dropping resistor R5. The 306 is a excellent amp, I run two of them one channel of each amp for each speaker.
A further nice group of mods can replace zeners with the outstanding LM329 devices. TR2 then can be biased properly allowing for sufficient current for the J503, rather than excess current via the former zener at D3 and its dropping resistor R5. The 306 is a excellent amp, I run two of them one channel of each amp for each speaker.
Last edited:
Can we just add, placing a 100nF cap across a zener is a pretty (very) dumb idea. There are much better topology changes to a resistor-zener arrangement that actually work. The clue is a zener impedance is (broad-brush) 5ohm to 15ohm, BTW, suggested LM329 is ~ 1Ohm :- ) and you are fighting this with the 100nF reactance at a given frequency, which at say 100Hz (120Hz) rectified mains ripple is huge, hence achieves nothing. The amp works on current anyway, looking at voltage is the wrong direction. I have to question people like DADA who recommend such a school-boy approach to electronics. That is, it looks like a voltage rail, put a cap in it. The gain reduction with 27R, in lieu of 9R1, you are correct, it will drop the gain. However it also increases the bandwidth considerably, from around 40KHz to around 190KHz,. This would need a full intermodulation and stability check, and not just into a resistive load. I do like a high bandwidth, but it needs solid Eng and I am not sure this has been done here.
I've done most of the mods with no downside and measurements were good.
I agree about putting caps across zeners and I didn't do this.
I agree about putting caps across zeners and I didn't do this.
Which proves my point, and disproves LesW’s.
Possibly you have confused my claim with someone else’s.
Having just purchased a near-mint, original QUAD 306 - not yet landed - I am aware that I may need to address the amp's extremely high sensitivity, partnering with our 93dB-sensitive loudspeakers.
I have read where owners have simply increased the amount of the amp's (original design) negative feedback (NFB), in order to reduce the overall Voltage Gain.
Whilst I am no EEE, based upon similar experiments in NFB that I have conducted on other amps, I am loathe to go down this path, having learned that the amount of NFB has far-reaching impact on wider circuit parameters; and changes the sound of the amp - rarely for the better, IMHO.
QUESTION: So on the basis that I would like to hear the QUAD 306 - as much as possible, as the designers originally intended it - is there any consensus on how best to usefully reduce this amp's extremely high level of sensitivity - for best sound - without messing with the original levels of NFB?
(Yes, I have read this entire thread, but the answer to my question is not sufficiently clear...)
I have read where owners have simply increased the amount of the amp's (original design) negative feedback (NFB), in order to reduce the overall Voltage Gain.
Whilst I am no EEE, based upon similar experiments in NFB that I have conducted on other amps, I am loathe to go down this path, having learned that the amount of NFB has far-reaching impact on wider circuit parameters; and changes the sound of the amp - rarely for the better, IMHO.
QUESTION: So on the basis that I would like to hear the QUAD 306 - as much as possible, as the designers originally intended it - is there any consensus on how best to usefully reduce this amp's extremely high level of sensitivity - for best sound - without messing with the original levels of NFB?
(Yes, I have read this entire thread, but the answer to my question is not sufficiently clear...)
The 306 meets consumer line level standards. AS ALL AMPLIFIERS SHOULD in fact is 1.18 x on the upside. The level you experience from a CD player is always below 0.316V RMS if you realize the level is in the media you play, and not in the equipment. The 306 is best used with a passive attenuator, so best of both worlds, in that your source component can deliver pristine audio quality and the 306 amplifies that same ability. IMHO it is the best power amp you can buy - just connect it with a passive.
I changed the NFB with no audible difference. I also changed the old psu electrlytics to 10kuF.
These small updates don't changed the character of any amp.
However if you still want to reduce the sensitivity, the only way is a reistive potentisl divider on the inputs. Assuming your preamp has a typical output impedance of under 1k, then a 4k7 in series and 2.7k to ground will reduce it by about 3.
These small updates don't changed the character of any amp.
However if you still want to reduce the sensitivity, the only way is a reistive potentisl divider on the inputs. Assuming your preamp has a typical output impedance of under 1k, then a 4k7 in series and 2.7k to ground will reduce it by about 3.
No you just need to use the 306 with a passive volume control (attenuator) , inferring no change to its input resistance components or to R13 on each channel. It should dawn on anyone doing so that this is the way ALL amplifiers should be used to maintain the source component capability, namely that the input sensitivity is essentially the same as consumer line level and other than being a great amplifier explains why the OP and everyone else using a 306 ..is in awe
Last edited: