If you still can make a choice, a OPA134 could be a better solution, slightly less nois then the TL family, and not the risk Marcel noted. This risc is real, especially when you convert the circuit to non-inverting. If memory serves well, it has to do with so called common mode voltages.
Ok. I will not do the non-inverting unless I get different opamps. Thanks for the advice both of you. I will make a post after I have done one of the boards and done a comparison.
I went with the changes that were on the Dada Electrical website in the end. They seemed like a good compromise after I had read Ludwig, Snook and Nick de Smiths papers. I am hoping I don’t need to do the psu caps as they are not original and have been upgraded in the past.
My components arrive today and I will mod at least one board. I was looking for another pair of board so I could do a dcd 3 or 4 modification in slow time and came across this for £155 including postage. A Quad Quad 405 4x100w build by an ex Quad guy. I bought it not knowing what is inside but it works apparently.
I have done one of the boards. It is about 80-90% quieter on idle which means my ear is right next to the tweeter to hear anything, this is very good. My NAD C658 preamp won’t balance enough so I need to sit closer to the right speaker. I can’t tell the difference, perhaps the modded board sounds a little smoother. I will see if the wife can hear the difference. C17’s on the clamp almost caused a problem and my clamp board is on its last legs!
Thanks for the pictures, I think your wife will notice! Also when you have done the other board, maybe it comes more clearer. Some advice after I looked at the pictures. Place R30 and R31 on the copper side, prevents heating up components. Remove C3. Why are these 10uF BP so big? Check R35 and R36, they look damaged. I see you use the copper 'bus lane' between the elco's as speaker returns and connection of the Tx centre tap. A 'better' way is connecting the speaker returns and the Tx centre tap to the central chassis point at the rectifier. The wire with the eylet from the centre of the elco's is ok. But this maybe a lot of work in your case. But maybe the lenght of the wires permit. The ratio behind this advice: the loading pulses between the elco's could influence the voltage level at this point, also it could inject a lot of noise/rubbish in to the ground system.
I don’t know why my 10u’s are so big but they are staying, there are too many tracks lifting on the clamp board. I did one wire to the rectifier earth and the rest are all at 0v between the caps. I have the wrong voltage cap at CD1 so I need to change that and R35 R36 do look damaged so I will change them. It makes a noise when I turn it off (CD1).
The size of the BP's don't matter, I was just wondering. I think with CD1 you mean C5? If it is a higher voltage than neccesary, no problem. I saw what you did with the ground wires, I was just suggesting, it could be improved. The switch off noise is the lowest with the standard TL071 op-amp.
On the Dada page it say to decouple D1 with a 100n and decouple D2 with a 2k7. Decoupling D2 will help reduce the noise at switch off, I have not done this yet. What will removing C3 do? I will get some TL071’s if I can’t reduce the noise enough. Thanks for the advice Joost.
The thing is that the input pins, output pin and supply pins are allocated to the same pin numbers for almost all single op-amps, but what is done (if anything) with the remaining pins depends entirely on the op-amp type. So when you don't remove C3 when using another op-amp, you either have a 3.3 pF capacitor dangling between pins that have no contact with the op-amp chip, which is perfectly harmless, or you connect a 3.3 pF capacitor between two internal nodes, which may or may not cause instability.
In the case of the OPA134, pins 1 and 8 are offset trimming pins. Those are bound to be connected to bias resistors of the first stage or to current mirror degeneration resistors of the first stage. Connecting 3.3 pF there could very well reduce stability, it could also be that the effect is negligible, there is not enough information to be sure.
In the case of the TL071, one pin is an offset trimming pin and the other is either a pin with no internal connection or a "Do not connect" pin, depending on which brand's datasheet you read. 3.3 pF between 1 and 8 should be harmless when pin 8 has no internal connection, but I have no idea what happens when pin 8 is a pin that is not to be connected externally.
In the case of the OPA134, pins 1 and 8 are offset trimming pins. Those are bound to be connected to bias resistors of the first stage or to current mirror degeneration resistors of the first stage. Connecting 3.3 pF there could very well reduce stability, it could also be that the effect is negligible, there is not enough information to be sure.
In the case of the TL071, one pin is an offset trimming pin and the other is either a pin with no internal connection or a "Do not connect" pin, depending on which brand's datasheet you read. 3.3 pF between 1 and 8 should be harmless when pin 8 has no internal connection, but I have no idea what happens when pin 8 is a pin that is not to be connected externally.
Does that mean that the LM301 won’t like this mod and explode? I removed C3, put a 3K across D2 and it is almost silent when I turn it off. I did the other board but forgot to change the opamp! That channel is not working, I hope it is just the 301 and nothing else. It is too late for me to do anything now, it can wait until the morning. Marcel I did not know those pins were used differently on different opamps, I thought they were all operating in the same sense.
The LM301 requires that capacitor for stability, and the required value depends on what feedback you have around it. It could very well be oscillating without C3.
The ultrasonic signal generated by an oscillating op-amp will be suppressed by the low-pass filter between the op-amp and the rest of the amplifier, but oscillations usually also cause DC shifts. A DC shift can trigger the clamp circuit if it is large enough. The clamp circuit should survive that (as should everything else), but it had somehow turned into a short circuit on the QUAD-405 I once repaired.
All in all, there should not be any damage, but just to be sure, with the amplifier off, I would do a resistance/continuity measurement between the output terminals of the channel with the wrong op-amp to check whether the clamp has turned into a short circuit. If not, I would check if the DC feedback capacitor C2 has become a short circuit and whether the dumper output transistors Tr9 and Tr10 have a collector-emitter short. You can do all that while they are in the circuit, no need to desolder anything.
If they are all OK and the fuses are also OK, replace the op-amp and continue with what you were doing.
The ultrasonic signal generated by an oscillating op-amp will be suppressed by the low-pass filter between the op-amp and the rest of the amplifier, but oscillations usually also cause DC shifts. A DC shift can trigger the clamp circuit if it is large enough. The clamp circuit should survive that (as should everything else), but it had somehow turned into a short circuit on the QUAD-405 I once repaired.
All in all, there should not be any damage, but just to be sure, with the amplifier off, I would do a resistance/continuity measurement between the output terminals of the channel with the wrong op-amp to check whether the clamp has turned into a short circuit. If not, I would check if the DC feedback capacitor C2 has become a short circuit and whether the dumper output transistors Tr9 and Tr10 have a collector-emitter short. You can do all that while they are in the circuit, no need to desolder anything.
If they are all OK and the fuses are also OK, replace the op-amp and continue with what you were doing.
Last edited:
FS1 keeps blowing. I replaced Tr9. The clamp board is still working, it tested ok and I transposed the boards to be sure. C2 is ok. Not sure what to check next.
Check the work you done on the board when renovating it, whin it worked it must have something to do with the work you have done, check for component switch and solder work. Check the component and copper side against the working board. A picture could help.
Ok Joost, that sounds like a plan. I did go over everything I have done but will do so again. I will post pictures if I can’t find anything.