405man said:
Just a side note: I measured my 405 working into my main speakers, at a barely bearable sound pressure level (not bearable for anyone but me), and the voltage was +/-25V in the peaks (very bassy tune) -i.e half-way to the rails. OK, the speakers are fairly sensitive, and the room in only 20sqm, but I don't think clipping would be an issue unless cranking it up a lot.
I had a clipping indicator on the first 405 clone I built over 20 years ago and it was surprising how often it clipped, You really need at least an oscilloscope to check and if you are operating from much less than +/_50V you will clip the output stage which will be unstable and slow to recover.
Stuart
Stuart
405man said:
I used a scope, and had glitch detect on, as it's digital and isn't the fastest updating scope (altough it is battery powered, which is comforting when hooking up to the output). The scope does 100MHz though so _should_ not have missed anything.
BTW, I was planning to add clipping detection to the amps, as they play very cleanly at very loud levels (so are easily cranked up).
An observation:
The modded 405-2, with new big 22000uF EPCOS caps, has plenty of low end power now. Mids still seem a bit recessed compared to my memory from pre-mods.
Luckily I also have a 405 mk1 to compare with, as a point of reference. Although different, I expect the 405-2 to be at least as good as the mk1 in all areas.
Things i will try include paralleling the big caps with some smaller, for lower ESR at higher frequencies. Perhaps dropping in the same OP-amps as in the mk1, in case the cirquit allows it, just to try to track down the reasons for the differences.
The modded 405-2, with new big 22000uF EPCOS caps, has plenty of low end power now. Mids still seem a bit recessed compared to my memory from pre-mods.
Luckily I also have a 405 mk1 to compare with, as a point of reference. Although different, I expect the 405-2 to be at least as good as the mk1 in all areas.
Things i will try include paralleling the big caps with some smaller, for lower ESR at higher frequencies. Perhaps dropping in the same OP-amps as in the mk1, in case the cirquit allows it, just to try to track down the reasons for the differences.
Update:
As mentioned above I updated my 405-2 with massive 22000uF caps. I then only listened to the older 405 instead due to a recessed midrange in the 405-2..
Today I put two 1000uF caps in parallel with the 22000uF ones, and that made an absolutely massive difference to the sound! Before, it sounded as if a loudness compensation circuit was active, but this is now history. Before, bass was prominent to the point of dominating. Now bass is actually fairly lean, but not anemic.
There must be something going on with frequency dependent ESR.
Soundstaging (more like a huge bubble of a soundfield) is great now. Seeing into the recording and depicting it is easy, and many studio creations seem a bit slow in "action/activity" since it's so easy to hear exactly what they have been doing. No effort needed.
I still feel dynamic contrast could be slightly better, but the amp boards are still connected with push-on contacts. I will solder all that later after they are in monoblock configuration. I might also move the 1000uF caps onto the boards, if at all possible. (or 500uF, as they will now be 4 instead of 2)
Finally, OP-rolling is a possibility, altough I'm suprised how good it can sound with the old OP-amps (both in the 405 and the 405-2).
As mentioned above I updated my 405-2 with massive 22000uF caps. I then only listened to the older 405 instead due to a recessed midrange in the 405-2..
Today I put two 1000uF caps in parallel with the 22000uF ones, and that made an absolutely massive difference to the sound! Before, it sounded as if a loudness compensation circuit was active, but this is now history. Before, bass was prominent to the point of dominating. Now bass is actually fairly lean, but not anemic.
There must be something going on with frequency dependent ESR.
Soundstaging (more like a huge bubble of a soundfield) is great now. Seeing into the recording and depicting it is easy, and many studio creations seem a bit slow in "action/activity" since it's so easy to hear exactly what they have been doing. No effort needed.
I still feel dynamic contrast could be slightly better, but the amp boards are still connected with push-on contacts. I will solder all that later after they are in monoblock configuration. I might also move the 1000uF caps onto the boards, if at all possible. (or 500uF, as they will now be 4 instead of 2)
Finally, OP-rolling is a possibility, altough I'm suprised how good it can sound with the old OP-amps (both in the 405 and the 405-2).
I have checked the components and they all appear to be in the right orientation, with reference to circuit diagrams for the 405 rather than the silk screening on the PCB (which was wrong, but I knew that). I'm in the middle of constructing the +/- 50v supply at the moment but can't help thinking I might be on a hiding to nothing. What am I missing I wonder? Once the supply is built it will be much quicker to power the unit up in my lunch hour and bench test it, so hopefully I might get the answer then. With my lack of experience I am not too hopeful.
JonLawes said:
If this fails for one reason or the other, you may want to look into the "Quad 405 DCD mod3 - PCB" and tag along there! Since several of us will be building from scratch there will be plenty of opportunity for support. Should be a superior amp too, as it incorporates most known improvements.
JonLawes said:
BTW!!
Try to power up with less voltage than normal, either by using a bench supply with amp limiter (as well), or a variable transformer. Also use the other safe guards mentioned here (I never used to when I was younger, but then I used the output resistors to make toast..).
Consult someone who knows more than I do about running the amp at lower voltages. Measure output voltage (0) before hookingup speakers. Use cheap speakers first.
I tried to start it with 30v but the results were not good. I may have another go and check the op-amp rails yet, problem is its time consuming booking the power supplies out of stores and setting them all up in my lunch hour!
JonLawes said:
OK, my original 405s normally stay around 0-2 mV.
Too low voltage, or current starving, could of course give all sorts of effects, depending on topology. Hopefully you found a voltage and current that stabilised it at a good working point.
I'll have to see what happens once my supply is built. We use Farnell 0 to 30v variable supplies here so I wasn't able to get up to nearly what is required. I used both outputs of a supply to give me +30 and -30v, which I confirmed with a fluke.
I tried an AD845 in place of the TL071. I added 0.1uF bypasses to the OP supplies first. The AD845 did not sound very nice, but perhaps I would have needed to change something else to get the best out of it -the 3.3kOhm resistors on the Vcc lines, for example. Or adding bigger local bypass caps.
- Status
- Not open for further replies.