For big amps, I prefer to use separate protection that is detected at the output. This is because it includes: Voltage offset, temperature, output short, internal amp breakdown, and oscillation. Any one of these factors can effect the amp and its load. For us, since we MUST use E-I protection due to the fact that we operate at 50-100V rails, and the output devices need it. Everything else is added for almost no extra cost.
AndrewT said:
The resistor and capacitor values are calculated from filter table coefficients for an 8-pole Butterworth response.
An 8-pole filter (-48dB/octave) is probably overkill. I came up with this circuit 10+ years ago, just after I first learnt how to design active filters from filter tables after concluding that typical circuits relying on a single pole RC with a time constant of anything up to several seconds for sufficient LF attenuation are next to useless - rail fuses are often likely to blow due to the heavy DC current flowing through the burning woofer voice coils before the DC fault protection operates.
This is the first time I’ve simulated it. I’ve always used a 10Hz f3 and it works fine.
Response time is proportional to the cut-off frequency.
With a 10Hz corner frequency the 5V threshold time is actually 75mS. With capacitors twice as large for a corner frequency of 5Hz the 5V threshold response is 150mS.
Cheers,
Glen
Attachments
chascode said:
Circuit boards do sound worse than hard wired, and for a variety of reasons. I would like any folks who I may compete with (when I/we finally get into the manufacturing of audio gear 'proper') to please continue to use circuit boards. Thank you.
I use Teflon boards for my best stuff. Expert layout is most important. Most layouts are not professional, unfortunately.
Quote
"
Hi Bonsai,
Regarding offset correction and input stage imbalance, see my post #9842 and let me know if you agree.
Cheers,
Bob"
Good explanation Bob - I agree.
I am in the process of removing the blocking cap (1000uF 16V in parallel with a 1uF stacked foil) from my amp, so this discusssion is particularly relevant. A complication in my case is that the lower leg is 220 Ohms and I am also moving to a fully balanced input via a diamond buffer in a amongst a few other upgrades.
I purposely kept the feedback resistor low ( 4.9k made up from 2k2 and 2k7) to reduce the bias current errors (fully compl. balanced input) because I am runining the diff amps at 5mA per side. Input side of my amp is DC coupled and Rin is 10k Ohms.
"
Hi Bonsai,
Regarding offset correction and input stage imbalance, see my post #9842 and let me know if you agree.
Cheers,
Bob"
Good explanation Bob - I agree.
I am in the process of removing the blocking cap (1000uF 16V in parallel with a 1uF stacked foil) from my amp, so this discusssion is particularly relevant. A complication in my case is that the lower leg is 220 Ohms and I am also moving to a fully balanced input via a diamond buffer in a amongst a few other upgrades.
I purposely kept the feedback resistor low ( 4.9k made up from 2k2 and 2k7) to reduce the bias current errors (fully compl. balanced input) because I am runining the diff amps at 5mA per side. Input side of my amp is DC coupled and Rin is 10k Ohms.
KBK pointed out that point to point wiring sounds better, in general. I tend to agree, and that is why you see so many wires strung like a clothes line in the Blowtorch preamp. It was better to route the signal in the air, itself, rather than making a long trace on the circuit board. We can hear the difference.
Not sensitive to elmag. fields generated inside the box? It is difficult to me to believe, I have always had opposite experience (long airwires = worse sound and more induced emi by cap. coupling).
supply voltage was something around 24V and the speaker impedance was 8R. Transformer was small too.myhrrhleine said:
I don't know how long the speaker was mistreated.
The voicecoil wasn't small, but the wire was really tiny and the wire was burnt only in small area.
Maybe there was a 'hot spot' there and the speaker would have taken more power, if the voicecoil had the chance to heat up slowly.
But this is pure speculation of mine.
regards
PMA, the 'long wires' are not on purpose, but better than an equivalent length of 'wire' on a circuit board. When the 'box' dimensions were first made, we had high hopes of filling it up with added boards and switches. We made it too big for most client's needs, but just big enough for a full accessory package, including, a phono stage, an output buffer board and balanced input switching.
However, we used the BEST Teflon covered pure silver wire that we could find, and went out of our way to keep the wires separate from each other. Our Xtalk wound up to be about 100dB at 20KHz, so what is the problem?
Now, IF you want to see an example of what you want, then look at the internal pictures of the Parasound JC-2 preamp. There we use almost NO wire, this is also true with the JC-80 preamp.
There is room for discussion with that design, and it was designed by the SAME people who made the CTC Blowtorch. Think of it as a Porsche Boxster compared to a RUF 911.
However, we used the BEST Teflon covered pure silver wire that we could find, and went out of our way to keep the wires separate from each other. Our Xtalk wound up to be about 100dB at 20KHz, so what is the problem?
Now, IF you want to see an example of what you want, then look at the internal pictures of the Parasound JC-2 preamp. There we use almost NO wire, this is also true with the JC-80 preamp.
There is room for discussion with that design, and it was designed by the SAME people who made the CTC Blowtorch. Think of it as a Porsche Boxster compared to a RUF 911.
John,
thanks for the explanation. I know the internal pictures of the JC-2 preamp, as well as the full schematics. Soon I will have a chance to see the preamp and to listen to it.
thanks for the explanation. I know the internal pictures of the JC-2 preamp, as well as the full schematics. Soon I will have a chance to see the preamp and to listen to it.
PMA and everyone else. The Blowtorch example is not a universal sample for everyone to copy. The extra long internal wires were a 'necessary compromise' in order to keep the components spaced away from each other. Moreover, we used unshielded teflon silver wire of the highest quality that we could find, and spaced it in the air itself. Please remember, there is NO AC 50-60Hz within the control enclosure itself and the outside aluminum case is both thick and virtually sealed from the outside. I would not say that it is air or water tight, but it has no vents or open gaps for air circulation. Still, the silver-on-silver switches can sulfide, but they are easily cleaned by just moving the pot or switch back and forth a few times. If you go back and re-read the negative comments on the internal wiring at the beginning of this thread, you might note how I initially thought about the criticisms of this approach.
John, I really do not intend to argue. I can see that the connectors are insulated, so HF has a chance to get inside the box along signal wires and then capacitively coupled. That was my concern, not 50/60Hz.
There isn't much of a slot, just enough to put the individual connectors through. What frequency would get in? Mega Hz, Giga Hz? Even so, we can't stop everything, without other compromises. For example, we did not provide shorting caps on the unused inputs. That is a way in for RF. We must assume that people do not live in extremely high RF environments, such as might be necessary in some industrial environments.
Yes, I was showing Nelson my first complementary differential input based power amp that I originally built 40 years ago.
john curl said:
I was 11 years old then, and was happy because I finally could find one NPN transistor for my first bass guitar amp! It was 4W class A amp made on Germanium transistors, quasi-complementary design, 4W output (huge!), made in a toolbox.
- Status
- Not open for further replies.