I know from experience with my Nait 5i these amplifiers have a significant turn on transient and although I don't have the circuit for mine also have inbuilt short circuit protection.
From memory the latter was not in your .asc version of your clone. There are relay-delay dc protection circuits which for the muting operation use a diode connected to the secondary of the transformer which feeds an RC network to the base of a primary transistor to the pont of being on the cusp of conduction but drops immediately the mains is turned off. This transistor supplies a second transistor which energizes a relay coil.
From memory the latter was not in your .asc version of your clone. There are relay-delay dc protection circuits which for the muting operation use a diode connected to the secondary of the transformer which feeds an RC network to the base of a primary transistor to the pont of being on the cusp of conduction but drops immediately the mains is turned off. This transistor supplies a second transistor which energizes a relay coil.
Member
Joined 2009
Paid Member
Your RIAA stage has a 470R buffer followed by a 10uF cap. So OK there however there is a time constant there as far as charging the latter is concerned.
My suggestion is to safeguard both sections against unwanted noise in your system by a power amplifier output dc protection relay-delay with sufficient time constant to cover both needs.
I have used D.Self's mute arrangement. He detects mains AC after an isolating transformer and if a missing half waveform allows the hold up capacitor to drop out the mute relay disconnects. The circuit can easily be adjusted to require two half waves or even three before dropping out the mute relay.
I prefer the one missing halfwave, i.e. 10ms and then the drop out time of about 2ms to 5ms giving a maximum instant OFF delay of <15ms.
For me 100ms is too long.
I prefer the one missing halfwave, i.e. 10ms and then the drop out time of about 2ms to 5ms giving a maximum instant OFF delay of <15ms.
For me 100ms is too long.
Member
Joined 2009
Paid Member
There is also the RIAA feedback loop, at DC the feedback is very high impedance and has to charge up the large feedback capacitor before the input stage is at the proper bias conditions (~two Vbe drops above ground). The Shunt regulator takes a few seconds to increase after power-on and reach the final rail voltage then a few seconds afterwards the RIAA is stable. On power-down any abrupt change in the rail voltage is communicated to the output through the low PSRR of the phono amp. So we want to cut-off the output fairly quickly.
fyi - My power amplifiers have input protection capacitors (except TGM7 which is DC all the way through) and DC protection on the output but turn-on delay for power-amps is relatively short.
Some people use relays with SPDT contacts, for muting:
Naturally you can gild the lily by including additional resistors, using DPDT contacts instead of SPDT, installing Self's ternary relay drive circuits "to save precious electricity", and so on.
The relay's swing arm connects to the output jack signal pin.
The relay's normally-closed contact connects to the output jack ground pin.
The relay's normally-open contact connects to the amplifier output signal.
In this arrangement the output is solidly shorted to ground at all times, except when the muting control circuit decides it is safe and reasonable and wise to energize the relay coil. In particular, when AC mains is not present (e.g. unplugged), the relay coil is not energized and the output is solidly shorted to ground.The relay's normally-closed contact connects to the output jack ground pin.
The relay's normally-open contact connects to the amplifier output signal.
Naturally you can gild the lily by including additional resistors, using DPDT contacts instead of SPDT, installing Self's ternary relay drive circuits "to save precious electricity", and so on.
Member
Joined 2009
Paid Member
Board#3
I'm thinking more about Board#3 now - the idea here is to foster some experiments with active cross-overs. I could go out and buy a DSP based thingy but I'd miss the traditional learning possibility with a simpler discrete approach. I do DIY speakers, full range single driver and I'd like to extend that to a full range with supporting woofer with an XO around 300Hz (say). This doesn't need a fancy XO because the drivers I will work with can be chosen to have well behaved responses. A sealed box woofer would benefit from bass-boost.
I also have some commercial speakers, two-way and 2.5-way floorstanders that might be interesting to experiment with using an active XO in the range 2kHz to 3kHz. This is less interesting to me than the first option though.
I don't have very much pcb space available so it should be simple. I am wondering if I can realize a simple set-up, a) low-pass with optional bass boost, b) high-pass.
I'm thinking more about Board#3 now - the idea here is to foster some experiments with active cross-overs. I could go out and buy a DSP based thingy but I'd miss the traditional learning possibility with a simpler discrete approach. I do DIY speakers, full range single driver and I'd like to extend that to a full range with supporting woofer with an XO around 300Hz (say). This doesn't need a fancy XO because the drivers I will work with can be chosen to have well behaved responses. A sealed box woofer would benefit from bass-boost.
I also have some commercial speakers, two-way and 2.5-way floorstanders that might be interesting to experiment with using an active XO in the range 2kHz to 3kHz. This is less interesting to me than the first option though.
I don't have very much pcb space available so it should be simple. I am wondering if I can realize a simple set-up, a) low-pass with optional bass boost, b) high-pass.
Attachments
Perhaps the Sallen-Key family of filters might appeal to you, since their active portion is just a gain-of-positive-one buffer. Buffers made from discretes can be quite small; just a source follower or emitter follower. D. Self devotes 15 pages of SSAD2 to unity-gain buffer designs and manages to build them with as few as one, and as many as seven, transistors.
Member
Joined 2009
Paid Member
Indeed, and I was thinking of re-using the simple Diamond buffer for that. It could be fairly simple, in the spirit of the Nelson Pass B4. However, to do a nice job of bass EQ for sealed woofers - I have not yet seen an implementation of a biquad that didn't involve an op-amp for gain ?
Mind you, it is my chance to use an op-amp, which I've never done before and for bass EQ, even a 741 has enough slew rate
Mind you, it is my chance to use an op-amp, which I've never done before and for bass EQ, even a 741 has enough slew rate
Last edited:
Member
Joined 2009
Paid Member
Does this mute automatically turn off again or is it reset by the power on button.
Gareth,
Thrust me you do not want to add a fixed boost bass circuit. I take it that is the schematic on Elliot Sound Products deigned years ago by Linkwitz. If you have a 10 watt amp driving the majority of your speakers and you have a 20 dB gain at the low frequencies, the bass amp must be capable of a 100x that i.e. 1000 watt????
Thrust me you do not want to add a fixed boost bass circuit. I take it that is the schematic on Elliot Sound Products deigned years ago by Linkwitz. If you have a 10 watt amp driving the majority of your speakers and you have a 20 dB gain at the low frequencies, the bass amp must be capable of a 100x that i.e. 1000 watt????
Member
Joined 2009
Paid Member
Neither actually. I think I lost the plot many posts earlier!
The phono RIAA is based on the NAD but I've modified the circuit at the output to provide lower output impedance. This exact design has not been published before as far as I know. But I think it's fine to refer to it as the 'NAD' design since the key DNA from the NAD is all there. We're not using Op-amps for the phono. I've no op-amps on Board#1 (phono) or Board#2 (tone) which are both discrete and with +/-30V rail voltages. I am considering op-amps for Board#3 (active XO). However, it's not a given and a fully discrete Board#3 in the same vein as the other two is a possible outcome.
My interest in the bass-boost (yes, the biquad such as that shown on Elliot's site) is for bi-amping. The whole of Board#3 is an active XO for bi-amping experiments. I have more amplifiers than I have speakers. I'm not looking to have the capability of 20dB gain for the bass but there will still be a much higher power requirement for a sealed woofer bass-extension than the mid-treble drivers. There's not much music down at 30Hz so I don't think I need a 1kW amp for the occasional peak although it would be the right idea on-paper. In reality I don't think I need it. It's worth reviewing the numbers though, if I can get good results with 100W to 150W on the bass driver that would be good.
Last edited:
a) Electronic low frequency boost - What is the X.max of the bass driver unit you intend to use. Vance Dickerson quotes the case of an example by Marshall Leach who used a driver with 7mm max whereas the box driver combination (Qts 0.707 F3 35Hz) would require near enough to 19 mm for 20 Watts power input to boost the level by 8dB at 24Hz.
Looking at some speakers in the Dayton range increases in X.max go with increase with cone size. Only the very largest ones have 19mm capacity excursion. You could rule using these out on grounds of box size needed to house them.
Needless to say running a smaller driver hard will increase the heat dissipation and operation outside the linear range will result in distortion.
From your projected graph I think it advisable to further consider the prospects of success with this approach.
Member
Joined 2009
Paid Member
Actually, I don't take credit for this graph, it's a graph by Linkwitz who invented this biquad approach - I thought it was so famous that his graph would be recognized! I would have to arrange for some switchable gain/fc options for my implementation of the circuit.
At this point I don't have an active speaker project that would use this circuit, I'm just thinking ahead as I've been interested in the idea of bi-amping - a) because I have lots of amps! b) because it might produce a better result. I was originally thinking of using Board#3 to make an optical compressor - they seem like interesting things. But as I don't know if it would ever be used the active XO seemed like a better plan for something to include.
There is a thread that has inspired some of my interest in a woofer + full-range bi-amped speaker project which does suggest a driver choice but there are a number of other options: http://www.diyaudio.com/forums/full-range/273524-10f-8424-rs225-8-fast-ref-monitor.html
At this point I don't have an active speaker project that would use this circuit, I'm just thinking ahead as I've been interested in the idea of bi-amping - a) because I have lots of amps! b) because it might produce a better result. I was originally thinking of using Board#3 to make an optical compressor - they seem like interesting things. But as I don't know if it would ever be used the active XO seemed like a better plan for something to include.
There is a thread that has inspired some of my interest in a woofer + full-range bi-amped speaker project which does suggest a driver choice but there are a number of other options: http://www.diyaudio.com/forums/full-range/273524-10f-8424-rs225-8-fast-ref-monitor.html
Re - Linkwitz - Wireless World June 1978 page 71 - I have that issue. Linkwitz later forayed into planar speakers with active crossovers and driver eq.
I have used active crossovers and thought them to have resulted in improvements over my earlier DIY speakers and demonstrated these to a group of fellow enthusiasts who used to gather on a monthly basis. I saw active crossovers as a short cut to getting this aspect correct however I think good results are attainable with conventional two way designs.
Even with modelling programs it can take a while to get these correct with tuning and damping of the speaker cabinets. This is easier by far if you are replicating a design where someone has done all the work - such as with the example you have posted.
A while ago you mentioned it had been a long time since you spent much time listening to music. Build them as see how good they are before you decide what to do about using an electronic crossover.
Member
Joined 2009
Paid Member
Good thought - to be honest with myself, I don't currently have a real need for much functionality beyond Board#1 and #2!
However, I don't expect I'll be doing much listening to music for awhile still, so I should try and finish up this project based on good intentions - I'm atrocious at finishing my projects and I take advantage somewhat on a little peer-pressure here to get it done!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.