I am starting to design an active crossover to work with naim 42.5 and a pair of 90.2's.
The speakers are tannoy m1's @ 2.5KHz 12dB per octave.
Any last second advice on op-amps or any other areas?
Regards
Kieran
The speakers are tannoy m1's @ 2.5KHz 12dB per octave.
Any last second advice on op-amps or any other areas?
Regards
Kieran
Kieran
There are many options for op-amps etc, by why not go discrete and use the Naim Naxo circuit. Details can be found here:
http://www.neilmcbride.co.uk/naxocircuits.html
Geoff
There are many options for op-amps etc, by why not go discrete and use the Naim Naxo circuit. Details can be found here:
http://www.neilmcbride.co.uk/naxocircuits.html
Geoff
xover
Thanks for the link.I have decided to use op-amps because there are less conections and simple to write a report on compared to tranys.
I am now interested in what you said and if I am sold the idea i will continue with it.If i made the trany one i could make a dedicated PSU and use it on other naim kit????
Thanks for the link.I have decided to use op-amps because there are less conections and simple to write a report on compared to tranys.
I am now interested in what you said and if I am sold the idea i will continue with it.If i made the trany one i could make a dedicated PSU and use it on other naim kit????
Kieran,
A discrete crossover circuit need not be particularly complicated. A Sallen-Key circuit using followers has minimum parts and minimum connections.
Grey
A discrete crossover circuit need not be particularly complicated. A Sallen-Key circuit using followers has minimum parts and minimum connections.
Grey
Geoff,
thanks for that link.
Looks like a very interesting circuit worth to take a closer look at.
Interesting way Naim connected the transistors, haven't seen anything like it before.
Janne
thanks for that link.
Looks like a very interesting circuit worth to take a closer look at.
Interesting way Naim connected the transistors, haven't seen anything like it before.
Janne
Ok what order is it and are the 2 groups of four tranys replacing op-amps??I admit I am leaning towards making it but i need to make it 2.5kHz and understand it a bit more.I can handle electronics so throw it at me
Kiernan
If you redraw the schematics in a more traditional layout form you will find them easier to understand. In each set of four transistors, one ZTX384 and the ZTX214 form a compound pair emitter follower, with a constant current source made from the other two ZTZ384.
The first set of four transistors and associated components is a unity gain input buffer. This is followed by a passive 6dB/8ve filter stage and a 12dB/8ve active filter utilising the second set of four transistors in place of an op-amp.
The crossovers are therefore 3rd order 18dB/8ve and have been designed primarily for use with Naim SBL and Linn Kan speakers but I see no reason why they should not work well with other two-way designs. A search on the Naim site might elicit more information regarding the use of the Naxo crossover.
Geoff
If you redraw the schematics in a more traditional layout form you will find them easier to understand. In each set of four transistors, one ZTX384 and the ZTX214 form a compound pair emitter follower, with a constant current source made from the other two ZTZ384.
The first set of four transistors and associated components is a unity gain input buffer. This is followed by a passive 6dB/8ve filter stage and a 12dB/8ve active filter utilising the second set of four transistors in place of an op-amp.
The crossovers are therefore 3rd order 18dB/8ve and have been designed primarily for use with Naim SBL and Linn Kan speakers but I see no reason why they should not work well with other two-way designs. A search on the Naim site might elicit more information regarding the use of the Naxo crossover.
Geoff
Geoff
Thank you I will try and dig out more info and take on the challenge (then again I am half way down a very nice red wine).Joking apart It makes sence to use the circuits due to the fact it is designed for Naim, all I might do is put the regs in the same box and see if I can upgrade any parts.
Nice wine
Thank you I will try and dig out more info and take on the challenge (then again I am half way down a very nice red wine).Joking apart It makes sence to use the circuits due to the fact it is designed for Naim, all I might do is put the regs in the same box and see if I can upgrade any parts.
Nice wine
Kieran
Naim uses a cross-coupled buffer used extensively in video circuits, an improved version is used by Levinson. The circit can be improved by current sourceing the emitters of the input pair.
Another approach is to use two complementary j-fets which simplifies the circuit down two active devices per buffer.
Jam
Naim uses a cross-coupled buffer used extensively in video circuits, an improved version is used by Levinson. The circit can be improved by current sourceing the emitters of the input pair.
Another approach is to use two complementary j-fets which simplifies the circuit down two active devices per buffer.
Jam
Kieran
Hi there again,
The ZTX384 / 214 combination is often known as a complementary or series feedback pair.
A hunt through some electronics books should help you understand it, it's a little more complex than just a follower, owing to the feedback, but it's not too difficult.
It offers exceptionally good linearity for a simple circuit.
Jam's comments are interesting - may take a look at this.
Andy
Hi there again,
The ZTX384 / 214 combination is often known as a complementary or series feedback pair.
A hunt through some electronics books should help you understand it, it's a little more complex than just a follower, owing to the feedback, but it's not too difficult.
It offers exceptionally good linearity for a simple circuit.
Jam's comments are interesting - may take a look at this.
Andy
Andy,
My mistake Andy you are correct. A four transistor cross-coupled buffer would work in the circuit which to me is still a better choice. My favourite is still the complemantary j-fet pair, partly due to simplicity and ease of offset correction.
I find complementary circuits work better than single ended ones when driving loads like filters (equal pull up and pull down), if a single ended buffer is used it has to be idled at higher currents to avoid asymetry.
Jam
My mistake Andy you are correct. A four transistor cross-coupled buffer would work in the circuit which to me is still a better choice. My favourite is still the complemantary j-fet pair, partly due to simplicity and ease of offset correction.
I find complementary circuits work better than single ended ones when driving loads like filters (equal pull up and pull down), if a single ended buffer is used it has to be idled at higher currents to avoid asymetry.
Jam
Cheers everyone this is a good brainstorming session.I am going to have to study my SELF ON AUDIO book for help on the impovments mentioned ie levinson.
Kieran
Kieran
Hello Geoff , Grey , Janne , Jam , Andy ,
Ok I have studied the drawings and can see the complementary feedback pair (ztx384/214)with the other two and the 12k and 68r res as a contant current source.
Just to pause there.....
The CFP I have found notes on but the constant current source in that configuration I can't find anything.Can someone fill in the gap please???
There are three 6dB/Octave filters making 18dB/Octave.
At the input and output there are a 10uF cap and 100k Ohm res in a filter looking arangment. Are they??
At the input the two 330k res are a pot div.
What is the 4.7k Ohm/470pF filter for at the input (after the pot div)?????
The neg feedback is not clear to me because it looks like 3 filters in sucsesion but with the middle one not to 0v but as feedback instead it has lost me.
Last bit now the 2.2k res(input)and 10k res(output) before the ztx384 of the CFP. Why.And can Geoff or someone send me the spiced version of the schematics or EWB version Please.
Kieran
Ok I have studied the drawings and can see the complementary feedback pair (ztx384/214)with the other two and the 12k and 68r res as a contant current source.
Just to pause there.....
The CFP I have found notes on but the constant current source in that configuration I can't find anything.Can someone fill in the gap please???
There are three 6dB/Octave filters making 18dB/Octave.
At the input and output there are a 10uF cap and 100k Ohm res in a filter looking arangment. Are they??
At the input the two 330k res are a pot div.
What is the 4.7k Ohm/470pF filter for at the input (after the pot div)?????
The neg feedback is not clear to me because it looks like 3 filters in sucsesion but with the middle one not to 0v but as feedback instead it has lost me.
Last bit now the 2.2k res(input)and 10k res(output) before the ztx384 of the CFP. Why.And can Geoff or someone send me the spiced version of the schematics or EWB version Please.
Kieran
Ins and outs are coupled with a cap: That is necessary, because the circuit runs with a single supply, you have to prevent a DC offset. 100k is to connect the cap to ground, not to leave it "open" to the outside world for several reasons.
The two 330k resistors set the bias point for the input cap - appr. halve voltage - and the reference voltage point for the input transistor.
All these do not have anything to with filters.
The actual input filter is the 4k7 / 470pf RC network - this is an essential part of naim design, I suppose.
The two constant current sources in the filter circuit are formed with two NPN transistors (ZTX384) and two resistors (47k and 68R).
Neill McBrides schematic drawing is actually a bit ODD and confusing.
Please check out his preamp drawings, than everything will be clear for you immediatly.
best whishes
Klaus
The two 330k resistors set the bias point for the input cap - appr. halve voltage - and the reference voltage point for the input transistor.
All these do not have anything to with filters.
The actual input filter is the 4k7 / 470pf RC network - this is an essential part of naim design, I suppose.
The two constant current sources in the filter circuit are formed with two NPN transistors (ZTX384) and two resistors (47k and 68R).
Neill McBrides schematic drawing is actually a bit ODD and confusing.
Please check out his preamp drawings, than everything will be clear for you immediatly.
best whishes
Klaus
Superb. Thank for the notes on the input output parts and the confirmation on the other parts.
But can anyone fill in the gaps with regards the 2.2k res and 10k res?????
DC offset??
The feedback??
Please Kieran
But can anyone fill in the gaps with regards the 2.2k res and 10k res?????
DC offset??
The feedback??
Please Kieran
The 2k2 res. is part of the input circuit and the 10k is the input resistor for the output CFP. It is a part of the network already.
The filters are, as Geoff already mentioned, a passive 6db and an active 12db filter together. The output rail is feed back to the filter.
But I actually just follow the circuit paths with astonishment. These are very simple and very clever engineered circuits.
Neil McBride draw the schematics following the orginal PCB design, which Naim used for all crossover filters.
Klaus
The filters are, as Geoff already mentioned, a passive 6db and an active 12db filter together. The output rail is feed back to the filter.
But I actually just follow the circuit paths with astonishment. These are very simple and very clever engineered circuits.
Neil McBride draw the schematics following the orginal PCB design, which Naim used for all crossover filters.
Klaus
Nearly there, can someone explain the 6db passive then 12db active filter. Ican only see it as 3x6db passive!!!!
Help Kieran
Help Kieran
- Status
- Not open for further replies.