I am going to have to get my Autocostruire kit built….pronto!Delerius said:
Was this the stock kit, ie no mods? Also, could you tell me the details of the power supply you used please. I am excited by the thought that this kit might solve the problem of lack of bass on the SI amp, though I am not really sure why this might be. The 2020 kit is based on a slightly more powerful version of the SI Tripath chip and with the input cap mod the SI amp frequency response is flat to 20Hz, so flatness of response is not the problem with the SI bass, something else (power supply issues?) . The Naim 250 is no slouch in the bass department I gather, so to equal/better this amp is some achievement.
It's had no modification and was trialled using inferior QED speaker cable. I'll be using better cable when it's permanently in situ.
Critical to performance as I learned in my Naim days is the power supply. When I first received the Sonic Impacts I tried battery and cheap wall warts and suffered clipping. I then bought a 12V, 5A Power supply of EBay (or 3 to be precise) search for 'AC Adaptor DC Power Supply 12V 5A for Laptop/TFT Screen' ebay member lukeisabrennan. Next day delivery , great product and no more clipping!
Critical to performance as I learned in my Naim days is the power supply. When I first received the Sonic Impacts I tried battery and cheap wall warts and suffered clipping. I then bought a 12V, 5A Power supply of EBay (or 3 to be precise) search for 'AC Adaptor DC Power Supply 12V 5A for Laptop/TFT Screen' ebay member lukeisabrennan. Next day delivery , great product and no more clipping!
kujina said:
You could try www.hificollective.co.uk/componentshome.html - lots of quality bits and bobs, CambshireGordon
English Instructions?
Hi,
I finally was able to email Autocostruire and receive a reply. They seemed starightforward so I went ahead and ordered the 2020 amp. They shipped within about a day and I got the amp in about 3 days. Very good service indeed...
The kit looks very good and the board quality is high. The only problem is that they sent me instructions in Italian. I do realize that they are an Italian company but their web page has an English version.
Did any of you get instructions in English?
I tried emailing them but have not gotten a reply yet. I might be able to figure it out from the schematic but I would like to see if they have any warnings or gotchas in there.
Best,
Paul
Hi,
I finally was able to email Autocostruire and receive a reply. They seemed starightforward so I went ahead and ordered the 2020 amp. They shipped within about a day and I got the amp in about 3 days. Very good service indeed...
The kit looks very good and the board quality is high. The only problem is that they sent me instructions in Italian. I do realize that they are an Italian company but their web page has an English version.
Did any of you get instructions in English?
I tried emailing them but have not gotten a reply yet. I might be able to figure it out from the schematic but I would like to see if they have any warnings or gotchas in there.
Best,
Paul
Re: English Instructions?
Paul
I too found their ordering, turn-round and delivery service to be excellent - I ordered two kits which were delivered to UK within 5 days. Like you, I had Italian instructions - photocopied stapled sheets in which the photos and circuit diagram are difficult to see. I sent them an email last Friday 15th, aking if they could email me clean copy of the instuctions - or at least the page with the circuit diagram on. So far - now Tuesday 19th, I have not heard back. I find the Italian language reasonably straightforward - although technical terms are more specialised. There are some excellent colour photos on one of the web pages-
www.autocostruire.it/fc/audiodigit/class-t/1.htm
I did take liberty to 'correct' their advertising copy and I see that their web site page has been updated! Maybe if I get time I will do the same with the 'Instruzioni di montaggio".
Page 4 talks about different louspeaker loads (4/6/8 ohm). If my understanding is correct, the kit seems optimised for 6 ohm loads. I do have some difficulty with the table at the top of page 5. I think they are trying to give values for different loads, but each set of columns (4 / 6/ 8 ) has the same values apart for the last line (Q). In the piece of following text it is suggesting that for 4 ohm use just the 0.22 uF capacitor; only parallel with the 0.33 uF for 6 (and 8?) ohm loads. I plan to build more or less "as supplied" (see below) and connect some older 8 ohm speakers for testing.
I agree with you, that the quality of board and components are first class. I have almost finished building the first of my two amps - and have not encountered any problems. Start at the back with the Chip, smaller components next with larger ones last. One thing I may do on my second amp is to mount the 2 x 15 Ohm 2 watt resisters (RzL and RzR) on the underside of the board. That will leave a bit more room for the four pink inductors on the top surface and for proper seating of the two loudspeaker output sockets. It may also help with heat dissipation should these two resistors get hot. I will also add some heat sinks to the backs of the TA2020 chips.
I am wondering whether to take full advantage of the differing amplifer gains that seem to be available - see box diagram alongside the circuit schematic on page 8. Instead of soldering resisters RinL /RinR and Rfl /RfR in their respect places on the board, I am considering jumping these out with screened audio cable to a rotary switch, and putting in a full set of resisters. This will enable me to vary the input gain on the turn of the switch - could be useful, but not sure if this will introduce any instability or feedback problems - Any thoughts / advice anyone?
I hope to have my two amps up and running by this coming weekend. One will be used to drive my Quad ELS's, the other to drive my passive B&W sub used with the ESL's to 'fill-out' the lower octave. This arrangement will also allow the first amp to give all its power to the ELS's - and of course I will have the facility to be able to switch off the sub amp when/ if listening says it is not necessary.
Hope this is helpful
CambshireGordon
mourip said:
Paul
I too found their ordering, turn-round and delivery service to be excellent - I ordered two kits which were delivered to UK within 5 days. Like you, I had Italian instructions - photocopied stapled sheets in which the photos and circuit diagram are difficult to see. I sent them an email last Friday 15th, aking if they could email me clean copy of the instuctions - or at least the page with the circuit diagram on. So far - now Tuesday 19th, I have not heard back. I find the Italian language reasonably straightforward - although technical terms are more specialised. There are some excellent colour photos on one of the web pages-
www.autocostruire.it/fc/audiodigit/class-t/1.htm
I did take liberty to 'correct' their advertising copy and I see that their web site page has been updated! Maybe if I get time I will do the same with the 'Instruzioni di montaggio".
Page 4 talks about different louspeaker loads (4/6/8 ohm). If my understanding is correct, the kit seems optimised for 6 ohm loads. I do have some difficulty with the table at the top of page 5. I think they are trying to give values for different loads, but each set of columns (4 / 6/ 8 ) has the same values apart for the last line (Q). In the piece of following text it is suggesting that for 4 ohm use just the 0.22 uF capacitor; only parallel with the 0.33 uF for 6 (and 8?) ohm loads. I plan to build more or less "as supplied" (see below) and connect some older 8 ohm speakers for testing.
I agree with you, that the quality of board and components are first class. I have almost finished building the first of my two amps - and have not encountered any problems. Start at the back with the Chip, smaller components next with larger ones last. One thing I may do on my second amp is to mount the 2 x 15 Ohm 2 watt resisters (RzL and RzR) on the underside of the board. That will leave a bit more room for the four pink inductors on the top surface and for proper seating of the two loudspeaker output sockets. It may also help with heat dissipation should these two resistors get hot. I will also add some heat sinks to the backs of the TA2020 chips.
I am wondering whether to take full advantage of the differing amplifer gains that seem to be available - see box diagram alongside the circuit schematic on page 8. Instead of soldering resisters RinL /RinR and Rfl /RfR in their respect places on the board, I am considering jumping these out with screened audio cable to a rotary switch, and putting in a full set of resisters. This will enable me to vary the input gain on the turn of the switch - could be useful, but not sure if this will introduce any instability or feedback problems - Any thoughts / advice anyone?
I hope to have my two amps up and running by this coming weekend. One will be used to drive my Quad ELS's, the other to drive my passive B&W sub used with the ESL's to 'fill-out' the lower octave. This arrangement will also allow the first amp to give all its power to the ELS's - and of course I will have the facility to be able to switch off the sub amp when/ if listening says it is not necessary.
Hope this is helpful
CambshireGordon
Re: English Instructions?
There are a couple of issues to consider. First was the bit about setting the gain. The default uses a value of 68K for the feedback resistor Rf. I used the 20K, the lowest gain value sugested as I am using it with a CD player only.
There are 4x 0.22uF caps which you do not have to fit. Also, the coils are too big for the board to be honest.
Anyway, got it built no probs, fired it up and it works first time which I am rather pleased and relieved about. Listening to it in the background on cheap PC speakers for 30 minutes till I let it loose on the main ones. I will post my findings.
As for the Italian manual, Alessandro of Autocostruire sent me this link for the “English version”. http://www.autocostruire.com/audiodigit/class-t-e.pdf
Apologies to him as I am sure tried his best, but it really is the most hilariously bad bit of translation I have ever had the privilege of reading. I suspect he used one of the web based translation services. I challenge anyone to read it without falling off their chair laughing. Perhaps Andrea can send the translated version to Alessandro so it can become the “official” version.
mourip said:
There are a couple of issues to consider. First was the bit about setting the gain. The default uses a value of 68K for the feedback resistor Rf. I used the 20K, the lowest gain value sugested as I am using it with a CD player only.
There are 4x 0.22uF caps which you do not have to fit. Also, the coils are too big for the board to be honest.
Anyway, got it built no probs, fired it up and it works first time which I am rather pleased and relieved about. Listening to it in the background on cheap PC speakers for 30 minutes till I let it loose on the main ones. I will post my findings.
As for the Italian manual, Alessandro of Autocostruire sent me this link for the “English version”. http://www.autocostruire.com/audiodigit/class-t-e.pdf
Apologies to him as I am sure tried his best, but it really is the most hilariously bad bit of translation I have ever had the privilege of reading. I suspect he used one of the web based translation services. I challenge anyone to read it without falling off their chair laughing. Perhaps Andrea can send the translated version to Alessandro so it can become the “official” version.
Thanks
Thanks to CambshireGordon your thorough reply. Between your reply, the web photos, and Andreas' offer to translate I should do well. I do hope that Autocostruire decides to offer an English translation. They will certainly sell more kits via these forums if folks know that they can read the instructions.
Boy it is great having a through hole kit. I think that I will just sell my 41mz smd kit if I can find either an eagle or an eight year old to buy it...
Best,
Paul
Thanks to CambshireGordon your thorough reply. Between your reply, the web photos, and Andreas' offer to translate I should do well. I do hope that Autocostruire decides to offer an English translation. They will certainly sell more kits via these forums if folks know that they can read the instructions.
Boy it is great having a through hole kit. I think that I will just sell my 41mz smd kit if I can find either an eagle or an eight year old to buy it...
Best,
Paul
Re: Re: English Instructions?
Wow, Solar2 thanks a lot. That is a huge help.
Reading it is sort of a hoot. A little like reading some of the Japanese electronics manuals that first started showing up when I was a kid. I am tempted to do some editing and email it back to help them out...
Time to start burning some solder
solar2 said:
Wow, Solar2 thanks a lot. That is a huge help.
Reading it is sort of a hoot. A little like reading some of the Japanese electronics manuals that first started showing up when I was a kid. I am tempted to do some editing and email it back to help them out...
Time to start burning some solder
Re: Re: English Instructions?
Hello CambshireGordon,
I too have a kit and I agree with you, very good quality components, especially the pcb: densely packed, large ground planes on both sides, thick and well finished.
Instructions: as far as the yellow table at page 5 is concerned,
note that the values are correctly repeated in the three cases of nominal impedances, that is for every load you have three possibilities of capacitance (inductance is always the same!): ,22uF alone, ,33uF alone and the two paralleded ,55uF.
Since the cutoff frequency of the filter is determined by the LC couple, the three values are repeated for aech load.
What changes with speaker's load is the Quality factor (Q) of the filter, which is the inverse of the damping factor: a high Q means that the frequencies around the cutoff value will be enhanced and the filter behaviour in that range will be more unstable and prone to overshoot or oscillations.
The maximally flat response is obtained with a Q of 0,707, wich is what is used in Butterworth crossovers, but since our cutoff is far outside the audio band we can choose a higher Q and choose for the lowest cutoff frequency, for maximum effectiveness of filtering. Anyway I would not choose a Q higher than 1,2. (oops, sorry for the comas, in metric we use them as decimal separator).
Gain: I don't know if your idea of displacing the feedback network is good. First of all the circuit is very noisy and you risk to pick up digital EMI from the output and feed it back to the analog imput of the chip; even if you use very good shielded cable, consider that it has some relatively large capacitance (compared to the short traces) and you are going to modify the feedback topology of the circuit and it is hard to tell the effects.
Anyway, if you decide to experiment, please post your experience, because a switchable gain could be an interesting feature.
Enjoy your new, quality amp!
CambshireGordon said:
Hello CambshireGordon,
I too have a kit and I agree with you, very good quality components, especially the pcb: densely packed, large ground planes on both sides, thick and well finished.
Instructions: as far as the yellow table at page 5 is concerned,
note that the values are correctly repeated in the three cases of nominal impedances, that is for every load you have three possibilities of capacitance (inductance is always the same!): ,22uF alone, ,33uF alone and the two paralleded ,55uF.
Since the cutoff frequency of the filter is determined by the LC couple, the three values are repeated for aech load.
What changes with speaker's load is the Quality factor (Q) of the filter, which is the inverse of the damping factor: a high Q means that the frequencies around the cutoff value will be enhanced and the filter behaviour in that range will be more unstable and prone to overshoot or oscillations.
The maximally flat response is obtained with a Q of 0,707, wich is what is used in Butterworth crossovers, but since our cutoff is far outside the audio band we can choose a higher Q and choose for the lowest cutoff frequency, for maximum effectiveness of filtering. Anyway I would not choose a Q higher than 1,2. (oops, sorry for the comas, in metric we use them as decimal separator).
Gain: I don't know if your idea of displacing the feedback network is good. First of all the circuit is very noisy and you risk to pick up digital EMI from the output and feed it back to the analog imput of the chip; even if you use very good shielded cable, consider that it has some relatively large capacitance (compared to the short traces) and you are going to modify the feedback topology of the circuit and it is hard to tell the effects.
Anyway, if you decide to experiment, please post your experience, because a switchable gain could be an interesting feature.
Enjoy your new, quality amp!
Hello,
I've just uploaded the translated document (didn't translate the bill of materials and the tutorial on how to solder ) to my web disk, accessible clicking on the www button below.
If someone has a web space to publish it.. just do it.
Cheers
Andrea
PS Angelo, when you say "near Milano" what you mean exactly? I also live in the green Brianza...
I've just uploaded the translated document (didn't translate the bill of materials and the tutorial on how to solder
) to my web disk, accessible clicking on the www button below.If someone has a web space to publish it.. just do it.
Cheers
Andrea
PS Angelo, when you say "near Milano" what you mean exactly? I also live in the green Brianza...
ohenry said:
Okay, I got the air cores for about $30 per set shipped to the US. I installed them in two amp3 kits and they make a nice difference in tone. The bass has improved in quality and quantity. There seems to be more low end extension and the mid-bass has been enhanced as well. There are no noise issues.
The amp3's I used them on have larger power supply caps that are bypassed, and one has auricaps and the other has Sprague PIO's on the inputs. Otherwise, the amp3's are stock.
Another exciting find to make a good thing better.
Re: Re: Re: English Instructions?
Thanks Solar2 for the link to the English translation of the manual (post48)
Thanks dillo for the explanation (your post 53) which, along with the English manual, all now makes sense!
I have decided to experiment with variable gain switching - I will post my experiences later - the Amps are almost completed, but switching and casework still under way. I will only be experimenting with using three of the possible five gain settings - mainly because I could only source a 4 pole / 6 throw swith at £32GBP - which I thought to be a bit excessive for this experiment! Maplin does a 4P /3T for £1.29GBP. If vaiable gain switching is satisfactory, then I will consider investing in the more expensive switches; (I will need two, because I am building two amps - the second will drive my passive sub).
CambshireGordon
Thanks Solar2 for the link to the English translation of the manual (post48)
dillo said:
... ... Instructions: as far as the yellow table at page 5 is concerned,
note that the values are correctly repeated in the three cases of nominal impedances, that is for every load you have three possibilities of capacitance (inductance is always the same!): ,22uF alone, ,33uF alone and the two paralleded ,55uF. Since the cutoff frequency of the filter is determined by the LC couple, the three values are repeated for aech load.
... ... Gain: I don't know if your idea of displacing the feedback network is good. First of all the circuit is very noisy and you risk to pick up digital EMI from the output and feed it back to the analog imput of the chip; even if you use very good shielded cable, consider that it has some relatively large capacitance (compared to the short traces) and you are going to modify the feedback topology of the circuit and it is hard to tell the effects.
Anyway, if you decide to experiment, please post your experience, because a switchable gain could be an interesting feature.
Thanks dillo for the explanation (your post 53) which, along with the English manual, all now makes sense!
I have decided to experiment with variable gain switching - I will post my experiences later - the Amps are almost completed, but switching and casework still under way. I will only be experimenting with using three of the possible five gain settings - mainly because I could only source a 4 pole / 6 throw swith at £32GBP - which I thought to be a bit excessive for this experiment! Maplin does a 4P /3T for £1.29GBP. If vaiable gain switching is satisfactory, then I will consider investing in the more expensive switches; (I will need two, because I am building two amps - the second will drive my passive sub).
My kit came with input capacitors of value 2.2uF as detailed in the components list and circuit diagram. The make is unspecified, but they appear to be sourced from the same supplier as the rest of the polyester capacitors - little cream boxes with no manufacturer's name. Autocostruire say in their sales description that all components are of high quality. Can anyone assure us of the quality of these little cream boxes?kanaddict said:
CambshireGordon
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.