Home › ELEKTOR FORUMS › Audio & Video › PaX amplifier: invitation to discussion

ELEKTOR FORUMS

Please log in to post a reply or subscribe / unsubscribe to topics
	1 2 3 next

Topic: PaX amplifier: invitation to discussion
Author	Post
Elektor Editor 602 posts Power User	21-04-2008 13:37 Please use this forum topic to tell us what you think of the paX error correction amplifier published in the April and May 2008 issues! The author, Jan Didden, is ready to discuss issues and exchange opinions here. Jan
jan.didden 24 posts Occasional Visitor	21-04-2008 14:22 Well, one thing I am interested in is to hear from someone who has been using those Sanken darlingtons before. Any experiences with thermal stability (or not )? In my amp, there's no traditional Vbe multiplier, so no bias transistor on the heatsink for thermal feedback. That's totally unnecessary with those inbuild-diodes, and for me is a real advantage of these devices. Jan Didden Post edited by janneman01 on 21-04-2008 14:28
Shahriar 9 posts Popping In	25-05-2008 05:03 Dear Sir I will be difinetly build this nice amplifier. Because I am tired of those classical designs. But some Few points: 1- I Love Building active cross over systems, So I don't need High power amplifiers. 28VDC of power supply is enough for me. Maybe I could Redesign it for lower power but I would not be completely sure of results. so If it is possible for U, Put a lower power version too. i.e. just with a pair of output Transistors. No matter if U didn't put any PCB. just a schematic is enough 2- Please put a factor measurement table of your built prototype. i.e. Slew rate, damping factor, intermodulation distortion, signal/noise ratio and ... 3- All Famous brands Digital sources have Differential outputs But elektor never put an Amp design with XLR inputs. So if it is possible add a differential input or if U don't do that, guide us how to add a diff. stage. ThanX in advance Shahriar.Farmanesh Post edited by Shahriar on 25-05-2008 05:06 Post edited by Shahriar on 25-05-2008 05:07
jan.didden 24 posts Occasional Visitor	25-05-2008 19:54 Hello Shahriar, Niced that you want to built my amp! I'm sure you will be happy with the sound. I am also using active xovers, at the moment a heavily modified Behringer DCX2496, see my website for details if you are interested. You can lower the power supply voltage for lower power, but maybe 28V is a bit too low, I will check and let you know. I can give you the changed component values for a lower supply. I will also see if I can dig up any measurements I have. I must say after designing amps for 25 years I know what to measure and what to listen for, and that numbers don't tell the whole story, so I don't tend to measure all of the things you mentioned. Do you really need a balanced input stage? I can give you an example, but if you don't absolutely need it it is better to avoid an extra stage. Jan Didden
Shahriar 9 posts Popping In	30-05-2008 18:48 Thank you very much - Can I see Lower power design (at least components to change) in Next elektor Magazine, or to look for it in your website? - I would greatly appreciate it to see the balanced input stage you recommand! BTW, Have you seen this kind of Ballance input implemented by accuphase. I put it just to see, not anything else! http://www.4shared.com/file/48473266/61e625dd/p-3000_ea.html ThanX Shahriar
jan.didden 24 posts Occasional Visitor	31-05-2008 12:52 Hello Shariar, I must apolgize for the silence, but I have been travelling away from home. I will post the necessary changes for a lower power version either tomorrow or Monday. I had a quick look already and maybe a supply voltage of 28V is a bit too small; it would make some resistor values too small, but I'll see. Maybe 30VDC would be a good compromise? That would give you between 40 and 50W of output power. Best regards, Jan Didden
jan.didden 24 posts Occasional Visitor	01-06-2008 11:07 OK, I checked the changes necessary to run the paX amp at +/-30VDC for 40-50W output power. There are several resistor value changes and two zener value changes. I added a page to my website with the details, here: http://www.linearaudio.nl/pax-5.htm Let me know if you can work with that, and how it goes. Best, Jan Didden Post edited by janneman01 on 01-06-2008 11:08
Shahriar 9 posts Popping In	03-06-2008 08:52 Dear Jan Thank You very much. With lower power, Using two pairs of output Transistors is critical or can we use just One pair? If yes, write about what other components should be omitted! I can Guess, but Your opinion is superior! absolutely, Using two pairs can tweak Amp Parameters, but imagin when it is to drive just a Tweeter or midrange in an active cross over system! ThanX Shahriar
jan.didden 24 posts Occasional Visitor	03-06-2008 10:36 Shahriar, I have added the info to run with a single output pair to my website. Let me know if you need more info. Best regards, Jan Didden
Shahriar 9 posts Popping In	07-06-2008 10:29 Thank you very very much. Dear Jan when designing feedback amplifier PCBs, we should pay great attention to the feedback tracks. Which tracks are most important and should be kept them as short as possible in your design? I ask this because I will be probably design my own PCB for the lower power design as soon as I get free time! BTW how is the noise immunity of Error correction topology in comparison with feedback amps? Thanx again Shahriar Post edited by Thijs Beckers on 24-06-2008 16:17
jan.didden 24 posts Occasional Visitor	07-06-2008 11:39 I'm not sure it is so critical to keep it ultra-short. What IS critical is to connect it at the right point. The top of the feedback divider should be connected as close as possible to the speaker outpout as possible. Remember, it is the top of the feedback divider that will have the cleanest sound, that is where the distortion is nulled out. Also, connect the bottom of the feedback divider to the same point as the input ground, and to this point all other signal grounds like the load resistor of the AD844. Don't share these tracks with anything else. The star ground for the supply, where the local supply decoupling and the speaker return comes together, should be connected to the signal ground star via a small resistor like 1 ohms. See my layout as one way to do it. Jan Didden
jan.didden 24 posts Occasional Visitor	09-06-2008 15:27 I have changed the protection curve on the XL preadsheet on my website. The amplifier can now deliver a bit more power with complex loads. If you want to implement this change you need to change one zener diode value (two zeners) and some resistor values. Details here: http://www.linearaudio.nl/pax-5.htm Jan Didden
KioseusZnook 6 posts Popping In	02-07-2008 10:04 Hi Jan I too would be interested in adding an XLR balanced input to the amp, mainly because there'll be a big distance between the pre-amp and the PAX (driving surround speakers). What is your recommended design for this? Thanks Znook
jan.didden 24 posts Occasional Visitor	02-07-2008 12:46 Hi Znook, There are several ways to make the paX amplifier to accept a balanced input. The easiest is to use a balanced cable right up to the amp and then use an adapter from XLR to RCA. That way, you will keep the advantages of balanced operation - cancellation of common mode hum and noise - to a great extend. Second best is using an adapter cable as I describe on my website here: http://www.linearaudio.nl/bal2unbal.htm. The best in terms of performance would be a balanced cable and then a balanced to single ended conversion circuit either inside the paX box or as an external box close to the amp. This can be based on one of those well-known balanced line receivers from ADI or TI. My advice would be to check how options 1 or 2 works with some listening tests. Option 3 would be better for common mode supression but does introduce additional active circuitry. If CM noise is not a big problem, you could introduce more problems then you get rid off. Jan Didden
KioseusZnook 6 posts Popping In	02-07-2008 22:23 Jan, thanks for your input. Yes I've seen those XLR to RCA adapters and they are an option but I intend to make the amp/s the best that I can and would rather have everything 'built in' so to speak. To that end I'll probably use THAT's 1280s rather than the TI or ADI options as I believe these are a tad better. Anyways at the moment I'm just researching everything so once again thanks for your valued input. Znook.
jan.didden 24 posts Occasional Visitor	03-07-2008 13:42 Yes the THAT's will certainly work well here. You could use the exisiting supply voltage with a regulator like I use for the AD844's in the voltage amp. Jan Didden
KioseusZnook 6 posts Popping In	03-07-2008 20:43 Jan I note your post above regarding the changing of some component values with respect to the paX driving complex loads; could you clarify exactly which components these are with regard to the circuit diagrams printed in the Elektor magazines, and also to what value/s they should be changed to. Thanks. Also regarding L1; the article states 3 layers of 8 turns of 1mm5 magnet wire but at what start diameter? Other than that everything's looking good (bar that the Sanyo 2SA1208s appear to be getting in short supply...) Znook Post edited by KioseusZnook on 04-07-2008 00:07
jan.didden 24 posts Occasional Visitor	04-07-2008 08:34 Hello Znook, The changed values can be seen in the spreadsheet if you compare them to the diagram in the article, but I have just now also documented them here: http://www.linearaudio.nl/SOA-2.htm . For L1, I use a screwdriver as a core to wind it, and then afterwards slide it off. The inner diameter is about 5mm. This is not critical. Jan Didden
hadihf 30 posts Occasional Visitor	05-07-2008 07:32 Hello Jan, Can you please point to a good and cheap source of suitable enclosure for pAX. Regards Farooq
jan.didden 24 posts Occasional Visitor	05-07-2008 12:49 Hi Farook, Hah! Cheap and good? Difficult . I am also looking at one, I made my prototype myself but that takes too long and is also not cheap. I am looking at these: http://www.hifi2000.it/default.asp?id=104&mnu=104 The 'pesante' are models with heatsinks. Not really cheap but look very good. Jan Didden