| |||||||||||||||||||||||||||
| Moderators: cmoy | Welcome. Please log in. | ||||||||||||||||||||||||||
Registration is required to post a new topic or a reply. 
| |||||||||||||||||||||||||||
runeight  Headphone Council Joined: Mar. 8, 2002 |  Message [#260] posted on: 01-14-2007 02:36 PM CST (US).   |  |  So, Ferrari, how does it sound after a few days of breaking it in? |
| Dave_M HeadWizer Joined: Jul. 1, 2006 | Message [#261] posted on: 01-14-2007 07:00 PM CST (US). | | Still waiting on the PCB to arrive  Been thinking about the amp and reading up about DC servos. Does the DC servo in this amp feed back in to the amp in such a way so that any distortion from the servo will be fed back into the audio signal (such as it is with nearly all DC servos)? If that is the case, then I think the 10uF capacitors in the servo can affect sound quallity (if they are part of the audio signal path). And it would be better to design the servo around some high quality film caps instead (0.22uF - 0.47uF). What do you think? |
| runeight Headphone Council Joined: Mar. 8, 2002 | Message [#262] posted on: 01-14-2007 07:15 PM CST (US). | | All servos feedback to the audio signal somehow. Some more than others. This particular servo is pretty well isolated from the main signal path. You can see that it looks into two 3k3 resistors which are bypassed to ground by 1000u caps in parallel with 47n caps. The servo itself has very little AC on its output. This servoing scheme is better than the original, where the servo went back to the input of the output stage, and the one we worked on for a while, where the servo went back to the input. This is actually a very good servo since the servo controls the voltage sources at the input stage. However, if you really don't like the 10u caps, there is an alternate servo shown on the schematics that I have drawn that uses two 1u caps instead. Resistor values change in the servo. Go to the first page of this thread an open the latest schematic to see this. AMB designed the board to accept either then 10uNP caps or the 1u film caps. Just for guys like you. |
| Dave_M HeadWizer Joined: Jul. 1, 2006 | Message [#263] posted on: 01-14-2007 07:27 PM CST (US). | | Ah I did not consider the 1000uF caps in parallel with 47nF. I will build with 1uF in servo instead of the 10uF... just to be different if nothing else  I know of one amp where the DC servo is not part of the audio path in any way. But it has poor control range, so overall I dont think it is as good. Also, do you think 100nF will be ok for C7 and C16? |
seroxatmad Member Joined: Nov. 13, 2005 | Message [#264] posted on: 01-15-2007 02:34 AM CST (US). | | Hi Dave The Gilmore Lite/Dynamic had poor servo control but it is outside of the audio side if i understand it correctly. Plus no capacitors are present within the amp circuit. I've noticed pics on the officail site they now have trimmer pots per channel because the servo will not work/poor performance if the transistors are not very well matched. http://headwize...ilmore3_prj.htm http://www.heada...e_board_med.jpg John [Edited by seroxatmad on 01-15-2007 at 02:37 AM.] |
tomb Headphone Council Joined: Mar. 1, 2006 | Message [#265] posted on: 01-15-2007 09:13 AM CST (US). | | Is the TL081 opamp sufficient for the DC servo? That thing is really cheap (22 cents at Jameco) and widely sourced with even NTE equivalents. Pardon another transistor question, but I've seen an either-or scenario that subsituted 2N5087/2N5088 in the same positions as BC550/560, although with reversed pinouts. Would those work as well? |
| runeight Headphone Council Joined: Mar. 8, 2002 | Message [#266] posted on: 01-15-2007 12:05 PM CST (US). | | The Gilmore servo is no farther out of the audio path than the CK2III servo. It is controlling the tail current in a constant current source in a differential input stage. A CCS holds the current constant over a wide range of voltages so that as the signal varies the current in the diff pair stays constant. It does this by using active devices that introduce their own peculiarities into the changing voltage at the tail of the pair. The servo is in the audio circuit, although it is a few steps removed from the direct audio path. The CK2III servo controls the voltage sources for the input stage. It is also several steps removed from the direct audio path. However, the CK2III servo actually has a wide control range and can handle a fairly large mismatch in components. So you get the best of both worlds. Tomb, any high gain fet input opamp will do. Even better if it's pin/pin compatible. Yes, you can use the alternative transistors, but the bias points might change and this might change the behavior of the amp. |
Ferrari HeadWizer Joined: Apr. 29, 2006 | Message [#267] posted on: 01-15-2007 02:51 PM CST (US). | |
My first impression of the CK2III was very positive and it is still the same after a few days of breaking it in. Listening with my Beyerdynamic DT880 (250 Ohm), it is a well balanced amp with tight bass and lot of details in the mid and high without any sharpness. Even with my old AKG-240 (600 Ohm) at loud volume, the amp behaves very well without any sight of harshness or distortion. For a 2 channel amp, with the output adjusted at ~30mA, this amp sounds much bigger than it actually is. Initially, I plan to build this little amp for office use (compact and light), with my MP3 player (iAudio M3L) as source. But afterwards, I must say that this amp can also act in a home rig, with good CD/DVD players as source  .A note about my CK2III: besides the Toshiba 2SC2238/2SA968 as output devices, I also use the alternate servo with 1uF MKT caps instead of the 10uF bipolar electrolytics. I don’t know if the sound characteristics will change (much) if the ouput devices BD139/140 or MJE243/253 and the default servo circuit were used. Maybe is it for other builders to try it out [Edited by Ferrari on 01-15-2007 at 03:44 PM.] |
.amb Headphone Council Joined: Apr. 1, 2004 | Message [#268] posted on: 01-15-2007 04:04 PM CST (US). | | The CK²III website is now up. I still need to add a couple of illustrations and make a few minor adjustments, but it's there for you to see now.  |
gewa HeadWizer Joined: Sep. 22, 2006 | Message [#269] posted on: 01-15-2007 04:33 PM CST (US). | | AMB That's a very informative webpage Ti, great job.
Regards |
| dBel84 HeadWize Fanatic Joined: May 5, 2006 | Message [#270] posted on: 01-15-2007 07:02 PM CST (US). | | nicely done on the web page AMB..dB |
| amb Headphone Council Joined: Apr. 1, 2004 | Message [#271] posted on: 01-15-2007 07:09 PM CST (US). | |
You used an older schematic. We changed R18/R42 to 2.37K (2.4K) and R19/R43 trimpot to 1K to give a wider range of bias adjustment. You can try using the old values and see if you could dial up 30mA. If not, then change to the new values. |
mb3k Member Joined: Mar. 2, 2006 | Message [#272] posted on: 01-15-2007 08:14 PM CST (US). | | Excellent Amb, I've been waiting for the writeup to go up |
Mister X HeadWize Fanatic Joined: Jul. 23, 2003 | Message [#273] posted on: 01-16-2007 11:24 AM CST (US). | |
There seems to be one small problem with that one.... you got to order them in lots of 2500. |
| amb Headphone Council Joined: Apr. 1, 2004 | Message [#274] posted on: 01-16-2007 03:55 PM CST (US). | |
After thinking about it a bit, I think the 0.1uF 63V film caps should be used for C20 and C21 only. The Wima MKS2 datasheets says the're rated at only 40VAC, which I think is a bit low as diode snubbers, so for C25-C32 I think a better choice would be 100V multilayer ceramic caps. The multilayer caps can also be used for C20 and C21 too. I've updated the website parts list to reflect this. |
| Mister X HeadWize Fanatic Joined: Jul. 23, 2003 | Message [#275] posted on: 01-17-2007 12:13 AM CST (US). | | I agree that those Wimas are not a good choice for the C25-C32 snubber caps but that is my reaction to using film caps as snubbers anyhow so....... |
| jarpatus Member Joined: Aug. 3, 2005 | Message [#276] posted on: 01-17-2007 01:32 PM CST (US). | | Boards arrived few days ago, thank you very much amb & everybody else involved in this project! I also managed to build first version of this amp, picture attached. Had some bad troubles fitting some parts in like those 10uF bipolar electrolytes (they are... pretty... big!) and some plastic caps. Forgot also to order female molex connectors so wiring are not pretty. However, worked perfectly from first power up!
Cannot say anything about sound quality since I don't have any decent audio source at hand right now. Also I had to use cheapes quality polyester caps, don't know how much it matters. Regulators and transistors will get burning hot so I probably need to get few heatsinks until I dare to burn this one ine... Best regards, Attachment: C2535.JPG |
| Ferrari HeadWizer Joined: Apr. 29, 2006 | Message [#277] posted on: 01-17-2007 03:58 PM CST (US). | |
Jari, to avoid the big 10uF bipolar electrolytics you can use the alternate servo if you have some 1uF MKT onhand. I think it can also work with the 2,2uF MKT caps you have, the resistors value should be changed to 1,4M and 45K approximately. |
| jarpatus Member Joined: Aug. 3, 2005 | Message [#278] posted on: 01-17-2007 04:30 PM CST (US). | | I have another board left, I'll use alternative servo for it and perhaps I'll order some polypropylene caps for it. However I was wondering that could there be any resistor value problem or similiar, since my CK2III runs very hot? I have adjusted trimpots so that there are only 7mV across R23 / R47 and only then output transistors and regulators will stay cool enough for touching them for few seconds. Above that and they will heat up so much that they cannot be touched. Should that be normal (and me to get some heat sinks) or should I check everything? Best regards, |
| runeight Headphone Council Joined: Mar. 8, 2002 | Message [#279] posted on: 01-17-2007 04:39 PM CST (US). | | jarapatus, it would be good to heatsink the regulators. Do you have an oscilloscope available? If your output devices are running really hot, the O/P stage might be oscillating. You can only know this by looking at the output. If they are running moderately hot, then you are probably ok. Does it sound ok? Also, those 10uNP caps are huge. Most of the ones we had in mind for this should be very small and fit easily on the board. [Edited by runeight on 01-17-2007 at 05:01 PM.] |
| seroxatmad Member Joined: Nov. 13, 2005 | Message [#280] posted on: 01-17-2007 05:16 PM CST (US). | | My Boards arrived 8 days after posting...thanks amb...only 8 days to get to the UK so on sorting components out to order. Anyone else from UK with theres? Years ago when tesing designs and building JLH desings i killed a few headphones so would recommend buying some really cheap ones for inital testing. I have no scope to test mine when its built but does anyone recommend a cheap PC type osc scope addon i often see advertised? Other than that a visit to my local college/university may be required. Regards john |
| Page 1 2 3 4 5 6 7 8 9 10 11 12 13 14 > >> | Prev Topic | Next Topic | ||||
 (remove _nospam_ ) |
© Chu Moy, 2001.