| |||||||||||||||||||||||||||
| Moderators: cmoy | Welcome. Please log in. | ||||||||||||||||||||||||||
Registration is required to post a new topic or a reply. 
| |||||||||||||||||||||||||||
DIY Workshop » β22: A discrete, cascoded, fully complementary, pure class A amplifier (part 3) | Ava12 Member Joined: Jan. 19, 2009 |  Message [#260] posted on: 01-21-2009 03:20 PM CST (US).   |  |  Ok messured some voltages and first things first, the heatsinks of the working boards get about 40-50°C warm whereas the heatsinks of the other boards don't get warm at all. so hear some things that stick out, can you something out of this, I can't  messured should be q2-r6: 268,5mv ? 309.5mV q2-q6: 11,55v okish 11.48V q1-q5: 11,67v ok 11.48V q5-q6: 12,18v ok 12.15V q5-r9: 28,21v ! 24.34V q1-r5: 495,8mv ! 309.6mV r9-cr1: 28,54V ok 28.85V d1-r6: 126,4mV ! 206.2mV d2-r8: 212,7mV ! 264.0mV r14-r16:23,66V ! 25.27V d5-q13: 21,64V ! 13.33V d6-q14: 11,49V ! 13.33V Any thoughts on this? Please |
amb  Headphone Council Joined: Apr. 1, 2004 | Message [#261] posted on: 01-21-2009 07:28 PM CST (US). | |
You are measuring across the resistors themselves, right? (Not relative to ground...) What did you get when you measured the resistance of R34 and R35? I still think that one or both of them are partially burnt. There is really no other explanation for this discrepancy.
Is your D5 installed backwards? Or is it shorted? Measure the voltage across D5 and (with the power off) check the resistance across it. Check also that CR5 is not installed backwards. |
| wnmnkh Member Joined: Apr. 20, 2008 | Message [#262] posted on: 01-21-2009 07:52 PM CST (US). | | R34 has been exchanged. Let me replace R35 then.... |
| wnmnkh Member Joined: Apr. 20, 2008 | Message [#263] posted on: 01-22-2009 12:22 AM CST (US). | | No, nothing has been fixed even after replacing R35. I guess, it is time to give in and rebuild the whole board again, which is not that bad, I guess. Good thing is that HD800ds are a little more than a month away (or even more if there isn't reasonably priced balanced cable would not come soon) |
| housing HeadWizer Joined: Dec. 16, 2006 | Message [#264] posted on: 01-22-2009 05:22 AM CST (US). | |
IMHO you should fix it rather than rebuild it afresh. You definitely can learn something in the process of fixing it. |
| wnmnkh Member Joined: Apr. 20, 2008 | Message [#265] posted on: 01-22-2009 10:06 AM CST (US). | | I don't know, we just tried and measured every possible causes on this. By the way I think I fixed the pot after some rewiring. |
| Ava12 Member Joined: Jan. 19, 2009 | Message [#266] posted on: 01-22-2009 12:35 PM CST (US). | | These are some more messurements around D5. across D5: -9,15V ! -11,80V(working board) Q9-Q13: 16,70V ! 13,81V(working board) across D6: -12,11V -11,90V(working board) Q10-Q14: 3,824V ! 13,67V(working board) After this I changed D5 again, and same Voltages again. So I'd rule out D5 nor could I find any cold joints, and these that looked suspicious I resoldered. CR5 looks ok apart from the voltage drop from -31,67 across it. But this could be caused by the high voltage between Q9 and Q13, couldn't it? So could Q9 be damaged? |
| amb Headphone Council Joined: Apr. 1, 2004 | Message [#267] posted on: 01-22-2009 03:05 PM CST (US). | | Ava12, what is the output DC offset on your problem board(s)? Also, measure the DC voltage at the outer leads of R9 and R10, as well as R30 and R31 (jumper wires), all relative to ground. What do you get? Check D9 and D10 to make sure they're installed in the correct orientation and that you have ~12V across them. Power off, let all capacitors discharge, then use your DMM's diode check function to check all the BC550C and BC560C transistors. Search this forum about how to do that (I had posted about this technique before). |
| Ava12 Member Joined: Jan. 19, 2009 | Message [#268] posted on: 01-22-2009 05:27 PM CST (US). | | So I messured the parts and this is what came out, I will do the transistors tomorrow because it's quite late over here by now. dc off: 0mV R9 to gr.: 28,40V R10 to gr.: 24,24V R30 to gr.: -0,760V R31 to gr.: -3,028V across D9: -12,43 across D10: 12,33 Greetz Ava |
| kschong Member Joined: Oct. 16, 2008 | Message [#269] posted on: 01-23-2009 10:11 AM CST (US). | | Hi Amb. Another problem came up on the board again. Using a 3 channel b22 config. Transformer and power supply in one case, amp boards in another. Previously had a history of load induced stability, but it's been solved with a zobel network. Now, on my R channel board, R34 reads about 70mV, yet my R35 reads up to about 140mV! The funny thing is that the amp is still perfectly listenable! No peculiar volume change or noise. The heat sinks for Q22 and Q24 heat up really fast though. The reading on across R9 and R11 is ~4.4V, which is normal. The output DC offset is at under 1mV. This readings are the same with or without headphone. What could be done? [Edited by kschong on 01-23-2009 at 10:14 AM.] |
| amb Headphone Council Joined: Apr. 1, 2004 | Message [#270] posted on: 01-23-2009 01:42 PM CST (US). | | kschong, see previous posts regarding wnmnkh's amp. Measure R34 and R35's resistances to make sure they are still correct. When subjected to overload, metal oxide resistors fail by increasing its resistance, which could be an explanation. Check the voltages at the mounting tabs of the two inner MOSFETs (relative to ground). What are they? Any signs of smoking/burning parts? |
| kschong Member Joined: Oct. 16, 2008 | Message [#271] posted on: 01-23-2009 08:07 PM CST (US). | | I just gathered some measurements. My DMM not fine enough to measure down to 0.47. Shall get a better DMM. D9 14.95V Q21 -27.3 (relative to ground) R34 ~65mV No signs of burning parts. R34 and R35 looks normal. Say if I change the metal oxide resistors, I should change to another with a higher rating right? |
| amb Headphone Council Joined: Apr. 1, 2004 | Message [#272] posted on: 01-23-2009 10:04 PM CST (US). | |
You probably had your red probe on ground and black probe on Q21 tab while measuring this, resulting in the negative reading. At any rate. looks like your Q23 is internally shorted from D to S. Also check R32 to make sure it's still 75 ohms. Replace Q23 (and R32 if the resistance is not correct). I suspect that R35 is probably partially burnt as a side effect of this failure. The damage is probably caused by short circuited output while you plug/unplug your headphones. Be sure to set the volume to minimum while doing so.
A partially-burnt metal oxide resistor may not look obviously charred.
No. Keep the same rating resistor. Under normal circumstances the wattage is more than enough. In failure conditions, these resistors burn up by increasing their resistance (to the point of becoming an open circuit in severe cases), which could save the amp and your headphones from much worse damage. |
| johnwmclean Member Joined: Dec. 8, 2008 | Message [#273] posted on: 01-23-2009 11:29 PM CST (US). | | I’m just finalising the details with Jeff (GJA) with my balance kit. I have a query regarding gain, what do most people go for with this configuration as a safe default? |
| amb Headphone Council Joined: Apr. 1, 2004 | Message [#274] posted on: 01-23-2009 11:53 PM CST (US). | |
It depends on your intended headphones and how "hot" your source is. Keep in mind that in balanced config the effective gain is double that of the amp's actual gain. |
| johnwmclean Member Joined: Dec. 8, 2008 | Message [#275] posted on: 01-24-2009 01:01 AM CST (US). | | Thanks AMB, Headphones will be HD650. Source output impedance from my Cambridge Audio < 50 ohms |
| amb Headphone Council Joined: Apr. 1, 2004 | Message [#276] posted on: 01-24-2009 01:14 AM CST (US). | | If the Cambridge is a typical home CD player, then its output would likely be 2Vrms at 0dBFS (unbalanced). If the CD player has balanced outputs, then you should consult the manual to see what its output level is. You'll probably want to keep the amp gain at 5 or below. |
| housing HeadWizer Joined: Dec. 16, 2006 | Message [#277] posted on: 01-24-2009 01:49 AM CST (US). | |
A thick film resistor does so also. |
| alwayshungry Member Joined: Jan. 24, 2009 | Message [#278] posted on: 01-24-2009 04:49 AM CST (US). | | Hi amb, I have been planning my b22 and have done quite a bit of reading but I still have a few questions. (1) Regarding gain, I'm using a PS Audio DL III as source which has a fair amount of juice and plan on using balanced HD650s as cans. Is a gain of 2 per board (4 overall) adequate or should I try for 5 (2) For the potentiometer, is 50K a good amount of impedance? (3) Can you tell me which front power button you used on KRMathis' B22 build? (4) How do you suggest mounting LEDs on the front of the case? Thanks. |
Mister X HeadWize Fanatic Joined: Jul. 23, 2003 | Message [#279] posted on: 01-24-2009 05:13 AM CST (US). | | This was the switch----> http://www.mouse...0045%2f1E2BL012 |
| amb Headphone Council Joined: Apr. 1, 2004 | Message [#280] posted on: 01-24-2009 06:33 AM CST (US). | | alwayshungry, are you building a balanced β22 or unbalanced? If unbalanced, a gain of 5 or 8 are both good for HD650. For balanced, consider a gain of 2 or 5. 50K pot is good. For mounting LEDs, you can use a special LED holder/mounting bezel, or just drill an appropriate sized hole and glue the LED in place. |
| Page 1 2 3 4 5 6 7 8 9 10 11 12 13 14 > >> | Prev Topic | Next Topic | ||||
 (remove _nospam_ ) |
© Chu Moy, 2001.