Hi Ian . About slipping probes when measuring . I solder mine on sometimes . I bought some cheap probes which are fine ( $3 US ) . The plating takes solder nicely . It also gives you a free pair of hands to adjust things with . The clip on probes are expensive and often loose grip . I have melted a few ( load measuring ) .
I am surprised that R29 and 30 are different .They should be the same .
I am surprised that R29 and 30 are different .They should be the same .
Hi Nigel
Yes, I mention the slipping because those new to testing their work only think of the instruments but of course, when you have to concentrate on the probes, measurement and documentation and signs of trouble simultaneously, your juggling luck runs out and ZAP! - no output stage and usually no spares, money or clue how to troubleshoot either. I think a small bag or box full of assorted leads, small jumper leads, clips and connectors is a must for your audio workbench in order to stay out of this trouble.
It's a shame for the low cost of IC clips, alligator/croc. clips or jumper leads, considering the likely number of times DIYs need them. I splashed out for "Parrot" branded hooks for my benchtop DMM a few years ago. They look like having inadequate grip but the quality is very good and they actually don't often fall off. Before that I used cheap, ($2.00) medium sized IC clips for everything but as you say, too much current eventually melts the plastic and they do short out close pitched connections. The baby IC clips are not for much current of course, but for only $1, they sure find a lot of uses.
One of the best type to make is a good set of power leads to test the project. I made several from colour coded 5 and 10 amp flex that are terminated at one end with slotted spade terminals to suit the power supply and best quality shrouded alligators at the work end. A cable-tie keeps them together - No more shorts or drop-offs!
Speaking of Radio Shack (earlier post); they marketed a pair of nice Taiwanese IC clips that simply slid onto the probes - a great idea for infrequent use and if I could find mine.......
Yes, I mention the slipping because those new to testing their work only think of the instruments but of course, when you have to concentrate on the probes, measurement and documentation and signs of trouble simultaneously, your juggling luck runs out and ZAP! - no output stage and usually no spares, money or clue how to troubleshoot either. I think a small bag or box full of assorted leads, small jumper leads, clips and connectors is a must for your audio workbench in order to stay out of this trouble.
It's a shame for the low cost of IC clips, alligator/croc. clips or jumper leads, considering the likely number of times DIYs need them. I splashed out for "Parrot" branded hooks for my benchtop DMM a few years ago. They look like having inadequate grip but the quality is very good and they actually don't often fall off. Before that I used cheap, ($2.00) medium sized IC clips for everything but as you say, too much current eventually melts the plastic and they do short out close pitched connections. The baby IC clips are not for much current of course, but for only $1, they sure find a lot of uses.
One of the best type to make is a good set of power leads to test the project. I made several from colour coded 5 and 10 amp flex that are terminated at one end with slotted spade terminals to suit the power supply and best quality shrouded alligators at the work end. A cable-tie keeps them together - No more shorts or drop-offs!
Speaking of Radio Shack (earlier post); they marketed a pair of nice Taiwanese IC clips that simply slid onto the probes - a great idea for infrequent use and if I could find mine.......
Last edited:
Absolutely Ian .
My favourite UK supplier Rapid Electronics have 4 mm to clips at less than $1 US . As you say a moments carelessness and disaster . I now have Parkinson's ( think I always did ) . My friends say to me how can you work like that . I can and I do . I especailly have to be careful of these things . I laugh mostly . I have to almost like I am diving under water hold my breath sometimes to do a measurement . It will be when a friend says can you look at this and not in my workshop I get nervous , that makes it worse . Seldom do I ask them to do it for me . The point I make is with a little thought risks can be avoided . My worse thing is soldering anything . I cope . My work is not good enough to sell due to this . I have others do it so no problem .
My favourite UK supplier Rapid Electronics have 4 mm to clips at less than $1 US . As you say a moments carelessness and disaster . I now have Parkinson's ( think I always did ) . My friends say to me how can you work like that . I can and I do . I especailly have to be careful of these things . I laugh mostly . I have to almost like I am diving under water hold my breath sometimes to do a measurement . It will be when a friend says can you look at this and not in my workshop I get nervous , that makes it worse . Seldom do I ask them to do it for me . The point I make is with a little thought risks can be avoided . My worse thing is soldering anything . I cope . My work is not good enough to sell due to this . I have others do it so no problem .
Nigel, I'm impressed by your determination in such adversity. I have no family history of Parkinson's but live in an area where there are many sufferers who need this low-stress environment for their well-being. Some are tragically young.
Electronics has been a hobby and therapy for generations in my family - I can't imagine not being able to tinker with things that entertain and inform or be able to show others how to get involved and plan their own projects. It's so much fun but eyesight (my avatar) has become a bane. It stops me completely when I can't read tiny component markings or benefit from lens aids. There are a few disorders being treated, yet they pale in significance to Parkinsons. You have my deep respect for your contributions here and sincere wishes for an easing of the symptoms.
Electronics has been a hobby and therapy for generations in my family - I can't imagine not being able to tinker with things that entertain and inform or be able to show others how to get involved and plan their own projects. It's so much fun but eyesight (my avatar) has become a bane. It stops me completely when I can't read tiny component markings or benefit from lens aids. There are a few disorders being treated, yet they pale in significance to Parkinsons. You have my deep respect for your contributions here and sincere wishes for an easing of the symptoms.
Ian , thanks . Just in case others need to know for themselves I have made almost a full recovery and they let me drive now ( I'm 56 ) . What is has done is made me so happy to be alive . I wrote some Christmas cards which is as hard as it gets . The sad bit is I didn't build stuff for myself in the past . I am just about able to do that again. I feel a bit like the prodigal son , it is so good to be welcomed back to a life I thought had gone . I dare say it's not forever I get to do things again . I will go for whatever I get . Two Christmases ago I couldn't do up my shoes or shave ! My eyes also , I think that worse .
1. Output to ground reads -40 vdc.
2. Q5 is replaced. No voltage across it (E-C), reads 0. Could it be bad cap at C6?
3. I removed the pot, checked and it works fine.
4. 0.003 vdc
I stopped measuring at this point.
When I did the initial test run, heat sink never got warm even after on for more than 10 minutes.
Hi, good to see you have some results.
Unfortunately, 40VDC on the the output is real bad, particularly for a speaker.
Did you hear audio at any point after building it?
You can temporarily remove C6, it isn't essential to operation. I think you'd realise if that didn't
restore voltage across Q5 that it is either wrongly oriented or shorted again. I would look
around the other parts there at the bias generator circuit around Q5 for shorts, open solder
joints or PCB problems. Something is wrong there. Check R15.
A) Lets check the rail voltages at Q4,6 emitters. Approx. +/-40V to ground should be there.
B) Now check the voltage between both collectors. This should read as the bias voltage applied to
the output stage or Vbias. It should be around 1.8V when adjusted but could be between 0-40V
or more if things are wrong or unadjusted. (C) If this is the case, measure the individual C-E
voltages across Q4,5 to find which is the problem area. Let us know what you find.
'Need to know those Vbe readings in the output stage (Q9-12) to check the transistors are OK.
Unfortunately, 40VDC on the the output is real bad, particularly for a speaker.
Did you hear audio at any point after building it?
You can temporarily remove C6, it isn't essential to operation. I think you'd realise if that didn't
restore voltage across Q5 that it is either wrongly oriented or shorted again. I would look
around the other parts there at the bias generator circuit around Q5 for shorts, open solder
joints or PCB problems. Something is wrong there. Check R15.
A) Lets check the rail voltages at Q4,6 emitters. Approx. +/-40V to ground should be there.
B) Now check the voltage between both collectors. This should read as the bias voltage applied to
the output stage or Vbias. It should be around 1.8V when adjusted but could be between 0-40V
or more if things are wrong or unadjusted. (C) If this is the case, measure the individual C-E
voltages across Q4,5 to find which is the problem area. Let us know what you find.
'Need to know those Vbe readings in the output stage (Q9-12) to check the transistors are OK.
Last edited:
Yes, during initial test. It also had lots of audible radio signal due to exposed board.
R15 (27 Ohm) is fine.
Yes, + and - 41 vdc.
0 between collectors of Q4 and Q6. I think Q6 (2SD667) is defective because I pulled it out and ran some test but no tone from my multimeter no matter which pin is touched. If local Radio Shack doesn't have exact one, what are acceptable substitutions?
82vdc from C to E of Q4.
0vdc from C to E of Q5.
82vdc from C to E of Q9.
0.5vdc from C to E of Q10.
81vdc from C to E of Q11.
2.5vdc from C to E of Q12.
Last edited:
Hi,
I don't know about Ian, but I don't trust your measurements.
You can't have 82V across Q9 and additional 2.5V across Q12, if your power supply total is 2x41V.
I think the best way to get some meaningful help with this circuit is to write the voltages at all nods on the schematic and then to scan it and post it. That way would be easier to troubleshoot.
I'd suggest that you connect the black probe of your meter to the ground, and leave it there. With the red probe take all the measurements and don't forget the positive or negative voltage sign (take DC measurements only).
By the way, hearing radio is very bad sign. Something is very wrong with your amp.
Hi Ruwe, yes I agree that the voltage measurements are inconsistent. It would be better as things get more complex, to only use the supplier's schematic and post the node voltages on that, if possible. (the nodes are the the junction points on the schematic representing the PCB trace junctions). I actually asked for B-E voltages (Vbe) at the output stage, which should have given a quick idea of where to look next.
I think now that applying power is just going to take out transistors if they aren't already blown. I should have insisted first up, fitting a bulb tester but was satisfied by the indication (#1045) that it worked ok.
Can you safely build yourself a bulb tester that connects a 60W incandescent lightbulb in series with the the active mains lead to the amplifier, Evenharmonics? It's elementary wiring but it's at mains voltage and dangerous unless you take great care with the proper insulation, wiring and mains fittings. Even a table lamp can be used as the bulb and socket and wired into the power supply if need be but the idea is to limit the current such that parts don't get enough to blow, yet there is enough voltage to fault-find. There are literally hundreds of posts here urging its use with all new construction and mods. A glow indicates excess idle current and a fault.
The 2SB647/D667 transistors will be hard to source other than through Ebay which means they will be dodgy anyway. The original Naim parts are ZTX653/753 from Diodes Inc. Mouser sell them for $1 or so. Unfortunately, there are no cheap and easy substitutes there that supply the current at the speed and voltage. BC639/40 can actually work but would be unreliable as Vceo is only 80V. In all substitutions, pinouts will likely be different. Google the relevant data sheets for simplest info. from Mouser or Digi-key on-line catalogues.
If you replace transistors, be sure then to use the "tester" to check that it is not about to repeat the same fault. Just replacing failed parts still doesn't eliminate the reason they failed. This is a DC amplifier - a nudge at the output stage can even cause the input transistors to fail by the domino effect so be aware that problems can keep coming back.
I think now that applying power is just going to take out transistors if they aren't already blown. I should have insisted first up, fitting a bulb tester but was satisfied by the indication (#1045) that it worked ok.
Can you safely build yourself a bulb tester that connects a 60W incandescent lightbulb in series with the the active mains lead to the amplifier, Evenharmonics? It's elementary wiring but it's at mains voltage and dangerous unless you take great care with the proper insulation, wiring and mains fittings. Even a table lamp can be used as the bulb and socket and wired into the power supply if need be but the idea is to limit the current such that parts don't get enough to blow, yet there is enough voltage to fault-find. There are literally hundreds of posts here urging its use with all new construction and mods. A glow indicates excess idle current and a fault.
The 2SB647/D667 transistors will be hard to source other than through Ebay which means they will be dodgy anyway. The original Naim parts are ZTX653/753 from Diodes Inc. Mouser sell them for $1 or so. Unfortunately, there are no cheap and easy substitutes there that supply the current at the speed and voltage. BC639/40 can actually work but would be unreliable as Vceo is only 80V. In all substitutions, pinouts will likely be different. Google the relevant data sheets for simplest info. from Mouser or Digi-key on-line catalogues.
If you replace transistors, be sure then to use the "tester" to check that it is not about to repeat the same fault. Just replacing failed parts still doesn't eliminate the reason they failed. This is a DC amplifier - a nudge at the output stage can even cause the input transistors to fail by the domino effect so be aware that problems can keep coming back.
It goes away when I moved my hand over the board so I was thinking that it's picking up RFI from somewhere.
$1 part is in cheap category IMO.
I'll report back in a few days on bulb test. Thanks for the info.
Misprint on board
I discovered a few problems.
C1 and C5 were connected wrong. Photo on ebay is correct but the board received is reversed (misprint).
Q3 and Q7 were connected wrong. Again, misprint on board.
I'm not sure why they do that. To avoid copyrights maybe? They could have noted that in the instruction or something. Such is the risk of buying cheap stuff on ebay...
I discovered a few problems.
C1 and C5 were connected wrong. Photo on ebay is correct but the board received is reversed (misprint).
Q3 and Q7 were connected wrong. Again, misprint on board.
I'm not sure why they do that. To avoid copyrights maybe? They could have noted that in the instruction or something.
Such is the risk of buying cheap stuff on ebay...The polarity of C1 and C5 is best determined experimentally. Normally, it doesn't matter, and it would not affect the amp operation that much.
If Q7 is wrong then Q8 is probably also wrong. Both are in the protection circuit though. You can omit them altogether, during test and afterwards, if you don't need output protection. They kind of affect the sound quality as well.
Q3 is really a big problem, because it will put the amp output in disbalance (huge DC voltage at the output, like in your case). I hope the drivers and the output transistors survived.
Q8 is printed correctly. Bottom side of board shows the connections right. I noticed that couple areas have been altered compared to online picture. Again, this may be due to copyrights.
I checked with my DMM but sounded ok. Is there other ways to check them?
Possibly, it's just as in the past, where the board is only correct for the parts used when it was drawn up. Maybe they were ZTX653/753 for Q4,6. Later, packagers can't buy them cheap enough to be competitive and substitute Chinese 2SD647/D667 which work OK, but like I have said over and over - check the pinouts. You download datasheets to verify what they are in case just Googling "2SD667 pinout" or "2SD677 connections" for example, doesn't bring a quick response from the net.
Should there be problems like this? Well, here we are buying from Ebay stores and this thread is about the issues we endure in dealing with those sellers just to buy ultra cheap - if we are awake to them! I've learned from long-term experience only to buy from trusted sellers who have gained a good reputation - on Ebay or anywhere.
Sellers (Ebay stores) mostly buy the kits from packagers who just fill the bag with whatever suitable parts they can buy cheap. They obviously don't bin thousands of boards when they run out of part X and printing advice of changes usually confuses as much as informs DIYs since the distribution and languages are global. It falls to the seller who isn't always on the ball with his products or service. They could be operating entirely from a bedroom whilst good ones may even email late details with invoice and schematic but do we always check our email?
Sure, there are several different kits and PCBs, but apparently, a lot more sellers with the cheapest possible problem kits and lack of understanding than you can imagine. I'm afraid it's "caveat emptor" and we can only offer suggestions based on what you discover about your particular product.
You can give them the standard junctions test with the DMM set to "diode check". If you don't know how, you can google it. Usually that's enough. The transistors should be not in circuit for proper testing. Very seldom transistor that passes the junctions test will still be faulty. I've seen many, but never in audio circuits.
Thanks, I can't tell on that NAP140c image, but several other images already have logos and watermarks I've seen from other suppliers. Hmm...!
Well, is the board signed LJM as in the the pic? If that is the same, this is possibly the most common kit available and is likely OK, as ljm_ljm is a member here on DIYaudio and has posted some of his PCB revisions in this thread even. I have seen several of his NAP140 boards assembled by local kids and they work fine, when the correct parts are supplied. (Yes, I have seen the wrong parts supplied without notice too, so I keep a small stock of the small signal parts to help out.)
However, if there is an issue with his board marked with the same part numbers as those supplied, I suspect the problem is with the part itself or your reading of the connections.