My Aleph5 has been working fine for quite some time, however I was irritated by a slight buzz from the toroid, so I decided to replace the 650VA unit with 2 x 500VA units.
After turning-on the mains fuse blew, then on repeat power-up the left channel was gone. (I had to do some drilling and I think a filing may have found its way under the PCB).
Anyhow, I'm now at the hair-pulling frustration stage. Basically I get about 3-4V under the positive rail at the output.
Removing the PCB so all of the output devices are free they all check OK (measuring Vgs per Nelson's article). I have replaced all other semiconductors on the boards (even though they measured OK in situ ....... I was getting desperate!)
My questions, for those with better EE theory than me (which is >90% of the list):
1. Is it possible for the output devices to be blown but still give a correct reading on Vgs using the set-up in Nelson's article?
2. Can anyone think of another reason we might have 30V at the output?
Thanks
Mark
PS: And yes .....that slight buzz really wasn't that bad
After turning-on the mains fuse blew, then on repeat power-up the left channel was gone. (I had to do some drilling and I think a filing may have found its way under the PCB).
Anyhow, I'm now at the hair-pulling frustration stage. Basically I get about 3-4V under the positive rail at the output.
Removing the PCB so all of the output devices are free they all check OK (measuring Vgs per Nelson's article). I have replaced all other semiconductors on the boards (even though they measured OK in situ ....... I was getting desperate!)
My questions, for those with better EE theory than me (which is >90% of the list):
1. Is it possible for the output devices to be blown but still give a correct reading on Vgs using the set-up in Nelson's article?
2. Can anyone think of another reason we might have 30V at the output?
Thanks
Mark
PS: And yes .....that slight buzz really wasn't that bad
Hi Mark,
I´m feeling verry sorry for you
.
If you´ve got 30 volts plus on the output the fets are probably allright (otherwise it would be 34 volts)
If you have replaced all other semiconductors you must check the sense connections from the active current source. When this is disconnected in any way you will get your 30 volts.
hope this helps,
william
I´m feeling verry sorry for you
.If you´ve got 30 volts plus on the output the fets are probably allright (otherwise it would be 34 volts)
If you have replaced all other semiconductors you must check the sense connections from the active current source. When this is disconnected in any way you will get your 30 volts.
hope this helps,
william
The "thot plickens" ......
Measuring circuit voltages:
Z5 = 9V / R11 = 5.2V
Q1/Q2(S) = -5V to ground (cf +4V)
dV across R14 = 0
Checking R14 = open circuit !!!
I have never seen a metal film resistor blow without any visible sign of stress externally.
Checking other resistors, R19 is also open circuit.
Off to the shop tomorrow
You see something new every day!
mark
Measuring circuit voltages:
Z5 = 9V / R11 = 5.2V
Q1/Q2(S) = -5V to ground (cf +4V)
dV across R14 = 0
Checking R14 = open circuit !!!
I have never seen a metal film resistor blow without any visible sign of stress externally.
Checking other resistors, R19 is also open circuit.
Off to the shop tomorrow
You see something new every day!
mark
Pass DIY Addict
Joined 2000
Paid Member
Mark, sorry to hear of your troubles! While I was working on my a40, I managed to toast several Mills wire wound resistors that showed absolutely no outward signs of trouble. When they failed, they formed an open circuit also. I never thought to check them until I tracking down the source of some distortion...
Aaarrrrrrrggggggggghhhhhhhh !!!!!!!!
Thanks for those who replied. All is now well after 4 days of hell!
I ran around in circles unable to make head nor tail of the problem and wound up at the point where I had floating DC at the output, ranging from -2 to +3V.
All of the input voltages were OK - I-source / diff pair / R19.
Output FETs Vgs = OK. dV across the bias resistors also varying erratically over time, like the dV at the output. It was as though the sense circuit was having intermittent trouble picking-up input.
could there be a dry solder joint, or could there be a microscopic flaw in one of the PCB tracks.
In absolute desperation I resoldered every joint and tinned every track on the PCB so it is now a shiny silver ......... CURED !!!!!
I was about 5 nanoseconds from making a whole new PCB from scratch
cheers
mark
PS: Yeah ...... I've been watching the X-threads with great interest. Though, I'm cured from tinkering for a little while, I might just sit back (with a nice glass of red) and listen to some music
Thanks for those who replied. All is now well after 4 days of hell!
I ran around in circles unable to make head nor tail of the problem and wound up at the point where I had floating DC at the output, ranging from -2 to +3V.
All of the input voltages were OK - I-source / diff pair / R19.
Output FETs Vgs = OK. dV across the bias resistors also varying erratically over time, like the dV at the output. It was as though the sense circuit was having intermittent trouble picking-up input.
could there be a dry solder joint, or could there be a microscopic flaw in one of the PCB tracks.In absolute desperation I resoldered every joint and tinned every track on the PCB so it is now a shiny silver ......... CURED !!!!!
I was about 5 nanoseconds from making a whole new PCB from scratch
cheers
mark
PS: Yeah ...... I've been watching the X-threads with great interest. Though, I'm cured from tinkering for a little while, I might just sit back (with a nice glass of red) and listen to some music
Mark,
glad to read that you've fixed it
I'd like to use this thread to publish my problem :
Aleph 5 one channel completed and put on the test bench.
Problem is that the DC offset is about 100 mV. What's the best thing to tweak here ? The voltage over the source resistors is almost the same ( less than 40 mV difference)
Should i change the value of the base resistor of Q5 (Nelson's schematic)
Or modify R20 (1k) to another value....
Please help me out, i want to start listening asap
Grtz,
Nick
glad to read that you've fixed it
I'd like to use this thread to publish my problem :
Aleph 5 one channel completed and put on the test bench.
Problem is that the DC offset is about 100 mV. What's the best thing to tweak here ? The voltage over the source resistors is almost the same ( less than 40 mV difference)
Should i change the value of the base resistor of Q5 (Nelson's schematic)
Or modify R20 (1k) to another value....
Please help me out, i want to start listening asap
Grtz,
Nick
Yes, I used a 1K ohm resistor in parallel with a 2k ohm pot. Just be sure to set the paralleled resistance to 390 ohms before you fire up the amp. Then from there after warmup you can adjust the pot. to null the output. With the ability to adjust the output offset you could remove the capacitor from the negative input, I think it is C5, The only draw back is the initial output offset is larger until things warm up.
On the Aleph I built I used this pot. and one on the output current source to adjust for higher or lower bias.
BDP
On the Aleph I built I used this pot. and one on the output current source to adjust for higher or lower bias.
BDP
Hi Guys ...... sorry for being slack getting back to the thread.
Sonically it is very hard to tell any difference, but I haven't actually run a mono source from 1 channel to the other to compare.
Offset dV has been adequately answered above. My preference would be to re-match the FETs but the variable resistor option is quick and easy.
cheers
mark
Sonically it is very hard to tell any difference, but I haven't actually run a mono source from 1 channel to the other to compare.
Offset dV has been adequately answered above. My preference would be to re-match the FETs but the variable resistor option is quick and easy.
cheers
mark
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.