A quick question following on from from my hum / earthing problems....
If one diode in a bridge rectifier has failed open circuit, what voltage drop would you expect to see? Would 34 volts when expecting 37 be about right?
Also, would you expect to be able to detect any ripple with a multimeter on, say, its 5 VAC range?
Finally, I need to double check that I should actually be getting 37 volts. My amplifier is an Alpha 8 but the only service manual I can find is for an 8r. My maths suggest there should be just over 26 volts AC coming out of the transformer. Does that make sense?
Thanks
If one diode in a bridge rectifier has failed open circuit, what voltage drop would you expect to see? Would 34 volts when expecting 37 be about right?
Also, would you expect to be able to detect any ripple with a multimeter on, say, its 5 VAC range?
Finally, I need to double check that I should actually be getting 37 volts. My amplifier is an Alpha 8 but the only service manual I can find is for an 8r. My maths suggest there should be just over 26 volts AC coming out of the transformer. Does that make sense?
Thanks
The DC voltage across the reservoir cap would depend on the current drawn with an OC diode so no definitive answer to that.
Detecting ripple with a meter also has problems because some meters react differently to the underlying DC present so again no real answer for that.
If you measure the AC voltage applied to the bridge and multiply by 1.414 and then subtract around 1.4 volts to allow for the diode drop you will get a fair answer.
Its very rare for power diodes (as in a bridge) to go open circuit but I suppose it could happen. It would either be a known issue with the diodes used and that model (and as such a common and recognised fault) or else a 10000:1 chance of a "normal" one off failure.
Detecting ripple with a meter also has problems because some meters react differently to the underlying DC present so again no real answer for that.
If you measure the AC voltage applied to the bridge and multiply by 1.414 and then subtract around 1.4 volts to allow for the diode drop you will get a fair answer.
Its very rare for power diodes (as in a bridge) to go open circuit but I suppose it could happen. It would either be a known issue with the diodes used and that model (and as such a common and recognised fault) or else a 10000:1 chance of a "normal" one off failure.
That is what I had assumed with the 9V going the the remote volume circuit and "r" standing for remote. But apparently not according to my conversation with Arcam!
Mine doesn't have an "r" on the front but does have a remote volume.
I will check the AC later in the day. If it is 25 then that should give about 34 DC which is what I'm getting but the service manual I've got suggests 37!
I will check.
The Alpha 7 (covered by the same service manual) should be about 34 volts. Assuming that my 34 is in correct proportion to the AC and therefore not a diode fault, should I set the bias to the higher levels they indicate for the lower supply voltage.
They suggest 3 to 3.5 mV with 34 Volt supply and 2.3 to 2.8mV for 37.
Thanks again.
I'd have to study the manuals to be sure...
Unless you suspect a problem then it probably is best to leave as set. The adjustments once set shouldn't drift. And tbh supply voltage shouldn't come into the equation for bias. The correct bias depends on the circuit configuration and would be the same whether running on 25, 35 or 60 volts
Unless you suspect a problem then it probably is best to leave as set. The adjustments once set shouldn't drift. And tbh supply voltage shouldn't come into the equation for bias. The correct bias depends on the circuit configuration and would be the same whether running on 25, 35 or 60 volts
I'd have to study the manuals to be sure...
Unless you suspect a problem then it probably is best to leave as set. The adjustments once set shouldn't drift. And tbh supply voltage shouldn't come into the equation for bias. The correct bias depends on the circuit configuration and would be the same whether running on 25, 35 or 60 volts
It was wrong, in that the right channel was 2.8 (top end of suggested for the 8r) but the left was only 1.9. I've adjusted both to 2.7 but can't detect any audible change.
In addition to the earthing type problems I described in the other thread I feel the amp has somewhat thin, lifeless sound compared to my Alpha 5 in the same setup.
I can't detect any sign of the amp having ever been worked on apart from the output mosfets being obviously hand soldered. However, Arcam tell me they are put in by hand so that doesn't prove that they have been replaced.
Last edited:
You won't hear any difference once the bias goes above a milliamp or two. I'm guessing that 2.7mv is across something like a 0.22 ohm or similar which would give around 12 millamps. When you adjust the bias you (normally) need to make sure the amp is at normal temperature. If you check the bias when its totally cold it may well be very different. This is one instance where the procedure in the manual should be followed. Higher bias isn't always better and brings the risk of thermal runaway.
OK, thanks.
It was across two .22 ohm resistors in parallel from memory. I did follow the manual carefully and gave it time to warm up.
Presumably though there shouldn't have been a big difference between the two channels? There is no sign of it running hot but I will bring both down a touch next time I'm inside.
None of this solves my problems though! To recap, I have all kinds of ground loop issues (none of which happen with the Alpha 5 in the same place) plus, as far as I can tell, a thin and not very nice sound.
It was across two .22 ohm resistors in parallel from memory. I did follow the manual carefully and gave it time to warm up.
Presumably though there shouldn't have been a big difference between the two channels? There is no sign of it running hot but I will bring both down a touch next time I'm inside.
None of this solves my problems though! To recap, I have all kinds of ground loop issues (none of which happen with the Alpha 5 in the same place) plus, as far as I can tell, a thin and not very nice sound.
So why then does substituting an Alpha 5 in the same place, connected to the same cables and equipment work fine?
As far as I can see, they are a very similar design earthed in the same way.
Yes, I've confirmed two .22 ohm 3W metal oxide resistors in paralel and they say 2.3 to 2.8 mVolts across. I'm currently at 2.7 when it is warm.
Would ageing, drying out electrolytic capacitors give these audio problems?
Hmmm
You mean the thin sound ? I'm going to say no to that. Assuming there is no real oddball fault with the amp (a scope check would confirm this) then I'm thinking the sound is just the way it is.
Hmmm
Assuming there is no real oddball fault with the amp (a scope check would confirm this) then I'm thinking the sound is just the way it is.
I have an ancient scope and a valve based signal generator. Could I trouble you for a few clues on what to check, it is nearly thirty years since I've done any of this!
Thanks
No problem
First thing to check is that there is no ultrasonic or hf instability. You need to connect the scope across each speaker in turn (be aware that the scope is probably grounded and so the probe earth is also at ground potential. This means that if you connected or touched it to the positive speaker terminal or lead then a "short" would occur on the amplifier output). I sometimes add a 10 ohm or thereabouts in series with the probe ground wire. That stops mishaps and will stop another ground loop from occuring. It won't affect the measurements at all.
So you do that while listening to music and see if the trace is clean. Any HF oscillation should be obvious.
You can use a signal generator (or just a test CD of tones) and see how the amp performs into a dummy 8 ohm load. Squarewave testing needs a generator. You won't get anything like a squarewave from a CD player over about 100hz (and we worry about slew rates in amplifiers).
First thing to check is that there is no ultrasonic or hf instability. You need to connect the scope across each speaker in turn (be aware that the scope is probably grounded and so the probe earth is also at ground potential. This means that if you connected or touched it to the positive speaker terminal or lead then a "short" would occur on the amplifier output). I sometimes add a 10 ohm or thereabouts in series with the probe ground wire. That stops mishaps and will stop another ground loop from occuring. It won't affect the measurements at all.
So you do that while listening to music and see if the trace is clean. Any HF oscillation should be obvious.
You can use a signal generator (or just a test CD of tones) and see how the amp performs into a dummy 8 ohm load. Squarewave testing needs a generator. You won't get anything like a squarewave from a CD player over about 100hz
(and we worry about slew rates in amplifiers).No problem
First thing to check is that there is no ultrasonic or hf instability. You need to connect the scope across each speaker in turn (be aware that the scope is probably grounded and so the probe earth is also at ground potential. This means that if you connected or touched it to the positive speaker terminal or lead then a "short" would occur on the amplifier output). I sometimes add a 10 ohm or thereabouts in series with the probe ground wire. That stops mishaps and will stop another ground loop from occuring. It won't affect the measurements at all.
So you do that while listening to music and see if the trace is clean. Any HF oscillation should be obvious.
You can use a signal generator (or just a test CD of tones) and see how the amp performs into a dummy 8 ohm load. Squarewave testing needs a generator. You won't get anything like a squarewave from a CD player over about 100hz
Many thanks, I'll try that tomorrow and report back.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.