Hi folks, First post here so hello all.
I have an Arcam A85 which I quite like, but it won't switch on. The problem seems to be in the microprocessor controlled front panel PCB, as all the error signals are normal but the micro doesn't switch on the power supply relay. Only the 5V system is live. The on/off switch causes pull-down at the microcontroller as expected but the micro doesn't then switch on the mains relay or power LED.
Bit of a long shot but if anyone is parting out one of these would you send/sell me PCB L865?
I might try a new crystal, they are cheap enough.
If I try to replace the micro (hitachi H8/3084 - amazingly I can still get them), does anyone have ideas for firmware? I don't think firmware can be ripped from a working one so probably only Arcam has the original 🙁
Ultimately, if I can't resurrect it would anyone be interested in having it for parts or to build a new front panel for it? (you would need some digital know-how as the pre-amp uses digital busses for volume control and multiplexing). converting it to a power amp shouldn't be too hard but it wouldn't do what I want.
I've messaged Arcam but I think they only do factory repair these days, which will probably be more than the amp is worth.
I'll keep an eye on ebay for a spares/repairs one, but the only one on there at the moment also has a front panel fault. Any other ideas?
Tom
PS: service manual and circuit diagrams
https://www.vintageshifi.com/repertoire-pdf/pdf/telecharge.php?pdf=Arcam-A85-Service-Manual.pdf
I have an Arcam A85 which I quite like, but it won't switch on. The problem seems to be in the microprocessor controlled front panel PCB, as all the error signals are normal but the micro doesn't switch on the power supply relay. Only the 5V system is live. The on/off switch causes pull-down at the microcontroller as expected but the micro doesn't then switch on the mains relay or power LED.
Bit of a long shot but if anyone is parting out one of these would you send/sell me PCB L865?
I might try a new crystal, they are cheap enough.
If I try to replace the micro (hitachi H8/3084 - amazingly I can still get them), does anyone have ideas for firmware? I don't think firmware can be ripped from a working one so probably only Arcam has the original 🙁
Ultimately, if I can't resurrect it would anyone be interested in having it for parts or to build a new front panel for it? (you would need some digital know-how as the pre-amp uses digital busses for volume control and multiplexing). converting it to a power amp shouldn't be too hard but it wouldn't do what I want.
I've messaged Arcam but I think they only do factory repair these days, which will probably be more than the amp is worth.
I'll keep an eye on ebay for a spares/repairs one, but the only one on there at the moment also has a front panel fault. Any other ideas?
Tom
PS: service manual and circuit diagrams
https://www.vintageshifi.com/repertoire-pdf/pdf/telecharge.php?pdf=Arcam-A85-Service-Manual.pdf
If the standby supply is okay, in spec, no excessive ripple, and if there is nothing obvious wrong with other rails (short circuits etc), could you just force the standby relay to power up?
Assume it is RLY200C, if you could feed a couple of volts into the junction between R212 and TR203, through a resistor (1k or so) you might be able to get it to power up and see if the microcontroller is completely dead...
Assume it is RLY200C, if you could feed a couple of volts into the junction between R212 and TR203, through a resistor (1k or so) you might be able to get it to power up and see if the microcontroller is completely dead...
Great minds think alike 🙂
I wanted to get listening this afternoon, so I put a link on the digital board to supply 5V from SK1 pin 14 to pins 10 and 11. This closes relay 200 and also the output relay for the 1st set of speakers. It now works as a power amp, except that the fault detection system is ignored.
Audio quality seems fine, there is some hiss but I think that's just because the source (an HTPC) is running at 10% volume and that cripples it's SNR.
There's still no life on the front panel though - no lights, no display, the buttons do nothing.
The standby supply seems ok, steady at 4.98V on a DMM but I don't have an oscilloscope to check for ripple (can't check if the crystal is running either).
I wanted to get listening this afternoon, so I put a link on the digital board to supply 5V from SK1 pin 14 to pins 10 and 11. This closes relay 200 and also the output relay for the 1st set of speakers. It now works as a power amp, except that the fault detection system is ignored.
Audio quality seems fine, there is some hiss but I think that's just because the source (an HTPC) is running at 10% volume and that cripples it's SNR.
There's still no life on the front panel though - no lights, no display, the buttons do nothing.
The standby supply seems ok, steady at 4.98V on a DMM but I don't have an oscilloscope to check for ripple (can't check if the crystal is running either).
Personal experience over many many years generally says the uP won't be at fault or at least it will be the last suspect. Only when all other possibilities have been exhausted would I consider looking at swapping it.
Is the reset line OK?
The 24C02 memory chip wouldn't be above suspicion either in a general sense.
When faced with problems like this I always used to around every pin on the uP with a scope and see if what was present seemed at reasonable. Data is meaningkess on a scope but incorrect levels stand out a mile.
Even things like the keyscan for the front panel needs checking in case anything like a diode or tact switch is leaky and shorting two lines all the time thus inhibiting the uP.
Is there a standby light as well as a power on indicator? If there is and that is not lit then that is something else to check out.
Its a case of looking for clues but the uP is way down the list at this point.
Is the reset line OK?
The 24C02 memory chip wouldn't be above suspicion either in a general sense.
When faced with problems like this I always used to around every pin on the uP with a scope and see if what was present seemed at reasonable. Data is meaningkess on a scope but incorrect levels stand out a mile.
Even things like the keyscan for the front panel needs checking in case anything like a diode or tact switch is leaky and shorting two lines all the time thus inhibiting the uP.
Is there a standby light as well as a power on indicator? If there is and that is not lit then that is something else to check out.
Its a case of looking for clues but the uP is way down the list at this point.
As Molly said good to probe right on the microcontoller pins, even without a scope. If your multimeter has sharp probe points you can learn a lot if there is a dry joint somewhere.
OK, I'll find my self a needle to tape to the probe, remove my link, and see what I can see.
I found a thread suggesting this amp can do this if a fascia button is stuck Arcam A85 fault
None of mine are stuck but I'll be paying particular attention to their levels
it's a 100 pin QFP, so I may be a while 😛
I checked the reset pin earlier and it's OK.
There is a combined power/standby green/red LED but it's operated from the micro not directly from the 5V digital supply. It's off like everything else but that's normal for this amp when in set standby at the front panel - it only seems to be lit if the amp is on (green) or switched off with the remote (amber).
I found a thread suggesting this amp can do this if a fascia button is stuck Arcam A85 fault
None of mine are stuck but I'll be paying particular attention to their levels
it's a 100 pin QFP, so I may be a while 😛
I checked the reset pin earlier and it's OK.
There is a combined power/standby green/red LED but it's operated from the micro not directly from the 5V digital supply. It's off like everything else but that's normal for this amp when in set standby at the front panel - it only seems to be lit if the amp is on (green) or switched off with the remote (amber).
Fortunately I have some spare time with the lockdown...
I've been round the IC and most pins have expected voltages, all button pins idle high as expected. There are a few pins where I don't know what to expect though.
The Programming pin 75 (MD2) seems to be oscillating, I can get a DC reading of 5.8V and an AC RMS of 1.3, but I can't get a frequency measurement. This pin is connected to 5V through diode D2 and I think it's intended to be pulled up to 12V by the programmer to set boot mode, and sit at 5V the rest of the time. (nothing is connected to the programming header). The pin also plays a role in setting the micro's mode (internal/external ram etc). I'm tempted to try bypassing D2?
I'm not sure what the circuit with Z7 is supposed to do, it seems to monitor audio output from the pre-amp. Pin 78 shows 2.4V and 96 is at 0.3. Given the amplifier is essentially off, the comparator could go either way and I can't see this mattering?
The other thing that doesn't seem right is that quite a few of the unconnected pins seem to show 2.2V, while I would expect them to be set to 0 or 5 if the micro had booted successfully.
I get 1.3 and 1.9V at the crystal (XTAL/EXTAL)
Several data or clock pins leading to the pre-amp and digital display show about 2.2 to 3.7 V, but I don't know if they are idling at these values or if I'm getting an average of a running clock / data stream.
The data/clock lines to the memory are at 5V, it's an I2C bus, so this is the default state if the bus hasn't been initialised.
I'm using a good quality handheld meter (a fluke 115) if that helps anyone interpret things. The AC voltage measurement rolls off from somewhere above 1KHz, and frequency measurement goes up to 50KHz
I've been round the IC and most pins have expected voltages, all button pins idle high as expected. There are a few pins where I don't know what to expect though.
The Programming pin 75 (MD2) seems to be oscillating, I can get a DC reading of 5.8V and an AC RMS of 1.3, but I can't get a frequency measurement. This pin is connected to 5V through diode D2 and I think it's intended to be pulled up to 12V by the programmer to set boot mode, and sit at 5V the rest of the time. (nothing is connected to the programming header). The pin also plays a role in setting the micro's mode (internal/external ram etc). I'm tempted to try bypassing D2?
I'm not sure what the circuit with Z7 is supposed to do, it seems to monitor audio output from the pre-amp. Pin 78 shows 2.4V and 96 is at 0.3. Given the amplifier is essentially off, the comparator could go either way and I can't see this mattering?
The other thing that doesn't seem right is that quite a few of the unconnected pins seem to show 2.2V, while I would expect them to be set to 0 or 5 if the micro had booted successfully.
I get 1.3 and 1.9V at the crystal (XTAL/EXTAL)
Several data or clock pins leading to the pre-amp and digital display show about 2.2 to 3.7 V, but I don't know if they are idling at these values or if I'm getting an average of a running clock / data stream.
The data/clock lines to the memory are at 5V, it's an I2C bus, so this is the default state if the bus hasn't been initialised.
I'm using a good quality handheld meter (a fluke 115) if that helps anyone interpret things. The AC voltage measurement rolls off from somewhere above 1KHz, and frequency measurement goes up to 50KHz
I suspect you are seeing averages of the data on many of these pins as you suggest. A good scope is always the best diagnostic tool to try and figure out what is going on.
I don;t know the reason for Z7 and why just the left channel is sampled but as wired the comparator with its equal value resistors could (as I see it) go either way under no signal conditions depending on R tolerances, however the feed from pin 98 modifies that behaviour.
If you are curious then just dab a resistor across those one of those 100k's and flip the state of the comparator. Pin 78 is sat where you would expect it it were an input and its voltage is just reflecting the equal value 100k divider across the 5v rail.
No easy answers to anything like this 🙁
I don;t know the reason for Z7 and why just the left channel is sampled but as wired the comparator with its equal value resistors could (as I see it) go either way under no signal conditions depending on R tolerances, however the feed from pin 98 modifies that behaviour.
If you are curious then just dab a resistor across those one of those 100k's and flip the state of the comparator. Pin 78 is sat where you would expect it it were an input and its voltage is just reflecting the equal value 100k divider across the 5v rail.
No easy answers to anything like this 🙁
I seems a bit odd that the micro would be trying to communicate with components that it hasn't powered up yet, but I'm starting to think I'm not going to get very far without a scope. I'll keep prodding it for a bit longer and see if anything comes up, and look through schematics for the pins that are not 0 or 5 V to see what pull-up and pull-down arrangements they have.
Fortunately, I have access to a couple of nice modern tektronix scopes at work, so might try and hook one up to the amp once our landlord allows us back in the building.
In the mean time, here's a few pics of it, cos I assume you all like that sort of thing 🙂
Pre-amp at the back left, power supply back right, and output stages on the same PCB as the power supply.
All digital interface, including infinitely rotating volume encoder and momentary switches. It includes a green vacuum fluorescent dot matrix display that is a fairly good colour match to the front panel LEDs. If anyone else want to take the front panel off one like this, don't forget to pull off the volume knob and remove the nut behind or you will pull the front off the encoder.
When new this amp had expansion cards available for MM and MC phono (on the right) and a 7.1 surround sound pre-amp (in the bigger panel).
And I know it's taken as given on this forum, but it never hurts to mention: This amp, like many others, has hazardous live conductors exposed as soon as the main case lid is removed - always be careful
Fortunately, I have access to a couple of nice modern tektronix scopes at work, so might try and hook one up to the amp once our landlord allows us back in the building.
In the mean time, here's a few pics of it, cos I assume you all like that sort of thing 🙂
Pre-amp at the back left, power supply back right, and output stages on the same PCB as the power supply.
All digital interface, including infinitely rotating volume encoder and momentary switches. It includes a green vacuum fluorescent dot matrix display that is a fairly good colour match to the front panel LEDs. If anyone else want to take the front panel off one like this, don't forget to pull off the volume knob and remove the nut behind or you will pull the front off the encoder.
When new this amp had expansion cards available for MM and MC phono (on the right) and a 7.1 surround sound pre-amp (in the bigger panel).
And I know it's taken as given on this forum, but it never hurts to mention: This amp, like many others, has hazardous live conductors exposed as soon as the main case lid is removed - always be careful
mmmmm... really like a nice vacuum fluorescent display.
Just one thought, some Arcam kit I have seen has been packed full of Nover branded electrolytic capacitors. Think they are sourced from the distributor Anglia, but they are still far east low quality parts. If you see any in your unit might be worth checking or just replacing.
Just one thought, some Arcam kit I have seen has been packed full of Nover branded electrolytic capacitors. Think they are sourced from the distributor Anglia, but they are still far east low quality parts. If you see any in your unit might be worth checking or just replacing.
Pretty chunky build. I picked up an Arcam Solo Neo at a charity shop and the build quality and chassis work was pretty impressive.
Mine seems to be a mix of rubycon and ELNA cerafine in the pre-amp, with quite a few smaller polypropylene caps as well. The power amp gets samwah and some more rubycon, with BHC Aerovox for the big (10mF) reservoirs.
TBH I'm not inclined to swap anything in the amp itself until I get the digital side working again, just in case I introduce a new problem.
It's noticeably better built than the older Arcam Alpha 8r I also have. The case is completely rigid and everything seems good quality. I guess that's what one should expect for an amp that was £700 + when new, and it certainly contributes to me wanting to fix it. At ~20 years old it doesn't feel life expired at all.
well, I'm pleased to announce this amp is back in business, Though I didn't definitively fix anything 
When probing round the micro, I found the RC5 pin (pin 87) was idling low. I didn't think anything of it, high or low seemed equally valid to me, as the PCB is configured with a mixture of idle-high and idle-low inputs.
On working through the schematics, I wan't sure that was right and started probing around Q3 and Q4. This was somewhat confusing and I realised that while probing the circuit had changed state and pin 87 was now high. I hit the power switch and the front panel came to life. It's since been plugged/unplugged, moved, etc and has started reliably about a dozen times.
It's now back in it's usual home and has been playing fine for an hour or two, and the remote works as it should.
If the problem reoccurs I know where to look. If not I might re-run all the solder in the RC5 circuit to be on the safe side (my best guess is that I've disturbed a bad joint with my probe, but I can't seem to get it to fault again so I don't know where exactly).
Thanks for your suggestions and help folks, I might hang about a bit and see if I get an urge to make/mod anything audio related...
Tom
When probing round the micro, I found the RC5 pin (pin 87) was idling low. I didn't think anything of it, high or low seemed equally valid to me, as the PCB is configured with a mixture of idle-high and idle-low inputs.
On working through the schematics, I wan't sure that was right and started probing around Q3 and Q4. This was somewhat confusing and I realised that while probing the circuit had changed state and pin 87 was now high. I hit the power switch and the front panel came to life. It's since been plugged/unplugged, moved, etc and has started reliably about a dozen times.
It's now back in it's usual home and has been playing fine for an hour or two, and the remote works as it should.
If the problem reoccurs I know where to look. If not I might re-run all the solder in the RC5 circuit to be on the safe side (my best guess is that I've disturbed a bad joint with my probe, but I can't seem to get it to fault again so I don't know where exactly).
Thanks for your suggestions and help folks, I might hang about a bit and see if I get an urge to make/mod anything audio related...
Tom
Good news it's working again, at least it isn't a serious problem.
Just searched the service manual for Q4 and it throws up three transistors, not the best service manual!
If it's part of the remote control circuit, with Q4 being the drive to the microcontroller with a pull up resistor, sounds like it could be a dry joint on the pull up resistor or the pin on the microcontroller. If it was the transistor failed as a short it would be permanent short to zero, and if the transistor was open/dry joint it would be permanently pulled high by the resistor.
Just searched the service manual for Q4 and it throws up three transistors, not the best service manual!
If it's part of the remote control circuit, with Q4 being the drive to the microcontroller with a pull up resistor, sounds like it could be a dry joint on the pull up resistor or the pin on the microcontroller. If it was the transistor failed as a short it would be permanent short to zero, and if the transistor was open/dry joint it would be permanently pulled high by the resistor.
Well done 🙂 and another vote for 'its never the big chip' 😉
I nearly suggested that you should switch the unit off, quickly whizz around all the pins to make sure no large voltages remain and then to press some aluminium foil over the chip to short every pin out and remove any residual charge that could be holding any logic in a wrong state.
I couldn't see any backup cap or battery either (there wouldn't be with NVM or non volatile rom) as they can be a source of trouble.
I suppose a dry is possible... only you know by looking.
Nice one though
I nearly suggested that you should switch the unit off, quickly whizz around all the pins to make sure no large voltages remain and then to press some aluminium foil over the chip to short every pin out and remove any residual charge that could be holding any logic in a wrong state.
I couldn't see any backup cap or battery either (there wouldn't be with NVM or non volatile rom) as they can be a source of trouble.
I suppose a dry is possible... only you know by looking.
Nice one though
yeah it's not my favourite ever, There's almost no info about the digital side, no block diagram for the micro. Plus the bias setting is different on the schematics than in the text. The fault finding guidance is very limited. I might almost say the service manual for my older Arcam 8r is better, but then it is a much simpler amp.
It's briefly misbehaved again and no level of mechanical prodding helped, so I'm moving away from the dry joint theory (plus they all look OK as far as I can see). I measured 200mV at the output of the IR receiver/demodulator RX1 and 100mV at the base of Q3, so I'm thinking the IR receiver might be to blame (should be active low, idle high). Then it started working again so hard to be sure. £7 on eBay for the correct part or more like £2 for a modern equivalent so not a great loss if I change it and it doesn't fix it.
It's playing some Mumford and Son's right now so no rush to fix. 🙂
Ha! I was starting to think about that sort of thing. I think I think of lingering charge or persisting latchups after an over night power cycle a bit like the way you think of failed micros though: they can explain all sorts of faults, but it never is!
(and to be fair, my experience of replacing micros is usually directly related to something I did to them, not a spontaneous fault. Putting a 100 lead TQFP down 90 degrees was one of my finest achievements 😱 )