Allen & Heath Saber... Again!

[Rob Flinn]

Since you seem to have faults in every part of this console if I were you I would try to approach this a bit more logically than just swapping stuff out.

Personally I would do something like this

1. check the Power Supply
2. remove all channel modules & get the master section going.  If you have an option for a stereo return use that to put a signal of some decription into the master section & until you have that working don't worry about the channels.    Replacing the feed resistors because they're near a burnt section is a bit of awaste of time in my view.
3. Look at the channels one by one.

In my experience when a console has been stored like you say yours has then it is worth while looking  at the caps first.    There is no need to up the voltage rating.  Some people replace double the capacitance of the the audio path caps, which halves the low frequency cut off point for that point in the circuit.  Power supply decoupling caps are normally kept the same value because making them a lot bigger potentially can blow your PSU with the increased inrush current at switch on.      It is worthwhile getting 105deg caps.  People talk about certain caps sounding better than others, but the reality is if the caps are shot in your console anything is going to sound better than what was in there.   

Quite often if you recap the next area to look at is mechanical, i.e switches & connectors.        I restored a desk that was in quite poor condition & there were no blown transistors or op amps apart from on some VCA boards.  Once I had worked out the symptoms compared to which op amp on the VCA board was blown it was quite simple to get all the VCA boards going.    I didn't need to replace any resistors. 

It is quite useful to have some sort of scope, it doesn't need to be fancy just something to trace a sine wave.    Using a scope & a sine wave from Pro Tools (other DAWs are available)  it is easy to find the area of the fault.

As people have suggested randomly changing components for no reason at this point is probably not the way to go.

 

[bluebird]

+1

1. check the Power Supply
2. remove all channel modules & get the master section going.  If you have an option for a stereo return use that to put a signal of some decription into the master section & until you have that working don't worry about the channels.    Replacing the feed resistors because they're near a burnt section is a bit of awaste of time in my view.
3. Look at the channels one by one.

This is the only practical way to get through this.  And also you can get help easier from forum members if your only working on small blocks at a time. Diagnosing problems in a fully loaded console is a headache for even an experienced tech.

 

[Monte McGuire]

I'll also recommend checking the power supply, but I want to add some detail.

What you ideally want to do is power the device up and make sure that each IC is getting the proper power supply voltage, measured directly at its pins. This is an end-to-end test that will make sure that everything along the way to each IC is working properly. Sometimes, socketed chips will fail because the socket fails, and this will test the PSU pins of that socket. Sometimes, a PC board IO connector has a problem, or the PCB has a cracked PS trace, and this will test that as well.

You need to be very careful not to let your meter probe slip and short the power supply voltage to any other conductor. If that happens, it's easy to blow stuff up, creating a new problem. So, be careful, and use a probe with a sharp and small enough tip. An alligator clip on the meter's ground probe will make it simpler to quickly test a lot of terminals.

You therefore need a multimeter of some sort to determine if the supply voltages are right, and to be able to adjust them if you can or if you need to. You can rig up an oscilloscope to do this, but a multimeter is simplest and probably most accurate. You don't need a ton of digits of accuracy, but you want to get a meter that you can trust well enough. Not sure what you can afford, or what you have, but having too many multimeters is usually not such a bad thing ;-)

Best of luck, and again, I consider this to the the first thing to check. If there are problems, work backward and try to see where the fault is and isn't. Also, if there's a fault, don't leave the box on forever while you poke at it. If you find obvious problems, isolate those parts of the circuit, fix it, and then try again until it all works.

Then, once the supply voltages to each chip are all fine, see what else is broken. In a console with op amps used as linear amps, check that the op amp output pins are at a reasonable voltage between the + and - supply rails. If they're stuck a volt or two away from a supply rail, the amp is saturated and not working properly, so the fault might be at that stage. Or, maybe upstream. Another op amp trick is to measure the voltage between the + and - input terminals. If it's measurable at all, the amplifier is not working right. Feedback is supposed to reduce that voltage difference to nearly zero. This is best done with a two channel scope and a pair of probes, set to add the two channels together with one channel inverted (a poor man's differential probe). That way, you can test the circuit with a signal passing through, and be able to tell when an amplifier isn't working, either because the IC is dead or the feedback network has failed.

Best of luck!

 

[Rob Flinn]

If you read the Sabre manual it has pretty comprehensive fault diagnosis procedures in it.  There is a whole page which says what Monte has just written.  There is also a page which says if the battery on the mute board is dead you will get some pretty random mute stuff going on when you power the console up.

 

[psb_87]

Since you seem to have faults in every part of this console if I were you I would try to approach this a bit more logically than just swapping stuff out.

Personally I would do something like this

1. check the Power Supply
2. remove all channel modules & get the master section going.  If you have an option for a stereo return use that to put a signal of some decription into the master section & until you have that working don't worry about the channels.    Replacing the feed resistors because they're near a burnt section is a bit of awaste of time in my view.
3. Look at the channels one by one.

In my experience when a console has been stored like you say yours has then it is worth while looking  at the caps first.    There is no need to up the voltage rating.  Some people replace double the capacitance of the the audio path caps, which halves the low frequency cut off point for that point in the circuit.  Power supply decoupling caps are normally kept the same value because making them a lot bigger potentially can blow your PSU with the increased inrush current at switch on.      It is worthwhile getting 105deg caps.  People talk about certain caps sounding better than others, but the reality is if the caps are shot in your console anything is going to sound better than what was in there.   

Quite often if you recap the next area to look at is mechanical, i.e switches & connectors.        I restored a desk that was in quite poor condition & there were no blown transistors or op amps apart from on some VCA boards.  Once I had worked out the symptoms compared to which op amp on the VCA board was blown it was quite simple to get all the VCA boards going.    I didn't need to replace any resistors. 

It is quite useful to have some sort of scope, it doesn't need to be fancy just something to trace a sine wave.    Using a scope & a sine wave from Pro Tools (other DAWs are available)  it is easy to find the area of the fault.

As people have suggested randomly changing components for no reason at this point is probably not the way to go.

Thanks a lot for the tips.  So I have removed all the modules and am only concentrating on the master module atm.  I have already recapped the whole thing using 105 degree caps, keeping the voltages and capacitance the same as the originals except on the 10/40 caps I upped the voltage to 50 because that was all that was available, which is no big deal.

I have also socketed all the ICs and put all brand new ICs in them, the exact same spec as the originals.

I have a multi-meter.  I am not sure how to check the power supply? it all seems to be working properly though.  The reason I think this is because:

1. Before I removed the all the modules the input channels and sub groups all worked fine when soloed (or in check mode on this desk) there was no hum and audio was clean.

2.  The lights on the front of the supply are all lit.

3. I opened the supply up to take a look and it all looks good - granted this doesn't really mean anything.

I would have thought if there was a problem with the power supply then it would effect ALL channels, not just the master L & R?

However, I would still like to know how to check the power supply safely.  It's a +/-16v 48v supply.  Any tips for a novice would be appreciated.

Like you say the priority is to get the master module up and going - I'm pretty sure I know what the problems are on the other channels, apart from the gain pot activating the mute LED... that one has me puzzled.

Thanks a lot for your help, I am relying heavily on people like you to help me through this process.

 

[psb_87]

I'll also recommend checking the power supply, but I want to add some detail.

What you ideally want to do is power the device up and make sure that each IC is getting the proper power supply voltage, measured directly at its pins. This is an end-to-end test that will make sure that everything along the way to each IC is working properly. Sometimes, socketed chips will fail because the socket fails, and this will test the PSU pins of that socket. Sometimes, a PC board IO connector has a problem, or the PCB has a cracked PS trace, and this will test that as well.

You need to be very careful not to let your meter probe slip and short the power supply voltage to any other conductor. If that happens, it's easy to blow stuff up, creating a new problem. So, be careful, and use a probe with a sharp and small enough tip. An alligator clip on the meter's ground probe will make it simpler to quickly test a lot of terminals.

You therefore need a multimeter of some sort to determine if the supply voltages are right, and to be able to adjust them if you can or if you need to. You can rig up an oscilloscope to do this, but a multimeter is simplest and probably most accurate. You don't need a ton of digits of accuracy, but you want to get a meter that you can trust well enough. Not sure what you can afford, or what you have, but having too many multimeters is usually not such a bad thing ;-)

Best of luck, and again, I consider this to the the first thing to check. If there are problems, work backward and try to see where the fault is and isn't. Also, if there's a fault, don't leave the box on forever while you poke at it. If you find obvious problems, isolate those parts of the circuit, fix it, and then try again until it all works.

Then, once the supply voltages to each chip are all fine, see what else is broken. In a console with op amps used as linear amps, check that the op amp output pins are at a reasonable voltage between the + and - supply rails. If they're stuck a volt or two away from a supply rail, the amp is saturated and not working properly, so the fault might be at that stage. Or, maybe upstream. Another op amp trick is to measure the voltage between the + and - input terminals. If it's measurable at all, the amplifier is not working right. Feedback is supposed to reduce that voltage difference to nearly zero. This is best done with a two channel scope and a pair of probes, set to add the two channels together with one channel inverted (a poor man's differential probe). That way, you can test the circuit with a signal passing through, and be able to tell when an amplifier isn't working, either because the IC is dead or the feedback network has failed.

Best of luck!

Hi thanks a lot for the good advice, I really appreciate it.

I do have a multi-meter and would be happy to test the ICs as you suggest but I'm not entirely sure how I should do it? I have installed sockets for each IC and put brand new ICs in, replacing all of the old ones.  They're mostly TL072s and one 5532.

Could could lay out the procedure for me in a very simple manner (I'm still learning how to do all this stuff, which I think is pretty obvious) .  I guess the ICs themselves should be fine since they're all new, but this way I can see if they're doing their job in the circuit or not.

Thank-you for your time.

 

[psb_87]

If you read the Sabre manual it has pretty comprehensive fault diagnosis procedures in it.  There is a whole page which says what Monte has just written.  There is also a page which says if the battery on the mute board is dead you will get some pretty random mute stuff going on when you power the console up.

So I have been reading the manual quite thoroughly, it's just that as I am not a trained tech I am still trying to understand it all! I come from a background of making stomp boxes from kits or doing simple repairs on simple devices, so I've only ever had to follow 'paint by numbers' instructions.  When it starts talking about voltages and using scopes I'm lost.  I am still learning how to read a schematic properly.  But hopefully this project will teach me a few things

 

[Rob Flinn]

I do have a multi-meter and would be happy to test the ICs as you suggest but I'm not entirely sure how I should do it? I have installed sockets for each IC and put brand new ICs in, replacing all of the old ones.  They're mostly TL072s and one 5532.

There's not much point in replacing the i.c's unless you know one is blown.    You can check the psu rails by connecting the black probe of your meter to ground & the red lead to the rail you want to check.  Do it on a d.c range first to check that you are getting whatever the voltage should be for that rail.  Then check on a lowish a.c range to see if there is any ripple.  If you just have the master section in then you probably won't see much, but if the console is fully loaded it will be more if there is a problem.

It sounds to me like you need someone to guide you through this to get you started.  Where are you in the U.K  You may find there is someone closer to you than you think.

 

[psb_87]

I do have a multi-meter and would be happy to test the ICs as you suggest but I'm not entirely sure how I should do it? I have installed sockets for each IC and put brand new ICs in, replacing all of the old ones.  They're mostly TL072s and one 5532.

There's not much point in replacing the i.c's unless you know one is blown.    You can check the psu rails by connecting the black probe of your meter to ground & the red lead to the rail you want to check.  Do it on a d.c range first to check that you are getting whatever the voltage should be for that rail.  Then check on a lowish a.c range to see if there is any ripple.  If you just have the master section in then you probably won't see much, but if the console is fully loaded it will be more if there is a problem.

It sounds to me like you need someone to guide you through this to get you started.  Where are you in the U.K  You may find there is someone closer to you than you think.

Okay, so I've attached the schematic for the PSU.  Is it just the +16-16+48 I need to check?

I have been looking around for someone locally who can help me but there's no-one here who can do it. (I live out in the country side of west wales)  well there is one guy who is more than capable of doing it as he builds and designs mixers for a living (he designed the Mackie D8B) but frustratingly every time I ask him to help me and offer to pay him for his time he just tells me that he's incredibly busy and doesn't have a spare moment.  I've asked him multiple times already to no avail.

All the other people that I've spoken to are based really far from me, and I really need someone who can come to me rather than me bringing the mixer to them.

 

[psb_87]

...Here's the schematic

 

[scott2000]

...Here's the schematic

You'll want to be careful checking voltages. You can very easily damage something or hurt yourself.

Once you figure out what you want to test, by someone telling you or by you deciding to test voltage somewhere, make sure you are probing the right spots......

Look around for any helpful info or videos on this......  One wrong probe can create more headaches or worse. You don't need to be scared, although it doesn't hurt to have some healthy fear, just respectful of this fact and don't get impatient .Definitely  try to be level headed in your approach......I'm sure you know this already....Sometimes I forget ....

.......One thing I've done ...........With power off........is  use the continuity tester to make sure that I'm looking at the schematic compared to the actual board correctly. This trace looks like it connects to that on the schematic, let me see if the continuity test confirms this...... With power off........

It looks to me like you have some test points in the power schematic  so  it should make things a bit easier.......

Good Luck!

 

[psb_87]

I do have a multi-meter and would be happy to test the ICs as you suggest but I'm not entirely sure how I should do it? I have installed sockets for each IC and put brand new ICs in, replacing all of the old ones.  They're mostly TL072s and one 5532.

There's not much point in replacing the i.c's unless you know one is blown.    You can check the psu rails by connecting the black probe of your meter to ground & the red lead to the rail you want to check.  Do it on a d.c range first to check that you are getting whatever the voltage should be for that rail.  Then check on a lowish a.c range to see if there is any ripple.  If you just have the master section in then you probably won't see much, but if the console is fully loaded it will be more if there is a problem.

It sounds to me like you need someone to guide you through this to get you started.  Where are you in the U.K  You may find there is someone closer to you than you think.

So I've made a video trying to show the problems with the master module.


https://www.youtube.com/watch?v=RoraYxUJ0c4


Not sure if it's useful or not but take a look so you can at least see what I'm talking about.  You might recognise that sound and be able to identify the problem?

I start with the oscillator at 1Khz and then change it to 100Hz later on.  Throughout theres a weird sputtery sound which happens whether the oscillator is on or off.  This only happens when monitoring the master L & R bus.  If I pull the cable from the monitor module which connects to the master L & R then it stops, or if I select a different source to monitor it stops as well.

Could there be anything on the monitor PCB which would be causing the problem? It seems to be functioning exactly as it should.

It seems like the Master L & R are not getting enough juice, the LEDs and switches seem weird, but they are definitely doing something.  The fader is also functioning.  There's just no audio coming through.

I've changed all the chips, installed sockets of all the chips, changed all the caps, a few resistors and a couple of the BC214 transistors, there doesn't seem to be any shorts or any bad solder joints... but the modules seem to behave exactly the same as before all of those 'repairs'.  I'm starting to wonder if it's something to do with the inserts... but that wouldn't account for the weird power issues that module is experiencing.

All of the other channels are receiving power fine, all of the groups are working perfectly, routing any of the input channels to the groups works, and I can monitor the audio by soloing the channels.

Any ideas?

 

[psb_87]

You'll want to be careful checking voltages. You can very easily damage something or hurt yourself.

Once you figure out what you want to test, by someone telling you or by you deciding to test voltage somewhere, make sure you are probing the right spots......

Look around for any helpful info or videos on this......  One wrong probe can create more headaches or worse. You don't need to be scared, although it doesn't hurt to have some healthy fear, just respectful of this fact and don't get impatient .Definitely  try to be level headed in your approach......I'm sure you know this already....Sometimes I forget ....

.......One thing I've done ...........With power off........is  use the continuity tester to make sure that I'm looking at the schematic compared to the actual board correctly. This trace looks like it connects to that on the schematic, let me see if the continuity test confirms this...... With power off........

It looks to me like you have some test points in the power schematic  so  it should make things a bit easier.......

Good Luck!

Hey Scott, thanks for the tips! Take a look at the vid in the link above and see what you think?

 

[scott2000]

Not helpful to me. Sorry. Too many variables involved  to even begin to tell what's going on from the video for me at least....

Was that the oscillator trying to make noise????

Nice looking board though..... I'll keep an eye out for any info.......

Keep the faith..... You'll get it working if you don't give up on it...... Sometimes breaks are good though..... You'll get information and thoughts during breaks.....

They arrive like butterflies.......

 

[psb_87]

Since you seem to have faults in every part of this console if I were you I would try to approach this a bit more logically than just swapping stuff out.

Personally I would do something like this

1. check the Power Supply

Hi Rob, I checked the power supply today.  I'm getting just over +16v and just over -16v (16.34) which I'm guessing is good.

0v is 0V

+48 is just over at +49

Assuming that in order to check ripple I just need to set my multimeter to V~ I got a reading of 34V on the +16 but 0 on the -16 - what does that mean?

So could the PSU still be at fault?