Hi All,
about one year ago I bought a 2nd hand Sugden A21a power amp circa 1999.
I used it with great pleasure for about 6 months with a pair of Davis Acoustics bookshelfs. Then I moved it to my main set with two big Tannoy Gold and...hum! Hum on both channels, with right worse than left. Thought something went faulty during moving but, connecting Davis back and with my ear stuck against the woofer, the hum was there! The difference in sensitivity made a so big difference in revelaing background hum noise.
Ordered a new set of capacitors for the A21a, all of medium-to-high quality, all Rubycon as the original, except for the four big 10000uF which are RS rebranded Cehmicon, but with 105°C max temp, instead of original 85°. I opted for all 105° because of the very high temperature reached inside the Sugdy. This A-class power amp is one of those audio devices that asks for new capacitors each few years if you keep it on many hours a day...
Once the capacitors arrived I recapped the two twin PCBs. Not a so big effort, the Sugden design is simple and disassembling was quite straightforward.
I also checked the BIAS current that should be 1,1 A thru fuse and perfectly tuned it with the single available trimmer.
Ok, work done, powered it up...damn! The right channel is less but still hummy, being the left now perfectly silent! 🙁
Half a day audition, not only confirms that right hum will be always there but even that, when everything is at regime temperature, the overall sound quality definitely improved. On the big Tannoys now the bass is way better controlled than before recapping. In complex I hear something more sweet and defined than before. If only the hum was not there...
So my request to the community: where should I look to identify the hum source?
It is not an air catched one, since internal signal cables from RCAs to boards are of the same lenght and in perfect shape. Even short circuiting inputs the hum is unaffected. Changing signal cable path or position inside the case does not affect that noise, nor touching caps or resistors with fingers/metallic screwdrivers.
It is defintely something in the circuit topology, but what and where?
Before starting replacing film capacitors too, I would like to know your opinion on this. I do not even own a schematics, which is not easy to be found on the net.
Thank you
Have a nice day
Raf
P.S.
Of course, writing emails to Sugden resulted in an absolute silence...
about one year ago I bought a 2nd hand Sugden A21a power amp circa 1999.
I used it with great pleasure for about 6 months with a pair of Davis Acoustics bookshelfs. Then I moved it to my main set with two big Tannoy Gold and...hum! Hum on both channels, with right worse than left. Thought something went faulty during moving but, connecting Davis back and with my ear stuck against the woofer, the hum was there! The difference in sensitivity made a so big difference in revelaing background hum noise.
Ordered a new set of capacitors for the A21a, all of medium-to-high quality, all Rubycon as the original, except for the four big 10000uF which are RS rebranded Cehmicon, but with 105°C max temp, instead of original 85°. I opted for all 105° because of the very high temperature reached inside the Sugdy. This A-class power amp is one of those audio devices that asks for new capacitors each few years if you keep it on many hours a day...
Once the capacitors arrived I recapped the two twin PCBs. Not a so big effort, the Sugden design is simple and disassembling was quite straightforward.
I also checked the BIAS current that should be 1,1 A thru fuse and perfectly tuned it with the single available trimmer.
Ok, work done, powered it up...damn! The right channel is less but still hummy, being the left now perfectly silent! 🙁
Half a day audition, not only confirms that right hum will be always there but even that, when everything is at regime temperature, the overall sound quality definitely improved. On the big Tannoys now the bass is way better controlled than before recapping. In complex I hear something more sweet and defined than before. If only the hum was not there...
So my request to the community: where should I look to identify the hum source?
It is not an air catched one, since internal signal cables from RCAs to boards are of the same lenght and in perfect shape. Even short circuiting inputs the hum is unaffected. Changing signal cable path or position inside the case does not affect that noise, nor touching caps or resistors with fingers/metallic screwdrivers.
It is defintely something in the circuit topology, but what and where?
Before starting replacing film capacitors too, I would like to know your opinion on this. I do not even own a schematics, which is not easy to be found on the net.
Thank you
Have a nice day
Raf
P.S.
Of course, writing emails to Sugden resulted in an absolute silence...
I haven't used them as you have to register, so be careful, but try this Download the Sugden A-21-Mk2 manuals for free - Hifi Manuals .
Try the amps in total isolation, that is to say with only speakers connected and shorting plugs on the inputs.
I though the A21 was an integrated amp... no ? yes ?
I though the A21 was an integrated amp... no ? yes ?
Yes, integrated - with just 2 BJTs for the tone/volume controls and another 2 for the Phono stage.
It's a little too basic in the late 1960's style, to expect stellar noise/distortion performance but then I've never tested any.
Given the date of ca. 1999, I think the OP may be referring the second edition for which a full schematic isn't available, as he says. It is probably electrically similar but the updating will make it hard to locate the features from the old schematic. sbrads link is still fine for that.
http://www.sugdenaudio.com/pdf/a21a_series_2_specifications.pdf
This SE version does look quite different!
http://www.sugdenaudio.com/pdf/a21a_series_2_specifications.pdf
This SE version does look quite different!
Yep Ian! My A21a is a power amp, without pre stage, which was sold apart.
Here is a picture of it:
https://neophonics.files.wordpress.com/2012/03/sugden-original-a21a-3.jpg
It is an evoution of the first model from sixties. Topology has changed, as reported here:
The Original Sugden A21
Link suggested by Sbrads could be pointing at the right one, let's see...
Thanks
Raf
OK, so I guess it looks like this inside:
Of course, that doesn't give us any clues to the schematic, but we have a visuals at least. I can't believe how many darned variations of the A21 model number there are - just so they can keep that iconic old reference number as a marketing attraction.
Though it does look very simple, there appear to be a few wiring layout differences between the amplifier boards that could be worth checking.
Raf, please make a sketch in paint about your amp's GND layout:
Draw all GND path. And measure the resistance between the main GND points on both channels
and between them as well (RCA input, Signal GND on PCB, PS main GND, Speaker return, chassis, ...).
Draw all GND path. And measure the resistance between the main GND points on both channels
and between them as well (RCA input, Signal GND on PCB, PS main GND, Speaker return, chassis, ...).
Ian:
it is the same device but with slightly different wires layout (a later sub revision?)
In detail:
- the PCB in the centre , which is an anti-bump delayed switch, has screw terminals instead of directly soldered wires;
- the relay is a commercial grade transparent OMRON G2R of which I had to clean contacts, since music had glitches at very low volume levels;
- signal wires are coaxial instead of red/black twisted pairs;
- the overall wiring is way more well-ordered in my case;
The power PCBs look the same, even if from side-view only 🙂
Cortez:
I'll provide that but will take some day. During working days I have almost no time for my hobbies...
Thanks all
Raf
it is the same device but with slightly different wires layout (a later sub revision?)
In detail:
- the PCB in the centre , which is an anti-bump delayed switch, has screw terminals instead of directly soldered wires;
- the relay is a commercial grade transparent OMRON G2R of which I had to clean contacts, since music had glitches at very low volume levels;
- signal wires are coaxial instead of red/black twisted pairs;
- the overall wiring is way more well-ordered in my case;
The power PCBs look the same, even if from side-view only 🙂
Cortez:
I'll provide that but will take some day. During working days I have almost no time for my hobbies...
Thanks all
Raf
Those variations are not great and I imagine there could have been more between manufacturing lots as various cost efficiencies were realized with subsequent batches.
As you are not the first owner of this amp, it's always possible that someone has attempted to improve the appearance inside with neater wiring. Perhaps the screw terminals were also an addition to make removal easier and the relay may have already been replaced. These are not big changes but the placement of the wiring could be critical to the level of induced hum.
What I would consider first, is that the right side channel is where the power switch leads run to the front panel. Move that switch wiring away and towards the amplifier to see if there is any change in hum level. I would also return the signal wiring to a twisted pair for the right channel, to check that the shielded wire was actually a good idea.
Ian: I've checked the AC wires topology. Tryed to move them away from the right channel PCB. I even removed the AC plug in order to have it as far as possible with no results.
I also removed the power button from its position with no avail.
Ground topology is very simple: the AC GND is directly connected to the metal case with a screw. To the same screw are connected two wires connected with RCAs ground. That's all.
In my opinion the problem is on circuit itself, but where to look? Having schematics is mandatory.
Writing emails to Sugden is useless... I'll try with their facebook page.
Bye
Raf
I also removed the power button from its position with no avail.
Ground topology is very simple: the AC GND is directly connected to the metal case with a screw. To the same screw are connected two wires connected with RCAs ground. That's all.
In my opinion the problem is on circuit itself, but where to look? Having schematics is mandatory.
Writing emails to Sugden is useless... I'll try with their facebook page.
Bye
Raf
Well you did at least try, perhaps you know a little more about the amplifier now. 😉
Most small manufacturers will have exclusive service agents who will not break confidentiality agreements for supported products by showing schematics publicly. Usually, the manufacturer will just refer you to their agents if they have them or direct you to return the product to them. I doubt you will find schematics specific to this version that detail the wiring as well as you might hope for though. The original and later version hand drawn circuits didn't tell you much either.
At this point, given that there are now screw connectors, I would simply swap R & L channel boards outputs to the relay board, then power leads etc. to prove to yourself where the fault lies. The circuit of the PCBs will be the same as all A21 models so that will not be a great problem to follow from old circuits for which you already have links. If that proves the boards are no different, you then recheck the grounding by lifting each ground connection to see where a wiring duplication may have caused a ground loop etc. For example, one input ground can sometimes be a problem if both are grounded at the wrong point.
The speaker grounding point also influences this and it should be returned to power supply star ground, as the inputs grounds probably also were originally, if that point is still identifiable. Perhaps experiment with this for the inputs. There are some general fundamentals in this link and more amplifier wiring principles elsewhere in the articles section here: Earthing (Grounding) Your Hi-Fi - Tricks and Techniques
I am curious though, when looking at the image of the unmodified amp. There is no ground wire taken from each input to chassis ground as you describe for your amplifier. I would look at that connection to make sure it does not just form a ground loop by duplicating that of the amplifier PCBs.
Class A amplifiers always have more obvious hum issues because they draw high current all the time and this is even when there is no sound at all, so they seem more noisy. A class AB amplifier may seem quiet with no signal but be just as noisy when there is sound. You will then hear the sound + dynamic hum as distorted sound instead. Consequently, hum is more obvious but easier to trace with class A but we need to be much more careful with the wiring to ensure low enough background noise.
Most small manufacturers will have exclusive service agents who will not break confidentiality agreements for supported products by showing schematics publicly. Usually, the manufacturer will just refer you to their agents if they have them or direct you to return the product to them. I doubt you will find schematics specific to this version that detail the wiring as well as you might hope for though. The original and later version hand drawn circuits didn't tell you much either.
At this point, given that there are now screw connectors, I would simply swap R & L channel boards outputs to the relay board, then power leads etc. to prove to yourself where the fault lies. The circuit of the PCBs will be the same as all A21 models so that will not be a great problem to follow from old circuits for which you already have links. If that proves the boards are no different, you then recheck the grounding by lifting each ground connection to see where a wiring duplication may have caused a ground loop etc. For example, one input ground can sometimes be a problem if both are grounded at the wrong point.
The speaker grounding point also influences this and it should be returned to power supply star ground, as the inputs grounds probably also were originally, if that point is still identifiable. Perhaps experiment with this for the inputs. There are some general fundamentals in this link and more amplifier wiring principles elsewhere in the articles section here: Earthing (Grounding) Your Hi-Fi - Tricks and Techniques
I am curious though, when looking at the image of the unmodified amp. There is no ground wire taken from each input to chassis ground as you describe for your amplifier. I would look at that connection to make sure it does not just form a ground loop by duplicating that of the amplifier PCBs.
Class A amplifiers always have more obvious hum issues because they draw high current all the time and this is even when there is no sound at all, so they seem more noisy. A class AB amplifier may seem quiet with no signal but be just as noisy when there is sound. You will then hear the sound + dynamic hum as distorted sound instead. Consequently, hum is more obvious but easier to trace with class A but we need to be much more careful with the wiring to ensure low enough background noise.
Last edited:
Good point Finch!
At first sight I didn't notice the GND differences between my version to "yours". Since this A21a (may be others too?) is totally dual mono (except for power transformer) , I see no point in linking the input GNDs to a single point before entering the power stages. More likely these two grounds are already "put together" in the preamp...
So this could be the trick. We are forcing the two dual mono stages to a local GND reference level without real need.
I'll go for that in the next try and will also "play" with grounds before stating that the DX PCB is faulty.
PCBs are twin but specular: it is not the same print but they are mirrored each others; it won't be that easy swapping them inside the case. More easy to swap the connections at relais board and see if it makes some difference at all. 🙂
Thank you for now, I'll update you in the w.e.
Bye
Raf
At first sight I didn't notice the GND differences between my version to "yours". Since this A21a (may be others too?) is totally dual mono (except for power transformer) , I see no point in linking the input GNDs to a single point before entering the power stages. More likely these two grounds are already "put together" in the preamp...
So this could be the trick. We are forcing the two dual mono stages to a local GND reference level without real need.
I'll go for that in the next try and will also "play" with grounds before stating that the DX PCB is faulty.
PCBs are twin but specular: it is not the same print but they are mirrored each others; it won't be that easy swapping them inside the case. More easy to swap the connections at relais board and see if it makes some difference at all. 🙂
Thank you for now, I'll update you in the w.e.
Bye
Raf
Can you confirm that the bias adjustment was done after the amplifier had reached normal operating temperature and the current drawn by either channel was identical in those circumstances?
Base emitter voltages differences between the temperature sensing transistors could make a difference between channels if the adjustment was done cold.
I found adjustments would get out of step during warm up of a Class A amplifier I built and it needed monitoring throughout to get it right. Hopefully your circuit will be less finicky than mine.
Did you test by shorting the input connection to chassis or just across the input terminals?
If the signal input earths do not connect to the power amplifier chassis you could have a floating input signal.
The feedback loop of a power amplifier is an input point and the wiring to it should be short. The longer it is the more susceptible it will be to pick up and amplify radiated or induced magnetic fields from a transformer or capacitor charging currents.
In the last respect the charging pulses will be significant since each module draws 1.1 amps continuously.
To minimize these effects the connected pairs of wires should be tightly twisted together - you may like to think about that later, it is not in evidence from what I can surmise from the image on page 1 of this thread. The wiring is also long and somewhat tenuous so the routing could be critical anyway.
In the meantime, if your input signal earths are floating you can take each to chassis earth via a resistor of 10 ohms.
Milanraf,
Your two channel amplifier has two inputs.
Each input is a TWO wire connection.
These TWO wires (or a coaxial cable) must run from the Input socket to the PCB input pads as TWO wires soldered at BOTH ends.
This requires four (4) wires to be soldered at the input sockets and four (4) wires to be soldered at the PCB pads for your two channel amplifier.
Check your input side and see if all eight (8) soldered connections actually exist.
Next:
check that each TWO wire connection has the TWO wires close coupled all the way along the route from the input socket to the PCB pads.
Close coupled could be a coaxial cable using the core and the screen as the TWO connecting wires, or a twisted pair of TWO insulated wires, or a figure of eight insulated pair of TWO wires.
Your two channel amplifier has two inputs.
Each input is a TWO wire connection.
These TWO wires (or a coaxial cable) must run from the Input socket to the PCB input pads as TWO wires soldered at BOTH ends.
This requires four (4) wires to be soldered at the input sockets and four (4) wires to be soldered at the PCB pads for your two channel amplifier.
Check your input side and see if all eight (8) soldered connections actually exist.
Next:
check that each TWO wire connection has the TWO wires close coupled all the way along the route from the input socket to the PCB pads.
Close coupled could be a coaxial cable using the core and the screen as the TWO connecting wires, or a twisted pair of TWO insulated wires, or a figure of eight insulated pair of TWO wires.
Last edited:
This ambiguous and my intended meaning was the supply wiring from the transformer to the rectifier bridge and capacitors on board each power amplifier module. This wiring is long and passes heavy bursts of current to each module onboard supply. These wires should be tightly twisted.
Lets not be confused. The origin of the pic I posted at #8 appears to be misunderstood. It is not the OP's amplifier which, if you read his posts, clearly states It is "neater by far" and shielded wire is used for the signal input leads. In fact, I believe by his description that it has been extensively rewired and "upgraded" with some likely errors in the process.
Having looked inside one of these, I recall the familiar Sugden standard of hand wiring in the pic which looks original to me.
Having looked inside one of these, I recall the familiar Sugden standard of hand wiring in the pic which looks original to me.