I'd really sincerely appreciate help - I've just bought a lovely seldom-seen amp model (Crafter DSP1) - about 10 or 15 years old. I bought it in a noisy environment, and discovered very conspicuous hum once using it at home.
I managed to find a schematic, which is attached here.
I have long experience with passive guitar circuits, and some with simple stomp/active circuits, but very little with mains audio electronics.
The hum is irrespective of the volume control and starts before the output relay switches on. With the amp loud, the hum is drowned out and the amp sounds very good, but quiet playing is spoiled by it. I estimate the hum at 100Hz (comparing it by ear with a tone generator). UK mains is 50Hz.
As a best-guess before looking for help I speculatively replaced the two big 4700Uf power supply caps and the two additional 470uf caps - no improvement, nor from changing the four rectifier diodes :-(
Plugging a jack into the fx return makes no difference to the hum. None of the amp controls makes any difference. The PCB is in good condition, and with a careful look I see no obvious solder bridges, past repairs, scorching or disconnections.
Of course the symptoms lead me to figure the issue is after the preamp stage, and that it is mains-related hum (post-rectifier?) but I`m really not sure how best to further troubleshoot this.
Please does anyone have any other suggestions (or any encouragement!). I have several basic multimeters to hand, could make an audio probe, but have no oscilloscope. I'll really be very very grateful for help... I love this amp, my first one specially for acoustic guitar, but the hum is really disappointing. I can`t easily afford to get this professionally fixed, and also know both how thrilled I`ll be to fix it with my own hands, and how much it would make it `mine` ;-)
Thank you,
Rob
If you put your DMM on AC volts and connect it to the output of the amp (turned on) what reading do you get?
Using the send and return try and see if the hum is in the preamp or power amp or both.
Do you hear the hum in the headphone output?
gbono,
Thank you for the guidance: here are answers and more...
1. Speaker output:
0.2v DC
0.4v AC
2. Send/Return
There is no hum on the fx send. Signal via return has hum. Nothing affects hum level.
3. Hum in headphone output?
Yes
Also
Power amp & pre amp power supply rails (A, B, C & D on the sch):
(I'll double check these tomorrow. A couple of components seems to have their labels swapped schematic vs pcb!)
A: 11.6v DC, 25.0v AC (!)
C: 10.8v DC, 23.2v AC (!)
B: -25.8v DC, 0v AC
D: -24.8v DC, 0v AC
I'd be very very grateful for advice!
Thank you again...
If there is no hum at send (J3) then the issue is not in the preamp if I read the schematic correctly.
Recheck voltages at A,B,C,D and +- VCC. Make sure center tap is grounded....make sure your orientation (polarity) of the replaced filter capacitors C55/56/57/58 is correct.
Remeasured rail voltages,pretty much confirm yesterday's measurements:
A: 12.0v DC, 25.8v AC (!)
C: 11.2v DC, 24.1v AC (!)
B: -26.6v DC, 0v AC
D: -25.5v DC, 0v AC
additionally:
VCC- : -33.1v DC, 0v AC
VCC+ : 32.9v DC, 72.8v AC (!)
The caps C54-C58 and the diodes D6 to D9 are oriented as they were originally (I took photos!) and in keeping with markings on the PCB.
Centre tap is grounded.
In case it is useful I've attached an image of the front and back of the actual circuitboard superimposed!
Sincere thanks.
Quote
A: 12.0v DC, 25.8v AC (!)
C: 11.2v DC, 24.1v AC (!)
B: -26.6v DC, 0v AC
D: -25.5v DC, 0v AC
additionally:
VCC- : -33.1v DC, 0v AC
VCC+ : 32.9v DC, 72.8v AC (!)
Disregard your VAC readings
when there is also DC present, i.e. your multimeter
can not read ripple.
Cheap multimeters do not have a *real* AC metering circuit, but use the same DC one , just add a diode in series.
And since you are chopping half the sinewave off, they multiply reading by slightly more than 2X .
It does work, sort of, to measure pure AC, such as wall voltage or transformer taps and even the output of an amp (if no DC present) but give it +10V DC and it will think it is 22 to 24 VAC :duh
And give it -10V DC and it will claim "0 V" ... because diode is reverse biased.
So you can't measure ripple with your present multimeter and it's not sensitive or accurate to measure hum at the output.
To know which is which, those who can't usually have only 2 VAC scales, typically 200VAC and 750VAC or so typically go for around U$10; while those which can are much more sensitive , usually have scales of 2VAC or 200mVAC because they have a specialized AC circuit inside, also can read ripple properly if set to VAC, and typically cost U$40 up.
JMF,
Thank you for that information: I had no idea. I looked out an old analogue meter that has 4 ranges (10v AC and 50v AC as well as higher ones) but it reads the same so guess it's working in the same manner as my (yes, cheap) DMM.
I will very seriously consider buying a better DMM, although since very little of my (limited) electronics work involves mains circuits I have not come across this issue before and perhaps seldom will again: that said, my enjoyment of this field has grown a great deal since I originally turned to it out of neccessity, and I now take on challenges (such as this one) that I would not have touched before.
Is there any possibility, despite these readings being quantatively incorrect, that someone here could offer a suggestion as to what failure might be giving rise to these readings? Are there useful tests I can do with my existing meters or by swapping out components? I certainly can't get another meter for a week or so and am longing to fix this amp if at all possible...
In the meantime, you can put a non-electrolytic capacitor in series with your meter probe to block the DC when measuring AC. Something like .1uf with a sufficient voltage rating for the highest voltage you will measure, I would use 600V if you have it.
Yes, try it.
Many old analog meters have an extra probe jack labelled "output" , which has exactly that: a .1 or .047uF x 400 or 600V meant to measure audio levels ("output") on tube plates, which of course have 100 to 300VDC on them.
For measuring audio levels with any cheap multimeter I built an "AC probe" , like old VTVM tube meters had.
Old intelligent tricks still work, of course.
This one , labelled "RF probe" ,actually reads audio very well, and is calibrated for 10M input voltmeters (Tube or Digital) :
(http://www.n5ese.com/rf_probe_schem.jpg)
The 1N34 Germanium diode allows it to read very low voltages, down to tens of millivolts, but stands maximum 30/35V peak signals.
I built basically the same, but lowering impedances 10X , so used a .047uf cap, a 1N4007 diode and no series resistor ( I'm not interested in RF but Audio) so mine reads peak voltage, only problem with silicon diodes is that they chop 0.7V from the reading, so I can't read guitar pickups and such, anyway it's a useful tool to trace audio and measure ripple.
If you build the original one, remember cable from probe to meter must be shielded.
I don't have any high voltage caps but will pick a couple up tomorrow, and try re-measuring.
I sincerely appreciate the very educational advice.
I'm still very eager for any suggestions as to what might remedy this issue: beyond the four power supply caps and four rectifier diodes which I've replaced, I'm at a loss. Of course I'll re-re-check what I've done...
Please do 2 tests for me:
1) lift the top leg of C46 (.1uF) ... any change in hum/buzz?
2) short across R60 (47k) with a short piece of wire ... any change in hum/buzz?
Please post results.
Note: "top", "left", etc./ refer to "as seen on the schematic" ; in the actual PCB they may point anyway so follow the track to check what the suggested end connects to.
JMF (& G1),
Thank you, I really appreciate your help.
Lifting the top leg of C46 (top as on schematic, the leg towards R83) does not seem to change the hum at all.
Bridging R60 made some difference whilst C46 still connected, maybe 50% less hum. Seemed not to make a difference with C46 top leg disconnected, although not entirely easy to be sure. There is still plenty hum in each configuration.
I've tried using a DMM with a series cap (metallised polypropylene) and also to build the probe with cap and diode. I'm embarrassed to say I'm then metering 0v in both cases DC and AC.
I'm curious about the transformer output between the secondary coil and D6. Following that, I find D3 seems to leave clean DC in the FET2 / FET3 circuit, but the relay circuit meters some AC either side of all its diodes. This being at all significant is pure idiot speculation on my part, given my lack of expertise and of equipment, but I'm trying!
I will be very very grateful for more guidance.
Do I smell a ground loop issue here? :-\
Maybe common (circuit ground) is connected to chassis in more than one place. There is a mounting hole right next to R86 (top left in pic) try leaving the screw out when remounting pcb. If it improves hum/buzz then a rethink of the mount will be needed if hum gets worse then not the issue,, worth a try though.
I've fixed hum issues like this before by just installing fiber washers on mounting holes.
Ground plane hum issues can be a nightmare to track down as there are often many places a circuit can be grounded.
Phil.
Thank you Phatt.
The design seems to have been quite thoughtful as regards earthing. When I tested all the earth connections in the schematic for resistance, they all connected to the same single earth point on the board, a soldered wire between the rectifier diodes and the large electrolytics, which connects to the mains input earth and the chassis. There are two mounting holes like the one you point out: neither has PCB track under it. The bodies of the potentiometers are earthed through their mounting to the chassis, but are isolated from the circuit.
This issue is beyond my skill level to resolve without expert help, the main clue is evidence of substantial ripple on the positive DC rails. This seems to induce hum in the power amp stage directly, as the hum starts before the relay triggers at power up and is unaffected by lifting a leg of the coupling capacitor bringing signal into the power amp.
I'd be delighted to be wrong about everything - I'm just longing to fix the darn thing (rehearsed without it last night!) and eager to learn as much as possible along the way. There's no way I can fix this without some guidance: I'm a hard worker but am much more familiar with a 9v battery by way of power supply :-/
Ok clears that up, Then other things to consider;
EMF radiation from the transformer can induce noise onto the signal especially if the Transformer is close to sensitive circuitry. Older gear often had steel bell ends on power transformers but these days you are lucky to get a copper strap shield.
The transformer can also vibrate form poor mounting or even the winding itself can cause audible buzz, though unlikely as ripple on DC rails.
Also your unit has an onboard DSP,, hum? Can anyone recall a recent issue posted about a huge hum/buzz on a amplifier which had a DSP board.
In that case I recall the supply to the DSP section was not filtered well and I believe it was resolved by additional filtering to the DSP. I'll have another look when time permits and see if I can find it,,just thinkin out loud.
Phil.
Phatt,
I think I read that myself regarding the DSP introducing noise! Thank you for thinking of it.
In this case I'm not so sure: there is no hum in the send of the send/return loop, which I believe is post-fx. There is hum before the signal-path relay is triggered at power on. The power amp still hums when the coupling capacitor of its signal input has a leg lifted (C46 on the schematic).
What is interesting is that there does seem to be unrectified AC tapped into the FX, relay and FET2/3 subcircuits, so maybe something there could be leaking. Just for my education, can anyone explain why there is a power connection from both unrectified (pre-D6) and post-rectified power (after R75)? All 3 subcircuits seem to be part of that power loop. As I mentioned earlier, there seems to be lots of ripple within the relay circuit power, the FETS seem to get clean DC via D3. I'll measure the fx circuit power in situ when I get the chance.
So the anomalies are hum in the power amp and a nasty ripple on the +ive power rails: if the hum has a cause in one of the above subcircuits it seems it is being induced in the power amp via the power rails rather than noise in the input signal.
I've retraced the rectifier and smoothing capacitor circuit today and they appear to be in agreement with the schematic. The key components seem to meter as one would expect.
Instead of lifting C46, ground it's end that connects to Fet3. Any effect on the hum?
The unrectified AC that is feeding the relay and Fets is for turn on muting.
When you say there is large or nasty hum on supply rails, please provide numbers you have measured.
Thank you! Grounding that end of C46 increases hum. Short recording attached.
As yet I only have the measurements I posted earlier in the thread which I have been warned by JMF are untrustworthy since I have a cheap DMM which probably cannot measure ripple. I'm puzzled that attempting to use a capacitor in series with the probe simply yields 0v on all scales and measurements. I hope I haven't invented a whole new kind of stupid.
I'll order a better DMM today if i can find a worthwhile one I can afford which will offer better ripple measurements.
Would anyone encourage me to buy, or warn against, a UNI-T UT39C http://www.ebay.co.uk/itm/291301645135 (http://www.ebay.co.uk/itm/291301645135) or a Mastech MS8268 http://www.amazon.co.uk/dp/B000JQ4O2U (http://www.amazon.co.uk/dp/B000JQ4O2U) ? I could spend a bit more if need be, but can't go 'pro'!
I sincerely appreciate your help with this amp...
Yep a pretty bad hum/buzz. :( >:(
I wonder if you could tempoarily bypass the relay and mute? That seems to be part of the issue.
I'm thinking,, lift output side of R27 (Preamp output) while also lifting Input side of C46 (poweramp input), then bridge the two with a wire.
Of course you still may need to kill power to both relay and mute to be sure they are not introducing hum. Of course making sure you keep power at IC7 to run FX circuit.
All that likely easier said than done ay? :-X
Phil.
Had time so I dug into the picture and schematic 4U.
My suggestions would look something like this. I've marked with a red cross each part to lift.
Just stand R27 up on end and a wire across to C46. (I think it's C46) A couple of parts are hard to read so double check.
Phil.
Thank you! I'll take a run at it tomorrow...
Opps, forgot about the Q of meters.
This one is better; Mastech MS8268 (Higher CAT rating :tu:)
The UT39C has the lower Cat rating and is fine if you never need to work on mains voltage, obviously Co$t goes up with the better units.
Phil
1. The mod to take the relay out of circuit left the hum unaffected. I have left it in place in case of any suggested checks/tests. Thank you so much Phatt for marking it out for me.
2. Using an iPad app and a USB audio dongle I grabbed a trace of the hum waveform via the headphone output - image attached
3. I bought a more sophisticated DMM: UNI-T UT60E http://uni-trend.com/UT60E.html (http://uni-trend.com/UT60E.html) - true RMS AC
This meter does not report AC on the + rails VCC+, [A] and [C], although both my basic DMM and Analog meter do - none of the meters show AC on the -Rails
RESULTS:
RMS DMM:
-VCC: -32v DC, under 0.1v AC
B : -25.7v DC, under 0.1v AC
[D] : -24.7v DC, under 0.1v AC
+VCC: +32v DC, under 0.1v AC
[A] : +11.5 vDC, under 0.1v AC
[C] : +10.8v DC, under 0.1v AC
Simple DMM:
-VCC: -32v DC, negligible AC (due to diode bias?)
B : -25.8v, negligible AC (due to diode bias?)
[D] : -24.8v DC, negligible AC (due to diode bias?)
+VCC: +32v DC, 71.1v AC
[A] : +11.6 vDC, 25.5 AC
[C] : +10.8v DC, 23.6 AC
Analog Meter:
-VCC: -32v DC, negligible AC (due to diode bias?)
B : -25.5v, negligible AC (due to diode bias?)
[D] : -24.5v DC, negligible AC (due to diode bias?)
+VCC: +32v DC, 90v AC
[A] : +11.5 vDC, 26v AC
[C] : +10.6v DC, 24v AC
Incidentally the Hz range on the meter reads 50Hz against speaker/headphone output.
That waveform does look like something a bad cap could cause. Do you see that kind of waveform on any of the supply rails?
I don't have an oscilloscope, only an app that can plot an audio waveform
:-/
I hear tell of kludges using voltage dividers, I don't know if anyone here would recommend it... ?
I've changed the two 4000uf, and two 470uf caps plus C60 and C49. Any other suggestions?
Thank you for the thought y'all have been putting into this, I really appreciate (and need!) the help.
Now that you have Fets and Relay out try lifting R59 this will kill power to FX but the amp should still work. If you still have hum then maybe a dead diode in the bridge Rectifier.
Phil.
Thank you again! Just checking in to say my lovely, much used top-of-the-range butane soldering iron has stopped working and is going back to the (excellent) manufacturer (Antex) in the hope they can revive it. I'm asking them to send me a cheaper model as a spare, but I'm out of action for a couple of days :-(