Hi just joined this forum as I'm having trouble with my amp and thought someone might be able to give me some insight into how to fix it. I recently bought a Marshall valvestate as not working (only £20 so worth a shot). The overdrive channel was very quiet but would sometimes cut back in at regular volume, splutter a bit and then cut out again. The pots were noisy too so I sprayed some cleaner in them and replaced the ecc83. This fixed the problem and it started working fine, except for when the volume would occasionally drop as it did before the valve was replaced. Hitting it always worked to make it cut back in though. A couple of days ago during a rehearsal there was a popping sound from the speaker and the amp cut out entirely. It powers on okay but the only sound I can get from it is a low hum/buzz from the speaker which remains constant regardless of what any of the controls are set to (volume and gain controls neither reduce nor intensify the hum, it is just one constant volume). Plugging a guitar into it makes no sound at all and also does not affect the hum. Touching the tip of the jack lead whilst the other end is plugged into the amp also makes no sound and I've tried a couple of cables. I've checked the bottom of the pcb for any dry or cracked solder points and can't seem to see anything wrong. Any ideas as to what could be wrong with it?
First thing, is the speaker cone moving hard forwards or backwards when the amp is switched on, or staying roughly in the rest position? If it is moving to one end of its travel and staying there it must be disconnected to avoid damaging it while other tests are carried out.
Second thing; take a known good lead and bridge the Effects Loop Send and Return - any change?
Try feeding a signal into the Effects Return, say from an MP3 player or the like - any response?
Try feeding the Effects Send into another amp - any response?
These test will help narrow down which section of the amp is playing up.
Hi, thank you for your response. I just checked your first suggestion and yes the speaker moves back hard when the amp is switched on. Bridging the fx loop send and return makes no difference at all. Plugging an MP3 into the fx loop return does nothing until the MP3 player is turned up past a certain volume, where the amp begins to crackle. The crackling intensifies slightly as the gain is increased and is worse on the boost channel, but disappears when the 20db cut is engaged. When plugging the fx send into another amp it just produces the same buzz as the valvestate. Does this mean it's a problem with the speaker? Should I disconnect it? Thanks for your help.
Absolutely - disconnect the speaker before it gets damaged.
This is a classic symptom of a failed output transistor. With the speaker disconnected, measure the voltage on the output connection (to chassis/ground). The two supply rail voltages aren't marked on the "pwr" circuit but the cap voltages of 50 volts suggest they will be 30-40 volts. When you measure the voltage on the output I'm sure you will find that it's something of this order, either +ve or -ve, and the polarity will give you a clue as to which transistor(s) have failed.
At this point you will need to make up a limiting lamp lead for testing after the faulty transistors have been located and replaced. See;
http://www.ssguitar.com/index.php?topic=2093.0 (http://www.ssguitar.com/index.php?topic=2093.0)
Post what you find, and please confirm the actual supply voltages you measure across C61 and C62 (2200uF/50V, in power supply).
In the 8080 combo rail voltage is slightly more than +/-40V (usually 42).
On the 4100 head, which is the exact same schematic but meant to drive a cab or two, voltage is somewhat smaller (+/-36 to 38V).
Output transistors can be replaced by TIP142/147 .
okay, I've measured what you said and I'm getting +46.7v across C61 and -46.9v across C62. But on the output I'm only getting -8.3v.
Pretty sure I'm measuring it correctly but then I'm a bit of a novice so just to double check it's + to the speaker output wire and - to the chassis (ground)? Sorry if I'm being an idiot. Thanks again for the help though.
- 8 or 9V DC at the speaker terminals is *bad*.
Normal is less than 0.1V DC , go figure.
No, you are not being an idiot, those are good measurements.
We have good supply rails, but the offset on the output tells us there is a problem in the power amp section, so we are on circuit 8080-pwr.gif (I'm appending pc0689p.pdf which is clearer).
Now this one is a bit tricky because the output devices are not normal transistors but a dual composite called a Darlington where the driver and main output transistor are in the same package and testing them is not quite as straightforward as normal transistors.
What seems to be happening is that the "lower" transistor TR10 is pulling the output down a bit, but not fully as one would expect with a shorted transistor in the TR10 position, so I'm inclined to suspect that the opposite "upper" transistor TR8 has gone open circuit.
If the board has component identifiers printed on it, find the diode D5 and measure the voltage on each end (w.r.t. ground).
Let us know what you find.
{JM - you got any suggestions how to tackle this one?}
Well, step by step.
I'd start by measuring all transistors (with amp unplugged from the wall outlet) for shorts.
We are looking fror gross shorts here, so we can start by measuring them still on circuit, needlessly pulling all them for measurement will do more harm than good.
Builda lamp bulb limiter, you'll need it later.
The original Marshall 8080 power amp is somewhat tricky, don't know what they tried to achieve, but it's still repairable, of course.
Later Msh amps went back to more conventional circuits.
Google how to measure transistors for shorts, I guess Rodd Elliott has a good tutorial.
No, not Rod that I can see, but here are some others;
Making Electrical Measurements Part 3:
Testing Diodes and Transistors
http://www.elexp.com/t_test.htm (http://www.elexp.com/t_test.htm)
Meter check of a transistor
http://www.allaboutcircuits.com/vol_3/chpt_4/3.html (http://www.allaboutcircuits.com/vol_3/chpt_4/3.html)
Rudimentary transistor test (diodes and TO3 power transistors)
http://www.youtube.com/watch?v=wFHAGd_Tcvw (http://www.youtube.com/watch?v=wFHAGd_Tcvw)
Google how to test transistors
http://www.google.com/q=how+to+test+transistors (http://www.google.com/q=how+to+test+transistors)
Okay, I've gone ahead and replaced both TR10 and TR8. For what it cost I thought it was worth it just to be sure. However, the problem is still there with no change whatsoever. I've read elsewhere that bad filter caps can cause similar problems. Could this be the case? If it looks like there's component damage that'll be difficult to isolate or that I can't fix without taking it to a tech, I may give up on it. There's only so much money I can put into it before it becomes false economy, becuase I've decided that regardless of whether or not I can get it working I'm still going to sell it afterwards as I simply don't trust it enough to gig with given the problems it's had. I'd rather sell it, make my money back and get something newer and more reliable. Whether or not I can fix it will be the difference between selling it for parts or repair, or selling it as working, with the latter obviously being preferable as I'll make more money from a working amp (I obviously won't sell it without mentioning its history). But yeah...if you have any other suggestions let me know. Thanks.
Firstly; you must feed back your findings and measurements to us so we have some idea of what is going on at your end. There is no point in randomly replacing components unless they actually test faulty.
Secondly; for problems in solid state output stages making up a limiting lamp and using it as directed when bringing up the amp after repair is not optional (unless of course you enjoy replacing brand new transistors you have just fitted).
http://www.ssguitar.com/index.php?topic=2093.0 (http://www.ssguitar.com/index.php?topic=2093.0)
Thirdly; no, this is dead silicon, not leaky caps; it's just a matter of locating it all.
Okay, now, given that you have already tried the amp without a limiting lamp, what voltage do you have on the output now that you have replaced the main power transistors?
Did you test the ones that came out? With what result?
My thoughts here are that this is not an output stage problem, I suspect TR8 TR10 are doing what they are told.
As Roly said, filter caps won;t put that DC on the output. They will make loud hum if they are bad, but not DC. DC on the out will ALSO make loud hum. And so will a bunch of other things.
You had about 46v on the main rails, close enough. But you also have to check the low voltage supplies. This power amp is driven by an op amp, and that needs both +15 and -15 up aqnd running. Pins 4 and 8 of the IC. Is it there?
If one of your 15v rails has failed, it will leave the IC lopsided power wise. It won;t be able to center itself and thus center the output at zero. FInd R101 R102, 5 watt resistors. SHould have that 46v at one end and 15v at the other on each one. One will be negative and the other positive.
It may be a coincidence, but if one of your 15v rails is missing, your op amp will try to center on what is left, and that would be just about 8 volts.
And along with the op amp, TR5 TR6 need the 15s so THEY can center as well.
TR4,5,6,7 are cheap common types, and wouldn;t hurt to replace them, but if I recall your outputs were not really shot, so likely did not damage those smaller ones. I could be wrong, but slapping the whole deal over to a rail is a more likely result of one of those failing, not 8 volts.
Right now, my money is on a bad zener.
Okay Enzo R101 and R102 aren't giving me what you say they should. R101 is giving me +47.3V on one end and +16.2V on the other. On R102 I have exactly -47V on one end and -0.08V on the other.
Regarding the op rails, which IC is it that I need to measure pins 4 and 8 on...Is it IC8B?
That's what Enzo suspects.
A shorted Zener killing the -15V rail.
INCREDIBLE marksmanship, as usual.
Replace that Zener, check whether you recovered your -15V rail and whether your amp now works with little output offset.
Don't even go to the IC yet.
Your power supply is missing the -15v rail.
R102 has -47v on one end and next to zero on the other (where -15v belongs). Now either R102 is open, so the voltage can;t get to the othger end, or the -15v is shorted to ground. This is easy to determine.
The other resistor R101 will naturally get pretty warm during operation. COmpare R102 to it in terms of heat. If R102 is open, it will be cold. But if the reesistor is intact, but the zener or other item on the -15v end is shorted, then R102 will be VERY hot.
And it is also possible the zener is shorted and that caused the resistor to open. SO if you find the resistor open, you still have to check for shorted -15v.
About the op amp ICs. They are powered by +15 and -15, but they have no ground connection. They sometimes fail and short the two power supplies together. This makes them hot, and the two supplies will BOTH be very low or at zero. But a failed op amp IC really doesn;t have a way to short just one power rail to ground.
SO my main suspect still remains the zener.
I forget which one is -15, but it is the one connected to R102.
Sorry I got R101 and R102 the wrong way round when I gave you my measurements, it's actually R101 that's giving me the -0.08V reading not R102. And yes R101 seems to have shorted, it's VERY hot. So I'm guessing it's ZD1 that needs replacing. And that's a 15V 1.3W zener? Guess I'll order some. How exactly does a zener fail? Is it something that can wear out gradually? I'm just wondering if it would be worth replacing both of them as if one has failed the other could have have problems too. yes?, no?
Quote from: Enzo
It may be a coincidence, but if one of your 15v rails is missing, your op amp will try to center on what is left, and that would be just about 8 volts.
Very well spotted. :dbtu:
Quote from: JamesAlexander93
How exactly does a zener fail?
That is a very good question. There are a number of failure mechanisms in solid state devices, but the best bet here would be heat, and thermal cycling. Semicons generally don't like getting too hot since it tends to reactivate physical/chemical processes used during their manufacture. Thermal cycling is a bit easier to understand since it results in physical stresses that change from cold to hot to cold and physically "work" the various parts of the semicon, the dice itself, its interconnects, the encapsulation, and in particular the seals between the lead-ins and the encapsulation.
Should you replace both? I have to say that Zener diodes in general are not my favorite devices. This may have more to do with designers pushing them to the max than anything inherent to the devices themselves, but generally they do seem rather failure-prone, so I'm inclined to say "yes", renew with fresh Zeners in both positions.
Thanks roly, I'll replace both. I'm a little unsure as to which diodes to buy though as I have two schematics for the 8080 power amp and one tells me the zeners are 15v at 3W and another tells me they're 15v at 1.3W. Do you know which one is correct?
Probably the original ones were the 1.5W type and after a certain history of problems (like yours ;) ) they upgraded to 3W .
Higher power capability never hurts.
Okay, 3Watt ones seem to be a little scarce would it hurt to go up as far as 5Watts? Or is that too much?
Scarce? First place I always look is Mouser, and they have:
15v 3W 1N5929 for 37 cents each, and 1500 of them in stock
15v 5W 1N5352 for 36 cents each, and almost 15,000 of them in stock.
Take your pick. For 36 cents, replace them both. Watch which way they go in.
Can you tell what parts are in yours? Your amp may have 1.5W zeners, and maybe the 3W on more recent drawings is the step up you need.
Are your existing ones mounted right down against the circuit board? When you install new ones, leave teh wires long enough for thgem to sit up in the air a little - half an inch maybe. That allows more air around them for cooling, PLUS the extra wire length will act as a small heat sink - honest.
How do they fail? All manner of things can go wrong with them, including random component failure. Heat is not their friend, but they run hot by their nature. SOme other component failure COULD have exposed them to some unwanted voltage or current, and stressed them. Those are matters of why. As to what goes wrong inside them, they can suffer any failure a regular diode can, plus their zener voltage can wander way off. They can short, they can open, they can get intermittant, they can even get noisy.
As to figuring out what was wrong with the amp, despite what my wife says, sometimes I get the right answer.
Quote from: Enzo on September 06, 2012, 04:02:04 PM
As to figuring out what was wrong with the amp, despite what my wife says, sometimes I get the right answer.
Clever Man if you can use your Wife as a zener diode tester!!! :lmao:
Mine is Obviously not working well,, as it gives false results all the time. :loco
ps; I had my SS Laney work horse go dead a few months back while playing and sure enough dead shorted Zener diode for opamps. :grr
The drop resistor was hot as hell,,,which was my clue and sure enough two new Zeners and everything was fine. :tu:
Replaced a few suspect caps while inside it.
Phil.
Quote from: phatt
Wife as a zener diode tester!!! :lmao:
{Ooooh no, I'm not going anywhere
near that. :-X }
The power rating of a zener, like a resistor, is how much power dissipation the device can handle without being damaged. The only problem you are likely to strike is that as the power rating goes up so the devices tend to get physically large (and in this case lead thickness). So apart from that there is no reason why you couldn't use 5 watt zeners if that is all you can get. Going
lower simply means it will overheat and burn out in a hurry.
Quotefalse results all the time
Well, that's *good*. :o
You can ask and then do the exact opposite, with excellent results. :tu:
The problem is when the tester (electronic, human or female ;) ) gives random answers, which average 50/50% or thereabouts.
Hmmmm...
Yes, my wife is like a zener diode. If things go her way, they flow just fine, but when things go against her way, she resists them, until it gets to the point she gives in. Things just avalanche from there.
Quote from: J M Fahey
human or female
female =/= human ... I often wonder. :loco
My wife's Subaru Liberty has developed an occasional and total loss of power for a second or five that is driving her, her mechanic, and therefore me, stark raving nuts.
Ha, techo-pun by
Enzo :lmao: (just to let you know Zener/avalanche wasn't wasted).
SWMBO (http://www.urbandictionary.com/define.php?term=swmbo) is more your 1N4007 - gets a bit hot and bothered if her PIV is exceeded.
Quote from: Roly on August 30, 2012, 07:52:43 AM
First thing, is the speaker cone moving hard forwards or backwards when the amp is switched on, or staying roughly in the rest position? If it is moving to one end of its travel and staying there it must be disconnected to avoid damaging it while other tests are carried out.
Second thing; take a known good lead and bridge the Effects Loop Send and Return - any change?
Try feeding a signal into the Effects Return, say from an MP3 player or the like - any response?
Try feeding the Effects Send into another amp - any response?
These test will help narrow down which section of the amp is playing up.
Hi Guys
Newby to this forum. I found it because I have the same amp problem.
I tried
"take a known good lead and bridge the Effects Loop Send and Return" and it made huge difference so can you help me now. What is my next step please?
Hi chezqui and Welcome, :tu:
So you now are faced with 2 options;
1/ Leave the short lead in place which has fixed the problem and likely cost close to nicks. :dbtu: :dbtu: :dbtu:
OR
2/ Spend a lot more money, time and effort pulling it all apart and replacing with a new switching socket which will more than likely have the very same design flaw as the original part,, which means it will happen again down the track. :'( :'( :'(
A majority of the socket designs that are used in most guitar amps nowadays are just crap and are a very common fail point. The cost cutting is now so extreme that you are lucky to get more than a few years of service before a socket fails. The plastic itself is no longer plastic the molded plastic sockets have filler mixed into the plastic to make it go further hence they become very brittle and crack. the metal tab contacts are only half the thickness of the older types. And the list goes on,,,,All in the name of $$Profit$$
So Brand Name now is just meaningless marketing BS. :trouble :grr :trouble :grr :loco :duh
So obviously my advice would be leave it alone; The short patch lead is the most reliable fix. :dbtu:
You can buy electrical cleaners which may help,, spray and insert the plug in and out many times to help clean and dirt and oxidation from the contacts but I've found the patch cord to be the best fix. :tu:
Phil.
Great answer, thank you. Spookily I have taken the leads out and it seems to be OK at the mo.
Still some frying bacon going on but we can maybe discuss that later. :dbtu: