Hi All,
This is my first post as I just discovered this forum today. I have been in the repair of TV, video and broadcast equipment for the last 30 years or so. Recently a friend asked me to repair a HH MA150 amp which was totally dead. After checking it out I found both 3.15ma fuses were o/c. I found one of the output transistors s/c so I ordered a pair and replaced both. This restore life to the amp but it sounded distorted. I checked both + and - HT supplies, the + and - 15V supplies which all checked out OK. I checked out all the other transistors on the driver board as many are dc coupled to the o/p transistor that was s/c. All transistors checked out ok. The following day when I switched on the amp to have another go at the distortion fault I was presented with a loud hum sound. At this point I checked the 0 line which should read 200mV max. This was totally 'off the radar' so I switched off the power to avoid a puff of smoke and a loud bang. I also discovered that if I switch on and off the main power switch 5 or six times the unit will come on with NO him present. I suspect capacitor trouble so I checked the electrolytic s and found that all have a high ESR reading compared with new ones of the same value. Could this be the problem as the amp is over 30 years old. Maybe a block change of the caps is what is needed.
I welcome any comments from other members.
Thanks in advance.
Use your scope, either the power supplies are clean and free of ripple or they are not. This is a guitar amp, so close enough is fine.
Whenever output transistors blow, or any transistor for that matter, always check the associated resistors. They can go open or be damaged.
Posting the schematic or linking us to it would be a help.
Solid state amps don't require a load, unlike tube amps, so always work without a load until we know the amp is stable and not making DC. Run a sine wave through the thing, preferably into the power amp in jack or the effects return jack, if the amp has one of those. Look at the waveform at the output. How is it distorted? Or is it clean? In other words symmetrical clipping is a different issue from one side clipping. Missing half of the waveform is different than ripple modulation.
FIlter caps so worn that they no longer filter can cause problems, but just a little higher ESR is not going to cause gross symptoms.
Loud hum is either DC on the output or it is a total loss of filtration on one or more supplies. I doubt you need all new caps, but couldn;t hurt to resolder the main filters.
Thanks Enzo for you reply,
I will bite the bullet and make room on the bench where my scope and signal generator reside.
The main caps are mounted on the chassis and are well soldered and wired to the main board. When I saw the esr on the small 6.8uf caps all over the place I thought that it would be no harm to replace them. Your suggestion to use the eco return socket is a good point as it rules out a lot of the pre-amps etc. It's the 'intermittent' element that experience has led me to think capacitor trouble. As I stated earlier turning the unit off and on a few times will bring the unit on with no hum present and zero volts on the o/p to speaker socket. All the bias resistors checked out ok on cold test but as you know components can act funny when working voltage is applied.
I will keep you posted. Thanks for now.
Quote from: philly on April 15, 2016, 06:33:15 AM
turning the unit off and on a few times will bring the unit on with no hum present and zero volts on the o/p to speaker socket.
A bad solder connection or dirty connector could also give these results.
Does it respond to tapping on the board with a chopstick?
And for that matter, ball up your fist and whack the thing, or use a rubber mallet if you are queasy, that will expose loose connections if it reacts.
If you have good reason to suspect a part, then replace it and find out. But just replacing a lot of parts "as long as I am in there already" is a bad idea. Make the amp work right FIRST, then go back and update parts as you see fit. Every extra part you change is another opportunity to introduce more issues. "Can't hurt?" Well, really, it can.
Thanks guys for getting back to me,
I checked for dry joints, flexing pcb, etc as 30 years on bench repairs would have conditioned me to check for the simple and obvious things first. Power supplies are all OK and smooth. Most of the transistors are dc coupled so measuring voltages under a fault condition can be hard work. The o/p to the speakers swing to -60v ( sometimes +60v ) when the unit is powered up. The dc biasing setup has some caps which have high esr value so I have ordered replacements and will replace them all as cost is insignificant.
I will try and post the schematic here for your comments. At worst it may be useful to some other forum members.
Did we remove the load? 60v on the output won't draw current without a load, so we can continue to troubleshoot while it runs that way. Is that the case here?
That sure is a busy looking schematic.
I don't think your outputs will DC because of a little ESR. Look at C8, if it even shorts, all that does to the amp is put it maximally cold, so crossover distortion increases, but it won't upset the DC balance. And C7? Well if it shorted it would cause issues, but just a little ESR there, and it isn't going to create DC problems. With the output slammed over to a rail, you have a gross problem somewhere, not a nuance.
D1 is the heart of your bias circuit, but it is also a division point for the repair. IN a working amp, whatever voltage is on that diode, that is the voltage that will appear at the output - more or less. Let's say within a volt or so. SO if you have 60v on the end of D1, then the 60v on the output tells me the circuit to the right of D1 is likely OK, just doing its job.
The amp works then by having the signal yank D1 up and down to the music via Q2. Is the Q2 base looking like it is at appropriate levels to explain its collector? Or not?
Power supplies are OK? Fine. +/-60 (I assume) for the main rails, and +/-15 for the preamp. But did you notice the 741 op amp does not run on those 15v rails, it has its own 15v supplies derived from the main rails. See ZD1, ZD2 above and below it? Are both of those present right on the chip pins? A missing 15v there will throw the whole thing off to one end. I probably would expect less than the full 60v offset in that case, but it merits checking.
I urge you to resist replacing a number of parts "just because" until the amp has been repaired. Once it is stable and centered, THEN you can update parts.
Thanks for your reply,
I am not in the workshop at the moment as it is night time here on this side of the Atlantic. From memory, D1 and the resistors connected to it are different in the amp. D1 does not exist and there is a pot fitted which I think replaces R15 on the schematic. Both 15V +, - supplies are OK and I Isolated the op amp in case it was driving Q1,2 and 3 mad. Am I correct in thinking that Q 2,3 are for biasing / crossover setup.
With nothing connected to the o/p socket the volts swing to -60v. With a wire wound resistor connected it the voltage may swing to +60v.
With 60v of either polarity, you do not want a load on it. An 8 ohm load will try to draw almost 8 amps if you do.
If you can somehow test the op amp in isolation, so be it, but you cannot remove it from the amp and have the amp work. A solid state amp like this is one big loop, and if you break that loop anywhere it unravels. if pins 4 and 7 have good 15, then OK, and pin 6 tells the rest of the circuit what to do.
Sure D1 can be replaced with a single transistor circuit as seen in a zillion other amps, but regardless, the whole point of that diode or a transistor there is to maintain a voltage "space" between the bases of opposing drivers, Q6,7. The voltage at D1 with respect to ground dances around with the music, but the voltage it sets between the bases of Q6,7 is constant.
Q2,3 are your VAS, or voltage amplification stage. Everything from D1 (or equivalent) to the right is just a current amplifier, the voltage is already at full level by then. Clearly the signal voltage leaving the op amp is not remotely as large, so Q1,2 amplify the signal voltage. Q2 then controls what the bias thing does. Q2 does not set the output stage bias, that needs to remain at whatever D1 (or...) set it to.
If the thing can snap over to EITHER 60v rail, then chances are none of the transistors are bad. Open resistors come to mind, and broken connections.
Thanks Enzo,
Your explanation has clarified things a little more for me. I think the problem may be drive to Q1 which is connected to the 741. There is a - voltage there at the base of Q1( holding it off) and the collector is at full + potential and the emitter is close to zero ( - .7V). I have checked Q1,2, and 3 with my component checker all looks good with zero leakage, gain etc. Checked the bias resistors R9, 10, 11,8.
All measured ok, I even replaced one of these that lookes like it got stressed in the past. I am all out of 741's at the moment.
Do you have a TL071 or even a 5534 you could drop in there as an experiment?
And before we doom the 741, have you verified the resistors around it? R2,33,5,8? The feedback is important to the whole thing, loop remember.
And if you have not done so, look right above us here at the top of the section, where all the sticky threads sit. One is titled Book ABout Solid State. It was written by our own Teemu and is an excellent resource. It is there free to download from the author. I highly recommend reading it.
Hi Enzo,
Thanks for getting back to me once again. I was a little slow my end as I was locked out of this site due to my IP being mistaken for a spammer.However this is an update on the fault.
I replaced the 741 which restored the center voltage to 0 so things are now looking good. I put some guitar through it and it now does not have the distortion which it previously had after I replaced to o/p transistors. I have not injected a sine wave or scoped the o/p yet but things sound ok, however one thing I am wondering about is R13 (1k, looks like a 2W) feels hot to the touch. Q2, 3 feel cool. This amp has a variable resistor fitted between R14 and R15 which I guess is for setting up the zero or cross over point. The voltage across the resistor is 35v and the current I measured is 40ma. (1.4 watts) ????.
I just noticed that if i leave the amp powered up for about 10 mins or so the resistor cools down. I think I should change the caps at this stage. Some due in tomorrow.
Keep you posted and once more many thanks for all your help.
Update.
I injected a signal and all is not right. The o/p of the 741 pin 6 looks good at the base of Q1 but the collector is 'spikey' ( well nothing that looks like a sine wave). I will check the bias resistors later as I have to go out on some service calls.
I think we already decided they replaced the diode with a transistor for bias, your variable resistor for R14,15 would be bias adjust. Monitor mains draw current, and adjust that to the end of minimum current. That is not the correct setting, but it removes overbias as an issue while we repair.
Sine wave in, is the output clean when unloaded?
Hi Enzo,
To update on my progress. I was using the aux i/p at the rear of the amp while testing and I think I was hitting the amp with to high a level. I replaced the caps that had a high esr value and the resistors that looked like they had suffered from excessive heat in the past although they measured ok. Using the pre-amp i/p with a low level produces a good looking sine wave at the speaker socket with no clipping on the peaks. Should I check the bias adjustment before I put the lid on this one. If so what is the proper procedure for this adjustment.
Thanks for all your help.
Yes, important to test things under realistic conditions. I don't know what your test signal was that was too hot, but generally a line level signal works, but mostly as a functional check, I just plug a guitar into the effects return. If that comes out clean and strong, things are likely OK.
Of course check the bias. You can run a sine wave through it, doesn't have to be loaded, and look for crossover distortion. None is the goal. if you start with the bias all the way cold, there will be a crossover notch in the waveform. As you turn the control warmer, the notch will disappear. At that point, and further turning will just increase idle current and amp heating.
A quick and dirty way to do it is to monitor mains current draw. Start with the bias cold, and slowly turn it warmer watching current. The moment the current starts to rise, stop, and back off a hair. That's it.
A cold bias makes crossover distortion because the output transistors are not quite ready to conduct. But once the notch goes away, then the V+ and V- sides seamlessly stop and start at the same instant. If you keep turning it warmer, then both sides conduct at the same time for a certain part of the waveform. (or continually at zero signal) The more you warm it up the more current it lets flow.
Hi Enzo,
Scoped the sine wave on the o/p and it looked fine at the cross over point so I left well enough alone.
I played a guitar through it and it sounds good. I also left it on over night with no signal i/p. Today I played music through it for about an hour and it sounded good. The o/p tranny's feel cool to the touch which I presume is a good sign. The amp belongs to a guy who has been gigging for over 50 years. It hasn't seen a duster cloth since it left the factory but it is now ready to go back into action. Repair workshops are becoming a rarity these days so I was happy to help him out. At least the old stuff doesn't have any surface mounted components.
Thanks for all your help.
I just rebuilt the power amp stage of one of these on Friday and the correct schematic of the amp is the Laney KB80 .It uses 2N3773 as Output devices and Tip 31 /32 as drivers . Laney took over H/H .The power amp modules of the H/H MXA 100 ,MXA150 plus the Laney KB80 .The main difference is the output devices 2N3055 FOR the !00 and 2N3773 for the 150. The other difference is the Secondary voltages of the mains transfomer . Be sure to check the longtail pair TR3 and4.
Thanks Graham,
My old 100w VS combo uses 2N3055 as O/P's. I must compare the specs of both the 3055 and the 3773. The latter might be a better replacement if I ever need to replace the 2n3055's in my 100w.
By the way, I got the MA150 running ok with help from Enzo and others on this site. Replaced all the electrolytic caps that had high esr readings while I had the unit on the bench. It's back gigging again.