Acoustic model 450 repair blog

[vertamps]

Iv had an acoustic model 450 sitting in the closet for 5 years that i will be repairing and keeping this post updated with the progress.
Power supply cap'

10 minutes later its down to 84 volts, leakage test was fine and its holding charge like a boss.

[vertamps]

Just replaced all the aluminum electrolytic capacitors on the amp minus the 7800uf power supply cap. I had purchased the 2000uf output coupling cap 5 years ago and replaced it.
*Schematic deviation: Contrary to the schematic the output coupling cap was placed befor the switchable output inductor 'L501 on my amplifier, i dont see how it could hurt.

[Enzo]

The order there doesn't matter. It is an inductor and a cap in series with each other and in series with the load. So the order in which they sit doesn't make a difference.

[vertamps]

I thought so.
The problem with the amp before the recap 'if i remember from 5 years ago it was distorted on all settings. I should have tested it again befor working on it, but owell.

This thing is LOUD. However, all eq set to mid-level and with the distortion switched off my humbucker equipped guitar playing through it there is clipping when the volume knob is at noon/ mid point.
* Also with no signal going through the input the amp its heating up the chassis around the ouput transistors.
I will pull and check them.

[Enzo]

They won't be faulty, they will be improperly biased.

Look on the schematic in the power amp, ther is a small area called "View A" and that is the bias circuit. If you have board number 170045, it should already be adjustable. Do you have small trimmer controls on the powr amp board? The drawing has voltages all over it. In most cases the absolute voltage is not the issue, what matters is the relationship between two points. Q404, Q405 - note their base voltage is about 2.4v different. That difference is what matters, not whether it is 44.2 or 43.2. The smaller that voltage difference, the cooler the amp will run. Of course too cold and you get crossover distortion.


Oh well indeed. it isn't a good idea to "fix" something if you don't know what it does wrong in the first place. A good rule of thumb is FIX an amp first, then go back and update whatever parts you plan to replace just for age.


Om amps of this age I expect to replace all the small electrolytics in the signal path. The small 1uf, 4.7uf or similar caps.

[vertamps]

My board says 170051-A, no trim pots. Thank you for pointing that out on the schematic and the info, it would have taken me some time to figure that out.

Iv found so many bad capacitors in old gear that on my own stuff i replace them because of restoration. I also bought caps in bulk from Allied Electronics 'excessive inventory' and need to use them. I had a Peavey amp with several electrolytics in the signal path that were bad and it would have saved me time to have replaced them all at once than trace on the oscilloscope through each section just to find another bad one (which is what happened).

The next step is to check the voltages and then decide to mod the board for adjustable bias and symmetry.
BTW, whats your paypal so i can donate?

[g1]

Near the middle up at the top of the page, there is a green "donate" button.

Edit: Maybe you want to donate to Enzo . If so, send him a PM.

[Enzo]

My paypal? Nah, just donate to this board. Look top of the page for the green DONATE button.

And when someone comes to you wanting to know something you can tell them, help them out. Good Karma.

[vertamps]

Ok, after this post i will keep it to the Acoustic 450 update.
-$20 donate to start.
Could there be a future addition to this forum for a synthesizer section? There are plenty out there that need service. I belong to a facebook based synth repair forum but this would be a better avenue with the saved history and search functions.

[Enzo]

Long ago I worked on piles of synths, but I rarely see them in this shop any longer. Either they don't break, or there is someone in the area servicing them. Could be you. This forum is mainly amps and pickups. We do discuss synths when they come along, usually in either the Music Electronics section or maybe the "other amps" section. Our members always surprise me the expanse of things they know.

[vertamps]

On the lowest bias voltage with the trim pot at 1.7ohms (short) Q405's base is 47.8vdc and q404's base is at 46.3vdc with no heat coming out the back.
With q405 at 46.3vdc and q404 at 45.3vdc there is some warmth on the heat sink/chassis but defiantly not hot, and working fine.
There is distortion in the waveform but i have not installed the Symmetry pot yet.
* side note i have a mismatch of power transistors: Two are ON semiconductor 2n3773G's the other four are 480037's.

Also understanding why the voltage difference sets the bias is probably above the context of this thread.

[Enzo]

If you don't make yourself understand the biasing of this amp, how can you know if what you do to it is a good idea or not?


The symmetry control is merely an adjustment to center the output in the power supply. Your amp is a single sides power supply of 92vDC. The symmetry control sets the output right at half that, 46v. If your 92v is something else, then set it for half whatever it might be. However, the only real time that should impact distortion is when you clip the power amp. By centering, any time we clip it ought to do so symmetrically. Anything short of clipping the output, and it won;t really care if it is offset from the middle a bit.


Look at any basic amp. The output centers on the power supplies. In this case a single supply and the output at half that. On most amps, we center at zero which is halfway between V+ and V-. Now think about a plain old power amp with dual supplies. The outputs take turns conducting to make the positive and negative portions of the output signal. We don't want both halves conducting at the same time, that just shunts the two power supplies together. If both sides are turning on, the amp runs hot. On a dual supply amp, the symmetry control sets the output to zero.

Think about the output transistor. Its emitter connects to the speaker, just as the one on the other polarity does. Using the zero volts at the output as a reference, think about what turns on the transistor. You need to turn on the base to emitter junction. That happens when there is about a half volt on the base. So as the signal from the amp starts to go positive - and the same thing will occur on the negative side in turn - it has to go up half a volt before it starts to turn the output on. If we left it at that, we'd have a section cut out of the waveform. So we bias the base up to half a volt positive, and now when the signal starts to go positive, it is already on the edge of conducting, so the output starts immediately. What I just described was crossover distortion and how we cure it.

Now imagine both sides, the plus side base is sitting at +.5, and the minus side base is at -.5. The bases are 1v apart. If I adjusted them so they were 2v apart, 1v on each, then both sides would already be conducting heavily from that whole volt on their base. That is running hot. And if I adjusted down so there was zero volts between them (both sitting at the same voltage), then we are back at that unbiased crossover distorted cold setting. That is why the voltage difference matters.

To add to it, we don;t directly feed the output bases. We use smaller transistors to drive the bases, and sometimes even smaller ones to drive the drivers. To get each added layer of drive, we have to add another eemitter to base junction voltage, so a +.5v on the output, then a +1v on its driver. And if we add a predriver, then yet another half volt, 1.5v on the predriver base. Two 1.5v offsets makes a 3v difference.

My numbers are not exact, just illustrations. That is why your schematic says 2.4v difference between the predrivers instead of my 3v. But this matches your findings. At one end, we have a very low voltage difference, and the amp runs cold, turn it up to a larger difference, and the amp will run hot.

You mentioned a setting that was warm but not hot, but had a little distortion. That setting had 1v differnce. I suspect the distortion you are hearing is crossover distortion, as your amp is still underbiased at that 1v level.

[vertamps]

So the symmetry adjustment should change the 'virtual ground voltage of 46vdc or half of 92vdc? Currently when the amp is cranked to clipping the waveform is clipping on one side a decent amount before the other. Checking the 92vdc and the supposed to be 46vdc should confirm this with the 46v hopefully being off.

I am going to check the Base to Emitter voltages on the Power and Predriver transistors to better understand how much they are turned on at different settings of that bias pot.
* Are Q401 and Q402 acting as signal mirrors? I will be going through with the oscilloscope to understand the signal trail, how it gets to Q405 and Q404.
It does look like that MJ341 is supplying the signal to both sides. I have only gotten into transistor theory within the last 6 months, i have learned alot but this looks more complicated than what i have usually seen and tried to understand.
-Vertamps.

[Enzo]

Scope your power supply to get an idea where zero adn 92v are, then look at your clipping signal. Is it clipping by running into one of those limits? Ther are other things that can cause clipping.


Run a sine wave through the amp, scope it at the output. Now CAREFULLY adjust the bias control, watching the effect on the crossover notch in the waveform.

Q401 402 are the differential pair. q401 takes the input, Q402 sees a sample of the output and serves to correct distortion.

MJ341 is Q403, your voltage amp stage - VAS.


Your output stages sit there, the bias circuit keeps them ready to conduct. The point of that bias circuit is that no mater where the signal leads them, it maintains that same couple of volts space between those opposing bases. If I grab that bias transistor and shove it positive, then the output will follow. Think about it. By doing that I made the plus side conduct more heavily and at the same time turning off the minus side. And vice versa.

And you are right, the MJ341 does feed the signal to both sides. Q401 controls Q403's base, and the collector of q403 is connected to that bias circuit. Q403 then yanks the bias circuit up and down to the music. The output dutifully follows.


Teemu K has written a book for you, it is over at SS Guitar for free download:
Book about solid-state guitar amplifiers

I suggest you read it. It explains SS amplifiers. The analysis I made above and that I use for troubleshooting assistance is not really very official in terms of theory, But it is a seat of the pants sort of approach. His is more formal.


This may or may not help. I learned my electronics on tube circuits back in the 1950s. I knew a little about transistors. I later found myself learning logic. I got pretty darn good at relay logic, then I learned TTL and CMOS circuits. It was only later that I went back and l;earned a lot more about transistors. But I still internalize or intuit transistors as if they were logic elements. Logic works on 1 and zero, also called high and low. So I parse a circuit by thinking if this goes high then that goes low. That works for me to understand a simple common emitter circuit. Like Q403. I don't like to compare tube and transistor circuits, but the basic configuration is not unlike the input stage of a Fender. MY rule of thumb is this: if you pull the base of a transistor towards its collector, the collector is pulled down towards the emitter. A high on the base makes a low on the collector so to speak. So when I look at a string of transistors, I start at one end and think, OK, if I pull this high then that goes low which makes this go high which makes....


I should point out that my bias discussion was kinda generic, your amp is not "complementary" It is "quazi-complementary". Your lower side outputs are NPN, same as the upper side. My discussion covers those predrivers, which are really in charge of the outputs anyway

[vertamps]

The symmetry voltage/virtual ground was off by 0.5 volts. I will skip the symmetry pot installation.
The crossover distortion is visible and adjustable with the bias pot.
* I was however confused as the treble pot at noon was introducing hi order harmonic distortion/clipping that i assumed was crazy crossover distortion the bias was not getting rid of! Turning down the tone controls got me a clean sine-wave on the the output making crossover distortion adjustments visibly accurate.

I will be sharing that PDF book on other forums and taking in as much from this 450 as possible.