Valve operation part 7 (filters)

In the attempt to make some sense out of passive components and their function in valve operation, we’re going to come across a lot of capacitors, so we need to know what they are, what they’re for, and what they do.

The ‘what they are’ bit is straightforward enough. A capacitor is two metal plates that have no electrical connection between them. The stuff that seperates (insulates) them so that they don’t short together, is called a dielectric, and there are a lot of different materials used for this. Mica, polycarbonate, polyester, polystryrene, are a few.

If we put a d.c. voltage across a capacitor nothing (apparently) happens. It behaves a lot like an open switch; but not quite. So what’s the point of the two circuits above, which should according to that, do nothing at all? Well, although a capacitor won’t pass d.c. it does pass a changing voltage electrostatically, so a signal from, say, a guitar pickup or a mic will be passed through the capacitor, but d.c. will not. This is one of the prime functions of a capacitor in all electronic circuitry, but especially in valve circuits.

The schematic above is of two triode valves, the output of one feeding the input of the other. It could be two halves of, say, an ECC83. This circuit arrangement is called a CASCADE circuit. The idea of this arrangement is the basis of all preamp circuits whether valve or solid state of some sort. The first valve (v1) amplifies the input by some factor, lets say 20 times; and then the second valve v2 amplifies that again. So if they both had the same gain of say 20, if for instance we put in a voltage of 1 volt onto v1 valve the output would be twenty times that, so 20 volts. Then that 20volts is applied to the input of v2 which is multiplied by the gain of that valve so the output would be 20×20 volts which is 400 volts, give or take. (Note that it’s not 20+20 which would be 40. It’s 20×20)

This all seems fine, except that the d.c. voltage at the output (anode) of v1, for it to operate might be anywhere between say, 100 and 300 volts, but the voltage at the input of v2 would normally need to be very low, almost zero in fact. This is where we get to the point (at last). We can’t put the output of v1 straight onto the grid (input) of v2, because the 100+ volts would be very likely to destroy the valve. So we make this connection via C2 which will block the d.c. we can’t have and couple the a.c. signal that we need to pass onto the second valve. In this function C2 would be called a coupling capacitor. The reason C1 and C3 are there is to block any d.c. that might be at the input or output.

Going back to our thread and the first diagram. Without going into values of capacitive reactance and such, in general, whereas the resistance of r2 will stay the same whatever signal we put onto it, the situation with C1 is different. The higher the frequency the easier it will pass through the capacitor (that’s any capacitor). The resistor will have some resistance value in ohms. The capacitor has a property called impedance, that we can think of as being the same thing. The resistance of a resistor is represented by a letter R (or r). The impedance of a capacitor is also in ohms but is referred to as Xc. The big difference between R values and Xc values is that, depending on the frequency of the signal applied R will be a constant and Xc will vary.

This property means that the circuit at the top behaves as a HIGH PASS FILTER. That means that it will pass high frequencies and tend to block lower ones. If we double the frequency onto C1, Xc1 (that’s it’s effective impedance) will be halved, and the output increased.

That was hard going. Time for tea. The next post will deal with low pass filters and how these are used in tone/treble/bass controls.

Yes, it’s another Studiomaster 1208!

These things seem to be queing up to be difficult. On this one everything looked fine; but the effects section didn’t work at all. All the display stuff came up as it should, but no effect. Sometimes this is simple. The reverb fader is a simple wet/dry signal mix, twin track fader to deal seperately with left and right effects. It comes out with a couple of screws, then you take off the hard-wired cables (making sure you’ve made a note of where they’re from!!!!!) and replace it. Shouldn’t take much over a half hour.

This wasn’t that. There was a tell-tail symptom extra to all that. At switch on, the red peak led for the reverb section (just over the left side of the graphic and above the green ‘reverb on/off ‘ led) came on and stayed on.  Whenever you get led’s staying on, the first thing to look for is LT (low voltage) power supply rails down, or a ground fault. 

In all the Studiomaster powered desks (not the new ones; they’ve got nothing to do with the original Studiomaster) the DSP internal effects section consists of the digital processor which is soldered to another pcb, this being the power supply for the DSP board. This small assembly (about 4 inches square or so) is screwed to the underside of the graphics pcb and is situated under a steel screening plate (although these tend to go missing), the assembly being connected by two multiway IDC ribbon connectors (or SIL connectors in earlier models). The screening plate comes off with five screws which reveals the DSP assembly.

It’s likely that the actual DSP board itself is ok if it is putting up a readout on the display, but not for certain.

The main thing to look out for on this power supply pcb, (that’s this bigger one with discrete components on it. The other has flat pack chips on it.)  is the power transistor that is mounted straight onto the board, which is fairly obvious being a big-ish (TO 220 case) component. It will also show signs of heat on the pcb under it. This is a complicated power supply and has several digital and analogue supplies and grounds, but there are two dead giveaways on it. If you switch the desk on for a few minutes, and then switch off and put your finger on that transistor (its a TIP 30c ) it should run warm/hot. If not, the power supply won’t function at all. You might change the TIP30c and also a resistor which is stood off the board by a centimetre or so (it’s a fusible 18 ohm) but there is a lot to go wrong with this power supply. Still, you might get lucky.

Studiomaster Horizon 1208

We’re back on Studiomaster powered mixers. Here’s one with the same fault the last one we looked at. i.e. it gives a thermal overload warning for one channel (sometimes both channels) on the readout after switch on. The one we looked at last turned out to be the main processor chip faulty (an 80C) and these are eprom devices which have been programmed for these particular models. The chances of a replacement are zilch, I would guess.

The problem is that the dsp on the desk becomes unusable ( the fault warning continuously strobes and you can’t read the dsp settings, so you don’t know what it’s doing) and it will sometimes shut down the output  altogether. So the problem with the microprocessor chip (the 80C) isn’t repairable, sadly. Although the fault was the same, the cause of it was different, and in this case, fixable. So good news!

On the power supply pcb (that’s the one fixed to the bottom of the chassis in the middle) there are two yellow led’s and these are marked ‘Thermal fault’. In normal operation, they should remain unlit. They should also flick on very briefly at switch on. There is a chip situated quite close to them which is an LM393 (this a comparator, but is used as a dual drive for the led’s, as it has a higher current rating than the 4558′s or TL072′s which are just about everywhere else in the unit. If the two thermal led’s either don’t go out, or are slow to do that, the LM393 is likely to be faulty, and will give the thermal error readout.

Valve Operation Part 6

We were leading up to valve characterists, graphs etc., in the last post, but I realised that more needs to be discussed about the components associated with the valve, before bringing in that stuff. So it’s on to passive components, bearing in mind that ‘variety is the spice of life’. It could also be said that nothing is worth having that didn’t cost a lot of effort and commitment. Not sure where that might leave us.

In the above diagram there are two symbols that you may find a lot of,  in schematic diagrams. They mean the same thing i.e. a resistor.

Below those are three symbols for capacitors, two lower ones being alternative representations of a polarised capacitor, and under those is the symbol for an inductor, which is basically, just a coil of wire, on a former of some sort.

It’s these passives that make the active components work. The difference between an active and a passive component is that a passive will always give less out than you put in, and this is called ATTENUATION; and an active component (could be a valve, a transistor, a chip) has the property of GAIN, which means that you get more out than you put in, potentially. Back to passives.

   

 Circuit A and Circuit B are exactly the same, functionally, but look different in the way they’re laid out. Circuit C has a variable arrangement of r1-r2, whereas the other two circuits have fixed resistors, but is otherwise the same. All these arrangements (also called NETWORKS) are examples of POTENTIAL DIVIDER networks, and are what amounts to a VOLUME CONTROL in an amplifier.  

The POTENTIAL DIVIDER is so called, because it divides whatever comes in onto the INPUT if we take the OUTPUT from the position marked. So far so good. Whatever voltage we might put in across the INPUT and GROUND will see the total of the two resistances r1 +r2, and that is the same thing as R. Before we get totally ravelled up in letters, we’ll call r1 = 100 and r2 =100. So R =200. Seems a lot simpler like that, but is the same thing really. Anyway, if we put a 1 volt input into it, at the OUTPUT we’ll see a half a volt; 0.5 volts if you like. Why is that then?

Because the POTENTIAL DIVIDER divides the INPUT according to the ratio of r1 and r2 and puts it out to the OUTPUT.

The above diagram tries to put that last sentence into a graphic form. The output voltage is that which we see between the OUTPUT and GROUND. Half of the input has been dropped across r1 and the rest of it appears  across r2, and that is our output. It only works like this if r1 and r2 are the same. If r2 gets bigger then so does the voltage across it, which is the output.

Going back to circuit B.                                              OUTPUT=INPUT x r2/R.             What this means is that the OUTPUT is some fraction of the INPUT determined by the ratio of r1 to r2.

Let’s say r1 is 100 and r2 is 200. Then what we’d expect to get out would be the fraction r2/R of our INPUT which we said was 1volt. R is the sum of both resistors, so R=100+200=300. The fraction of the INPUT we’d expect to get at the OUTPUT would be the 200 of r2 divided by the 300 of R; so 200/300; or 2/3.  After all that then, the OUTPUT would be 2/3 of the INPUT, 2/3 of a volt with the INPUT at 1volt.

This might make it a bit clearer. The diagram above represents the kind of thing you  might find on your amp as a volume, or gain, or master control; and they would usually be a rotary device, even though it looks like an in line thing. It could also be a fader on a mixer; although it would look a lot different it is the same thing. The arrow is the slider that is pushed or rotated and can therefore vary the ratio of r1 and r2 infinitely. If the 0utput slider is set to the top, the INPUT and the OUTPUT would be the same. That’s fairly obvious; and if at the bottom you would get no OUTPUT at all. Between those extremes the fader can give you any fraction of the INPUT at the OUTPUT.

We’ll maybe do something on filters (tone controls) in the next post. There are a lot of similarities between the potential divider networks and filter networks. Time for tea; I’ve earned this one.

Dynacord Powermate 600

It’s been hard work finding any information, schematics etc on this beast. The site I use for schematic sourcing are Free Information Society, who have heaps of schematics for older gear, and a lot of American stuff. Brilliant site. eserviceinfo have a lot on there site which is also excellent. The Dan Rudin Vault deserves a good looking at; also wonderful. They’re all free, and printable schematics. Highly recommended, all of them. Schematic Heaven is another one, but tends to be biased towards British stuff, Marshall, Vox, etc.

Anyway, all that was a waste of space on this occasion; until a new one came up out of the blue on eserviceinfo. I’d been looking for this for months. On to the job.

It didn’t take much to work out that the cause of the fault (which was both power amps shorted) was that some kind person had shoved a screwdriver through the grill at the back, which had shorted one of the voltage rails to ground.  A bit of sabotage or ill-advised ‘help’, possibly? I’ve seen this sort of thing before, and apart from blowing d.c. fuses to bits, the output devices can get away with it; but not often.

There are a number of upsetting design issues with this amp, that aught to be better understood by those who might be tempted to buy them. This is the second one I’ve seen, and the first big problem waiting to happen lies in the way the draft cooling is routed. It seems to be routed by accident. If you look through the grill at the back, you can see the heat sink (so not difficult to short that with a screwdriver; but I’ve been there.)  and over that is a box-kind of arrangement with the fan mounted in it. There are two problems with this. One is that the draft has to turn a corner to blow over the heatsinks. This forces dust to collect on the output devices. The two I’ve seen both a had a carpet of dust a half inch thick over the heatsinks. That is a desperately bad idea. It seems to be designed like this because it’s dead easy to put together (so, cheap), but the fan is not going to cool very much when a heat sink is in this sort of condition; and if the amp is used in any normal environment, it is definitely going to finish up like this. But it gets worse.

What happens to all this dust that gets past the heat sinks? As there is completely no ducting system in the amp, it gets blown over all the rest of the components. So far as the semiconductors and circuitry are concerned, these amps are very similar to the old Studiomaster Powerhouse 300-300 amps. But from every other angle (apart from the fact that the casework looks a bit prettier) the Powermate 600 is light years behind the physical design of the old workhorse of the Powerhouse range, which is has now been around for 20+ years, I suppose. I can’t imagine many Powermates being around in 2030; but maybe that’s the idea.

Trace Elliot Super Tramp stereo Twin Quad Chorus

I think it’s a sign of old age. This fault was so obscure as to have been buried under the Sphinx when it was built and came out when it saw I wasn’t very well. I used to relish those challenges but they seem more and more to be a pain in the arse. Anyway, onward and upwards……..

The customer had bought it recently (ebay) and it sounded great till it had been running for twenty minutes or so, when it set up a monumental hissing. (They used to keep snakes in the Sphinx, y’know. Just maybe???) No, that way lies insanity. The hiss was in no way modified by any settings on the front panel of the amp. This, we think, is obvious. Especially when the scope showed up some sort of negative breakdown of random frequency. It has to be some component failure on the negative side of the power amp. This is a stereo amp each amp driving one of the two 12″ speakers, the right hand side has the chorus modulation and the left, the straight signal. This is very like the Roland Jazz Chorus amps of the 70′s, which were a great range of quality bipolar amps. I digress.

The fault only occurred  on the right hand side, the straight signal side.

Another problem with this amp was that there had only been something like six ever made, in the Trace Elliot custom shop; so no schematics anywhere! 

Well, it wasn’t the power amp at fault, but somebody had thought that before me, because the whole of both power amps had been rebuilt. I bet he was pig sick when all that work made completely no difference. The fault was in the pcb that generated the chorus effect, and provided the preamp-ed output to the power amp pcb. To cut a long and arduous story a lot shorter, there were two p-channel jfet transistors on that pcb, and one of them was breaking down. They were 2SJ177 and as I didn’t have a schematic I still don’t know what they were there for. They’re mostly associated with switching in these applications. How that might help anybody, I’ve no idea; but you never know.

Valve operation part 5

From the stuff put out on the last post, it would seem that the valve operates like a glorified switch. So it can; but it can also operate as a lot of other things also. So the grid needs a negative voltage on it for it to regulate the current through the valve. We should say about here that the direction of the arrows from the cathode to the anode denote ELECTRON FLOW. This is in the opposite direction to conventional current flow (that stuff flowing from positive to negative) and this is because electrons are negatively charged, and so do the opposite.

The odd thing about saying that the grid needs a negative voltage on it, is that there are rarely any negative voltages around in a valve amp; not in this bit anyway. The crunch phrase is ‘Negative with respect to the cathode’ and that means that raising the voltage on the cathode is the same thing so far as the valve is concerned, as dropping the voltage on the grid.  This is about where that piece of mumbo-jumbo BIASING comes in. This is a term that is used (almost entirely in the areas where people are trying to take money out of your pocket and put it in theirs) as an issue of quality. Class A, for instance, is bound to be better than Class B, or Class AB; isn’t it? Well, it might be, but that is completely not what biasing is about. And whatever they tell you Class A is very rare beast. Why? Because to come up with say, a 15 watt class A amp, your transformers and power supplies would certainly need to be as big as a hundred watt plus, Class B amp.  More of that another time.

The way the biasing works in our little preamp, is simple, but, like most simple things has to be exactly right. It’s analogue, and whatever you might think about the analogue /digital  thing, the simple fact is that every digital piece of equipment is exactly the same as every other piece of similar digital equipment, and any analogue anything at all, hardly ever is. Your own preferences are what they are, but I like things that are always different from day to day, hour to hour. Your Blackberry (or Banana or whatever) is EXACTLY THE SAME AS EVERYBODY ELSE’S. My little Fender Tremolux amp isn’t even the same as any other Tremolux. I like that. But then I also like records. Back to baising.

The little cathode resistor which you can see more clearly in the drawing in the last post, raises the voltage of the cathode, depending on a number of issues, but the main one being its value. If it was 1000 ohms, the voltage on the cathode would be slightly more than if it were, say 680 ohms. This sets THE D.C. BIAS. which sets the  standing current (or flow of electrons) through the valve. The standing current being a set current flow very similar to a battery supplying current to a bulb.

The next bit we need to have a run at, because it involves mutual conductance graphs and the like. And if you’ve been brilliant enought to stick with it this far, I really don’t want to stuff it up for you. Time for tea.

Valve Operation Part 4

This  is close to the sort of thing you might find in the preamp of a valve amp schematic. On the left are passive component symbols, and on the right is what could be half of an ECC83 valve with those passives connected in such a way for it to function as a preamplifier.

At the top is a wire with HT+ marked on it. This stands for High Tension positive, and could be anywhere between say 100volts and maybe 300 volts. Not a good idea to put your finger on it. This comes from the d.c. power supply. At the bottom is where the ground terminal is connected. It is the voltage difference (or potential difference if you like) between these two terminals, along with the heated cathode that makes the valve do anything.

The heater is a completely seperate element, with a seperate winding on the mains transformer. It’s often fed as raw a.c straight from the transformer. In an ECC83 this voltage is 6.3 volts rms, and carries a hefty current of around 0.3 amps. The ‘E’ in ‘ECC83′ stipulates the 6.3 volt heater supply. So, apart from looking nice and making the valve work, the heater voltage has no other function in electronic terms.

In general terms, the values of the various passives are, give or take, 100 Kohms (100,000 ohms) for the anode resistor; 1 Mohm (1000,000 ohms) for the grid leak resistor; around 1 Kohm for the cathode bias resistor. The Bypass capacitor might be 47uF. This is not particularly important at this stage, but it just gives some idea.

If we put some volts onto the heater, apart from lighting  up and looking pretty, it heats up the cathode. What happens to it (the cathode) is the stuff of circuit theorems and some fairly complex maths; but the stock explanation is that it gives off electrons. As we live in these paultry three dimensions and have no idea of what happens outside that, that’s probably the best we can manage.

Anyway, sticking with that idea as we can’t really extend any further, the rest of the explanation goes something like this.

The electrons thrown off the cathode by the heat energy supplied to them, form a cloud around the cathode surface. This is called the space charge and is explained by the thrown ball effect. If you throw a ball in the air it gets to a point where it runs out of energy and gravity pulls it back. Do that a lot of times (Like, billions per second) and you get a cloud of balls. And that could well be a load of balls, but how should I know? It’s not gravity that pulls your balls back (?) it’s the positively charged cathode because the negatively charged electrons are attracted back to it.

So, we have a cloud of negatively charged electrons floating around the cathode. So what? Well, we now have to remember that the anode has a very heavy positive charge on it (100 volts plus, say) and this is a powerful force of attraction for the negative space charge. So it is pulled up towards the anode. Now then, electron flow means current flow, so the valve is now a conductor. If we swop the HT+ to be negative, then the anode will repel the negative electrons and nothijg will flow. So we’ve got a diode rectifier.

What does the grid do, then? If it isn’t connected to anything, nothing at all. But if it has a voltage that is negative when compared to the cathode, it will repel the negative electrons that are busy flowing to the anode. And the valve will be cut off. So it operates exactly like a water valve (tap, fawcett, whatever). The voltage needed to cut the valve off is very little and is dependent on the valve itself.

So a little voltage to the grid can CONTROL a substantial current through the valve. That’s why it is called a CONTROL GRID.

I think there’s a lot of stuff there, s0 rather than hand anybody a headache (especially me) I’ll get a cuppa, and get some more in the next post.

Valve Operation Part 3

We finished part 2 with me having a cup of tea. But that’s how they all finish. Anyway most of the actual post tried to explain the way a diode (like a GZ 34 or whatever) works. In Part 1 there was a picture of an ECC83 valve in a fairly sad state; but the bits inside had been marked as cathode, anode, grid, etc.

The intention here is to sort out the difference between the GZ34 and the ECC83. Apart from being called something completely different of course. We’ll start by getting one thing out of the way. The ECC 83 is a double triode valve and as each half of it is identical we’ll just explain the one half.

The BIG difference with the ECC83 triode as compared with any diode, is that it has a control grid, and it’s this that allows this valve to convert your microscopic output from a guitar pickup (a few hundred millivolts rms, usually) to something usable by the rest of the amp which is probably 10′s of volts. So what comes out can easily be a hundred times what went in. So we get something for nothing? If you believe that, you need to look a lot more at the small print. Madison Avenue is really only interested in what you have in your pocket, and not at all in doing you a good turn.

No, surprise, surprise, we don’t get something for nothing. This increase in voltage output comes from the power supply. Having established that, how does it do it?

The first thing to get to grips with is the way a schematic diagram actually works. This kind of diagram is totally not bothered about where things physically are. So a schematic diagram of the North Pole and the South Pole might put them next to each other, because they’re both cold. Not much good as a map, but at least you know they’re both cold.

The drawing above puts the anode at the top, the control grid in the middle, and the cathode and heater at the bottom. All it tells you is that the grid is between the anode and cathode, and that is the way it works. Nothing about how far apart they are nor even which way up it is. Below is a schematic diagram of the sort you would actually find in, say, a Marshall amp.

In this, we are shown two halves of an ECC83. Here they’re shown more or less together, but the two halves of the valve might well be quite seperate on a drawing. Where it is is more to do with where it is in the circuit, than the fact that both halves are in the same glass bottle. 

The a1,a2 stuff shows the way these connections are put onto the valve base, via the pins. There have been many different valve base patterns but this valve fits into a B9a base, which is a nine pin base of a particular size and is shown below, but not to scale. The function of the valve base is to connect the valve to its various components.

 The  valve base is always shown as you would look at it from the wiring side. The mysterious connection marked ‘ct’ is a centre tap connection to the heater. More about that in due course. Time for tea.

Carlsbro Cobra 90; 4 channel pa amp

This amp didn’t work. More to the point, it looked like there should be something coming out of it, but there wasn’t.

There were a huge number of these sold, probably in the ’80′s, and there are still a fair number around. The Marlin series were very similar but generally a higher power output. Whereas the Cobra put out around 90 watts rms, the Marlins were between 150 and 300 watts, which is a good whack for a non-forcedraft cooled amp. Just a chunky heatsink on the back. They were mono, bipolar amps with two speaker outs and designed to run down to 4 ohms. Much of the fabrication was done by hand with good quality pcb’s and components. Most of these components are still not difficult to get hold of, so they’re generally still repairable.

This one had been in a shed for the last ten years, or it looked like it, anyway. Two things to know about these amps and also the Marlins, is that FX send /returns and also the headphone out (where it’s got one) are in the very latter part of the preamp circuit, and if the switching gets clagged up behind the jacks, you’ll get nothing out of it. The remedy is simple and you don’t even have to open the amp up. Spray all the jack inputs and outputs  (with the amp unplugged!!!!!!!!!) get a jack plug when you’ve done that, and whack  it in and out of the sockets. This will clean up the switching contacts which is a big cause of failure on a lot of amps of this period.

If you want to do a real job on it (with the amp unplugged!!!!!!!!) take one of the end cheeks off the case and slide the top panel off. From there you can get to all the pots, and treat them to a spray out. They’ll probably need it.

That was all that was wrong with this one. It cost him £25 and he was as happy as Larry. If only they were all like that.