Trying to fix loss of volume in a 1978 Randall RG300 Head (SS)

[Enzo]

1N914. Or use 1N4148. Mouser has both for as little as 2 cents each.

If you think D3,4 are bad, just unsolder one end of each and lift it from the hole.

Earlier I gave a description of the signal path up to C15. You now mention C8 and the master volume, though they are a ways apart. Did you trace a signal step by step? And if so, where did it lose signal?

What meter are you using? DOes it have a diode test function? That is the thing to use, an ohm meter doesn;t give reliable diode and transistor tests.

And beware good/bad cap testers, they will often show you when caps are bad, but they also often do not show reliable indications of good.

[ParthaD]

Diodes were bad but some more fixing needed..

Well I disconnected the diode D3 that tested bad but had to do the other one D4 too since it turned out to be bad too, it was testing good in-circuit and with the meter voltage but I guess not at operating voltages. How do I know that they were bad.. because I got back the sound that I was getting prior to that when it abruptly disappeared... but not the full volume yet. Have to do some more signal tracing but getting close....

thanks for the diode part numbers Enzo. You had asked about my DMM.. I have an old Radio Shack Micronta that does not have a diode testing function so all I was checking was continuity.

[J M Fahey]

Dear Partha.
You don't follow simple instructions meant to help you.
1) You did *not* unsolder the wires connected to the top lug of the master pot. That was asked to leave the power amp free to be used as an audio tracer.
Since you did not, the clipping diodes D3/D4 were always there, clipping every signal you throw at them. Also there was an audio path to the same point, clouding where really did any sound come from, and if any of them had ben shorted , they would have killed al signal rendering the tracing method useless. Well, *now* they are shorted, most probably because of something you misconnected there, they were alive yesterday.
2) You are messing with D5/D6 , the power amp bias diodes. One millisecond after you open that bias string the power transistors and probably the drivers too will go to the SS Heaven . That the fuse will burst some tens of milliseconds later will not help you. That this particular amp uses a circuit breaker, which is even slower will help even less, although it will probably avoid your house catching fire. For that alone, circuit breakers should be mandatory, or at least, internal hidden, unaccessible-to-user, mains fuses, rated 150% the "official" needed value.
3) The reviews posted refer to a different Randall model. This one is, as I stated before, a sort of SS Twin , two non switching channels, reverb and Tremolo; but the ones reviewed are switching channel Metal machines.
The Doctor can't help you if you show up with somebody else's X Rays or blood and urine tests.
4) The Tremolo section is unrelated to any diodes.
5) Pull D3/D4 and follow exactly Enzo's directions; namely, trace your signal step by step and not at random.
Good luck.

[ParthaD]

Juan thanks for that note about D5 and D6. There was never any intention to mess with D5 and D6 anyway. And after D3 and D4 were removed I have loud sound as before and can proceed with the signal tracing which is what I was doing when D3/D4 decided to crap out on me. I still have the same sound from both channels as before right after I had connected the input directly to the master pot. Will update more later...

[ParthaD]

Found the problem (warning - lengthy post)!

As I've posted earlier, signal existed from both channels up to C8 where the signals from the amp's 2 channels mix. To those who are not aware, this is a unique amplifier in that both channels can be actually mixed (assuming the foot switch is not in use) i.e. you can mix the tremolo and treble boosted channel with the clean channel simply by varying the volumes on each channel and disconnecting the footswitch. But I digress...

I'd like to add something else here that is relevant to what I say later but it would of more interest to some of the readers who have some level of electronics understanding but are not as experts as some others. One of the unanswered questions I had posed on this thread was about AC/DC in a standard amplifier circuit - where was it AC and where was it DC? The guitar signal as it leaves the guitar pickups is a weak AC signal while the final output to the speaker is a strong AC signal. The DC portion is the mains power after passing thru the power transformer, rectifier, and filter caps. That is what powers the electronics which usually manipulates the original weak AC signal from the guitar. Next time you look at a guitar amp circuit (tube or SS) it will be obvious that as soon as the signal enters the amplifier it will be amplified somewhat in the preamp stage and taken thru the tone stack, i.e. volume, treble, mid, base. This is usually done by passing it through a set of resistors and caps in parallel which affects different FREQUENCIES differently, hence bass, mid, treble all have a different effect on the sound. The AC signal gets weakened after going through all of this and therefore has to be amplified again and again whenever such manipulations are taking place. Therefore, there are MANY such places where BOTH AC and DC exist *simultaneously* until they get "separated". That was not clear to me conceptually at all originally!!! Funnily enough it seems that when the AC is let through a cap and DC is blocked it is called "coupling" cap and the other way around it is a decoupling cap (usually on the power supply side).

That's a critical thing to remember for the following reason. In one of my earlier pics I had shown how I had simply connected two points - from just prior to the R9 resistor to the master volume pot and got a huge boost in volume compared to the earlier "broken" volume. I was actually VERY LUCKY that I did that at just THOSE points. Lucky because there wasn't a large DC voltage at that point which could have messed up some of the downstream components. At various stages in the circuit there are "coupling" capacitors which take away all the DC leaving only the AC signal passing through.

I realized this issue sometime during my analysis of the circuit and just some general reading. So instead of using this alligator clip jumpering method I found that I could do just as well using my DMM (digital multimeter) alone and with practically no risk (except of course inadvertently shorting something with the probe itself)! Connect the DMM black probe to chassis ground and the other to the point in the circuit you want to measure and change the DMM setting to AC or DC depending on whether you want to measure the audio signal or the DC volts. Basically the schematic for this amp shows both the expected signal AC voltages as well as the expected DC voltages at various points in the circuit - isn't that nice, I've not seen that on other schematics in the limited number that I've seen except for some text limited only to a test point or 2.

A DMM shows the RMS values of AC current but if you put in a continuously changing signal like a guitar loop recorded and played back through a looping machine, you can actually see the fluctuating AC voltages that change in time to the notes being played in the looper. It is also SO much safer since there is no danger of messing up the components! The looper in my case is connected to the channel 1 input jack just like a guitar would, not inserted using a DIY probe although that would be useful sometimes.

So that is what I did and found the problem almost right away. Right at C8 the guitar AC signal strength should be about 20mV (based on a 5mV input signal). The actual value is not important but the MAGNIFICATION in the next step IS!! This 20 mV signal goes into Q3 the TIS58 FET which magnifies the signal to about 450 mV, i.e. more than a 20x increase, as it goes into the next transistor Q4 which is an MPSA13. According to the schematic, Q3 should also have a DC voltage of -15 and 15V on the gate and drain of the FET.

There was NOTHING, i.e. 0 volts DC on the gate and drain - there was 23V sitting at the source but that was just coming from MPSA13 (which is connected to it at that point which has +/- 23 V sitting on it!!! There was also a signal right before Q3 at one end of R10 which showed that the signal was crossing C8 successfully.

But where to find a new TIS58, its an obsolete part, did some googling and found that the Fairchild 2N5457 is a substitute for the TIS58 and ordered that from Mouser. There is a twist here too... dug up the original datasheet on TIS58 and the gate on that is the leg in the middle while the source and drain are on each side, while on the 2N5457 (from its datasheet) the gate and drain are on each side while the *source* is in the middle. Hence while resoldering the new part I will have to take that into account.

So next step will be to replace Q3, and the diodes D3 and D4 and also the blown diode in the D5/D6 pair. Will report back after all that is done.

Thanks to all those who helped! Of course if anybody feels I have drawn an incorrect conclusion anywhere I am always willing to learn.. that's what this was all about, not just fixing THIS amp but getting at least a little understanding of the underlying electronics.

By the way I found the original designer of this circuit, Gary Sunda, at:

Gary Sunda

[ParthaD]

IT WORKED!!!

Just wanted to let you know that after soldering in the new transistor (2N5457) and replacing D3/4 and D5, everything works GREAT! Massive volume - its 300 watts after all!!!

Oh yeah the parts cost $0.13 for the transistor and $0.03 each for 3 diodes, for a total of 22 cents! Cannot believe that!!! But of course 51 posts too!!!

Cannot believe that its only been a month since I actually "read" my first schematic! Thanks again to all those who offered suggestions and especially to Juan and Enzo!

Now on to the next project... have got my hands on a 1972 Gibson G20 (ss) that has the most amazing tone but has a slight hissing issue. Will start a new thread for that shortly.

[J M Fahey]

Hi. Congratulations on your successful troubleshooting. You did very well in digging up the old obsolete datasheets and comparing them carefully to the modern ones, that saved you from many bitter hours.

[ParthaD]

Drain and Source/Drain on NPN JFET interchangeable?? (Not in my case!!!)

Thanks Juan!

Just a last item to note that might be useful in some future lookup.... while researching the replacement for the TIS58 NPN JFET for this amp, the modern datasheets had diagrams that were quite unambiguous about the pinouts for Gate, Drain, Source, however on the original one I was not sure if the diagram was a view from the bottom (pin side) or top. I am attaching the datasheets here. The literature seemed to indicate that the drain and source were interchangeable so I wasn't too worried. I also found an old MEF thread about this where the consensus was the mostly the same with a note from Enzo that this is not always the case.

MOSFET source/drain ... is there a difference?

As it turns out, when I first connected the 2N5457 I had it connected wrong, the symptom was a loud motorboating which went away when I re-connected it the other way around. On a general diagram the drain is usually shown at the top relative to the gate and the source is at the bottom. I somehow misread the schematic and thought that the *source* of the TIS58 was connected to the base of the MPA13 Darlington following the TIS58. It was actually the *drain* of the TIS58 that was connected to the base of MPA13.

So at least for this particular circuit the drain and the source of this transistor are not interchangeable.

[J M Fahey]

Hi Partha D.
Speaking of FETs (not MosFets), my experience agrees with Enzo's.
Most datasheets spell black on white: "drain and source are interchangeable", yet that's not what I test.
I guess they're interchangeable *when used as audio switches* and even, "maybe", when used as preamps with low +B supplies, say 9V in a pedal.
I'm sure of the first use, not much about the second, because I always "waste" those extra 5 minutes to get pinouts right.
I use Fets a lot, in "real" circuits, say 25 to 60V powered, not those contraptions favored by ROG and GGG , and there you see the difference.
I guess there must be some kind of avalanche effect involved, because "reverse-connected" Fets hiss and popcorn a lot.
Someone's else experience may differ, but that's what I've noticed.

[Enzo]

Well, MOSFETs are not JFETS and vice versa.

I can;t speak to situations where it matters, but for the most part in guitar amps, the JFETs can go in either way. Source to drain is more or less a resistor whose value is controlled by the gate.

WHat you CANNOT assume is that the legs are in the same order type to type. All TO220 transistors for example are BCE across the front, but you have to check the JFETs when changing type numbers. The gate could be on either end or in the middle.

Note on the TIS data sheet that the gate is in the center, and note on the 2N5484 sheet that the gate is on the end. One is not a drop-in replacement for the other. To sub them you will need to turn the part sideways and bend the legs. SO if you installed a 2N5484 and then rotated it 180 degres, you were not swapping source and drain, you were swapping the gate.


And I would also point out that while the part may have clearly identified source and drain leads, that is not the same thing as having source and drain interchangable in a circuit.

Look on your Fairchild 2N5484 datasheet, right under the SOT32 image. Says right on there source and drain are interchangable.

[ParthaD]

Slight clarification...

Enzo, I had caught the difference in pinouts in both the original TIS58 and the 2N5484 with that Gate on one end instead of middle on the 2N5484 issue right away. In fact I wrote about that 2 or 3 posts ago. So I was already bending the legs as you indicated. Therefore when I said I had to switch the legs I was only referring to the last 2 legs not the entire transistor. This was a great lesson to learn in practice since even though the datasheet said that the D and S were interchangeable they really were NOT for this amp!

[jtn0xvirus]

TIS 98 obsolete need repacement for RG100ES !!

Guys I am subscribing to this thread... need a help guys. I am making a Randall RG100es/ht amp for personal use. But one of the transistor Q8 and Q9 is TIS98 http://dl.dropbox.com/u/45053134/ran...-schematic.pdf , which i found obsolete . Any body knows how to fix this problem!! Please help me out guys.... please check the schematics

[J M Fahey]

You can use *any* general purpose NPN 40 or 45V 200mA or more transistor, just watch the pinout.
Here in Argentina we generally use European components, so I use there a BC547 , (which *is* available in the US, anyway) but I'm certain some US based colleague will suggest a more popular one.
Randall used what was widely available way back then, perfectly good substitutes are available now.
EDIT: since the part is dirt cheap, and you will pay postage anyway, order 10 or 20 of them, plus a similar quantity of the equivalent PNP's, just to have them in your parts box.
You can use them freely in *many* projects (preamps, fuzz, etc.). that´s what "general purpose" means

[jtn0xvirus]

TIS 98 obsolete need repacement for RG100ES !!

Using *any* general purpose NPN 40 or 45V 200mA or more transistor Does it effect the sound quality or its effects?? I dont want any variations in the sound..

[J M Fahey]

We´re talking SS here, not tubes.
Sound comes from the circuit design.
Contrary to tubes, where even 12A*7 "sisters" are *very different among themselves.
Even same code ones (say, 12AX7) vary a lot between them, already just by varying brand, and even within it, in different batches.
SS devices are *much* more consistent, and are mostly used in high negative feedback circuits which erase possible differences even more.
So, in a nutshell, sound will *not* change.