GK 1001RB no output

[lowell]

Hello,
I posted a while back bout this amp. I had decided to pass on the repair but now find myself digging through it trying to find the problem. I've actually gotten it to make some sound which is a start... however it quicky shuts down and a resistor goes up in smoke. The resistor is R27. I tested the speaker and it works. All of the output transistors measure differently in the circuit when I do the transistor diode test... I'm assuming that means that they need to be replaced, unless maybe they need to be tested out of circuit? Any direction on the repair would be oh so appreciated.

[lowell]

anyone? just checking

[Enzo]

I can't read the numbers well, where is R27.

Yes, if in doubt, pull the parts and test them.

[lowell]

R27 looks to be within the bias supply area.. not sure bout that though. when opening the pdf in adobe reader, click edit-line weights. that clears things up a bit. what a crappy document eh? really hard to read some of the numbers. Try the line weights thing and see if you might be able to decipher it... thanks!

[teemuk]

Are we talking about GK 1001RB mk II or just GK 1001RB? Two different amps, the schematic is for mk II. In it the resistor R27 is indeed the current compensation resistor of the bias circuit.

[lowell]

yes this is for MKII. What is the current compensation for the bias? Could you please explain? I'd appreciate it very much. Also, do you know why it might be frying?

[teemuk]

An ideally working VBE multiplier would behave like a constant voltage source. That is, the voltage drop over the circuit would remain “constant”, basically being affected only by the temperature’s effect to voltage drop of the base-emitter junction and so on. In real life the voltage over the VBE multiplier circuit is – however - also a function of the circuit’s input current – thus the bias varies according to input signal. (The bias voltage increases along with the input current). This is an unnecessary effect. The bias circuit is often bypassed with a capacitor to minimize this sort of variation and to make the circuit operate at DC level only. The current compensation resistor can also reduce the variation since it decreases VCE when the input current increases. This effectively creates more “steady” bias voltage regardless of the input current.

This is selfish promoting but you can read more about “crossover distortion compensation” from my book, which you can download (for free) from: http://www.thatraymond.com/downloads...ttala_v1.0.pdf. Beware; it’s a rather large file. Section 3.7 has graphs that depict the VCE versus input current behaviour of typical VBE multiplier configurations. Douglas Self also discusses VBE multiplier-based biasing circuits very comprehensively in his excellent book. In my opinion, all those tinkering with SS amps should read the latter.

----

What comes to failure of R27 I can only guess: The resistance of it is very small (182 ohms) so it doesn’t take a huge voltage drop over it to create a current flow that is high enough to cause enough power loss to burn the resistor. One of the components within the bias circuit could have gone bad – or one of the voltage amplifier transistors – or something else. So... I don’t know why the resistor is frying (if I knew I would have said it already) but hopefully you can figure it out. There are two test points indicating about 10V voltage potential between the collectors of the two voltage amplifier transistors. Do you get this reading and are the test point readings (+ and - 4.9V) OK as well? (I suspect that you will find something fishy here.) If the readings are very different it could lead you to the right direction.

I have the service manual for 1001RB II (with a better schematic) but it is so large that I won’t (or can’t) upload it here. If you need it just drop me an email to teemu[removethis]kyttala@gmail.com and I’ll send it to you.

[lowell]

ok cool! Yeah i'm buyin' that book. Any reading I can do on this I'm all about it. I'll check those voltages and report back. I sent you an email for that schem too.

thanks!

[Enzo]

If that resistor is burning up, there is only one path for the current to come through. That is through Q12,13, so they are instantly suspect. And to get current enough to those transistors, I then suspect the drivers and even the outputs. Did you ever pull the outputs and do real tests on them?

Check the output transistors and ALL the ballast resistors around them.

[lowell]

Enzo,
I'm assuming that these transistors will test as normal transistors do and that the diode test b/e and b/c junction is sufficient. Correct? I'm in the process.

[lowell]

Teemuk, I could not perform those measurements cause R27 was frying immediately.

ok... the output transistors are testing bad. Most diode tests I performed are short circuit. 4 or so of the ballast resistors are measuring open or higher than they should. Now, I know I should order all the output transistors, ballast resistors AND driver transistors. Is there something else I should be replacing in this situation? I already replaced Q12,Q13. Are there any other xsistors I should replace? Any diodes, zeners, caps?

Finally, can someone please shed some light on removing these pesky things from the double sided PCB? I tried one and pulled some traces up w/ it... oops. I am using soldering pump but it's not working too well.

[Enzo]

You already replaced Q12,13, but that resistor still smokes, so are the new Qs damaged now? Remove all those dead outputs and ballasts. Before installing anything else, blueprint the output stage. Go along with a meter and measure evey damn resistor in there, check every diode, do drops on every little transistor.

Make a copy of the schematic to scribble on - do not write on your original - and color highlight or circle each bad part you find. This way, the damage current path will be colored on - or circled - on the copy. This might suggest parts that escaped your attention. For example if there is a line of transistors and resistors from top to bottom that failed, you can infer that the failure current took that path. And if somewhere along the line of burnt out stiff, there is a lonesome "good" part remaining, but that current clearly had to flow through it - replace it. DOn't be cheap for a resistor or a transistor. Pay good attention to those resistors around Q12,13. Make sure that bias trim pot is OK, and not part open or burnt.

[lowell]

Ok,
replaced all output transistors, ballast resistors, both driver transistors, R61 b/t the drivers was bad so I replaced that, tested bias transistors and they seem fine since I last replaced them. I powered up w/ light bulb and it glows fairly bright. The original problem of R27 frying was not there but not sure if the lightbulb limiter kept things in check. I then did resistance check from collectors of output transistors to ground - seem ok. I checked EVERY single component between the - outputs and the current sources on the tail of the input. They are all fine. Pulled U4 (voltage source) and still 8K to ground. Ok now that I realize I did most tests on components by pulling them out and measuring collector resistance to ground on outputs. Now a realize that it matters which way the meter leads are because of the protection diodes on the + and - supply rails... will need to go through and check components again. Disconnected power transformer and After powering down from the first attempt to fire it up the + outputs collectors held their voltage for over an hour while the - outputs drained to near 0 volts within that hour. I'm lost here, feel as though I've checked EVERYTHING! Ok wait, I just realized something, I had the meter leads reversed... so the collectors to ground seem ok now however the thing is drawing too much current. With limiter there is + and - 30 volts on collectors and + and - 12 volts on inner sets. I measured the bias voltage, the schem says to get 10mv across J5 but I cannot go any lower than 50mv. The 2 test points of + and - 4.9V are actually measuring good.

I just plugged the amp direct to the mains and nothing burned or smoked, the voltages were same as above, however the power protection light (red) was on and would blink green for a split second and kept alternating mostly red w/ little green blips. That's where I'm at if anyone has any suggestions please by all means I'm stuck.

[lowell]

Ok scratch most of previous post because of my screwing up the leads of the meter while measuring outputs collectors to ground.

At this point nothing burns/smokes, however the power protection circuit is doing its thing and not allowing the amp to turn on thus the power light remains red and does not turn to green after turn-on delay. Anyone know what might cause this protection circuit to enact?

[lowell]

ok when the power light blinks green for that split second while the protection circuit is doing its thing, I monitored the mains current and it would jump to around 1.5amps when the light blinked. I then tested 1V of DC on the output speaker jack. Does this tip anyone into the right direction?