-
This is where I am at right now. I magnified the output and still detected some parasitic distortion. I tried increasing the resistance of R314 from 180 ohm to 260 ohm but it got a lot worse. Went back to 180 ohm and tried increasing the capacitance of C306 and it got worse. Just for the sake of seeing what happens I completely removed C306 and all parasitic distortion completely disappeared even under magnification. Currently Rbe (R307 and R309) are at 12 ohms up from the original 2.7 ohms when germanium outputs would have been installed. When connected to my iphone I can play at lower volumes and the sound is good but still at very low volumes the distortion returns. I also connected the amp up to the preamp and the same thing happens (though it is a lot better than before), hit a guitar note and as it decays the distortion is evident. I am thinking the outputs are not quite biased correctly and are turning on and off as the input diminishes. The Re resistors (R312 and R313) are still the original value at .68 ohms. Do these need to be changed to something else to alter current or anything. If I post the ecb voltages on both outputs with no signal would that help determine what's wrong with biasing? Currently MJ15016's are installed at Q302 and Q303.
I truly do appreciate any help and am slowly learning thanks to you guys.
-
Sorry, your post missed me ausman, the pull of bed was strong and rapid!
That sounds like good news, you are getting the hang of achieving reasonable stability in the feedback loop.
Your low level distortion must be crossover. The feedback loop should compensate it to some extent - but there are limits to the speed, especially around a transformer coupled stage.
Yes, you're on the right lines, the switch from Germanium to Silicon results in a big shift in required bias point - from around 0.2-0.3V up to 0.6-0.7V. Taking the original bias values... 31V with a voltage divider of (R308/R307 and the corresponding pair on the other transistor) 1k / 2.7R gave around 83mV. Your change to 11R increased the bias to just under 340mV. To get up to 0.6V you will need to go up to around 20R.
As I mentioned in a previous post, you need to be careful here. The transformer coupling means that there is no DC negative feedback around the amp. The output DC offset voltage will depend entirely on the bias of the output transistors. If one starts conducting before the other, then you will get a fairly large offset voltage. The amplifier must remain in Class B operation (both transistors off at idle) rather than Class AB (small output stage current running at idle). With perfectly matched transistors you could get closer but that ain't going to happen. You will always get some crossover distortion, you just need to get it down to the level where the negative feedback can bring it down to an unobjectionable level.
You're going to need to play around with the values of R307 and R309 a bit (I don't know if you have plenty of resistors to hand). Do it with feedback removed again.
You need to get each transistor to the point of near turn-on without actually hitting it. As I mentioned before, there is no temperature compensation for the output transistors (and they look to be sitting on a 'nice' steel chassis too). Their Vbe will drop by 2mV/'C so you need enough clearance that they don't turn start to conduct at idle when hot. The scheme must have worked well enough previously, because Germanium is -2.5mV/'C. It's a pretty primitive amp though.
I'd suggest experimenting with values between 18R (550mV) and 22R (670mV), 24R (726mV) is probably too high. You might also try staying at 15-18R and padding down the 1k with a parallel higher value low power resistor, probably the cheapest option. The scope trace will show you how close you're getting and output offset voltage and polarity will tell you when you go too far. Once you've done that, get the output transistors nice and toasty with a decent level into a resistive load for a while and check again. You don't want an offset to appear when hot, or the transistors to go into thermal runaway. As the old mechanical engineering adage goes, tighten up until the threads just strip, then back off half a turn!
Last thing then is to re-apply negative feedback and play with that. The more negative feedback you can apply and remain stable, the lower the distortion, and the less critical the Vbe resistor values.
Remember, it's an instrument amp, it's never going to be perfect at low levels, but it might get close. Just one other though - you could look at retrofitting more up to date amplifier modules, it depends how 'original' you want to stay.
-
Wow, thank you so much for the lengthy response! I wish you were my neighbour!
I will have to place an online order for resistors today since I don't have those values on hand. That's a bummer because I won't be able to report on any progress until next week. It's sad these days over here that you can't just jump in the car and run to the local electronic parts store since they are few and far between.
Why would the original bias design for the germaniums have been set so low at 83mV when the ideal range would have been 200 - 300mV, perhaps to create a larger buffer against thermal run away?
At the moment I am measuring around 320mV across the base emitter junction of Q303, is this the place to measure for bias voltage?
I did measure hfe of both output transistors on an Peak Atlas DCA55 before installation and they were closely matched but not perfect, can't remember the exact numbers though, around 220 and 210 I think.
So I will fine tune the correct resistor values with the feedback loop completely disconnected while at the same time keeping an eye on the dc offset. If I start to see a turn on condition at say 700mV should I aim for 600mV to stay in the safe zone so I don't creep into the thermal runaway area? and the theory behind padding the 1K is to give me a finer control over the biasing?
When reconnecting the feedback loop should I try adding back in the original c306 cap and see what happens there? What is that cap doing exactly, routing high frequency AC from the switching noise to ground through R305? By playing around with the feedback loop I really wouldn't know what I am doing or what I am looking for. Do you mean to lower the value of R314 to increase the feedback signal?
Thank you,
Damien
-
Haha, you're welcome. It's sometimes frustrating when you can't physically get your hands on something too.
Your measured 320mV (yes, that is the right place) is pretty close to what comes out of the basic R308/307 divider calculation, the supply voltages are unregulated, so almost certainly aren't exactly +/-31V (?). The resistors aren't very close tolerance either.
Regarding the original Vbe, they were clearly ensuring that the output stage remained solidly in Class-B. The problem is that the base turn-on voltage isn't a snap transition, it's a curve, with the voltage being specified at certain datasheet current parameters. The problem with converting such an old circuit from Germanium to Silicon is that the difference in Vbe threshold is a 'fixed' offset of ~400mV rather than a percentage change thing. Until you get past that offset, nothing's going to happen, all the time the transformer output is below that threshold you're getting nasty crossover distortion. It does make biasing tricky though.
The circuit is an old topology (pretty much a valve era conversion) to allow them to use two PNP transistors. It was very hard to manufacture NPN transistors with decent gain, particularly power ones, an NPN - PNP matched complimentary pair was pretty much out of the question.
Regarding tuning - yes, back off a bit from the start of turn-on to avoid temperature coefficient issues. At idle R307 and R308 (and their partners) form a DC voltage divide to set the Vbe bias (via the transformer secondary) the main factor in the choice of the 1k was to keep its heat dissipation at a sensible value (50% of power rating). If you parallel it with a higher value resistor then you can adjust the voltage divide ratio, eg. if you parallel it with 9k you will bring the value down to 900R and the 9k will only be dissipating 0.1W allowing a standard 1/4W part allowing you to play around a bit.
Yes, you could try the original C306, it's hard to know how much the extra crossover distortion was affecting the feedback loop. The ratio of R314 / R305 sets the feedback ratio at normal frequencies (yes reducing R314 increases the feedback, going to infinity removes it), C306 is there to decrease the ratio at high frequencies (by shunting R315) to maintain stability. The change to Silicon transistors means that the output stage frequency response is now higher (faster transistors) and higher gain. If you can find the datasheet for the original transistors then we can see the differences.
Ideally you would use a squarewave input to tune C306, tuning it for a clean waveform without significant ringing. This is an instrument amp so I don't think it will ever get that clean, but as long as it looks stable it should be ok. I noticed there is a stage in the Normal channel input circuit labeled "Distortion Booster" so I doubt that it matters too much!
-
You are one great teacher and I dearly thank you for sharing your personal time and knowledge. I hope others are also getting a lesson from this!
The parts are on order and I should be back at it next weekend. Until then....
Cheers -
A while back I worked on a guitar amp. of similar design, an Acoustic model 150. It also has transformer drive to the power transistors.
Original RCA 48-15 output NPN's were blown, replacing them by 2N3055 caused incorrect biasing and unwanted oscillations as in your case.
The oscillations must be stopped as they can severely heat the power transistors for nothing and cause incorrect biasing.
In my case oscillations were killed by changing C305 (B-C on transformer drive transistor) from 100pF to 560pF.
Getting the bias right was harder, I first tried altering the base resistors like you, but the output stage was much to sensitive to mains variations. The bias would easily vary from 200mA quiescent current to having crossover distortion.
The bias was modified b replacing R314 and R315 each by a series string of two SB140 diodes and a 1ohm resistor, R312,R313,R316,R317 left unchanged.
If that can help in any way...
-
Good to find someone with direct hands-on experience of doing the same thing!
Yes, diode TC compensation would be a more stable. I was assuming that if the resistive biasing worked sufficiently well for Germanium then it would hopefully be ok with lower TC Silicon too. I agree, the supply voltage variation could be an issue too.
The awkward problem with the OP's amp versus the one you worked on is that the speakers are DC coupled to the amplifier (yours has a speaker coupling cap and is single supply), making output offset voltage critical, it basically means that the amp must stay in zero bias state at idle - maybe the DC coupling is something that's going to need to change! Being single supply, there was no ambiguity about the polarity of the output coupling electrolytic on yours, the OP's might need a bipolar or two back-to-back.
The DC coupling also means that even a single diode in the output transistor bias circuit would probably be enough to cause turn-on. Maybe a single Schottky diodes in series with the 'B-E' resistor would be sufficient to provide partial Tc compensation, assuming that it is held in good thermal contact with the transistor packages.
Adding a B-E capacitor to the transformer drive transistor might well be a way of taming things, thanks. -
Thanks for the information.
I also have an Acoustic 150 and the matching 104 cab. Knock on wood, it is still working as good as new after touching up a few solder connections, and it is all original. I will print out your reply and stick it in the original owners manual I have, just in case!
Is there any information you need from the owners manual? It contains several fold out schematics.
Cheers -
Hi!
I have had many problems with germanium-transistor O/P stages over the years, both with getting the bias right and preventing thermal runaway!
What you can try is the following:-
1) Obtain a pair of BC327 medium-power output transistors from fleabay, etc (Hong-Kong SS8550 etc., will also do!), and twist the base/emitter leads together, connect a thin length of fine insulated connecting wire a few inches long to the paired leads, solder and sleeve the joint with thin heatshrink sleeving;
2) Solder a second length of similar wire to the first to the collector lead of the transistor, solder and sleeve the joint as before in step 1) above;
3) Obtain some Thermal Bonding Adhesive, apply a SMALL dot of adhesive to the top of the O/P transistor's case Q302, secure the Silicon Bias Transistor prepared in step 1 and step 2 above and leave to set;
4) Repeat Steps 1, 2 and 3 above with the second O/P transistor Q303,
5) Obtain a couple of 1.0 R 2W resistors, (these will take the place of R307 & R309) soldering ONE END only to the circuit-pad connecting to the junction of the driver transformer secondary winding for each half and the 1k 2W bias resistors R308 & R310 - see attached circuit-extract;
6) Connect the FREE end of each 1.0R 2W resistor R307 and R309 to the lead coming from the SINGLE COLLECTOR LEAD of the Bias Devices fitted in Steps 1 to 4 above to each O/P transistor case;
7) Connect the lead coming from the PAIRED EMITTER/BASE LEADS of each Bias Device fitted in Steps 1 to 4 to the original connection pads of R307 & R309 that connect to the +31V power supply and the LS mid-point of the O/P stage respectively;
DOUBLE-CHECK YOUR WORK!
9) Connect a DVM, set to a mV range, across R313 and switch on. A reading of no more than about 6-8mV should be obtained - if the current starts to rise rapidly, switch off at once and double check your work - you may need to fit a lower value in place of the 1.0R resistors fitted in step 6 above;
10) What you are aiming for is to get a current across EACH emitter resistor as close to the maker's value as possible - if Vox didn't give a figure in their Documentation, try for about 4-6mV across each emitter-resistor to begin with!
11) The Silicon Bias devices will have a negative Vbe/Vcb temp. coefficent of voltage approximately the same as the replacement O/P Devices you fitted, so some degree of temperature compensation should be obtained.
The above took me far longer to write and explain it than it does to actually carry it out, believe me!
The circuit-extract should make it clear, I hope!
Chris Williams
-
Sorry for the delay in responding, I only just now got back to working on the amp. Thanks for your input. The time and effort is very much appreciated. I will seriously consider your suggestions if I do run into thermal runaway problems.
-
Just got back to working on the amp. I ended up settling for 18 ohms for R307 and R309. Running a sine wave into it at very low levels produced no distortion that I could detect however as I started increasing the input level it started to distort at about 4V p-p at the output across an 8 ohm load. Removing the feedback cap C306 made the distortion go away. I am able to run the output up to about 50V p-p before clipping occurs. I ran the amp for about 20 minutes into the 8 ohm load and set the output to 10V p-p. Output transistors felt luke warm and the dc offset was virtually zero at about 20mV. So as it stands the circuit remains all original with the exception of C306 removed and R307 R309 changed to 18 ohms. When my son gets home I will have him test it with a guitar and report back. The pic shows the 10v output, it looks very clean to me.
-
Thank you everybody especially Gyro.
The amp appears to be working. My son tested it with guitar at moderate to loud levels and it sounded great. The heat sinks were only moderately warm to the touch so no thermal problems either. Well I came to learn, and learn I did! Still wondering though why C306 was causing distortion?? -
Thank you everybody especially Gyro.
The amp appears to be working. My son tested it with guitar at moderate to loud levels and it sounded great.
When was the last time you heard a father say that??? My father always yelled "Turn that damn thing down!!"
-
Gyro, you are now my sons hero.
He has abandoned his beloved Roland JC120 and is now married to this Vox. Most Dad's are saying turn it down, I am saying turn it up. It really sounds beautiful with an almost tube like sound along with the ability to emphasize every nuance and he fully agrees and more, it is almost dead quiet between passages.
Thank you much, cheers my friend!!!