Problems with Dragonfly's Kay tremolo

[svstee]

Layout here: http://aronnelson.com/gallery/main.php/v/DRAGONFLY-LAYOUTS_0/album19/album145/KAY_TREMOLO_001.jpg.html?g2_imageViewsIndex=1

I built this and when I tested it I had no tremolo, just a volume and treble loss. Took some voltage readings on the transistors, I'm not liking what I see.

Q1
E: 5.91
B: 0.00
C: 0.00

Q2
C: 0.14
B: 0.89
E: 0.14

Q3
E: 0.14
B: 0.63
C: 0.00

Q4
E: 0.67
B: 0.67
C: 0.00

Any ideas?

[Cliff Schecht]

All of your transistors are turned off. With any NPN transistor, they are biased "on" when the collector voltage is higher than the base voltage and the emitter voltage is at least .7 V below the base voltage.

[8bitapocalypse]

i had to change the 1k2 resistor on the last transistor to a 100k(!) trimpot to get this working with BC109c transistors.  i also got rid of the amp stage at the beginning and added a booster at the end so you have a volume control.

[oskar]

Oh lord!!!
I really need to get a grip on myself and read up on transistors...
What in earth are the trannies with the collectors connected doing?     

GGG schematics

[Cliff Schecht]

Hmm.. This is an interesting one.. I suspect that the base-emitter resistance of the transistor is being taken advantage here. The second transistor is acting as a "protection" stage between the first transistor and ground, to prevent the collector from going more negative than the base and emitter. This is an initial guess, so I'm gonna think this one through while I go run some errands. Be back in a little with a more definite answer!

[svstee]

Thanks!

[Dragonfly]

It's definitely a "verified" layout that has been built many times. You might reference the following threads for some suggestions.

http://www.diystompboxes.com/smfforum/index.php?topic=45493.0
http://www.diystompboxes.com/smfforum/index.php?topic=45668.0

[svstee]

I never doubted the layout, I had seen those other threads as well, so I assumed the problem was on my end. The transistor voltages confirmed it for me. I checked all the trace cuts, jumpers, polarity, and cap values especially, since those seemed like likely suspects for the problems I'm having. No luck, I'm thinking about going over all the resistors with an Ohmmeter to make sure values are right. Any other thoughts?

[Cliff Schecht]

Hmm.. This is an interesting one.. I suspect that the base-emitter resistance of the transistor is being taken advantage here. The second transistor is acting as a "protection" stage between the first transistor and ground, to prevent the collector from going more negative than the base and emitter. This is an initial guess, so I'm gonna think this one through while I go run some errands. Be back in a little with a more definite answer!

Upon further thought, it's actually a pretty simple circuit. I'm going to call the top transistor Q1 and the bottom transistor Q2. As the oscillator signal swings low, Q1 and Q2 start to turn off, presenting a high impedance to the Q1-emitter and 100k resistor junction. This means that no signal loading occurs and the full input sound (plus or minus any circuit gain) is seen at the output. As the oscillator swings high, Q2 starts to turn on and Q1 starts stealing signal away from the junction of the 100k resistor and the Q1 emitter. When the oscillator is at its high peak, the signal seen at the output is very small (nearly off).

There are two "unique" tricks used in this circuit that give it such an interesting sound. The first important trick is the fact that the tremolo circuit responds exponentially to its control signal, giving a tremolo that sounds "natural" to the human ear. The second trick used is that the oscillator has a large swing (~3 V), which drives the transistors in and out of saturation (as opposed to a more linear response) and gives a choppy sound. These two "tricks" are what gives the circuit it's unique chopping sound while still sounding natural to the human ear. Very cool circuit indeedy!

P.S. - It only took me 6 years after initially seeing this circuit (when I first started building pedals) to get to the point where I can actually understand what's going on completely .

[svstee]

What do you think my problem is, then?

[Cliff Schecht]

Your board doesn't work. Go through and double check every connection, every component value and make sure the solder joints are good. Then, if this fails, replace all of the transistors. If this fails...

[svstee]

I've triple checked everything but resistor values, would an extra digit on a resistor somewhere give me these problems?

[Cliff Schecht]

Yeah, that could throw off the biasing of everything. I'll run a simulation on this circuit later and tell you what the voltages should be.

[Cliff Schecht]

A couple of things to note before I give any values:
1. Many of the transistors in this circuit (well, all but the input one) will have continually changing collector and base voltages because of the nature of the circuit. Take a DC reading of the bias values will just give you the RMS value of the moving wave which isn't of too much help.
2. Are you sure that you don't have the collector, base and emitter pins backwards on all of your readouts above? With NPN transistors, your collector is more positive than the base is more positive than the emitter. Actually, looking at your readings, I'm sure you've got everything flipped .

From right to left on the schematic, I'll label the transistors Q1 (input transistor), Q2 (bottom of the connected transistors), Q3 (top of connected transistors) and Q4 (oscillator tranny):

Q1
C - 7.41 V
B - 0.62 V (with no input signal)
E - 0.0 V

Q2
C - connected to Q3 collector - will vary around +50 mV to -125 mV - set the tremolo to the slowest setting to measure this
B - Again set the trem to the slowest setting - 100 mV to 600 mV
E - 0 V

Q3
C - Same as Q2
B - Same as Q2
E - Only varies about 5 mV which may be hard to read with a DMM

Q4
C - with trem set to slowest - should vary from 0 V to around 3.5 V
B - varies from 685 mV to 710 mV (may be hard to read with your DMM..)
E - 0

Simulation isn't my preferred way to go about getting these values but since I'm at home and have no oscope, this is a good start. Good luck!

[svstee]

2. Are you sure that you don't have the collector, base and emitter pins backwards on all of your readouts above? With NPN transistors, your collector is more positive than the base is more positive than the emitter. Actually, looking at your readings, I'm sure you've got everything flipped

I double checked, I can't see how they could be. I'm looking at dragonfly's layout now, and my transistors are layed out the same as he shows.

[kurtlives]

Transistor pin outs can vary by manufacturer. You could be using a different brand than Andy.

I never go by what the pin out is on a layout.

[svstee]

At risk of being mocked, can I ask how to check pinouts? Can I do it with a multimeter without hooking up power?

[jacobyjd]

look up the datasheet for the specific transistor you're using--it'll show the pinout.

Besides that, it's usually not very likely that the Base will be anywhere but the middle pin, so the first thing to try would be to flip all the transistors around. Often the pinout is mirrored. 

[svstee]

I have an el cheapo DMM like this one;

If you look at the bottom left, you see a blue circle for doing something with transistors, but I have no idea how to use , specifically where to set the dial. can this be useful here? I don't remember where I got the transistors, they are 5088, either from futurelec, mpja, effectsconnection, or Ebay, so I,m not sure a datasheet will be of much use, since all that is written on them is "5088".

[kurtlives]

Ok the 2N5088 is a NPN silicon transistor with the pinout E-B-C. Good no awkward pin bending.

But how do we know which side is E and which is C? Set your DMM to read transistor gain and insert the transistor into the slots. I think the 2N5088 has a hFE of roughly 500. If you get in and around there you know you have the right pinout. If you get some weird reading like -4, 0 etc you know its in wrong.

The slots on the DMM are labelled E B C so then you know which leg is which.


Btw if you click more info beside the transistor on Futurlec it will give you the datasheet.