#1
The Dyna Comp in question:
http://oi43.tinypic.com/etyxx3.jpg
http://oi39.tinypic.com/10idwg7.jpg

Schematics:
Stock schematic
Mods (refer to stock schematic)
Ross Compressor vs. Dyna Comp (refer to stock schematic)

The IC is an original RCA CA3080E plastic DIP--everyone knows how legendary and rare these things are now. I don't read much into audiophool tone mojo until I experiment for myself so I set out to get some answers once and for all. I tried a modern CA3080E plastic DIP and a CA3080A metal can from 1995 (both made by Harris). I have yet to try the Rochester re-makes, which are supposed to be made with the original Intersil (formerly RCA, now Harris) substrate masks.

Notes about this particular unit:
There is supposed to be a 15k resistor on pin 3 of the IC, but this unit has a 16k here. All resistors except this one are metal film types and the pots' date codes are a year apart (1981 and 1982). This evidence leads me to believe that this is actually a very early Dunlop-made unit circa 1986. It is all-original, though I replaced the LED's current limiting resistor, the original wiring, and the old rotten foam was removed. I also turned the IC socket around, as it had been installed backwards at the factory (letting me think I had destroyed the IC when I re-installed it according to the socket orientation, only to find out a few years later that it still worked just fine--glad I kept it around)...

Anyway, here are the results as compared to the original chip:
The modern DIP doesn't give nearly as much of an effect as the old DIP or the metal can. I'd say the effect is about 75-80% as pronounced as the old DIP. Very noticeable difference, but still passable. It is also much lower-noise than the old one.

The metal can gives more of an effect than the new DIP but still not quite as much as the original; I'd say about 85-90% as much. The metal can would be the best substitute if you can't get an original RCA-made device (watch out, there are a lot of RCA-branded fakes out there). The metal can, not surprisingly, is the least noisy by a wide margin. These are the exact chips used in the '76 Reissue Dyna-Comps. They are also getting more and more rare.

The old DIP has the best effect but is quite noisy. It has that perfect slap-back clicky sound you'd expect from a Dyna-Comp. I love it.

A final note:
If you look at the schematic, you'll see a 27k resistor on the wiper of the Sensitivity pot. Lowering the value of the resistor will let more current into the Bias input (pin 5) of the OTA, which should, theoretically, increase the response of the effect, making up for the inferior sound of the new-production Harris chips. Do not let more than 1mA flow into pin 5 or the chip will be destroyed!

If there is enough interest in this then I will see if it's possible to experiment with the value of this resistor and its effect on the sound of the new chips, and I will post the results here.

Speaking of new mods, I also discovered quite by accident that you can turn a Dyna Comp into a distortion pedal. See post #8.
Last edited by Invader Jim at Sep 14, 2016,
#2
Good stuff Jim! I'm interested in the results...going to start working on pedals myself (moving from making amps to making pedals instead of the other way around!), I'd like to see what you get.
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#3
Man, the GB&C just isn't what it used to be.

I should be getting my gear back sometime this week so I'll be able to run some sound tests. I did some preliminary measurements and here's what I got.

Conditions:
Supply voltage: 9.2v (simulating a fairly new 9v battery)
Sensitivity pot: full clockwise (maximum compression)
Resistance of the 27k on the Sens. pot wiper: 27.2k (I really need a 4-1/2 digit DMM...)
No input signal

Under these conditions, the 3080s' pin 5 bias currents (Iabc) were:
RCA DIP: 248μA
Harris can: 250μA
Harris DIP: 249μA

It would seem that there is plenty of room for increasing the Iabc. The absolute maximum rating is 1mA, but I will try to keep it much lower than that; I am REALLY afraid of frying the chips so I'm trying to be overly cautious with the Iabc. I have 1 metal can and had 4 new DIPs but somehow 3 of them disappeared...

Anyway, that should give me enough of an idea of whether this will be a viable solution for the inferior sound of the newer Dyna Comps (do they still use 3080s?).

Stay tuned...
Last edited by Invader Jim at May 6, 2014,
#4
Jim,

Wish I could offer up more commentary - solid state just isn't my thing, I've really only worked with tubes. I'm slowly picking up the solid state stuff, but I'm at the point where if I wanted to I MIGHT be able to design a simple boost pedal using an op-amp. Might.

And on the tube front - I'm rigging up something of a homemade high-voltage breadboard for prototyping amps. 12V and 6V regulated DC heaters, regulated DC power supply, etc etc. I hope to be able to prototype a new tube amp a night (if I have my act together).

Life has been absolutely batsh!t crazy as of late - I'm currently sitting in a hotel in Jacksonville, FL after my 59th and 60th airplane flights of the year (I live in Wisconsin). LOTS of travel for work...I've come to the realization that haven't been doing some of my hobbies (like tube amp building) because I've been waiting for life to slow down some. I'm realizing now that it isn't going to happen, so I might as well get back in the game now. All the parts are purchased and are on hand...game on!

-Sorry for the threadjack, normal programming now resumes....

The only real way that I know of doing compressors is a variable voltage divider using an LDR and LED combo. Got some plans to try some things like that in a tube amp, but instead by fiddling with the screen voltages on a pentode as compared to straight dumping signal with an activated LDR.
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#5
It's cool, man. My threads never stir up any big commotion, anyway. I'm just doing this to put the idea and knowledge out there for anyone who may need it (and, if this becomes a big ground-breaking mod for the vintage guys, to prove I am the one who did it first )

I'm better with SS than with tubes, but I'm not that great at anything, really. Like a Jack of all trades, master of none. I've always had a difficult time grasping theory because it is all abstract concepts and I don't fare so well with abstract thinking. If I can't relate it to something I'm familiar with, it's hard for me to learn. It is almost crippling.

I can look at a schematic and spot simple stuff like gain blocks, buffers, and generic stuff like that, but a lot of stuff is lost on me. For example, I was reading R.G. Keen's Dyna-Ross article and I'm just thinking "What the hell is this sorcery?". Copying his articles down helps me understand them better because it forces me to think about every little word. I actually just finished typing the Dyna-Ross article out for myself and I understand how the thing works now.

It makes me wonder if this little mod will work as I thought it would. In case you didn't know, here's what is going on (or at least how I read it):

Q3, Q4, the 150k, the 10uF, and the two diodes convert the audio signal to a DC voltage level based on how high the output from the OTA is. Higher output=lower DC level. Q5 buffers this voltage and and the 500k pot and 27k resistor convert it to a current and feed it to the bias pin of the OTA.

Higher bias current=higher OTA gain. Turning the compression pot up all the way will max out the bias current, which causes a much more pronounced "clicky" sound as the level detector slews, trying to hurry up and lower the OTA gain.

With the new DIP, the clicky sound isn't as pronounced as the old DIP for the same compression setting. What I am trying to do with this little mod is increase the effect of the clicky sound to come close to the sound of the originals.

Now that I think about it, this may theoretically work after all. Does anyone else have any thoughts?
Last edited by Invader Jim at Nov 14, 2013,
#6
Hey Jim,

I am actually interested in this, I just need to actually get a chance to sit down and think about what is going on in the circuit again. Since I've been spending all of my days managing about 20 different electronic projects I just haven't really made myself look.

I'll take a look sometime in the next few days. I'm actually hoping to use one of the local PCB fabs to start making some PCBs so that I can prototype and test more stuff.
#7
hmm, interesting. i wish the IC wasnt such a pain to get to on mine, it sounds like changing it to the metal can would be a worthwhile change. im interested to see what you find on changing the value of that resistor. it would probably be an easier mod.
#8
I finally got around to experimenting with this. I'm sorry it took so long. I had a lot of crap to deal with before I could get back to experimenting and stuff.

So anyway, lowering the value of that 27k resistor does seem to make the new chips sound similar to the old ones (it's hard to a/b a circuit that takes a few seconds to "reset" itself). I say "similar" because each chip type sounds slightly different on the low notes and high notes. It's an amalgamation of things, kinda hard to peg down into simple terms. Basically each chip type has its own character, I suppose.

While I was trying to address the "character" issue, I thought about lowering the value of the 10u cap in the level detector. As I said earlier, this cap stores the voltage generated by the level detector, then Q5 buffers this voltage and and the 500k pot and 27k resistor convert it to a current for the OTA bias (Iabc). With a lower value, the level detector would charge it faster and the OTA would drain it faster, creating a faster response time. This way you'd get a faster attack.

I tried a 6u8 tantalum cap, but after a few minutes the Iabc slowly fell to only a few uA, then dropped out to 0uA. I think maybe the cap went bad. Tantalum caps will fail at the drop of a hat, so it isn't that far-fetched. Later, I will try an electrolytic type instead.

An interesting side note: removing the level detector cap turns this pedal into a distortion, with the Sensitivity pot becoming a distortion control. It sounds somewhat like a Distortion+ but the clipping isn't quite as hard. There's a bit of octave but it is subtle, not obnoxious and in-your-face like with a poorly-biased transistor circuit. As the Iabc is increased (lowering the value of the 27k resistor) the distortion amount increases and the octave gets more pronounced (in a bad way) and the effect starts to gate and sputter on the note decays. Wouldn't want to have the Iabc resistor lower than stock for this part.

Anyway, back to business... for the 27k resistor I went as low as 10k, giving an Iabc of 579uA for the new chip with an 8.56v supply at the board (I used a battery this time; I'll suck it up and use my power supply again and re-do the experiments).

The output is louder (higher Iabc=higher OTA gain), attack is harder, and the signal seems to actually become less muddy. Again, it is quite similar to the original chip and stock resistor value.

I'll continue the experiments later. So far this looks like a viable mod. I'm gonna see if I can find someone who can actually play and have them play-test the distortion mod. If anyone reading this has a Dyna Comp, new or old, I'd appreciate a verification. One data point a trend does not make.
Last edited by Invader Jim at Dec 7, 2013,