TL072CP Op Amp Differences

The variance in gain. Maybe it's just a weird compulsion/neurosis of mine…

The gain values of my batch of 5088s were all over the place, (bah, I deleted my graph—yeah, I'm weird) My batch of Fairchild 5088s were nearly uniform with a handful of outliers. Similar story with 3904s (no-name vs Motorola).

IDK, perhaps I'm overly concerned about numbers over things just working—the same way I badmouth unicorn tears vs "make it sound the way you want."

The wide spread makes me think poor QC. But that's just one opinion, and I'm not a seasoned expert.

I'd love to be schooled and set straight on where it does and doesn't matter!

<edit> formatting and spelling
 
Last edited:
If transistors are in-spec, then they will most likely work in the circuit. Some circuits need HFE to be in a certain range, either to get the biasing right or to control harmonic content. In the majority of Si circuits, whether the HFE is 100 or 500 doesn't matter. HFE is a measure of current gain. Most pedal circuits use BJTs as voltage amplifiers and voltage gain does not depend on HFE. Like Vf in a diode, HFE varies with current. There is no standard test current for HFE. The datasheets spec HFE at one or more collector currents and some datasheets plot how HFE varies with collector current. The transistor testers I've seen measure HFE at one collector current and it probably does not match the collector current in the pedal circuit.

For a particular wafer run, the specs of a given transistor are pretty tight because they were all made at the same time under the exact same conditions. Dozens or hundreds of wafer runs can be packaged on the same day and will have the same LDC. The only thing the LDC tells us is when & where the chip was put in the plastic package or metal can. Buy 100 2N5089s from Mouser and they will all have the same LDC. Buy 100 on Amazon, eBay, Tayda or Small Bear and you could end up with many LDCs. That is the cause of the variability.

HFE plays a minor role in a Silicon Fuzz Face. Q1 HFE mainly affects biasing. Q2 HFE affects tone. Either effect can be compensated by adjusting resistor values and/or adding a cap between base & collector on Q1 or Q2. Look at the Aldrin for an example of tuning the circuit to the transistors. Some (most?) pedal builders choose a particular part number based on price and availability. The EHX BMP is a prime example of this.
 
Chuck D. Bones said:
If transistors are in-spec, then they will most likely work in the circuit. Some circuits need HFE to be in a certain range, either to get the biasing right or to control harmonic content. In the majority of Si circuits, whether the HFE is 100 or 500 doesn't matter. HFE is a measure of current gain. Most pedal circuits use BJTs as voltage amplifiers and voltage gain does not depend on HFE. Like Vf in a diode, HFE varies with current. There is no standard test current for HFE. The datasheets spec HFE at one or more collector currents and some datasheets plot how HFE varies with collector current. The transistor testers I've seen measure HFE at one collector current and it probably does not match the collector current in the pedal circuit.

For a particular wafer run, the specs of a given transistor are pretty tight because they were all made at the same time under the exact same conditions. Dozens or hundreds of wafer runs can be packaged on the same day and will have the same LDC. The only thing the LDC tells us is when & where the chip was put in the plastic package or metal can. Buy 100 2N5089s from Mouser and they will all have the same LDC. Buy 100 on Amazon, eBay, Tayda or Small Bear and you could end up with many LDCs. That is the cause of the variability.

HFE plays a minor role in a Silicon Fuzz Face. Q1 HFE mainly affects biasing. Q2 HFE affects tone. Either effect can be compensated by adjusting resistor values and/or adding a cap between base & collector on Q1 or Q2. Look at the Aldrin for an example of tuning the circuit to the transistors. Some (most?) pedal builders choose a particular part number based on price and availability. The EHX BMP is a prime example of this.
Click to expand...
Thanks for the detailed explanation! I stand corrected.
 
I'm pleased(?) to report I've had my first ever failure of a TL072 fake! As in, even if they're rebadged 358 this thing completely failed on me.

Didn't notice I had popped a fake one into my madbean Degenerator until it was putting out odd, noisy, atonal static. I thought, is there a fake 072 in there? There was! (had the notch and no dimple on pin 1) replaced it and boom, it's working fine again.
 
I ran into an issue with a fake TL072 on my last build. I didn’t realize they were fake, I bought them off Amazon and used them in a Muzzle noise gate. The gate would make a strange fizzle or crackling noise when the input signal would be close to the threshold. After posting about the issue, a kind individual pointed out that I had been had, and I was using a fake TL072. After I swapped the fake out the noise was gone. Going forward I am avoiding Amazon for ICs, and sticking with a reputable seller. I bought a handful off of Pedal PCB!
 
ThatDude said:
I ran into an issue with a fake TL072 on my last build. I didn’t realize they were fake, I bought them off Amazon and used them in a Muzzle noise gate. The gate would make a strange fizzle or crackling noise when the input signal would be close to the threshold. After posting about the issue, a kind individual pointed out that I had been had, and I was using a fake TL072. After I swapped the fake out the noise was gone. Going forward I am avoiding Amazon for ICs, and sticking with a reputable seller. I bought a handful off of Pedal PCB!
Click to expand...
Is he Reputable? :LOL:
 
Chuck D. Bones said:
If transistors are in-spec, then they will most likely work in the circuit. Some circuits need HFE to be in a certain range, either to get the biasing right or to control harmonic content. In the majority of Si circuits, whether the HFE is 100 or 500 doesn't matter. HFE is a measure of current gain. Most pedal circuits use BJTs as voltage amplifiers and voltage gain does not depend on HFE. Like Vf in a diode, HFE varies with current. There is no standard test current for HFE. The datasheets spec HFE at one or more collector currents and some datasheets plot how HFE varies with collector current. The transistor testers I've seen measure HFE at one collector current and it probably does not match the collector current in the pedal circuit.

For a particular wafer run, the specs of a given transistor are pretty tight because they were all made at the same time under the exact same conditions. Dozens or hundreds of wafer runs can be packaged on the same day and will have the same LDC. The only thing the LDC tells us is when & where the chip was put in the plastic package or metal can. Buy 100 2N5089s from Mouser and they will all have the same LDC. Buy 100 on Amazon, eBay, Tayda or Small Bear and you could end up with many LDCs. That is the cause of the variability.

HFE plays a minor role in a Silicon Fuzz Face. Q1 HFE mainly affects biasing. Q2 HFE affects tone. Either effect can be compensated by adjusting resistor values and/or adding a cap between base & collector on Q1 or Q2. Look at the Aldrin for an example of tuning the circuit to the transistors. Some (most?) pedal builders choose a particular part number based on price and availability. The EHX BMP is a prime example of this.
Click to expand...
Wow. Very, very informative post, Chuck - thank you! I realize Ge may be a different story, but it's good to know that most Si circuits have a lot of leeway regarding hfe - the Fuzz Face being a slight outlier.
Chuck D. Bones said:
Germanium diodes is a whole 'nother discussion. They can be legit and still unusable.
Click to expand...
Hmm. I was going to get just a few (real) 1N34A from Tubes N More, as they seem like a pretty reliable place that would test vintage components - although I've yet to order from them. I guess I didn't think about leakage being a factor in Ge diodes, just transistors. Makes sense that it would still be a consideration... Small risk, but it'd still stink to pay 2 bucks a pop for unusable diodes. I wonder if AES vets their 1N34As well. Not sure where else to go, really - or if it's even worth it compared to say, a BAT41.
 
Depends on the circuit. If you're going to buy Ge parts, sooner or later you are going to get some stinkers. AES aka Tubes n More is one of the pricier sources. I have never paid close to $2 ea for Ge diodes. I just checked Small Bear and they are no bargain either. Again, part number is a secondary consideration. Which circuit are you building that calls for 1N34A diodes? In some cases, Schottly diodes are a decent sub.
 
Chuck D. Bones said:
Which circuit are you building that calls for 1N34A diodes?
Click to expand...
Good question! Ha… I’ve seen them pop up in build docs for several fuzz pedals, but I think any Ge diode would do (unless it’s crap). They’re mostly in pairs, so it seems like BAT41 etc might sound about the same in that situation…
 
You must log in or register to reply here.

Similar threads

J
Ceramic Capacitor Differences
Replies
1
Views
136
Smrtokvitek
Smrtokvitek
L
Non-inverting op amp buffer with gain control
Replies
6
Views
217
Sturdag Lagernathy
Sturdag Lagernathy
H
Could we build a Honey Bee with a dual op-amp chip?
Replies
5
Views
463
Fama
Fama
E
Rule of thumb for decoupling individual op amps vs a cap across the rails?
Replies
1
Views
110
Alan W
Alan W
W
General question on op-amp output resistors
Replies
15
Views
1K
Big Monk
Big Monk
Back
Top