The small signal devices won’t make a hill of beans difference. And they are being used in the commutation circuit, where they aren’t even in the main audio path.
OK got it.
You know there is another whole chunk of circuitry to these amps on another PCB called a "Driver PCB".
It starts with an NE5534 and has lots of BD420/ 421's on it.
What would be its function and any upgrades needed there?
You know there is another whole chunk of circuitry to these amps on another PCB called a "Driver PCB".
It starts with an NE5534 and has lots of BD420/ 421's on it.
What would be its function and any upgrades needed there?
Any idea where to find the ONSemi KSA1381 ?
Mouser has them on backorder not due till May and even then no guarantees.
A few Ebayers selling sets - but who to trust?
KSC3503 is OK its still in-stock with Digikey.
Thanks for your help with this upgrade BTW - its much appreciated!
Like I said - they are going EOL and in short order everyone will be SOL.
It’s going to be a far bigger problem in situations where 20 pF is an actual show stopper.
Toshiba’s current 230V TO-126 driver pair is useable, and in most situations an improvement over MJE340/50.
It’s going to be a far bigger problem in situations where 20 pF is an actual show stopper.
Toshiba’s current 230V TO-126 driver pair is useable, and in most situations an improvement over MJE340/50.
Last edited:
I messaged one guy here "finwbu" who has some stock - but still waiting for a reply.
Usually obsolete parts are available years latter - there are companies who specialise in obsolete parts - but its at a price!
Any idea about the "Driver PCB's" function?
Usually obsolete parts are available years latter - there are companies who specialise in obsolete parts - but its at a price!
Any idea about the "Driver PCB's" function?
Hi, these don't seem to be an upgrade, but I'd like to hear feedback if anyone has used and/or would use them. They appear to be the CDIL version of the ON SEMI 3281/1302.
The price of “companies that specialize in obsolete parts” is that they are not always the same. Don’t be surprised if you can get them later this year from several Korean outfits. Sometimes they are as good, other times not. The CDIL spec on that C3281 isn’t as good as the OnSemi - but at least they are not publishing pure copies of the original documents and not meeting spec. It looks like kind of in between On C3281 and the 21194 (beta curve looks like the older device types and min spec fT is only 10). I’ve seen similar data sheets in C4793 “clones” that look more like MJE15032’s from the curves. Others that are copies of the original. The Koreans are and even Chinese are coming out with all manner of driver transistor copies. They are not THAT hard to make, so KEC’s and UTC’s are pretty good. Since Profusion is their authorized distributor, and the “real” service industry and many OEMs use them - they have GOT to be at least toward the good end of the spectrum. If they weren’t they would quickly lose the endorsement. No one has ever managed to make a perfect replica of the original 2SD669 yet - the C3503 may prove just as difficult to get that low Cob. The UTC is good, but closer to the C4793 than the original Hitachi. Getting Cob down to C4793 levels (20-ish) doesn’t seem to pose too much of a challenge - but it’s too much for a VAS stage in a 3ppm amplifier. In this “monstrosity” it would be fine.
Now is the C4793 an upgrade from MJE340 here? Maybe. It is faster. Capacitance is about the same, so just as much capacitive load for the op amp to deal with. LF353’s will drive C3503’s without ringing or singing, which is why I use them. If the amp used a regular current source VAS the lower you can get unintentional capacitive loading the better off you are. Op amps have a hard limit, above which they just plain oscillate (and you’re not there yet, because QSC directly drives MJE15032, and a lot of those in the field). Gain linearity is better, but how much of that is it using? Just having the extra stage stomps the QSC in terms of sound quality. But it increases complexity, and totally renders their original short circuit current limit scheme useless. They simply considered those features more important - as the applications don’t call for the last word in fidelity, for all the reasons I discussed above trying to discourage going down this rabbit hole in the first place.
Now is the C4793 an upgrade from MJE340 here? Maybe. It is faster. Capacitance is about the same, so just as much capacitive load for the op amp to deal with. LF353’s will drive C3503’s without ringing or singing, which is why I use them. If the amp used a regular current source VAS the lower you can get unintentional capacitive loading the better off you are. Op amps have a hard limit, above which they just plain oscillate (and you’re not there yet, because QSC directly drives MJE15032, and a lot of those in the field). Gain linearity is better, but how much of that is it using? Just having the extra stage stomps the QSC in terms of sound quality. But it increases complexity, and totally renders their original short circuit current limit scheme useless. They simply considered those features more important - as the applications don’t call for the last word in fidelity, for all the reasons I discussed above trying to discourage going down this rabbit hole in the first place.
Last edited:
Gemini 2.0 AI says:
Absolutely! That statement from the engineer provides a wealth of insight and significantly clarifies the function of the "Driver PCB" and the overall architecture of your Class G amplifier. Let's break down what "inside out," "Peavey's version," "linear class G upper rail" and "not switched class H" means, and how it connects to the Driver PCB:
Understanding the Terminology
- "Inside Out" Circuit: This refers to a specific amplifier topology where the output transistors are on the inside of the feedback loop, rather than the outside (as is common in many amplifiers).
- Traditional Topology: In a typical amplifier, the output stage is outside the feedback loop. This allows the feedback circuit to correct for errors in the pre-amplifier stages, but does not correct errors introduced by the output transistors.
- "Inside Out" Topology: By including the output transistors inside the feedback loop, the feedback circuit can also correct for non-linearities and other errors introduced by the output stage, which is what the "inside out" part refers to. It is intended to produce a more accurate and linear output. It's a more demanding design and a less common topology.
- Benefit: This design often results in lower distortion and better overall linearity. However, it requires careful design and compensation to ensure stability.
- "Not a Direct Copy of the QSC RMX": The QSC RMX series is a popular line of professional power amplifiers, some of which use a form of Class G/H topology. The engineer is stating that while the amplifier conceptually has similarities, it's not a direct copy, and it might have further enhancements.
- "Peavey's Version": Peavey is another audio equipment company known for its robust, high-power amplifiers. By referencing "Peavey's version", the engineer suggests that your amplifier design might have borrowed concepts or topology from the way that Peavey implements a similar design. It is likely to share the same "inside out" topology.
- "Linear Class G Upper Rail":
- Class G Operation: As we've discussed, Class G amplifiers use multiple power supply rails to improve efficiency, by switching in a higher voltage rail at higher output levels.
- "Linear Upper Rail": This is the key point. Instead of a switched upper rail (like a Class H amplifier), this Class G design uses a linear regulator to control the upper rail voltage. This has major implications for the driver stage, which does not need to switch anything, but only change the level of the upper rail, which is more like an ordinary bias voltage.
- Class H Operation: Class H amplifiers switch between multiple supply voltages based on the instantaneous signal level. This switching can cause artifacts.
- Benefit of Linear Upper Rail:
- Reduced Switching Noise: Because the upper rail voltage is actively regulated (and not switched), it has lower noise and does not generate switching artifacts.
- Smoother Transitions: It avoids the abrupt changes in the output stage that may occur when switching between voltage rails, improving the audio fidelity of the amplifier.
Given these points, here's how we can now understand the Driver PCB's role:
- Input Signal Processing: The op-amp likely still handles initial gain, impedance matching, and basic shaping of the input signal.
- Linear Class G Control:
- Drive Signal Generation: The 7 outputs likely now send signals that are used to control the linear regulator that provides the upper rail, and not just switches that choose which supply to activate.
- Dynamic Bias for Upper Rail: The Driver PCB, then, is essentially in control of the voltage applied to the upper rail. This will depend on the output level of the current signal. The various outputs could also be used for bias for the upper rail transistors to make sure they are always operating at their peak performance.
- Reduced Switching Complexity: Because it's a linear Class G upper rail, there are no abrupt switching events. The Driver PCB's outputs control the level of the upper rail in a continuous and smooth manner.
- Biasing the output transistors:
- Class A operation The transistors on the Driver PCB will most likely be operating in Class A, since they are being used to control the levels of the power rails, rather than just as a gate driver.
- Adjusting bias: They will likely be adjusting the bias for the upper rail transistors, so that the operation is always smooth, with minimal distortion.
- Inside-Out Feedback Loop: The Driver PCB provides the signal that influences the output transistors which are within the feedback loop of the main amplifier (the "inside out" topology). It drives not directly the transistors, but instead drives the circuits that control the rails, which then influences the operation of the output transistors.
- Audio Input: The input signal is pre-amplified and shaped by the op-amp on the Driver PCB.
- Driver Control Signals: The Driver PCB generates drive signals for the linear upper rail and provides biasing for the power output stage.
- Linear Regulator: The linear regulator (likely on the Modulator PCB) then provides the upper supply rail for the power stage based on the control signal from the driver board.
- Power Stage: The "inside out" output stage on the Modulator PCB amplifies the audio signal, taking into account the upper rail voltage.
- Feedback Loop: The overall output is monitored by the feedback loop, and corrections are made to the input signal, including any errors by the output transistors.
- Clarified Function: The Driver PCB is not switching between fixed voltage rails, but instead is actively controlling the upper rail voltage in a linear fashion.
- Refined Topology: We now understand the amplifier is a Class G type with a linear upper rail, a variant of a more common Class H implementation.
- Specialized Purpose: This means that the Driver PCB isn't just providing a drive signal, but it's a critical part of the linear voltage control of the upper rail, which sets this topology apart from the QSC or many other Class G or Class H amplifiers.
- Better Audio: The linear upper rail has benefits for the quality of the audio, which is not the case with most switching Class G/H implementations.
- Identify Linear Regulator: Look for the linear regulator on the Modulator PCB and see how the driver outputs are connected to it.
- Trace the Transistor Path: Trace the signal paths to see how the outputs of the driver control transistors or supply the bias for other transistors.
- Analyze Bias: Analyze the bias points on the Modulator PCB to understand how the Driver PCB interacts with it.