The amp is way, way, underbiased as it is with only 2 diode drops for the 3 output stage Vbe junctions - this is why the diodes do not have to be on the heatsink. The output stage idle current should be near zero at all times. It is a class B output stage not AB.
Thermal runaway is not the problem with this design.
Where do you plan to order your parts from? Do you plan to get samples from OnSemi or Fairchild?
The MPSA56 and MPS8598 (and MPS8099 comp) are still active parts at OnSemi they are fine. I would
toss the BC parts and get the real compliment MPS8099 for Q6 and Q12. These have a low Vbe drop
to work well as a protection transistor.
Let me think about the VAS and the rest.
This amp is essentially the old RCA Quasi Comp reference design with higher rail voltages and a
doubled up output stage. It is widely copied and a proven design, works fine in a complementary
version also.
Thermal runaway is not the problem with this design.
Where do you plan to order your parts from? Do you plan to get samples from OnSemi or Fairchild?
The MPSA56 and MPS8598 (and MPS8099 comp) are still active parts at OnSemi they are fine. I would
toss the BC parts and get the real compliment MPS8099 for Q6 and Q12. These have a low Vbe drop
to work well as a protection transistor.
Let me think about the VAS and the rest.
This amp is essentially the old RCA Quasi Comp reference design with higher rail voltages and a
doubled up output stage. It is widely copied and a proven design, works fine in a complementary
version also.
Last edited:
This topology is similar to a 1974 RCA reference design for 70 Watts. All the outputs and drivers were beefed up and VAS stage is changed to a cascode using an RF output device. Compensation is probably wrong if you use faster modern devices. Yes it is biased to Class B which IMO can excite any resonance in the output devices with bad layouts to oscillate. My approach would be target for 120W at 6 ohms output and replace all modern audio devices along with a single quasi output pair and TO-220 drivers using P-P wiring. Bias for mild class AB and revisit the RCA compensation scheme and VAS.
Here is one option I'm considering, the DOGC-H amp
http://bas.elitesecurity.org/dogcE.html
It was refered to on page 3 of this thread:
http://www.diyaudio.com/forums/soli...omplementary-power-amplifier-j-diamond-3.html
The higher power version uses 48V rails. The transistors used are MJL4281A/MJL4302A with MOSFET drivers of IRF630/IRF9630. My power supply has 56V rails and it looks like these transistors would handle this voltage without issue.
My questions with this approach are:
1. Do any part values need to be changed if I boost to 56V and if so, what needs changed?
2. Should I use a more modern finned heat sink? or can I move the board to one side and attach all the transistors to only one existing heat sink so the transistors may all be board mounted? (existing heat sinks are just heavy cast aluminum box sides with mounting flanges for the TO-3 transistors. They have no fins and are just a heavy mass.)
3. Are there any other issues with this option I may be overlooking?
Thanks for everyones help.
http://bas.elitesecurity.org/dogcE.html
It was refered to on page 3 of this thread:
http://www.diyaudio.com/forums/soli...omplementary-power-amplifier-j-diamond-3.html
The higher power version uses 48V rails. The transistors used are MJL4281A/MJL4302A with MOSFET drivers of IRF630/IRF9630. My power supply has 56V rails and it looks like these transistors would handle this voltage without issue.
My questions with this approach are:
1. Do any part values need to be changed if I boost to 56V and if so, what needs changed?
2. Should I use a more modern finned heat sink? or can I move the board to one side and attach all the transistors to only one existing heat sink so the transistors may all be board mounted? (existing heat sinks are just heavy cast aluminum box sides with mounting flanges for the TO-3 transistors. They have no fins and are just a heavy mass.)
3. Are there any other issues with this option I may be overlooking?
Thanks for everyones help.
do you want to use the existing amp PCB with minimal changes to get the thing working again, hopefully with more reliability. Thinking of possibly not using lower than 6 ohm woofers makes this design easier. The fast simple choice is keeping the original PCB and going closer to the original 1974 design with cheaper, better, and more rugged modern devices OR redesign a PCB from scratch. Unless someone knows an amp PCB / kit that is available for your PS and heatsink?
I'm very seriously considering a new PCB and considering the DOGC-H design if no one sees an issue running it at 56V rails. His webpage includes a PCB layout and It looks like a no new design approach to getting the poweramp up and running again.
Since I am willing to etch a new PCB and add modern heatsink, I'm open to other tried and true designs that would utilize the existing 56V rails.
I would also consider a rebuild using the current PCB and leads to transistors layout if I can do so without compromising too much power for reliability. I'm not interested in compromising to the point of sacrificing the ability to run 4 ohm load. I usually run two 8 ohm cabs for bass and the rebuild must be able to do that.
Since I am willing to etch a new PCB and add modern heatsink, I'm open to other tried and true designs that would utilize the existing 56V rails.
I would also consider a rebuild using the current PCB and leads to transistors layout if I can do so without compromising too much power for reliability. I'm not interested in compromising to the point of sacrificing the ability to run 4 ohm load. I usually run two 8 ohm cabs for bass and the rebuild must be able to do that.
Cascode????? RF output device, what are you talking about????? This design has none of those features! Nothing to do with this design, in fact take a look at C1 and C10 killing any HF gain! This thing is way overcompensated with that large Miller cap also.
I can't see the heat sinks in the pictures, is that just the chassis or are there fins on the other side? Edit: I see you mentioned no fins earlier, yes you should use better heat sinks. Let us know what you see as options in terms of internal vs. external mounting, etc. The more heat sinking the better. Let us know if you plan for a fan, you'll need one if the sinks are internal and there is not a good path for natural convection. Edit: it would probably be wise to include a fan and run it off of a thermal switch, especially if you want to go for as much power as possible into 4 ohms or lower.
No I would not use that design.
Do you have a scope for this work, to support testing it?
Actually, the relatively low supply voltage and low impedances make it easier to get a lot of power since you will not run into SOA issues as fast. I think you'll be able to do 250W into 2 ohms fairly easily and around 150 maybe 200W into 4 ohms depending on how much the supply droops.
Q8 is D44R3 is an RF transistor, I have some here, and there are gold plated FWIW. What is the function of Q12 then?
HF gain what are u talking about. The max gain at C1 and c10 is unity the outputs are usually 2N3055 and similar or ~ 800 KHz devices. This is an early 70's design and parts.
Do you have a suggestion for a 56V rail design since you don't like the DOGC-H amp?
I have no problem putting a fan in the unit. There are vent louvers in both the top and bottom panel but I doubt it is enough. The picture MB7.jpg that I posted earlier shows the inside of the chassis and the heat sink box sides. There is plenty of room for a fan to mount in the back panel to draw air thru the chassis. Careful planning would allow airflow over an added internal heat sink. Another option would be to cut thru the back panel and put the heat sink external on the back of the box. That might actually be as easy or easier than cutting for a fan. I have access to several fully equipped machine shops and modification of the chassis isn't an issue.
Q12 is the VAS protection transistor, it senses current in the emitter resistor and clamps the base drive. It is required because the VI protection in the output stage essentially shorts the VAS to the output load when triggered which would destroy the VAS without Q12. It is always good to have in case of input overdrive.
D44R3 crosses by Motorola to TIP47 which is a 10 MHz part, not an RF part.
ECG crosses it? I'd check it with a data sheet before I'd claim that.
What protection, Q12 it's not even connected to the base, that's what throws me here?
The protection never works anyway right? This thing could be made to work with what's there, with minor improvements. I don't like it when everybody just guts the thing before them. Who knows this maybe a valuable piece of audio history?
could be like refinishing antique furniture in my thinking. LOL
Last edited:
And most on the internet are nuts or wannabes!
Did I say ECG cross? No I said Motorola cross, and the TIP47 is what is used in the newer revision of the design ... no need for me to redesign it the manufacturer already did. Now, you seem to be unable to figure out that the first revision of the schematic is missing a dot. Take a look at the next revision for the dot but please stop wasting my time with your misunderstandings. Do you really think that would be a cascode with it connected to the Miller cap?????
The owner wants to gut it because it has a reputation of blowing up ... what don't you understand about this????
Since you are a smart guy it should be real easy to figure it out and fix it with some needed, parts replacements, re biasing, and a couple of tweaks. Got nothing to loose by one go around here.
So I'm wasting your valuable internet time. LOL why such hostility? do you have anger issues? I guess we now know which of your 2 internet catergories you fall under.
You need to read the thread; I am suggesting parts subtitutions yet you offer the insult that engineers will always redesign. Trust me, I have done enough design work in my time thanks.
It is the OP who wants a completely new design, he has a right to do that doesn't he, or does he need your approval?
You are wasting my time because it is clear that you are not reading the thread, no hostility just a plain fact.
Last edited:
Sometimes an alternate view is mistakenly perceived as being against the person. I thought I was plainly speaking of the OPs options as I saw them, thinking I am entitled to an opinion even if it goes against yours or the OPs view of the world. nothing more nothing less.... but please don't start twisting things around telling me I'm forcing the OP to do something against his will or that I'm completely ignoring his wishes. If you really believe that, then you are the one not following the thread. I'm sorry that you feel I'm wasting your (more?) valuable time. PS yea I missed that revised schematic thing, my mistake.
Most I see are self important all knowing legends in their own minds that apparently need to show the world how clever they think they are.
Last edited:
I come on here to help people, that is plainly obvious by my contributions all over but in particular the recent Adcom thread. You come with the insults about what engineers need to do because I pointed out a mistake you made. Why not just admit the mistake and move forward instead of trying to turn it into some absurd contest. I have no need to prove anything, I've been at this a very long time. So, now can we get back on the subject at hand ....
Move forward. I hope your contribution will be fruitful. I reserve further comment about the "recent Adcom thread", to which after so much time on a single amp, we will never know the cause for said repair. Pity you feel annoyed by my presence. I think all you want is kowtow somehow. Why should I apologize for remark about engineers that I know is true.
Last edited:
Am I supposed to miss the fact that you are trying to push my buttons bringing up your engineer insult again? I'm old enough to spot that quite easily, LOL!
So what is it you are "just" a designer and not a degreed engineer is that what compels you to insult real engineers? After all, you would not insult your own kind if you were one, you are driving the point home that is for sure!
Why don't you do something productive and show us your upgrades to this design ... you know the saying put up or ....
Something for the OP to think about. You might have noticed that there is a lot of protection in this amp, and that the HF gain has been really knocked down by C1 and C10. These are probably due to the fact that it is a bass amp where it is likely to be overdriven often and should not fail or make nasty buzzing sounds or produce speaker damaging output in overload. Even if the intention is to play a bass amp "clean" it is likely that you will overdrive the amp for brief periods of time and it should not do nasty damaging things.
So choosing a replacement design might not be as simple as just finding a Hi Fi amp design to build. I think it is probably best to just upgrade this design which is based on the proven RCA design. Certainly look at other solid state bass amp designs with a reputation for good sound and reliablitity if you know of any. If you want to try a general purpose SS amp I'd go for simplicity and low NFB so that it behaves well in overload.
Is this inline with your thinking? I will finish the redesign of this amp if you choose to go that way, don't know if I'll have time to analyze the protection circuit.
Do you have a scope, dummy load, and signal generator to test this amp once it is rebuilt?
So what is it you are "just" a designer and not a degreed engineer is that what compels you to insult real engineers? After all, you would not insult your own kind if you were one, you are driving the point home that is for sure!
Why don't you do something productive and show us your upgrades to this design ... you know the saying put up or ....
Something for the OP to think about. You might have noticed that there is a lot of protection in this amp, and that the HF gain has been really knocked down by C1 and C10. These are probably due to the fact that it is a bass amp where it is likely to be overdriven often and should not fail or make nasty buzzing sounds or produce speaker damaging output in overload. Even if the intention is to play a bass amp "clean" it is likely that you will overdrive the amp for brief periods of time and it should not do nasty damaging things.
So choosing a replacement design might not be as simple as just finding a Hi Fi amp design to build. I think it is probably best to just upgrade this design which is based on the proven RCA design. Certainly look at other solid state bass amp designs with a reputation for good sound and reliablitity if you know of any. If you want to try a general purpose SS amp I'd go for simplicity and low NFB so that it behaves well in overload.
Is this inline with your thinking? I will finish the redesign of this amp if you choose to go that way, don't know if I'll have time to analyze the protection circuit.
Do you have a scope, dummy load, and signal generator to test this amp once it is rebuilt?
Yes let's get back on topic. I have absolutely no desire to keep this amp for it's "Historic Value" it was given to me as a non working piece of junk with a bad reputation and I love tinkering with electronics and I play bass guitar so I see it as something I can make useful to me. Most of my electronics tinkering has been with tube amps and solid state preamps/processors. I've not done much if anything with solid state power amps and that's why I'm asking for suggestions so I can learn as I make this boat anchor useful. I'm also not wanting to make this major lab project. I'm sure there are some 56V rail designs that would work just fine with this power supply and if some chop and fit is needed, I'm not afraid to do that either. Had only the power transistors been fried, I would have been more willing to spend 30 bucks on a set of transistors and see how long it ran before blowing up again but with much of the existing power amp board needing parts pulled and replaced, I'd just as soon make it a better, more modern amp. My quest now, is to find what existing 56V circuit is the best fit.
Pete, you've already been helpful in helping me figure out exactly what all was damaged in the amp and I look forward to your further suggestions.
In reading over this post from you, I can sort of see your point about design specific to bass guitar.
If you really think we can make the old layout work reliably and also reuse the pcb, I'm game since it wouldn't take much cash to give it a try.
I do have a signal generator, dummy load and an old scope but I just recently got the scope and haven't come up to speed with using it yet. It's an older tektronics scope.
Another design I have looked at is the ESP project #68 which is also designed for low frequency.
300/500W Subwoofer Power Amplifier
This design puts the transistors on the board and would require me to come up with a new heat sink. I built several of his mic preamps a few years ago and was pleased with the design.
Thanks again for your help.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.