Running a stereo amp in parallel instead of bridged

Status
Not open for further replies.
Uh, no. Tony is guilty only of being stubborn. 🙂

OK, stubborn guilty I may be, but guilty of assuming a different parallel personnae, I will never be. I am what and who I am.

BTW, I see you as tombo56.

In the vast majority of forums I'm at, I go by TonyE. That's my name, trust me, and I wasn't happy having to use a different user name, it's too much to remember, like passwords and pin numbers.

Oh, I see too that you are quoting people that I'm ignoring... mostly for my sanity. I made the mistake of "enabling" their comment. Oh well. The less I know of the insults upon my head, the more I maintain my sanity.
 
Last edited:
OK, one last comment.

Some of you have made the observation that current is "drawn" by the load, not "pushed" by the source.

That is indeed one way of looking at it, but not entirely the only model that is plausible.

If you think of the flow of electrons, you can see a simile to flow dynamics. Imagine Bernoulli and his water pipes. The water is being "pushed" downstream and the flow (the load) controls the flux of water by virtue of its cross area. (Water can not be compressed). Sort of like a faucet.

The flow of electrical charges, as some one already described earlier as dQ/dt, is thus controlled by the load via a resistance. But it is also limited by the ability of the source to drive power.

So, you can see two equivalent models to describe the flow of charge, current. Both are equally valid. Indeed, I posit that so long as the flow of power is limited on the source side, the most correct way of looking at it is to think of the flow of charge as being "driven" by the source.

After all, the electrons flow from the source back to the source. The load never provides any charge... Hmm... the electrons coming back to the source have to be at a lower energy level too... because they yielded some of their energy to the load to do some kind of work and/or heat.
 
Another issue with Parasound Zamps could be that they can’t provide more than 5W continuously, without overheating. At least, that was valid for V3 according to this ASR test:
https://www.audiosciencereview.com/...ds/parasound-zamp-v-3-amplifier-review.11872/

Hmm... I've used them for years. At one point I was using one bridged (!!) for the center channel in the HT and two more for the sides and rear surrounds. Never had any issues.

I've also run them in two previous office systems into minimonitors, 89db/watt with a 16 ohm load, with no issues. But listening in the near field configuration, about three feet from each, I never had to use up lots of power. Never had them fail.

The idea of the Burson opamps is intriguing... I have been buying those things for years... V4 and V5. Have some Vivids and Classics on the shelf. They do change the sound quite a bit.
 
I think you should do it and find out for yourself. You will probably learn a lot.

Part of the reason people like Fahey and I know so much is we've designed a lot of circuits and make some mistakes along the way. Mistakes really drive a point home.

Learning is not a mistake. I predict you will be motivated to try something else once you find out the pitfalls for yourself. I remember many times when my designs didn't work as intended. I always figured out why eventually. It made me a much better designer in the long run. I can now see the problems before I build the prototype, every time.
 
Another issue with Parasound Zamps could be that they can’t provide more than 5W continuously, without overheating. At least, that was valid for V3 according to this ASR test:
https://www.audiosciencereview.com/...ds/parasound-zamp-v-3-amplifier-review.11872/

That link illustrates perfectly what I said. Those channels are not very well matched; gain and distortion profiles are markedly different. Parallel them and you will have problems.

I experimented with triple parallel LM1875 chips. These chips are current limited so there's a good reason to parallel them. It worked great and I got over 40 watts continuous into a 4 ohm dummy load. But it was a "clean sheet" design with hand matched resistors, not a commercially built unit.
 
...

Learning is not a mistake. I predict you will be motivated to try something else once you find out the pitfalls for yourself. I remember many times when my designs didn't work as intended. I always figured out why eventually. It made me a much better designer in the long run. I can now see the problems before I build the prototype, every time.

Thanks

Yeah, I used to design digital circuits when we did it by hand and at the very beginning of FPGAs, nowadays it's all done by RTL or VHDL. It's become all pure logic and timing relationships. Nobody cares about designing an analog (de)multiplexer anymore, it's cheaper to put in a core, or two, and send it down in a packet of sorts.

I have noticed that nowadays the R&D design budget will skew heavily towards the analog side, stuff like the signal conditioning IO devices and specially, big time, the power supplies.

BTW, I have smoked a lot of electronic stuff in my life, when people see me they think I look younger than I am... that's because I've spend the biggest part of my career in electronic/computer labs kept at 68F/50% humidity. Mistakes made in good faith are not mistakes, only learning experience!

"Duh! what was I thinking?"

Thanks for the input.
 
Why?

How else can I match them?

I suppose the more High End the amp is, the better matched the components and channel behavior would be. But having two channels share a circuit board would minimize many of the possible differences. At the very least the channels would have very closely matched gain characteristics, huh?
Every component has a tolerance, many manufacturers (even so called high end) use +/- 1% resistors in there feedback network (the thing that sets the gain). That means 2 "identical" amps can have a 2% difference in output voltage. With a 10 volt output thats a .2v diff between outs. If the output impedance of the amps is .02 ohms you will have 10 amps flowing from one amp out into the other amp output.
 
Every component has a tolerance, many manufacturers (even so called high end) use +/- 1% resistors in there feedback network (the thing that sets the gain). That means 2 "identical" amps can have a 2% difference in output voltage. With a 10 volt output thats a .2v diff between outs. If the output impedance of the amps is .02 ohms you will have 10 amps flowing from one amp out into the other amp output.

Which is solved by a 1 ohm in series output resistor... which now lowers the difference to ~0.02%.. at the risk of wasting a lot of power on those resistors.

Feedback is required for gain in op-amps, but not necessarily for discrete components. Huh?
 
And the DF went out the window.
Also the real part becomes a bit more important than the imaginary part and we get less phase shift and more predictible loads . We loose more energy through resistive dissipation , but every Zobel has its damping resistor, every rise time needs to get settled and you get that in Time and time means heat dissipation whether it's in a real resistor or in a biased transistor junction. Every oscillator needs a damping resistor of some sort to settle down if the feedback can't do it in Time. As long as we are in Class AB domain matching capacitances earlier in the game is more important and the Parasound Z has an advantage the OP didn't even care to assess for diy purposes: is using a pretty low output capacitance power transistor so I'd bet on lower phase shifts at higher frequencies.
Now that I'm fully ignored by Miss Bohr Rutherford von queen drama Eisenstein I can recommend studying Kenwood L-A1 quadrive system to those who can afford to diy inside their 10 identical slow amplifiers output stages and Kenwood km-x1000 thx to those who can afford to use one transformer per supply rail 🙂
https://physics.stackexchange.com/q...aning-of-the-imaginary-component-of-impedance
 
You missed that primary school class when they taught the Energy Transformation, didn't you? Naughty boy....

I put a fan on top of my electric vehicle figuring that I could generate so much energy at 80mph that I could go forever... I found Physical Chemistry to be the most boring course I've ever taken... pfffft. They have some issue with eternal energy... maybe they're related to that guy over at ASR?

With no need whatsover of any negative feedback... afterall who needs negative feedback? I get enough in these forums? I don't want my circuits to have any feedback, none whatsoever, I need no peer reviews because that's supposed to be naughty. Besides I have no peers. And as the one guy wrote, I need to have this forum published so I can get my umpteenth doctorate on some obscure academic field of science... maybe polysorbothanic neo modern astrophysics...

Positive feedback would be nice, true, but I doubt I'll get it. Mostly I just get lots of resistance to my EMF that it discharges my well being over an absurd energy half time.

So, I just march my electrons flows in parallel, which oddly are in a negative flow of energy, as energy and charge flows are opposite to each other, since mostly we look at energy holes travel up the pipe. So, twin pipes are always more charistmatic and quadruple energetic.

BTW, what did that fellow mean by DF above? I guess I missed that class in Social Forums.

I suppose I need to order parts, I still don't know what scope to get though. Likely I will take a month to do the experiment. But I must add a fire extinguisher and a bench in the garage before I explore inline output resistors below 0.2 ohms.

In the meantime enjoy 2nd harmonic distortion in your stereo. I do. Even if it always comes with a blue LED.

So, I'm gonna do a parallel amp experiment. Let's see what happens. Why not?
 
Last edited:
Status
Not open for further replies.