Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 29, 2024, 03:05:28 AM

Login with username, password and session length

Recent Posts

 

Thermal limits chipamps

Started by phatt, September 04, 2014, 09:12:57 AM

Previous topic - Next topic

phatt

Hi  all,
Just thought I'd share this for the benefit of chip amp builders.

Seems the die area in poweramp chips is not really up to the task.
It's not just about a bigger heat sink.

http://sound.westhost.com/articles/guitar-amps.htm#s3

3/4 down page he mentions those LM1875 and like chips, between the red graph and the deep fried poweramp chip picture.

Phil.

Bear

So how would one go about designing for a more bullet-proof chipamp?  My noob reading and data sheet studies are making me think shooting very low on the voltage range for the opamp in question would help reduce the on-die thermal stress.  So, say the infamous Vox Pathfinder/Cambridge TDA-series chip upgrade "now it's louder!" Mod that does not increase volume--well, might going from a 20 watt part to a 50 watt part reduce the failure rate of that component?

I guess single-railing the supply when it enables non-insulated heat sinking can help heat transfer, but it won't really help that on-die issue.

phatt


Sadly from what I've read I don't think you can design a better chip amp. I guess it's a bit like a 4 lane hiway into a one lane bridge which would certainly slow down traffic flow and cause a bottle neck, in this case heat transfer flow rate is severely limited. Hence those old TO3 case power transistors are better able to cope.

A while back I purchased a small LM3886 kit with the aim of driving a 4 Ohm speaker but I failed to read the fine print so I figured some reading was in order.
Yes obviously power chip amps are convenient and simplifies a build but extracting the claimed wattage from them maybe a big problem at higher voltages.
I can only assume that Restricting the low frequency response in the design may help but how much may prove difficult to work out.

As to mounting directly to heat sink (No insulator) Yes you can do that but you then need to isolate the whole heat sink from the chassis which then introduces a lot of other issues.
Phil.

Roly

Current.

P = E I

but

P = I2 R

Losses rise directly as the voltage, but as the square of the current.


Groovy


Functional

If you say theory and practice don't agree you haven't applied enough theory.

J M Fahey

Quote from: Roly on September 05, 2014, 11:19:41 AM
Current.

P = E I

but

P = I2 R

Losses rise directly as the voltage, but as the square of the current.


Groovy


Roly, are you sleeping well?

Worried about something?

Take care, friend :)

teemuk

#5
Yes, there are limits. One can either accept them or completely ignore them. Don't forget that you can always parallel or bridge amplifiers or increase the current handling with additional discrete transistors. You can compensate the limits of the small die area by means of design.

But of course if you just stick a single TDA7293 into a way too small heatsink located inside a chassis with moderately hot ambient temperature, and then claim a 100W output power rating you can just blame yourself for frequent chip failures.

And overall, I do believe the Rod Elliot article paints a somewhat unoptimistic image of chip thermal handling. Those chips are used in countless applications ranging from cheap car radios and computer speakers to professional active monitors and yes, guitar amps. Picking a burned Marshall board as an example is kinda misleading because Marshall for some reason seems to be pretty good on making amps where power chips fail while thousands and thousands other manufacturers seem to get the same chips to function just fine and with decent reliability.

Pritchard amps, I believe, use LM3886 chips and their designer actually preferred them over the discrete circuits due to wide range of built in protections. His former discrete design had to be ridiculously overbuilt to meet equal performance regarding overall reliability.

As a side note, the die area issue mostly refers to modern power chips where everything is integrated on a single die. Old chips, such as those old Sanyo's, were built differently: sometimes they even housed the power transistors as discrete devices so overall die areas of the critical points were usually larger. Such details are of course hidden by the chip enclosure.

phatt

Thanks Teemu, It's good to hear other points of view.
I agree many (thousands) power chips seem to run fine so it's rather confusing for those out in the dark world of advertising hype claiming insane wattage from mere toy amplifiers.

I've had a few run ins with chip amps the last one was a small 4 input pa type unit claiming 300 watts touting 2 small speakers and a dedicated sub speaker, effectively a Biamp setup.

Once I fixed the broken pads from a poorly mounted bridge rectifier assy it ran fine until I turned the knob on the sub which is a separate control knob,,, this resulted in instant destruction melting pins on one of the TDA7294 chips.
The unit is a *Skytec MP8* claiming 300watts (see pics) 3 power chips on a large heat sink.
One chip powers the satellite speakers and 2 in bridge mode for sub.
The sub is a 10 inch 4-Ohm wobbly cone hifi type woofer which are nearly always low is SPL so they take a lot of power to drive.
Now unless I'm sadly mistaken it's only a fool who bridges amps then tries to drive a 4-Ohm load. Surely,,,8-Ohm would give it a chance and 16-Ohm would be safe?

The heat sink weight is only out done by the supply tranny,, a huge copper donut which I doubt would sag much under full power.

This brings me to a Q,, and some amateur observations.  :-\
Forget the bridge stuff for the moment, just a standard LM3886 unit driving a 4 Ohm load.
I'm wondering IF one were to use a slightly de-rated transformer (same voltage, less current capacity) which would sag under load and by automation back away from instant smoke?

OR if stuck with an over sized Transformer;

One of the Lenard pages has some interesting thoughts on introducing sag on supply rails but I'm assuming he is referring to discrete power amplifiers.

http://education.lenardaudio.com/en/13_guitar_amps_3.html

The Q is; Would something like this work on powerchip amplifiers as well as discrete poweramps?
I notice that a lot of the small bedroom amps that run LM1875 and similar chips without issues often come with quite small supply transformers which may sag a bit and stop them melting down.

Also;
The Lenard page implies that sag in supply rails makes it respond much like a Valve power stage which might be interesting to pursue. Power stage compression is a real bonus in valve amps and the idea maybe helpful in more ways than one for chip amps and discrete.

Long ago I did have a small guitar amp (simple 7 transistor discrete power stage) and it actually sounded very realistic while running the mains through a 60 watt light bulb limiter.
As I look back now I can see that with the aid of a hotter preamp section it would have come closer to this.
https://www.youtube.com/watch?v=ZkGCvLstPrE

This is my ultimate aim, to have the broad dynamic range like above without the need for a dozen pedals on the floor. (and of course all SS, wink)
Many SS amps sound good but few have the dynamic response like the link above.

Sorry if I've tried to cover too much in one post,,,,but if I don't spit it out I'll forget the plot.
Phil.



Roly

Quote from: J M Fahey on September 05, 2014, 11:00:29 PM
Roly, are you sleeping well?

Worried about something?

Take care, friend :)

No.  Yes.  I'll try, thx.
If you say theory and practice don't agree you haven't applied enough theory.

phatt

#8
Well thought I'd put my scrambled thoughts onto the bench and test something for real instead of talking about it. ;)
I tested the *Lenard* sag/powerstage compression concept and it certainly holds promise. :dbtu:

Using ESP basic power amp as test bed; http://sound.westhost.com/project03.htm
(using Fig 1a as reference to my circuit)

Full supply voltage after rectification is 30-0-30 VDC
I inserted 7-Ohms in supply rails, adding the extra filter caps and increased .47-Ohm emitter resistors to 2-Ohms.
OK it kinda works, supply rails sag down to 25 Volts and does a little compression but nothing like the Lazy J Amp in that video link.  :'(

**For the benefit of those who wonder, some hunting says; LazyJ Amplifier is a clone of the 5E3 fender circuit which is a cathode biased Amp and likely modified to reap a better sound than stock.
The power stage of these Amps are known to compress due to cathode biased setup and I'd bet compression is increased further by a rather large value screen resistor.
I did build a 5E3 years back as a test circuit but it just did not have enough power for my needs with 6V6**
Well I've learned a fair bit since then so maybe this time I might actually understand what I'm doing.  :duh **

OK I could go back and rebuild that 5E3 circuit but where's the fun in that? xP

Back to the SS power amp sag,
with the Values above still in place I then inserted a light bulb limiter on the mains as this is known to sag and compress the output signal.

As I've already created a fair bit of sag maybe rather than just increasing resistor values which did not really improve the result I thought I'd try a bit of mains sag as well.
With a 100Watt bulb limiter on the mains it really starts to compress, some tube freaks call this effect bloom of the note as the power stage compresses.

Obviously just sagging the mains is not enough,, you need to raise the emitter values as well,, then it starts to sound like a Valve power stage compressing.

The problem then becomes,,, this power amp is still quite loud and I have to keep the level
low as I now live in suburbia and I don't wish to annoy my neighbors on a monday night. :-X
So I'm using a very basic attenuator which does allow me to play hard and achieve much the same effect as a hot rod 5E3. (really need to run it without attenuation but that will have to wait) But I'm certainly getting closer to lighting that cigar. :) :)

What I did note before I packed up for the day is that when playing clean there is crossover distortion but as I increase guitar volume it comes clean then the wonderful Bloom at full volume :dbtu:,,, by then it's hard to know if there is crossover distortion. I'll look into the bias issue latter,, for now I'm just happy that it works much better than I thought.;)
The Attenuator seems to be part of the cause of Xover issue. Hum?

Unless someone knows more this trick obviously leaves chip amps out in the cold and lonely world of clean as you can't access the emitter resistors. bummer. :(
Might be a few days before I can make more progress but if anyone is interested I'll get a schematic drawn up with what I have so far.
Rock 'n Roll

ps; A picture of the actual Amplifier I used to test this circuit resides here;
http://www.ssguitar.com/index.php?topic=3344.0

Phil.

J M Fahey

Problem is that SS amps are so tightly "controlled" by typical huge feedback (think 60dB), that rail voltage does not affect gain at all, so you lose 80% of what sag does.

In pentode power amps sag affects screen voltage which in turn affects gain.

Since tube amps have little NFB to begin with (usually around 6 to 8 dB) stage gain differences are much more apparent.

Even more in no NFB amps, of which there are more than a few.

And Marshall/Tweed Bassman become "no NFB" at mid/high frequencies, because of Presence.

So in your experiment you are only left with variable attack clipping, simply because when overdriven the squarewave envelope amplitude falls with falling power rails.

Roland in their newer Blues Cube 60 uses something similar, but in the preamp.
They have a "Tube/SS" (rectifier) switch, which actually enables a transistor to clamp preamp rail voltage so it creates Sag.

This happens before the Master Volume, so they get this effect at any power level.

Clever chaps.

mexicanyella

#10
Interesting thread, even to a guy like me who's about 27,000 light years behind you guys in terms of electronics comprehension. Sag is a big deal to me; I'm very dependent on an amp having a certain response "feel" to do what I do. So I can relate on that level.

I will say that, as an electronics-using caveman, my own sag needs are currently pretty well met by running a compression pedal after a Rat II, which is set with the gain very low and the output very high so it isn't compressing the signal too much on its own, just adding a little grit and harmonic content. Maybe one of these days I'll have learned enough to approach this by actually building something the way you guys have, rather than just plugging things in in a certain order! Carry on...

Also...I have GOT to try one of those Blues Cubes one of these days.

Roly

{Gotta love that test rig Phil  }

Doug Self wrote a whole article as part of a series in Wireless World a while back dealing with the optimum value for emitter resistors.  Turned out that 0.2 ohms is close enough to ideal in just about all cases. 

http://www.douglas-self.com/ampins/ampins.htm   (can't find it)


Me?  Sag?  I'd be thinking of one of those bass-player compression boxes, lamp and LDR...
If you say theory and practice don't agree you haven't applied enough theory.

phatt

Quote from: J M Fahey on September 08, 2014, 02:52:36 PM
Problem is that SS amps are so tightly "controlled" by typical huge feedback (think 60dB), that rail voltage does not affect gain at all, so you lose 80% of what sag does.

In pentode power amps sag affects screen voltage which in turn affects gain.

Since tube amps have little NFB to begin with (usually around 6 to 8 dB) stage gain differences are much more apparent.

Even more in no NFB amps, of which there are more than a few.

And Marshall/Tweed Bassman become "no NFB" at mid/high frequencies, because of Presence.

So in your experiment you are only left with variable attack clipping, simply because when overdriven the squarewave envelope amplitude falls with falling power rails.

Roland in their newer Blues Cube 60 uses something similar, but in the preamp.
They have a "Tube/SS" (rectifier) switch, which actually enables a transistor to clamp preamp rail voltage so it creates Sag.

This happens before the Master Volume, so they get this effect at any power level.

Clever chaps.

Hi, Jaun,
            Thankyou for a prod in the right direction, I know I'm trying to invent a barbed wire canoe and I'm paddling against the tide.
At least this time no magic smoke escaped from the test transistors, for me that's an improvement. :lmao:
Big thanks for the Beautiful explanation of 5E3 workings,,, arrh!! now you have triggered my grey matter to work on another idea. (See it's your fault I get these crazy ideas  :P)

Here is the light bulb idea that fell out of my brain last night,
""There must be a way to get a saturated but dynamic effect using simple circuits?""

Ok how about this approach;
I've never liked compressors as they just turn the gain up and down, but what about a circuit that does the inverse? Turning up the gain at the input but at the same time turns down the output level?
I have no idea if it's ever been tried but I would assume a control circuit much like those used in compressors would be a starting point.
Unlike compressors which seem to try to avoid distortion, I want the distortion to go up and the output level to come down.
I might try a very basic manual form with a dual pot wired in reverse just to see if the idea works.
To the best of my understanding that makes more sense if one is trying to get that touch response effect between guitar and amplifier as demonstrated in that Joe Bonamassa link.
Meantime Home jobs need attention,, playtime later, sigh.
Phil.

phatt

Quote from: Roly on September 09, 2014, 12:47:54 PM
{Gotta love that test rig Phil  }

Doug Self wrote a whole article as part of a series in Wireless World a while back dealing with the optimum value for emitter resistors.  Turned out that 0.2 ohms is close enough to ideal in just about all cases. 

http://www.douglas-self.com/ampins/ampins.htm   (can't find it)


Me?  Sag?  I'd be thinking of one of those bass-player compression boxes, lamp and LDR...
Hi Roly, So you like my test rig? Well no point in making a pcb only to find you are only adding to land fill. :'(
Yeah I believe Rod Elliot has a lamp limiter idea on ESP site but that only does what a compressor does,, but as J M Fahey has touched on the Cathode biased 5E3 idea has a real big dynamic range and to the best of my understanding the only way that can work is if the output is kept at a set SPL while gain is greatly increased.

With some valve amps I've noted that at a certain point on the panel volume the output SPL remains at around the same level no matter how much further you turn the dial. These are the ones that are more likely to deliver the dynamics, set the panel level and the onboard guitar volume controls the distortion.

An example of the complete opposite of what I wish to archive sits on my bench this week, a Digitech RP80 pedal complete with cab sims and every imaginable effect known but after understanding how to drive this pile of crap I noted that it has ZERO dynamics so with any distortion setting you can't roll off the guitar volume and get clean.
Reminds me of my first midi keyboard which had no velocity sensing, Every note was set at 127.
My Mothers old pump organ had more dynamics than that crap toy I wasted money on.
Phil.

phatt

Quote from: mexicanyella on September 09, 2014, 12:35:22 AM
Interesting thread, even to a guy like me who's about 27,000 light years behind you guys in terms of electronics comprehension. Sag is a big deal to me; I'm very dependent on an amp having a certain response "feel" to do what I do. So I can relate on that level.

I will say that, as an electronics-using caveman, my own sag needs are currently pretty well met by running a compression pedal after a Rat II, which is set with the gain very low and the output very high so it isn't compressing the signal too much on its own, just adding a little grit and harmonic content. Maybe one of these days I'll have learned enough to approach this by actually building something the way you guys have, rather than just plugging things in in a certain order! Carry on...

Also...I have GOT to try one of those Blues Cubes one of these days.

Hi Mex,
Don't worry there are many ways to achieve a good sound and some of the modern amps are very impressive but most of the cheap stuff sold is not worth the money.
Reading forums like this will likely go a long way to help avoid the garbage. :dbtu:
Phil.