Vox Cambridge 30 Reverb Twin keeps blowing TDA2050

[vipersassasin09]

is this for the cambridge 30 r... not the twin?

[R.G.]

I took a quick look at the datasheet for the TDA2050A. It's rated for +/-25V absolute max on the power supply. The schemo for the Cambridge shows +/-27.1.

While things are not absolutely guaranteed to fail if you exceed their absolute max ratings, it's not a bad bet that some of the production run will give you trouble. And I personally would fire an EE for turning in a design that had the nominal power supply running over the absolute max for the parts without a very convincing argument from him/her about why the design was realiable in spite of the manufacturer's warning on his datasheet.

My opinion is that the Cambridge 30 is a poorly designed box. Sub in an LM3886 that will take the extra voltage, and for goodness sake, get rid of those silly +/15 and +/-9V zeners; stick in three terminal regulators for them and have a huge jump in reliability.

My anti-favorite for zeners is the Fender DeVilles. They use zeners for +/- 15V making, and the resistors for the zeners commonly burn *through the PCB* and *melt the solder off their leads*. Silly when the 7815 and 7915 are well under $0.50 each.

[Robert M. Martinelli]

Hi R.G.

You're absolutely right, some amps are poorly designed indeed IMHO,( BTW what you say about the Deville applies to the Hot Rod Deluxe and other FMIC amps as well, we've been discussing this kind of issue here many times in the past ).

As to the Cambridge, it's a mixture of poor design and cheap components ( e.g. the LT991X "trem roach" ) but, I have to say that, once all its inherent weaknesses have been removed by modding it, it's a pretty decent sounding amp, especially in its price range.....

As someone wisely said ( though I don't exactly remember who ) it looks like accountants have more control on designs than EEs have.....I believe this statement to have become sadly true.

I find it stupid to spoil/ruin a manufacturer's reputation ( built on years and years of hard work ) to save only a few bucks, but, unfortunately, in this world based on sheer appearance, that's just the way it goes....

I wonder what Tom Jennings or Leo Fender would have thought about it....

Cheers

Bob

[Steve Conner]

As someone wisely said ( though I don't exactly remember who ) it looks like accountants have more control on designs than EEs have.....I believe this statement to have become sadly true.

This is of course true, but why sadly? Left to their (or do I mean our) own devices, engineers would spend forever trying to create the perfect product. Putting a cap on the bill-of-materials cost is one way to force a reality check. Imposing a deadline is another.

In the last 10 years I've designed a lot of stuff that's considerably cheaper and nastier than it would be if I were building it only to satisfy myself. But it's sold well and made a profit.

I certainly wouldn't put 27v on a chip rated for 25 maximum, though. As an end user not constrained by BOM cost you can replace the poor TDA2050 by anything you like. I like RG's suggestion of the LM3886: this chip is used in hundreds of DIY hi-fi amps, so you can easily find information on how to hook it up. I've seen people just tack all the components on to the chip's leads without any sort of a PCB.

But if you really want to stick one over to the bean counters, how about a transformer-driven output stage from a Thomas Vox? :-)

PS: Wasn't Leo Fender famous for running tubes over their ratings? And Tom Jennings with his AC30 that you can fry eggs on top of. Good job they're more forgiving than semiconductors.

[teemuk]

But if you really want to stick one over to the bean counters, how about a transformer-driven output stage from a Thomas Vox? :-)

As if those weren't unreliable...

It's odd that this vintage circuitry stuff is now a big fad when back in the days people were screaming with joy to get rid of it. If Thomas Organ (or basically any other other manufacturer) had had access to chip amp designs of today they undoubtedly would have laughed at the idea of using a monstrosity such as interstage driven totem pole output - with Ge transistors.

Maybe after more than 40 years we can actually do something different if we want to use those interstage thingies... Think about it.

Nevertheless, despite the advancements in technology, some designs still are poor.

[Steve Conner]

As if those weren't unreliable...

Maybe after more than 40 years we can actually do something different if we want to use those interstage thingies... Think about it.

Well, there's another answer to the "What would Tom Jennings have thought" question. He'd have loved the Cambridge 30 because he already put the Vox name to far worse products

Inspired by RG's writings on the Thomas Organ amps, I've experimented with a more modern(?) version using an EL84 as the driver "transistor", a Champ output transformer as the interstage, and a pair of MJ15024/5 power transistors running off +/-50V. It gave over 70 watts and sounded remarkably good, to me at least. Like the original TV output stage it has a high output impedance and hardly damps the speaker: I had to add a damping resistor because the presence was too overpowering otherwise.

BTW, it was these experiments that convinced me that there's nothing fundamentally bad sounding about transistors.

[R.G.]

As if those weren't unreliable...

Oh, yeah!
I've pondered how to do some protection for those for a while. That was the origin of my "just turn off the power supply" protection scheme, which uses saturated high current power MOSFETs in each power rail to allow power to get to the output stage, plus any kind of protection sensing you like to whack the latch and turn the rail switches off when there is a disaster. It works pretty well, and also eliminates the issues with speaker relays if you use it for that kind of thing, as well as eliminating problems with fuses in the power rails if you sense current and use it to stop disastrous currents.

It's odd that this vintage circuitry stuff is now a big fad when back in the days people were screaming with joy to get rid of it.

Perverse, anyway. But then there ARE humans involved. Humans are fundamentally perverse. 8-)

If Thomas Organ (or basically any other other manufacturer) had had access to chip amp designs of today they undoubtedly would have laughed at the idea of using a monstrosity such as interstage driven totem pole output - with Ge transistors.

Nevertheless, despite the advancements in technology, some designs still are poor.

Yep. People haven't changed since the 1960s. Just because a wet-behind-the-ears engineer has better toys to play with today does not make him design any better.

Inspired by RG's writings on the Thomas Organ amps,

AACK!! Ohmigod!! I'm not sure I'd have ever picked out that first word! 8-)

[gbono]

Maybe after more than 40 years we can actually do something different if we want to use those interstage thingies... Think about it.

Mmmm can you say green, hemp based, etc, etc

And I personally would fire an EE for turning in a design that had the nominal power supply running over the absolute max for the parts without a very convincing argument from him/her about why the design was realiable in spite of the manufacturer's warning on his datasheet.

Many years ago when I worked at Hewlett Packard your design was reviewed by the Materials dept and every single component had to meet corporate derating guidelines - those guidelines went the way of the DoDo bird by the end of the 80's.

[Enzo]

Derating sounds good to me. I hear far too many examples of people thinking that getting away with something means it is good practice.

[Robert M. Martinelli]

This is of course true, but why sadly? Left to their (or do I mean our) own devices, engineers would spend forever trying to create the perfect product. Putting a cap on the bill-of-materials cost is one way to force a reality check. Imposing a deadline is another......

....I certainly wouldn't put 27v on a chip rated for 25 maximum, though. As an end user not constrained by BOM cost you can replace the poor TDA2050 by anything you like. I like RG's suggestion of the LM3886: this chip is used in hundreds of DIY hi-fi amps, so you can easily find information on how to hook it up. I've seen people just tack all the components on to the chip's leads without any sort of a PCB.....

PS: Wasn't Leo Fender famous for running tubes over their ratings? And Tom Jennings with his AC30 that you can fry eggs on top of. Good job they're more forgiving than semiconductors.

Hi Steve,
I used the word "sadly" because, even if I understand "real world" issues like "BOM constraints" and "deadline", I think that even if very important, these issues should not be considered so important to come to the point to exert a negative influence on a certain design.

I'd even bend this stiff neck of mine and understand if EEs and bean counters would be considered equally important, but, in today's world the latter have become the ones who take the final decisions, and I can't help myself in perceiving this as a negativity....

There must be a "sweet spot" somewhere in between....only, I think nowadays manufacturers are so "profit" and "budget" oriented that no one is really interested in finding it.

As to running ICs ( or any other component ) beyond their SOAs, this also seems to be a widespread behavior, maybe Enzo's right when he says that "getting away with it" has come to be considered good practice....Or ( and someone will say that getting old I'm getting nasty ) maybe it's only a practice aimed at creating some "aftermarket" business ( If the d@mn thing survives the warranty period ).

"O tempora, o mores", or should I say "mala tempora currunt".... ( approximately translated from Latin : "Bad times, bad behaviors" and "We live in bad times" ).

+1 on the 3886.... I actually used a 3876 ( 56 W ) in my C30, together with an oversized toroidal PT ( to properly feed the "current hungry" 3876 ), an oversized ( 8A ) bridge, PS "overfiltering", socket mounted 5532s and an VTL5C3 as the "trem roach". 56 W(RMS) are more than enough for almost every situation.... I didn't use the 3886 ( 68 W ) to be able to keep the original speakers, otherwise the mods would have cost me more than what I paid for the amp...

What you say about Leo Fender and Tom Jennings ( Dick Denney ) it's true, but we have to take into account that tubes in the old times were far more rugged and reliable than modern production ones; If a Mullard ( or Philips, or Telefunken ) 1960 data sheet said 12W Max PD I know I could have probably taken that tube up to 14 W with no problems at all; if a "modern" data sheet says 12W I know that "modern" tube will probably fail at 11 .

Cheers

Bob

[Steve Conner]

Corrupted by RG's writings on the Thomas Organ amps

Better?

Again, don't get me wrong, I don't think any amount of BOM constraints are an excuse for running a part over its absolute maximum ratings. That's just plain bad engineering. (Except tubes in an AC30, in which case it's rock'n'roll.) My gut feeling is that they wanted the "30" in the name to summon some AC30 mojo, so they needed 30 watts even if that meant pushing the power amp chip a little too hard.

I think this cost-driven engineering is part of what got solid-state its bad reputation. When manufacturers were trying to drive BOM cost down, the first thing to go was the tubes. So solid-state amps have mostly been cheap amps, with cheap cabinets and cheap speakers. In other words, cheap alternatives to tube amps rather than an artform in their own right. I often wonder what a boutique solid-state amp could be like, if it was built with the same care and attention that people here put onto high-end tube gear, and the current state-of-the-art knowledge in tone. Maybe we'll never know because the high-end MI market is too conservative to accept it.

Talking of which, Bob, how does your uprated Cambridge sound?

[Robert M. Martinelli]

I think this cost-driven engineering is part of what got solid-state its bad reputation. When manufacturers were trying to drive BOM cost down, the first thing to go was the tubes. So solid-state amps have mostly been cheap amps, with cheap cabinets and cheap speakers. In other words, cheap alternatives to tube amps rather than an artform in their own right. I often wonder what a boutique solid-state amp could be like, if it was built with the same care and attention that people here put onto high-end tube gear, and the current state-of-the-art knowledge in tone. Maybe we'll never know because the high-end MI market is too conservative to accept it.

Talking of which, Bob, how does your uprated Cambridge sound?

I absolutely agree, you'll surely remember we've been discussing about some "good sounding" SS amps in another thread; well, owning both a '64 AC30TB and an AC30SS, though neither as loud as the "original" nor similar in sound, I have to say I love the way the latter sounds.

Now, about my "uprated" C30, I still don't like the "tube" channel very much, I find it too "crunchy" ( probably some food-for-thought for another mod );I prefer to use an external OD stompbox ( a TS clone with upgraded components ) using the "clean" channel.

The main improvements I have been able to notice are about "clean" headroom ( thanks to the "spare" power I can count on now ), "dynamic" behavior ( thanks to the PS capability to deliver higher currents and the filter caps capability to support "dynamic" picking ), better trem performance ( partly due to the better opto, partly due to a careful calibration of VR11 and 12 ), and better noise ( hiss ) performance due to the use of 5532s where once 4558s were.

Cheers

Bob

[R.G.]

I'd even bend this stiff neck of mine and understand if EEs and bean counters would be considered equally important, but, in today's world the latter have become the ones who take the final decisions, and I can't help myself in perceiving this as a negativity....

There must be a "sweet spot" somewhere in between....only, I think nowadays manufacturers are so "profit" and "budget" oriented that no one is really interested in finding it.

Excellent point. I'd also fire an EE who consistently neglected economic considerations in his designs. The whole point of being an engineer is that you're trained to do the complete job, not just easter-egg in parts until you like the sound (or whatever). Engineering is a profession, and engineers should be professional. That's a much higher calling in many senses than just being capable at something or other. As a professional, you're required to do it all.

There is indeed a sweet spot in almost all designs; what artistry there is in electronic design lies in recognizing sweet spots and designing to them. Any fool can "design" with parts made of platinum, solid gold and hand picked/crafted semiconductors. Getting the same result out of copper and multiply-sourced jellybean parts takes a disciplined and alert outlook.

Corrupted by RG's writings on the Thomas Organ amps

Better?

Ah, thanks! Much better. 8-)

[Steve A.]

Excellent point. I'd also fire an EE who consistently neglected economic considerations in his designs...

You keep talking about firing people for making poor judgment calls. I believe that in cases like that they generally qualify for unemployment which usually comes out of your company's pocket. (That is how it is in California.)

It is much better to encourage someone to quit. Perhaps have a "friendly competitor" offer them a job at a substantially higher salary so that they give notice, and guess what- the new job was just eliminated in the latest bout of layoffs. Gee whiz, we are awfully sorry...

Steve Ahola

[vipersassasin09]

"+1 on the 3886.... I actually used a 3876 ( 56 W ) in my C30, together with an oversized toroidal PT ( to properly feed the "current hungry" 3876 ), an oversized ( 8A ) bridge, PS "overfiltering", socket mounted 5532s and an VTL5C3 as the "trem roach". 56 W(RMS) are more than enough for almost every situation.... I didn't use the 3886 ( 68 W ) to be able to keep the original speakers, otherwise the mods would have cost me more than what I paid for the amp..."

how would you modify the C30 to take the 3886? or can you just put it in there?