Transformer substitution on a audio power amp.

[beppe61]

Dear Friends,


I would like to get some opinions from you about a my idea.
I have a power amp of which I quite like the overall sound.
Unfortunately is not able to source the needed current to drive adequately my woofers.
The bass is too weak and soft, also out of control.
I think that the main culprit is an underdesigned the transformer.
My intention is to replace it with a nice and high current model.
I am thinking of something like a Talema 600-800 VA, a new very big diodes bridge and maybe new caps.
Is the modification worthwhile?
It would cost me just about 100 USD in the worst case.

Advice would be very much appreciated.

Kind regards,

beppe61

 

[Tweeker]

Be sure your problem isnt something like an undersized DC blocking input capacitor, or both.

 

[beppe61]

Be sure your problem isnt something like an undersized DC blocking input capacitor, or both.

Dear Mr. Tweeker,
thank you so much for your kind and valuable reply.
Actually looking at the schema I can see a 4,7 uF/50V electrolytic just before a 12K resistor to ground (a C-R high pass filter).
Is 4,7 uF too small a value ?

Moreover if you had then to upgrade the power supply would you start from the transformer or from the filter capacitors?

I look forward hearing from you.

Kind regards,

beppe61

 

[peranders]

The corner frequency of the filter is 2.8 Hz ( f=1/(2*pi*R*C) ) so this is enough when it comes to bass capacity.

What exactly are you not satisfied with? Max level of the bass? Do you drive the amp into clipping?

 

[beppe61]

Dear Mr. Peranders,

thank you so much for your extremely kind and very valuable reply.
Please let me explain better between the lines.

>

Originally posted by peranders
The corner frequency of the filter is 2.8 Hz ( f=1/(2*pi*R*C) ) so this is enough when it comes to bass capacity.

Thank you very much for your explanation.
Nothing to replace then.

What exactly are you not satisfied with? Max level of the bass? Do you drive the amp into clipping?

I will try to explain better.
I have connected two 5 ohm /100W power resistors at the output of the amp to run it in.
The heatsink of the amp got quite warm while the resistors remained cold.
The amp did not go into protection so it was working.
With the speakers connected the amp fails to reproduce with a good impact the lower end of the spectrum (let's say <100Hz).
After all I think that much depends on a underdesignaed power supply.
The first suspect is the transformer, that I have thought to replace with some high-current 600 VA from Nuvotem-Talema.
Maybe a nice pair of big caps should be a nice addition.
The power amp sports two pairs of nice Toshiba per channel.
I am sorry because all things considered I like the sound (bass response apart, of course).

Any your suggestion would be vey welcome and much appreciated.

Kind regards,

beppe61

 

[peranders]

I have connected two 5 ohm /100W power resistors at the output of the amp to run it in.
The heatsink of the amp got quite warm while the resistors remained cold.
The amp did not go into protection so it was working.

100 watts into a resistor gets quite soon, certainly if you use a sinus signal as input. Check first that you really get any output power/voltage AND also that you have 5 ohms!

How many VA has the transformer now and how much smoothing capacitance do you have?

 

[peranders]

The bass is too weak and soft, also out of control.

Before you dive too deeply into the amp, why don't you test your woofer with an another amp, just to see the difference? Maybe nothing is wrong really? It may be the woofer itself

 

[beppe61]

100 watts into a resistor gets quite soon, certainly if you use a sinus signal as input. Check first that you really get any output power/voltage AND also that you have 5 ohms!

Dear Mr. Peranders,

I have 4 power resistors of 10ohm/100W.
I have paralleled them two by two and got 2 power resistors of 5ohm/100W.
The light of the 1W was blinking but was not on steady.
The amp has a thermal protection but the correspondig light remained off.
As the price of the amp is quite low (220 euro) I think that some savings have been done on components, transformer above all.

Thank you very much indeed.

Kind regards,

beppe61

 

[peranders]

If you won't get more than 1 watts, why don't you increase the output? Turn up the volume "Full spätta"*) as we say in Sweden

220 EUR won't get you much but one thing to estimate the power is to measure the supply voltage. Theoreticly the max output power will be

Pmax = (U*U)/(R*2) where U is supply voltage and R is the load.

*) Full throttle but spätta is a flat fish

 

[beppe61]

Re: Re: Transformer substitution on a audio power amp.

Before you dive too deeply into the amp, why don't you test your woofer with an another amp, just to see the difference? Maybe nothing is wrong really? It may be the woofer itself

Dear Mr. Peranders,
the woofers are fine.
They worked nicely with two Albarry mono amps.
These monos however have a much heftier power supply.
I understand that they are known to be difficult to drive adequately.
They are current hungry.
The model is Dynaudio 24W75.
Maybe I should take a look also to the crossover.
I don't expect night and day changements.
I need just more bass, more impact on drums, pipe organ etc.

Thank you so much for your kind and valuable support.

Kind regards,

beppe61

 

[beppe61]

If you won't get more than 1 watts, why don't you increase the output? Turn up the volume "Full spätta"*) as we say in Sweden
220 EUR won't get you much but one thing to estimate the power is to measure the supply voltage. Theoreticly the max output power will be
Pmax = (U*U)/(R*2) where U is supply voltage and R is the load.
*) Full throttle but spätta is a flat fish

Dear Mr. Peranders,

just to clear things, I listened to a classical pieces with low frequencies with the Samson and with the Albarrys.
With the Samson it seemed like that a high pass filter (>100 Hz) was engaged !
A low note with the organ was very much less powerful on the Samson. and I mean very much.
Can the bass response of an amp be realted to the quality of its power supply in general ?
And if so, are caps more determinant than the trasnformer?

Kind regards,

beppe61

 

[peranders]

You may have a problem with "foldback current limitation" which means that the amp has difficulties to drive reactive loads. This has nothing to do with the transformer. The purpose is to protect weak output transistors. It possible that this isn't invoolved in your case, especially if you are listening at low levels. My believe here is that you won't get a new amp with a bigger transformer unless the present one is remarkable small.

Since 80-90% of the power is bass, the power supply has influence of the performance. A stiff power supply is good but costs also money.

Can't you mention the specified output power, estimate the VA rating (how big, physicly is it?) and measure the voltage?

 

[beppe61]

You may have a problem with "foldback current limitation" which means that the amp has difficulties to drive reactive loads. This has nothing to do with the transformer. The purpose is to protect weak output transistors. It possible that this isn't invoolved in your case, especially if you are listening at low levels. My believe here is that you won't get a new amp with a bigger transformer unless the present one is remarkable small.
Since 80-90% of the power is bass, the power supply has influence of the performance.
A stiff power supply is good but costs also money.
Can't you mention the specified output power, estimate the VA rating (how big, physicly is it?) and measure the voltage?

Dear Mr. Peranders,
I think you have pointed out the real problem when you say : "You may have a problem with "foldback current limitation" which means that the amp has difficulties to drive reactive loads".

The amp is specified at 130W/4ohm (but now I doubt this figure).
The speakers are 6 ohm nominal.
The tranformer seems a thing of 400 VA (not particularly small. Adequate sized).

I read once that the secondaries of a good power transformer must be of very low impedance.
Maybe can it be a case of a transformer not suitable for the purpose, with high impedance secondaries?

Maybe I am wrong but the driving ability of reactive loads of an amp is not related primarly to the execution of its power supply?
The available current comes always from the power supply.
Am I miss something ?
The output transistor are Toshiba.
I have to check the part numebr but they don't seem particularly weak.
At the output there are 0,8 mH paralleled to 10 ohm.

Thank you so much.

Kind regards,

beppe61

 

[AndrewT]

Hi Beppe,
give us some measurements to work with.
Vdc on supply rails Vac on output when driving open circuit.
Same two voltages when driving your 5r load.

Two parallel 10r/100W resistors give 5r/200W. This is a non reactive load and your 4 ohm capable amp should easily drive it. In fact a half decent amp should drive a load of half it's rated value (2r) without shutting down.

If your transformer is undersized then you can compensate a little by increasing the number and/or size of the main smoothing caps. However 400VA for a stereo 130W into 4r sounds just about right at +53% of max output power.

 

[beppe61]

Dear Mr.Andrew,

thank you so much for your extremely kind and valuable reply.
Let me please answer between the lines.

>

Hi Beppe, give us some measurements to work with.
Vdc on supply rails Vac on output when driving open circuit.
Same two voltages when driving your 5r load.

- I don't know which kind of signal I have to use.
When I did the test with the resistors I put a Prodigy CD in the player and then set the volume up to a point when the lights indicating about 2 V at the output remained almost always on.
As the load was 5 ohm/channel the power was > o = 1W.
I am not that expert to carry out the measurements you mention.
So I have some questions:
1) which kind of signal do I have to use.
A musical signal is not constant
2) I have to set the digital voltmeter to DC or AC for the measurement across the resistors?
3) I am afraid that if there is no load (open circuit) the relays disconnect the output.

> Two parallel 10r/100W resistors give 5r/200W.
- Yes I did this on each channel.

> This is a non reactive load and your 4 ohm capable amp should easily drive it. In fact a half decent amp should drive a load of half it's rated value (2r) without shutting down.
- The problem is that the resistors remained cold, indicating no current flowing through them, but the heatsink became quite hot.
And the amp was working because the light of the protection remained off.
In a similar test with a Panasonic digital receiver, the receiver became very hot but also the resistors !

> If your transformer is undersized
- maybe is not undersized but not well designed.
Maybe it is just not able to source good amount of current from the secondaries (high impedance secondaries) or something like that. A case of bad designed driven by economical reasons.

> then you can compensate a little by increasing the number and/or size of the main smoothing caps.
- That is indeed the other mod I have in mind.
the current caps are 10.000uF/63V made by Shawa o shaowa or something like that.
They are not Mallory or Cornell Dubilier for sure.
I am thinking to trace some nice 15.000 uF computer grade like Mallory or Sprague.
I understand that they are the best from a lot of point of views.
What do you think of this idea? or it is better to replace first the transformer anyway ?

> However 400VA for a stereo 130W into 4r sounds just about right at +53% of max output power.

- It is my feeling that the speakers are a quite demanding load for the amplifier.
I heard them sing properly with an english amp (Albarry monos) that had a very substantial power supply.

To end my ramblings, do you think that the mod should adress first the transformer or the filter caps?
My feeling is that the bottleneck is the transformer not able to source good current.
80 euro will buy me a nice 600VA Nuvotem-Talema high current toroidal.

I look forward to hearing from you.

Kind regards,

beppe61

 

[peranders]

I read once that the secondaries of a good power transformer must be of very low impedance.
Maybe can it be a case of a transformer not suitable for the purpose, with high impedance secondaries?

Maybe I am wrong but the driving ability of reactive loads of an amp is not related primarly to the execution of its power supply?
The available current comes always from the power supply.

I think you must first establish if there is any major difference between your amp and some other one with equal output power. As I see it your are thumbling in the dark a bit. If you now play at 1 W with a 400 VA and want to change it to 600 VA, it will hardly solve your problem.

A transformer is made always with a couple of trade offs. You usually optimize the weight meaning to use the iron core as much as possible and use a little copper as possible in the windings. You can't really make an low impedance transformer if you want to use the core to 100%. More VA => less impedance

The best thing you can do right now when you have such doubts is to get a good amp from a friend, let this friend connect the amps without telling which one you are listening to. Do a semi scientific blind test. The risk as I see it is that you tweak the amp broken in the hunt of the perfect sound. You could actually build yourself a Gainclone from me, BrianGT, Rod Elliott or somebody else to see if there is a difference. You could also buy a hypex module

 

[beppe61]

Dear Mr. Peranders,
thank you very much indeed for your always kind and extremely valuable reply.
I will try to answer to your question betwwen the lines.

Originally posted by peranders
I think you must first establish if there is any major difference between your amp and some other one with equal output power.

Definitely yes! Actually it was this test that started all my idea.
We compared this amp and two Albarry monos specified at 50W/8ohm and 90W/4 ohm (maybe a little conservative, don't know).
The bass with my speakers powered by the Albarry was cavernous and hugely powerful.
Actually it was the first time I heard that kind of bass from my speakers (the fs of the woofer is 33Hz anyway and probably they can arrive quite low if correctly driven).
In on other occasion I had the opportunity to hear a track of pipe organ with a 200W/8 ohm power amp and then with a low powered KRELL KSA50. eh, eh, eh, The "bigger" 200 Watter turned pale.
with the Krell I understood what high current, driving ability, ecc. mean.

> As I see it your are thumbling in the dark a bit.
Definitely yes again. I am not an expert in audio electronics.

> If you now play at 1 W with a 400 VA and want to change it to 600 VA, it will hardly solve your problem.
A transformer is made always with a couple of trade offs.
You usually optimize the weight meaning to use the iron core as much as possible and use a little copper as possible in the windings. You can't really make an low impedance transformer if you want to use the core to 100%. More VA => less impedance

Thank you very much for your extremely valuable explanation.

> The best thing you can do right now when you have such doubts is to get a good amp from a friend, let this friend connect the amps without telling which one you are listening to. Do a semi scientific blind test.

Actually I did this. I can assure the differences in the bass response were nothing but astonishing.
Incidentally, as these two monos have a red plexiglass front panel, I could see a big toroidal and two really big caps for each one. That also made me think.

> The risk as I see it is that you tweak the amp broken in the hunt of the perfect sound.
It is indeed a very real risk.
It is a pity because for the upper part of the spectrum, let's say above 100 HZ I quite like the sound of the amp.
I am asking for just a little more under 100 Hz.

> You could actually build yourself a Gainclone from me, BrianGT, Rod Elliott or somebody else to see if there is a difference.
You could also buy a hypex module

These are indeed nice options.
May I kindly ask you just two other question?
Which brand and model of transformers do you recommend to power your Gainclone project?
As a friend of mine has a impedance meter, what do you think could be a acceptable value for the impedance of the secondaries?
I ask you these because we actually measured two different transformers once, more or less with the same VA rating.
One measured about 1.5 ohm and the other 300 ohm (at 60 Hz).
I think that the first should recharge more easier the filter caps.

Am I saying a sillyness?

Thank you very much indeed again for your extremely kind and appreciated support.

Kind regards,

beppe61

 

[peranders]

A hint: Quote only the text you want to comment, otherwise you produce so much text.

If you use toroid transformers you have pretty many brands to choose from. Personally I prefer transformers from Toroid International AB, now changed name to Noratel International AB. A qaulity sign may be smooth windings, signs of good production technology.

About the measurements, 300ohms at 60 Hz, not very likely to be true. 1 A out => 300 V in voltage drop!

 

[beppe61]

A hint: Quote only the text you want to comment, otherwise you produce so much text.
If you use toroid transformers you have pretty many brands to choose from. Personally I prefer transformers from Toroid International AB, now changed name to Noratel International AB. A qaulity sign may be smooth windings, signs of good production technology.
About the measurements, 300ohms at 60 Hz, not very likely to be true. 1 A out => 300 V in voltage drop!

Dear Mr. Peranders,
thank you so much for your extremely kind and very valuable suggestion.
Maybe I am wrong about the impedance measures but I remember very well the difference was very remarkable indeed.

Thank you very much again for your kind and very helpful support.

Kind regards,

beppe61

 

[AndrewT]

Hi Beppe,
Buy yourself a CD test disc. It has lots of signals on it that can be used to test and/or evaluate your system. If you are good with computer sound cards you can generate test signals here but it will miss out part of the reproduction system.

Buy yourself a digital multimeter or two or even three. They are cheaper than the test disc, one more flexible and more expensive and two very cheap ones.
Now go and read about how to measure AC voltage and DC voltage. Alternatively get someone to demonstrate the correct and safe way to measure voltages.

By the way that impedance meter gave rubbish results. Either it was faulty or you used it incorrectly.
The resistance of the secondary windings of a power transformer are usually less than 1.0 ohm often a lot less. I said resistance not impedance.