There has been a lot written about the power supply for the Gain Clone amps. Typically they give about 50 watts into 8 ohms per channel. The power transformer recommended is about 300VA.
You will find that the Creek 5350SE amp rated at 85 watts per channel into 8 ohms has a 250 VA transformer ! The amplifier is highly rated and many guys are going really ga-ga over it.
Are we missing something here. If so can the enlightened let us know what gives. Does anyone know what the circuit ( of the 5350SE) looks like?
Does it mean a very high PSRR will also mean that one can get away with not only smaller power supply caps but also smaller transformers ? I find that hard to comprehend if ones needs good performance at the deep end.
Cheers,
Ashok.
You will find that the Creek 5350SE amp rated at 85 watts per channel into 8 ohms has a 250 VA transformer ! The amplifier is highly rated and many guys are going really ga-ga over it.
Are we missing something here. If so can the enlightened let us know what gives. Does anyone know what the circuit ( of the 5350SE) looks like?
Does it mean a very high PSRR will also mean that one can get away with not only smaller power supply caps but also smaller transformers ? I find that hard to comprehend if ones needs good performance at the deep end.
Cheers,
Ashok.
Actually a 160VA transformer would do just nicely for a GC.
At +-35V rails it would only slightly be limited into 4Ohm loads.
For power needs of your GC have a look here.
Commercial manufacturers would sell such an amplifier with the 160VA or even less (the less expensive models).
DIYers like to plan their transformer that it keeps up with every load, can play on a permanent basis, with sine waves on the input and has a safety margin to not heat up.
Also the regulation of a big transformer will be much better.
Most cheap commercial amps loaded with a dummy played at their rated max.output would not play for very long....
I work in the repair business and we repair quite a few commercial amps (Pioneer, Sony, etc.) with defect transformers.
Before I rip such an amp apart and use those transformers in my own projects I usually think twice.
In case of the Creek you can get away with around 200VA roughly calculating for an 8Ohm load so a 250VA is OK.
IMO
Jens
At +-35V rails it would only slightly be limited into 4Ohm loads.
For power needs of your GC have a look here.
Commercial manufacturers would sell such an amplifier with the 160VA or even less (the less expensive models).
DIYers like to plan their transformer that it keeps up with every load, can play on a permanent basis, with sine waves on the input and has a safety margin to not heat up.
Also the regulation of a big transformer will be much better.
Most cheap commercial amps loaded with a dummy played at their rated max.output would not play for very long....
I work in the repair business and we repair quite a few commercial amps (Pioneer, Sony, etc.) with defect transformers.
Before I rip such an amp apart and use those transformers in my own projects I usually think twice.
In case of the Creek you can get away with around 200VA roughly calculating for an 8Ohm load so a 250VA is OK.
IMO
Jens
With a 330 VA transformer you do get better regulation, but it also doesn't cost much more than a 200VA if they are bought new.
I think you are right- one person raises the VA a bit "to be sure"
and the next guy raises it a bit more....
The question comes when you have the transformer already or discover a surplus item. 300VA is certainly much more than you need for 2 channels
I have heard the Creek and it is VERY good IMHO.
I will take my gainclone to his house soon I hope for a comparo
I think you are right- one person raises the VA a bit "to be sure"
and the next guy raises it a bit more....
The question comes when you have the transformer already or discover a surplus item. 300VA is certainly much more than you need for 2 channels
I have heard the Creek and it is VERY good IMHO.
I will take my gainclone to his house soon I hope for a comparo
Just my 2 cents.
This Creek 5350 is quite shitty amp,like most amps at it's price.It offers lifeless and not very involving sound.Nowhere near well built GC.Actually it is very far , far away GC even the one passively driven.Not mentioning GC with good preamp.
I stopped reading reviews long time ago.
Bartek
Don't believe the hype
Guys, don't judge what you don't know, what you haven't heard.
Reading magazines is amuzing, but don't take those guys too serious.
I've been shocked too many times with much raved products.
Now... the GC sounds better with 1000 to 1500uf capacitance per rail.
That's why you need a big trafo.
This is very low capacitance for the standards of a 50w amp, and you can't do anything without mucking up the sound.
Well... actually you can.
If you regulate, you can use as much capacitance as you want before the regs, and IMHO you can use a smaller trafo.
And you'll get better sound.
Guys, don't judge what you don't know, what you haven't heard.
Reading magazines is amuzing, but don't take those guys too serious.
I've been shocked too many times with much raved products.
Now... the GC sounds better with 1000 to 1500uf capacitance per rail.
That's why you need a big trafo.
This is very low capacitance for the standards of a 50w amp, and you can't do anything without mucking up the sound.
Well... actually you can.
If you regulate, you can use as much capacitance as you want before the regs, and IMHO you can use a smaller trafo.
And you'll get better sound.
Regulated supplies..........
1. With very small supply caps , the supply impedance would be higher. Ripple would be higher and the effective DC of the power supply would swing up and down with the fluctuating load current particularly at lower frequencies.
2. With a regulated supply , the supply impedance will be very low , and there would be practically no ripple and DC voltage should be as steady as a rock ( just about ).
3. With a very large supply cacacitor the supply impedance will be low , the effective DC of the power supply will be far more stable than (1) and ripple will be much less.
So why does (1) sound as good as it it supposed to ?
Why does (2) sound even better than (1) when it is an extreme case of (3).
Does anyone have any plausible explanation ? Is it possible to see the effects of the supply on the music signal, like transients etc , when used with a real load. There has to be some meaningful explanation for the differences that people have heard. Has anyone got any scope shots as visible proof of the differences.
Cheers,
Ashok.
1. With very small supply caps , the supply impedance would be higher. Ripple would be higher and the effective DC of the power supply would swing up and down with the fluctuating load current particularly at lower frequencies.
2. With a regulated supply , the supply impedance will be very low , and there would be practically no ripple and DC voltage should be as steady as a rock ( just about ).
3. With a very large supply cacacitor the supply impedance will be low , the effective DC of the power supply will be far more stable than (1) and ripple will be much less.
So why does (1) sound as good as it it supposed to ?
Why does (2) sound even better than (1) when it is an extreme case of (3).
Does anyone have any plausible explanation ? Is it possible to see the effects of the supply on the music signal, like transients etc , when used with a real load. There has to be some meaningful explanation for the differences that people have heard. Has anyone got any scope shots as visible proof of the differences.
Cheers,
Ashok.
Well to go against the 'grain' In the unregulated GC I preffer large (10,000uF) capcitance over the small (1000-2000uF) option.
I did some measurments but took no photo's, with low capacitance I foud ripple to be approx 34% of supply, I found LF motorboating (could have just been down to my cap choice and layout) and 8-10% voltage sag of the supply rails at full o/p.
Adding capacitance improved matters and the sound reflected this. But going regulated put the amp, IMHO into a different league, I had HF stability issues but this was resolved with correct regulator bypassing. At full o/p the voltage is stable at all frequencies, there are now no HF or LF stability issues. With the low cap un-regulated supply the o/p signal is superimposed onto the supply rails at anything over 1W o/p, this becomes a larger problem as o/p rises and frequency rises, I have no figures. The regulated supply in comparison is way better in this respect, the o/p signal is only apparent and in any significat amount until you get nearer full o/p.
I did some measurments but took no photo's, with low capacitance I foud ripple to be approx 34% of supply, I found LF motorboating (could have just been down to my cap choice and layout) and 8-10% voltage sag of the supply rails at full o/p.
Adding capacitance improved matters and the sound reflected this. But going regulated put the amp, IMHO into a different league, I had HF stability issues but this was resolved with correct regulator bypassing. At full o/p the voltage is stable at all frequencies, there are now no HF or LF stability issues. With the low cap un-regulated supply the o/p signal is superimposed onto the supply rails at anything over 1W o/p, this becomes a larger problem as o/p rises and frequency rises, I have no figures. The regulated supply in comparison is way better in this respect, the o/p signal is only apparent and in any significat amount until you get nearer full o/p.
Re: Don't believe the hype
Hi Carlos,
When you regulate the PSU things may or can be different. Without regulation you need a big trannie to get sufficient regulation from the trannie indeed but not necessarily for the power needed. When you also use big smoothing caps, the RMS current in the trannie is a lot bigger than the average current due to the very low power factor. It is the RMS current that heats the trannie.
When you regulate electronically those limitations vanishes. You can use a much smaller buffer cap after the rectifier greatly improving powerfactor. You only need to have sufficient voltage in the “valleys” at full load so the regulator stays regulating and don’t starve off. If you do that properly me think a 100 VA trannie will suffice for a stereo pair of GC’s.
Cheers
carlosfm said:Now... the GC sounds better with 1000 to 1500uf capacitance per rail.
That's why you need a big trafo.
This is very low capacitance for the standards of a 50w amp, and you can't do anything without mucking up the sound.
Well... actually you can.
If you regulate, you can use as much capacitance as you want before the regs, and IMHO you can use a smaller trafo.
And you'll get better sound.
Hi Carlos,
When you regulate the PSU things may or can be different. Without regulation you need a big trannie to get sufficient regulation from the trannie indeed but not necessarily for the power needed. When you also use big smoothing caps, the RMS current in the trannie is a lot bigger than the average current due to the very low power factor. It is the RMS current that heats the trannie.
When you regulate electronically those limitations vanishes. You can use a much smaller buffer cap after the rectifier greatly improving powerfactor. You only need to have sufficient voltage in the “valleys” at full load so the regulator stays regulating and don’t starve off. If you do that properly me think a 100 VA trannie will suffice for a stereo pair of GC’s.
Cheers
These LM chips don't sound good with high capacitance on the (unregulated) PSU.
I did listening tests.
Sound becomes slower, bass is dominant and affects the midband and treble.
So, you only have two choices:
1. unregulated, with low capacitance (1000~1500uf per rail).
2. Regulated
The 2 is so much better...
I did listening tests.
Sound becomes slower, bass is dominant and affects the midband and treble.
So, you only have two choices:
1. unregulated, with low capacitance (1000~1500uf per rail).
2. Regulated
The 2 is so much better...
Does the input stage matter ?
I assume that all this power supply details are for straight forward Gain Clones , NIGC or IGC .
What about tube buffered or chip buffered gain clones ?
Do they behave just the same or do they benefit with Reg Power supply ? If so the 47 Lab's Gain Card is history and a new one has to take it's place. What's cooking in Nelson's lab ?
Cheers,
Ashok.
I assume that all this power supply details are for straight forward Gain Clones , NIGC or IGC .
What about tube buffered or chip buffered gain clones ?
Do they behave just the same or do they benefit with Reg Power supply ? If so the 47 Lab's Gain Card is history and a new one has to take it's place. What's cooking in Nelson's lab ?
Cheers,
Ashok.
I was not basing my Creek opinion on the magazine reviews.
A friend got one after comparing it to Musical Fidelity separates, and his previous amp. It really improved HIS system which had LS35a speakers with added woofers. This system was in HIS listening room, which was not like everyone elses. If amps in this price range are shitty, and my friend couldn't spend more, then
I guess he is just stuck with this quality. He isn't a DIY'er and not everyone is one or can be one...
Of course his system may have been shitty, his room shitty, our ears shitty, but I liked it at the time. This was before gainclones became popular, and gainclone kits available, so I haven't compared them. As I mentioned, I'm looking forward to the comparo. Like Carlos, I can't afford even the mid priced components. I 'm hoping my GC will blow the Creek away, so I won't be biased in favor of the Creek!!
A friend got one after comparing it to Musical Fidelity separates, and his previous amp. It really improved HIS system which had LS35a speakers with added woofers. This system was in HIS listening room, which was not like everyone elses. If amps in this price range are shitty, and my friend couldn't spend more, then
I guess he is just stuck with this quality. He isn't a DIY'er and not everyone is one or can be one...
Of course his system may have been shitty, his room shitty, our ears shitty, but I liked it at the time. This was before gainclones became popular, and gainclone kits available, so I haven't compared them. As I mentioned, I'm looking forward to the comparo. Like Carlos, I can't afford even the mid priced components. I 'm hoping my GC will blow the Creek away, so I won't be biased in favor of the Creek!!
Hello Ashok,
You have also relatively rich offer of explanations around. Voodoo oriented, sophisticated_theory oriented, “you are stupid ‘cause you don’t understand” oriented, “you can’t understand that ‘cause you are too stupid” oriented…
You take the one you like.
Pedja
Right.ashok said:
Right.
The range in which large capacitors have low impedance is relatively limited. The rest is correct.
If you ask me it is probably because smaller caps behave better at higher freqs, and you can use them because LM3875 has high PSRR at lower freqs.
You have also relatively rich offer of explanations around. Voodoo oriented, sophisticated_theory oriented, “you are stupid ‘cause you don’t understand” oriented, “you can’t understand that ‘cause you are too stupid” oriented…
You take the one you like.
I do not see much need to explain why the regulated supply is technically better, for the most of parameters at least. Of course, that might not necessarily mean it will sound better, but here it does.
That is wrong, (2) is not extreme case of (3).
Pedja
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.