airlink are generally well regarded in the uk.
To be honest, I think the 340a se is probably a pretty good place to start building a GC. It's £200 new, but that includes a case, all the inputs/outputs you need, a transformer, remote volume and input switching control etc not to mention an input buffer and a basic GC with standard parts. If the quality and feel of the unit is comparable to other modern CA kit, its a really good price and they usually come with a pretty nice and substantial feeling remote too.
Thinking about it, I can't think of a much better starting point for someone who wants to build a GC with all of those features that many DIY designs lack. Saves you a boatload of time too!
To be honest, I think the 340a se is probably a pretty good place to start building a GC. It's £200 new, but that includes a case, all the inputs/outputs you need, a transformer, remote volume and input switching control etc not to mention an input buffer and a basic GC with standard parts. If the quality and feel of the unit is comparable to other modern CA kit, its a really good price and they usually come with a pretty nice and substantial feeling remote too.
Thinking about it, I can't think of a much better starting point for someone who wants to build a GC with all of those features that many DIY designs lack. Saves you a boatload of time too!
Thanks.
Peter Daniel has kindly replied to an eamil I sent. He has confimed that I can easily modify the Power amp stages to the same configuration as his LM3875 config. I can also use his PSU with the addition of a new Trafo but I prefer the chipamp.com PSU due to its larger capacity and easy mounting. However, if the Mr Daniels PSU is better I will have to find a way to mount it.
I'll be asking his advice on the PSU.
If I do all this then I will be looking into a PSU re-design for the opamp, input selector and microprocessor, possibly local regulation for each section of a dedicated trafo. For the regs I'll be looking at various low noise options such as the Sercal Spower or PFM Teddyreg (I have some Teddyreg PCB's allready).
As I never use the tone controls Im thinking of removing all the circuitry here. That would leave me with 2 empty holes in the facia so how about a pair of dinky VU meters? Something about 1" dia and preferrably with blue backlight
. Where would I find these things? I've tried google but found nothing yet.
Peter Daniel has kindly replied to an eamil I sent. He has confimed that I can easily modify the Power amp stages to the same configuration as his LM3875 config. I can also use his PSU with the addition of a new Trafo but I prefer the chipamp.com PSU due to its larger capacity and easy mounting. However, if the Mr Daniels PSU is better I will have to find a way to mount it.
I'll be asking his advice on the PSU.
If I do all this then I will be looking into a PSU re-design for the opamp, input selector and microprocessor, possibly local regulation for each section of a dedicated trafo. For the regs I'll be looking at various low noise options such as the Sercal Spower or PFM Teddyreg (I have some Teddyreg PCB's allready).
As I never use the tone controls Im thinking of removing all the circuitry here. That would leave me with 2 empty holes in the facia so how about a pair of dinky VU meters? Something about 1" dia and preferrably with blue backlight
. Where would I find these things? I've tried google but found nothing yet.
Hi Mike,
Tone controls are not bad to have unless the tone circuit is poorly designed. You can make them switchable too. There are times when you may not agree with the recording engineer.
Just adding peak lights might be a good alternative if you are set about removing the tone controls. Or two LEDs, one for signal and one for clipping. Just a thought.
Finally, try to keep perspective on what actually makes a difference and what changes are only speculative or theoretical. Rectifiers really shouldn't affect the sound quality for instance. Not unless something is happening that shouldn't. You can always experiment later on a kit where you have complete freedom and control your variables.
-Chris
If the audio is sampled properly - no. You would be using a meter drive circuit, not just diodes and the meters (as the older Vu meters were often connected).
Tone controls are not bad to have unless the tone circuit is poorly designed. You can make them switchable too. There are times when you may not agree with the recording engineer.
Just adding peak lights might be a good alternative if you are set about removing the tone controls. Or two LEDs, one for signal and one for clipping. Just a thought.
Finally, try to keep perspective on what actually makes a difference and what changes are only speculative or theoretical. Rectifiers really shouldn't affect the sound quality for instance. Not unless something is happening that shouldn't. You can always experiment later on a kit where you have complete freedom and control your variables.
-Chris
Thanks. The tone circuit allready has a 'direct' switch which is always on. The tone control circuit isnt in the audio path but its drawing power and possibly dirtying the supply to the preamp section. Just a theory. I like to listen to music as the engineer intended it to sound anyway. I could just leave it there and disconnect the power to the opamp. But I do like the idea of annalogue VU meters with blue backlight, kinda like the Marantz power amps - purely from an aesthetic point of view. I had another little search on Google and couldnt find what I was looking for so I'll probably shelve that idea for now. Besides its whats inside that really counts.
I have decided to go for the new trafo but I cant decide which PSU. I read that the smaller capacitance of Peter Daniels approach gives a better midrange wheras the Carlos FM PSU seems to be better for bass. I need to get a good idea of how each would sound. The rest of my sytem is in detailed in the signiture below. A word on he TDL's, they are sensitive and quite revealing. Does this suggest the Carlos FM design might be better suited?
I have decided to go for the new trafo but I cant decide which PSU. I read that the smaller capacitance of Peter Daniels approach gives a better midrange wheras the Carlos FM PSU seems to be better for bass. I need to get a good idea of how each would sound. The rest of my sytem is in detailed in the signiture below. A word on he TDL's, they are sensitive and quite revealing. Does this suggest the Carlos FM design might be better suited?
Andrew,
I read up on both PSU designs and I cant see what your referring to. Maybe I missed it or havent seen the right thread. If you dont mind could you explain why you are leaning toward the Carlos FM design in regard to my speakers? Or at least post a link so I can find out for myself. Thanks.
Mike.
My suggestion would be to build the simple cheaper supply first and spend some time listening to it. Once you are familiar with it's sound build a regulated supply and have a listen to that for a while and see which one lights your candle.
At least you will get to listen to a working amp for a while and see where you want to go from there.
John
At least you will get to listen to a working amp for a while and see where you want to go from there.
John
Hi Mike,
Stay away from high capacitance in your power supplies. This will cause more trouble than it solves. I think Andrew and I strongly disagree on this point. Higher capacitance over what is normal will not deliver better bass. The higher capacitance units normally have greater inductance, this means less effectiveness at higher frequencies.
The tone circuit is not driving anything, so it will have a very close to constant current draw. No worries. Even if it did put a squiggle on your supply rails, this is nothing compared to what everything else is doing. This is what I mean by keeping things in perspective.
It has been suggested that you install local bypass capacitors. This will help more than a power supply upgrade. If you can get your hands on some 10 uF film caps, those will give you some great bypass effect. Also tack in some 0.1uF ceramic capacitors. Watch as some are rated for only 50V. I have found that the 300 ~ 500 volt caps are better. 100 volt units would be less expensive in your application. I view very low voltage capacitors (for a family) with great suspicion.
-Chris
Edit: Hi John,
Excellent suggestion. I second that motion.
Stay away from high capacitance in your power supplies. This will cause more trouble than it solves. I think Andrew and I strongly disagree on this point. Higher capacitance over what is normal will not deliver better bass. The higher capacitance units normally have greater inductance, this means less effectiveness at higher frequencies.
The tone circuit is not driving anything, so it will have a very close to constant current draw. No worries. Even if it did put a squiggle on your supply rails, this is nothing compared to what everything else is doing. This is what I mean by keeping things in perspective.
It has been suggested that you install local bypass capacitors. This will help more than a power supply upgrade. If you can get your hands on some 10 uF film caps, those will give you some great bypass effect. Also tack in some 0.1uF ceramic capacitors. Watch as some are rated for only 50V. I have found that the 300 ~ 500 volt caps are better. 100 volt units would be less expensive in your application. I view very low voltage capacitors (for a family) with great suspicion.
-Chris
Edit: Hi John,
Excellent suggestion. I second that motion.
anatech said:
Hi Chris,
Peter uses 1500uF per rail in his chip amps. I have tried this and did not like the result. It was fine at low volume, but bass response did suffer at higher volumes.
Isn't a better approach to use what can be considered the proper amount, not under-doing it or overdoing it?
I have used 3-4700uF caps per rail and find that to be very good.
Hi John,
I had no idea what Peter or anyone else uses. 1,500 uF is very undersized I would think. As you pointed out, obvious trouble starts at higher current loads.
The old rule-of-thumb used to be 1,000 uF per ampere of current draw. You would really want to look at how much ripple you have under load, also how much signal is riding on the supplies. This will then depend on what you are doing and seems like the most realistic approach. Using monster cap values will cause other troubles.
I once was asked to service a large Bedini amplifier. The caps ran from the front of the chassis to the rear. They were so large that there was high AC ripple on the supplies. The caps were not bad, that is a different wave form. At any rate, I did repair it. I couldn't believe the very poor build quality though! This amp was an excellent example of what happens when the filter caps are way too large!
-Chris
Absolutely!
I had no idea what Peter or anyone else uses. 1,500 uF is very undersized I would think. As you pointed out, obvious trouble starts at higher current loads.
The old rule-of-thumb used to be 1,000 uF per ampere of current draw. You would really want to look at how much ripple you have under load, also how much signal is riding on the supplies. This will then depend on what you are doing and seems like the most realistic approach. Using monster cap values will cause other troubles.
I once was asked to service a large Bedini amplifier. The caps ran from the front of the chassis to the rear. They were so large that there was high AC ripple on the supplies. The caps were not bad, that is a different wave form. At any rate, I did repair it. I couldn't believe the very poor build quality though! This amp was an excellent example of what happens when the filter caps are way too large!
-Chris
Hi Redshift187,
Read the rest of my post please.
There are industry norms you can look at to figure that out. Don't bother looking at products that make a big stink about the size of their capacitors. That would be an advertising gimmick. The intended application also may make a difference.
Let's just say that a preamp using 4,700 uF main capacitors is abnormal. That is excessive, wouldn't you think? There are no hard and fast rules and everything is a compromise. Greater capacitance values do have some negative effects. You have to balance these off with the positives. Electronic regulation in a preamp will do a better job than giant capacitors. In a power amp, regulating the voltage amp stages is far better than just using giant caps. The trade off is improved performance, lower shipping weight and extra cost in components for a reasonable capacitance against noisier supplies, heavier capacitors and more expensive capacitors. The only plus I can see here is the ad copy extolling the false virtues of higher capacitance. After all, they wouldn't spend the money if it weren't needed, would they?
-Chris
Read the rest of my post please.
There are industry norms you can look at to figure that out. Don't bother looking at products that make a big stink about the size of their capacitors. That would be an advertising gimmick. The intended application also may make a difference.
Well, yeah! Of course it is.
Let's just say that a preamp using 4,700 uF main capacitors is abnormal. That is excessive, wouldn't you think? There are no hard and fast rules and everything is a compromise. Greater capacitance values do have some negative effects. You have to balance these off with the positives. Electronic regulation in a preamp will do a better job than giant capacitors. In a power amp, regulating the voltage amp stages is far better than just using giant caps. The trade off is improved performance, lower shipping weight and extra cost in components for a reasonable capacitance against noisier supplies, heavier capacitors and more expensive capacitors. The only plus I can see here is the ad copy extolling the false virtues of higher capacitance. After all, they wouldn't spend the money if it weren't needed, would they?
-Chris
anatech said:
Hi again Chris,
This would have to do with ripple reduction, not so much for reserve power for bass transients, right? Still, how much improvement can be expected from a regulated front-end? Earlier in this thread you suggested that it should be a regulated separate supply from the output stage. Is this truly beneficial?
AndrewT advocates capacitance based on the expected load that the amp will drive - more for a lower impedance load such as 4 ohms as opposed to less for 8 ohms. This makes sense to me as the caps act as a short term storage battery, supplying the needed current for the low impedance. A high capacitance supply would be referred to as a "stiff" supply, correct?
Hi John,
-Chris
Yes and yes. But more importantly, the regulation reduces all noise on teh supplies. It shields the amplifying stages completely from garbage on the AC supply and sags during power peaks. From experience I can tell you that this works very well.
Yes. You need a higher voltage for the voltage amp supply so you have some headroom when the voltage sags and to drive the output stage to full power. The regulated output may only need to be 6 volts or less over the output supplies.
This is true, but Andrew and I differ in how much capacitance you need and how to determine what you need. I don't know who is actually right here. Maybe it's somewhere in between Andrew's ideas and mine.
Yes, however you only need to store enough energy to cover peaks. If you run maximum sine wave outputs, your capacitance will need to be higher. Real music is not like that at all.
Depends. At what frequencies? Larger capacitors have more inductance and other factors. I don't see any benefit from going over what is found to be an optimal capacitance in a particular design. I wonder if that can be found out by increasing the supply capacitance while watching the low frequency THD for a drop.
-Chris
Probably not. THD is only one part of the equation. As you said, it is all about finding the best compromise, and when it comes to bass, taste seems to matter much more than any in other area of sound reproduction.anatech said:
The fun of DIY is that you can find the compromise that suits your personal taste best. Commercial products have to find a very different compromise
Higher capacitance does not only influence the amount and quality in the lower regions. It also makes the bass notes lag behind the rest of the music. Most people will not notice that at all. If you play music in a band or an orchestra however, this effect can be very disconcerting in comparison to your live experience, where everybody strifes for perfect timing.
The Cambridge allready useds 10'000uf per rail and the only complaint I have is that the bass could be a bit tighter. Apart from that its fine. Dosent the VA of the trafo have an effect on bass too? What does a snubber do and how does that effect SQ?
A good compromise would be something like 4700uF 50/63v caps per rail with 100uf's at the chip. In this case either PSU would be suitable as they can accept a wide variety of caps and each can be used with or without the snubber. I believe Peters PSU is the most flexible but I prefer the mounting of the chipamp.com one....
A good compromise would be something like 4700uF 50/63v caps per rail with 100uf's at the chip. In this case either PSU would be suitable as they can accept a wide variety of caps and each can be used with or without the snubber. I believe Peters PSU is the most flexible but I prefer the mounting of the chipamp.com one....
The transformer in the CA will almost certainly be something like 120VA or as low as they can get away with. You would probably do better with 200VA or so.
My own LM3886 based amp uses 2x4700uF per rail, with 330uF and 100nF decoupling near the chip. For this reason I don't refer to it as a "Gainclone" as these normally have relatively small smoothing capacitors (2200uF per rail or so) on the same PCB as the amp itself. My arrangement is more like a conventional power amp. It has a 160VA transformer (Rapid were out of 225VA at the time!) and a regular 35A bridge rectifier bypassed with 100nF ceramics.
My own LM3886 based amp uses 2x4700uF per rail, with 330uF and 100nF decoupling near the chip. For this reason I don't refer to it as a "Gainclone" as these normally have relatively small smoothing capacitors (2200uF per rail or so) on the same PCB as the amp itself. My arrangement is more like a conventional power amp. It has a 160VA transformer (Rapid were out of 225VA at the time!) and a regular 35A bridge rectifier bypassed with 100nF ceramics.
What about using several smaller caps to make up a higher capacitance? For example 5 x 2'000uF per rail, or even a variety of sizes? Wouldnt this give the best of both worlds in terms of bass and midrange?
A good freind who knows my system has thrown this suggestion into the mix:
4700uF
4700uF
4700uF
470uF
47uF
Total = 14617uF per rail.
Any comments?
A good freind who knows my system has thrown this suggestion into the mix:
4700uF
4700uF
4700uF
470uF
47uF
Total = 14617uF per rail.
Any comments?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.