450VA/230Vac = 1.94Aac
Try T2A and use a 50r to 60r of soft start resistor, or 4series connected CL60
The UK has the EU harmonised supply voltage.
But in practice we still see a 240Vac average which tends to fall slightly during periods of high demand and increases quite a lot during periods of low to very low demand.
Expect around 245Vac during later evening listening.
I follow your logic but I'm surprised that Naim have gone with a T3.15A if it is inappropriate.
Finally finished construction tonight, see attached image.
I plugged it in (in series with a 60W bulb), switched on the power and.... completely uneventful, all was well!
There's about 7mV DC offset at each of the speaker outputs and I've set the bias to around 30mA.It's relatively late here now and I've only just finished so I've not really given the unit chance to fully warm up and listen properly (everyone else in the house is asleep), however the first impressions are of a very detailed sound.
I'm used to the sound of a genuine Nait 3 but, to my ears, this NAP 200 clone sounds a lot more detailed and controlled. I'm still using the Nait 3 as the pre-amp, and will be doing for a while, but I'm sure an upgrade will come at a later date.
All in all a success, I feel.
There's a little internal tidying and cleaning to be done yet, and I want to replace the pre-amp PSU reservoir cap (currently using the old one taken out of my Nait 3 during a recent service), but these things can wait.
It will be interesting to see how others judge their versions of the amp when completed.
Attachments
Thanks!
I used a custom transformer that I had built by a company in the UK - see this post for the spec. The intention was to match the part used in a genuine NAP 200 as closely as possible.
The case is one of these, I had it shipped over from Italy.
All PCB parts are as supplied in the kit, the output devices do appear to be genuine parts. I'm sure improvement by substitution can be made, but that can wait for another day.
I had to purchase additional components to populate the section of the board that powers an external pre-amp as these weren't included.
The output device-to-case heat transfer assembly was made out of a couple of pieces of aluminium bar cut to size - really not tricky at all. The plastic covers over the two pairs of transistors to the left of the board are just the end-caps off an HDMI lead.
Chris
Last edited:
Doesn't the particular design of the delayed fuse and the starting current peak/duration etc have a bearing on how it long it takes to fail? Users are faced with a confusing array of slow-blow types, some of which are scarcely different to plain wire so I think this is particularly important for DIYs.
At 8,200 uF, even the main power supply smoothing caps aren't exactly over-sized and I doubt the OEM transformer is rated more than 400VA total. For comparison, I believe the NAP140 toroid was around 230VA. I haven't seen the original nor read any specification, dimensions, mass etc. for a proper comparison but the usual presumption that bigger is always better for power supplies, is not often correct nor optimum for class AB designs, as far as SQ is concerned.
I thought it a good compromise. Keep a bag of fuses and don't think anything is wrong when you have to replace them. This one seems about right. The ceramic body helps safe removal.
MULTICOMP - MC000834 - FUSE, 5X20MM, CERAMIC, TIME DELAY, 3.15A | CPC UK
One thing I haven't tried is a 2R 5 watt resistor in series with a 2AT fuse as I think Andrew intended. The bit I haven't tried is a switch that is thrown a second or two latter accross the resistor to loose the possible degrading effects. If this works well I would be happy to draw a cheap tested relay circut using a NE555 timer to do it automatically. Personally I think a cheap 10A switch a better idea. A DPDT often is about the same price. Use both sets of poles to have a really good switch.
The reason the fuse will blow is until enough magnetic charge is built up in the iron core of the toroid the winding looks like a near short circuit ( could we get a resistance reading for the Tiger mains input? ). Often in books the inductance is drawn in series with this resistance to remind us not to think of it as pure inductance and that this will change calculations. The reality here is, as remarkable as it is we get away with ignoring it. That is most of the time. Toroids tend to exagerate this problem which also results in more noise. If you get big problems ESP Audio has a cure. I don't like his example as it runs electrolytic capacitors in reverse bias albeit at very small voltages. That is cured by using non polars. Also being that caps are cheap he uses too few. This problem is called DC on the mains. I am uncomfortable with that idea. What should be said is a mathamatical averaging of mains distortion shows a DC component. That is like you have a tug of war team and have 100 on one side and 97 on the other. All other things being equal the 3 people imbalance will move the rope the other side of 0V datum line ( middle of a scope AC wave ). This is the DC. Unlike our tug of war this contest is ongoing every 10 mS or less. Naim were troubled by this. Mr Verekers answer was " give them the money back ". Funny how many didn't want the money back and lived with the fact their electricty was low grade. I suspect with local power generation this will get worse.
Quad used an E and I type iron which usually is better, that is the cube style transformers seen from the 1920's onwards. Some 405's were dreadful of the E and I type. This was because Quad penny pinched the design. They used a special potting to make it quiet. If the potting wasn't done at the right temperature it went wrong ( bubbles I think ). Quad were always playing catch up to Naim in the 1980's. It was little things that let them down. Mostly they were very similar companies away from the publicity. Both would have been able to run the others company once little things understood. Now, who would dare say that? E and I usually cost more to do the same job and are much larger. The Quad was a not so specail 300 VA type being asked to give 400VA. In theory that is more than OK as music can not ask for that 400VA. In power testing it might. The potting I think was done to give the impression more money had been spent and a larger transformer used, Sorry Mr Walker if I'm wrong. If upgrading a 405 this Tiger type should be ideal if the correct voltage chosen. There is enough written about the Quad elsewhere to cover the other stuff. It inverts output phase as the number one thing people steadfastly ignore. Kick drums kick forward and it matters. If recorded " wrong " that was the sound the engineer heard so still correct. On most stuff the difference isn't vast.
What would have been great would have been a joint Quad/Naim speaker. It could have had like the car companies their own final versions. Naim got close to perfecting that. The costs went through the roof. British Aerospace were the only ones who could and would build it. It was the Quad original made to a very tight spec. If you like what it needed was to pass the sound barrier. 125 dB peak output was mentioned. This proved to be the problem. It was possible, but gave a short life. 115 db was no problem. It wasn't enough to justify the price. On a speaker like this 115 dB peak might be 95 dB average. As distortion is in the 0.05% region it would sound as if it lacked guts. Friends who heard them placed orders. At I think a then price of £25 000 a pair with no prospect of profit the project died.This was before anyone dared charge that price for sensible speakers. B&W seem to get business at that price now. Forget Audiophile products as the rich will always buy daft things.
I think this disproves the lack of accuracy Naim is said to have. Julians point was speakers matter most in making that statement. An interesting observation he stated this way. If a speaker has a dynamic range of 30db why think an amplifier needs minimum 80 dB? All the Naim negative traits vanish like summer had arrived when the electrostats. It was weird. Knowing this I will live with lower grade speakers and a Naim amp and say it isn't the fault of the amp that the speakers can't do it. Mordaunt Short MS20 is a cheap speaker that will work with Naim. A friend sold some for £60, that's bonkers. In some ways they better his LS3/5A. They are the basis of Naim speakers like SBL.
The Suns out so you got good measure today. Funny the Suns has gone in on that last word. Hey Julian, you must have gone to the good place. Who would have dreamed that! Me if asking.
I agree with your first suggestion if you are proposing to use Multicomp parts from CPC Farnell
As often the case with Farnell's technical specifications, they go only as far as saying "time delay" which implies the sort of pot luck you would have to take in a bargain store, or Ebay. The type required is also known as anti-surge rather than time delay and the construction is likely to be different to long delay types, which as the name suggests, are for protection against longer-term overloads and transformer overheating.Littlefuse's specs are at least referenced to UL burn specifications for the different categories of cartridge fuse: e.g 5x20mm Fuses - Littelfuse. The information may require a little investigation but at least there are distinct numbers relative to standard fast-blow fuses, that you can work with.
I've had a pair of Linn Nexus LS250 speakers with upgraded tweeters (015/5) partnering a Nait 3 for many years. Whilst the speakers are generally not rated very highly in stock form, indeed they were towards the bottom of Linn's range when current, they are a good combo IMHO and sound great in my listening room. I will be using these with the NAP 200 clone. "Work well" is a subjective term though.
I have used the Nexus. I don't think Linn liked them, that didn't help. I liked them. The Sara is a dream if the tonal balance can be lived with. I can. Linn Index as they improved over production are poormans Saras with some KEF excellence thrown in. My favourite cheap speaker with Naim.
Quad ESL 57 although not always 63's ( sounds dull and bloated). Epos ES14/11 if tweeter not too harsh for you.
Now a big surprise.Spendor BC1. It is a match from heaven. Spenser loved the NAP250. He said that would upset some people. He told me that weeks before he died. What a lovely man he was. I liked BC1's with Naim. Not everybodies ideal I know. They mimic the Quads although can not get right inside you like the ESL can. The Quads do many things wrong and BC1's do most things right.If it was a box ticking thing BC1 would better ESL's.
Royd in general. KEF 104 and anything KEF really. Mordaunt Short also if the real ones. Naims bough that driver nd made it in house. The SBL are very special.
Now a shock. Klipsche, the Forte 2. I don't doubt others. The Klipsche didn't seem to like a ANC valve design. That was well against what one would think.
Naim didn't say a world about the transformer. Holden and Fischer made them in Glasgow I think? H&F might have been mates with Linn. Very high quality and seen in big pro amps. The things you have should be worth any extra money.
Quad ESL 57 although not always 63's ( sounds dull and bloated). Epos ES14/11 if tweeter not too harsh for you.
Now a big surprise.Spendor BC1. It is a match from heaven. Spenser loved the NAP250. He said that would upset some people. He told me that weeks before he died. What a lovely man he was. I liked BC1's with Naim. Not everybodies ideal I know. They mimic the Quads although can not get right inside you like the ESL can. The Quads do many things wrong and BC1's do most things right.If it was a box ticking thing BC1 would better ESL's.
Royd in general. KEF 104 and anything KEF really. Mordaunt Short also if the real ones. Naims bough that driver nd made it in house. The SBL are very special.
Now a shock. Klipsche, the Forte 2. I don't doubt others. The Klipsche didn't seem to like a ANC valve design. That was well against what one would think.
Naim didn't say a world about the transformer. Holden and Fischer made them in Glasgow I think? H&F might have been mates with Linn. Very high quality and seen in big pro amps. The things you have should be worth any extra money.
Forgive the typo's, worse than ever. The CPC fuse has some spec. HRC usually a good sign as it means high rupture current.
Very interesting what you say Andrew. When designing something for work I used your ideas that that we looked at on another thread. The assumption usually is about 1/3 second. I agree with you to make it longer. I gave a generous 11 seconds and 4 seconds for smaller ones. I can make them what I like if we have any trouble. That is up to 24 seconds. The timings circa 1 minute overal.
If a light bulb ( lamp ) as resistor and DPDT swich would be excellent. Needs the right bulb. I did think of the idea also, but didn't see all of the many advantages.
Very interesting what you say Andrew. When designing something for work I used your ideas that that we looked at on another thread. The assumption usually is about 1/3 second. I agree with you to make it longer. I gave a generous 11 seconds and 4 seconds for smaller ones. I can make them what I like if we have any trouble. That is up to 24 seconds. The timings circa 1 minute overal.
If a light bulb ( lamp ) as resistor and DPDT swich would be excellent. Needs the right bulb. I did think of the idea also, but didn't see all of the many advantages.
howarthcd what did you used for the LTP transistor, the MPSA06 that were into the kit? Did you matched them in any way? 5mv DC offset is pretty good
Hi, no I went to no effort to match the parts. My emphasis was really on getting the kit up and running to see whether it actually worked and so I have an objective to investigate and improve in future where appropriate. I'm open to any suggestions.
That said, what would you recommend as a replacement for the MPSA06 and how should they be matched? Would you recommend, say, the BC239C? These are increasingly rare nowadays though.
How should such a pair be matched - e.g. what hFE and at what current for this particular amp? Are there any other parameters that should ideally be matched?
Chris
Last edited: