BC239C looks risky to me. It's rated at 25V VCEO and it's used with 40V across the C-E. Maybe Naim were selecting them, but I would not risk this unless I had measured them.
However MPSA06 would appear to have low gain when used in the low mA range (about 80) and no NF is specified. It seems quite unsuitable for LTP use.
BC547C with its legs crossed looks suitable.
2SC1815 also should be OK. Sadly mine are all -Y gain banding :-(
However MPSA06 would appear to have low gain when used in the low mA range (about 80) and no NF is specified. It seems quite unsuitable for LTP use.
BC547C with its legs crossed looks suitable.
2SC1815 also should be OK. Sadly mine are all -Y gain banding :-(
Last edited:
I was using BC327-40 the other day which is as far as I would dare go at 45V. BC337-40 could be interesting. Some people would see a common type like this as some sort of cheap wine sold as RED and no imformation. Nothing could be further from the truth. For my work I use tons of BC327-40 ( The PNP ). I use them in saturation as simple switches. They only loose 0.05V in my application. This saves my company needing to warn customers about very low voltages and drop out ( tested to 90/180 V AC and 135/270 ). I have to pull plenty of base current to do this. They don't mind. BTW most of my circuits will do 150/300. I am always sad to pull back to135/270V. OK at 300V they are not happy, the test is do they survive. 180 to 280V was what we learnt from a customer survey. It's bad, but is better than no electricty in these countries. USA isn't great and Australia tells me the same.
Search Results | CPC UK
One day I was desperate to tryout an idea which would be using 2N4403/1. A primitive transistor that by chance is OK for audio in the same way as dynamite can be used for fishing ( not far from true ). It has very low noise. I had a hunch BC327-40 would not be worse. I was right. Much like a BC557/ BC140 hybrid. I have to thank this paper below for what I know. Very unusal reading seeing as BC337 is not listed with low noise types. These almost power devices have very low base spreading resistance ( Rbb- ). A high gain, high current device always should be looked at. TV T0126 devices were used for MC amplifiers ( ie Lentek , Audiolab ). Philips became so annoyed they wanted the TV serial number before selling spares. It was costing them money to make small batches. That's when Hi Fi Folklaw is bonkers. No one needed the TV parts. BD139/140 was almost the same device. I dare say a set of 2SC5200/2SA1943 would be far better ( possibly not and have been told why ). No one ever seems to think that. There's a thought. See Quad MC stage. It looks impossible yet works OK unless you read Douglas Self who sees it as a high distortion circuit, get real when LP can be 10%? Always seemsed OK to me although Quad preamps not really. What ever you do never buy a Quad 44. A 34 is OK and can be vastly improved. Quad used almost the worst op amps of hi fi types. For a Naim Clone and home cinema it would be ideal and better than Naim. It inverts which means inverting the speakers. You will sell it when wanting better without trouble. Makes my wish I had one.
High gain, high current, wide bandwidth willl give better sound and lower DC offset ( not that DC offset matters at all if under 100 mV ). Some 2SA872A PNP I have are gain 500 and 120 V! One friend who studdied hi fi at advanced level said his boss considered low noise a big factor in power amp design. His observed valves have a pink noise spectrum and transistors blue. Even though very low in both is possible it isn't a non question. His feeling was if the noise is very very low in transistor the sound is warmer! Well it's not easy to prove and will do no harm to belive it as it can be had at zero cost. The noise in the external parts is very very hard to get right. One might even think nasty carbon resistors could be used. It is quite interesting to take a state of the art metal film out and put in a carbon film. Sometimes the noise is not really much worse and the maths sort of said that. What you might get is pinker hiss. The input resistor comes to mind. I am not saying it will be good. Don't assume you would hate it. Carbon compositon are fun. It's your choice. They are now common again ( pulse lasars need them ) and cheap. I bet 99% of DIY memebers will throw their hands up in horror. Did you try them? If shunted by the preamp sometimes they can work. The sound dull but completely without edge. Metal foil are great, and so is the price. they sound bright and even less edge!!!! It's a black art sometimes. Anybody would think it was for listening to.
http://www.janascard.cz/PDF/Design of ultra low noise amplifiers.pdf
Search Results | CPC UK
One day I was desperate to tryout an idea which would be using 2N4403/1. A primitive transistor that by chance is OK for audio in the same way as dynamite can be used for fishing ( not far from true ). It has very low noise. I had a hunch BC327-40 would not be worse. I was right. Much like a BC557/ BC140 hybrid. I have to thank this paper below for what I know. Very unusal reading seeing as BC337 is not listed with low noise types. These almost power devices have very low base spreading resistance ( Rbb- ). A high gain, high current device always should be looked at. TV T0126 devices were used for MC amplifiers ( ie Lentek , Audiolab ). Philips became so annoyed they wanted the TV serial number before selling spares. It was costing them money to make small batches. That's when Hi Fi Folklaw is bonkers. No one needed the TV parts. BD139/140 was almost the same device. I dare say a set of 2SC5200/2SA1943 would be far better ( possibly not and have been told why ). No one ever seems to think that. There's a thought. See Quad MC stage. It looks impossible yet works OK unless you read Douglas Self who sees it as a high distortion circuit, get real when LP can be 10%? Always seemsed OK to me although Quad preamps not really. What ever you do never buy a Quad 44. A 34 is OK and can be vastly improved. Quad used almost the worst op amps of hi fi types. For a Naim Clone and home cinema it would be ideal and better than Naim. It inverts which means inverting the speakers. You will sell it when wanting better without trouble. Makes my wish I had one.
High gain, high current, wide bandwidth willl give better sound and lower DC offset ( not that DC offset matters at all if under 100 mV ). Some 2SA872A PNP I have are gain 500 and 120 V! One friend who studdied hi fi at advanced level said his boss considered low noise a big factor in power amp design. His observed valves have a pink noise spectrum and transistors blue. Even though very low in both is possible it isn't a non question. His feeling was if the noise is very very low in transistor the sound is warmer! Well it's not easy to prove and will do no harm to belive it as it can be had at zero cost. The noise in the external parts is very very hard to get right. One might even think nasty carbon resistors could be used. It is quite interesting to take a state of the art metal film out and put in a carbon film. Sometimes the noise is not really much worse and the maths sort of said that. What you might get is pinker hiss. The input resistor comes to mind. I am not saying it will be good. Don't assume you would hate it. Carbon compositon are fun. It's your choice. They are now common again ( pulse lasars need them ) and cheap. I bet 99% of DIY memebers will throw their hands up in horror. Did you try them? If shunted by the preamp sometimes they can work. The sound dull but completely without edge. Metal foil are great, and so is the price. they sound bright and even less edge!!!! It's a black art sometimes. Anybody would think it was for listening to.
http://www.janascard.cz/PDF/Design of ultra low noise amplifiers.pdf
Correction to my earlier post:
BC547C whcih is B-C-E will fit the opposite way to MPSA06 E-C-B. No leg crossing requried.
2SC1816 will need its legs crossing.
BC547C is the same pinout as BC239C.
BC237C looks OK.
BC547C whcih is B-C-E will fit the opposite way to MPSA06 E-C-B. No leg crossing requried.
2SC1816 will need its legs crossing.
BC547C is the same pinout as BC239C.
BC237C looks OK.
Would the suitability of a device be affected by the higher current through the LTP devices in the NAP200 clone vs the NAP140 for example? Does the hFE of each of the pair need to be quite as high as it is in other models?
I'm still getting to grips with how the LTP works.
I'm still getting to grips with how the LTP works.
Devices specified for LTP duty usually have a max Ic, or Ie, of 100mA some can be as low as 50mA, but they are rare.
Higher hFE gives lower input offset voltage and lower input offset current.
LTP tail current for two devices can vary from 0.2mA to 20mA. Well inside the max Ic by a factor of 10 to 100.
More likely the problem could be max dissipation. To92 can vary from 300mW to 625mW max.
1.75mA Ie @ 40Vce results in Pq of 70mW. This is only 23% of 300mW. this could be doubled for peak transients and still be @ only 46% for a long term duration.
High Ie of 5mA to 10mA @ 40 to 50Vce will be a problem and this is where one would adopt a cascode. Particularly so if one were using a 150mW sot23 device.
I usually go to <50% power if I can. I then use T0126 if critical as a heatsink can be fitted. BD139 comes to mind. BC639 is possibly the same device.
A piece of tubing cures any leg crossing. I often put it on base to make it more obvious where I want it to go.
One can add emitter resistors and sleeve the whole thing. 1/8 watt are small. If staying in Naim style keep the resistors in a ratio.
A piece of tubing cures any leg crossing. I often put it on base to make it more obvious where I want it to go.
One can add emitter resistors and sleeve the whole thing. 1/8 watt are small. If staying in Naim style keep the resistors in a ratio.
There will be no change in LTP stage gain for reasons given in post #1902.
Higher gain transistors, BC547C or BC550C will increase the Miller capacitance between the collector and base of the feedback return transistor.
That will work against the design improvements to increase the LTP current delivery rate and reduce the constraint on the nfb return path.
This won't cause any harm but it might be a better amplifier if you stick to the NAP200 script than be influenced by aspects of the NAP140 design.
is this kit good to buy?
thx
Black Box ?clone Naim NAP200 Amplifier Kit DIY Power Amp Kit 75W 75W | eBay
thx
Black Box ?clone Naim NAP200 Amplifier Kit DIY Power Amp Kit 75W 75W | eBay
Nor by the NAP250, both the 140 clones and NCC200 being based on it, as there is no other published NAP design available. So even the popular clone kits have always had it wrong.
Incidentally, an output coil (damped with a parallel resistor) is only necessary if using unsuitable speaker cable. You don't even need NAC5 spec. cable - even a couple of metres of figure-8 lamp flex will avoid problems with any Naim model.
Please do read the last few pages of this thread. And if you need more info, read from further back. Maybe even from the start to get real background info.
This shows how little I understand about the working of this circuit as yesterday I ordered several of each of those BC* transistors to substitute the MPSA06s! 🙁
I guess the problem with this kit is that we don't know what exact parts are used in the NAP200. All it would take would be for someone with the genuine article to take a look and report back. I doubt anyone with the real thing is reading this thread though.
Chris
Andrew, you'd be correct to be auspicious. On about the 5th power up so far the T3.15A fuse blew. And then another one straight away. The third seems to have survived.
I've not been brave enough to leave the unit on permanently yet so I have been power-cycling it more than I ordinarily would.
Chris
Last edited:
I don't know if I have posted this in this Thread, but here goes on the way I select components for a soft start on 240Vac (UK mains power).
buy a 230:x+x Vac 300VA transformer
The maximum primary current driving the rated resistive load will be 300VA/230Vac = 1.3Aac
Since the rated VA is an output, then the maximum primary current will be slightly higher due to the efficiency of the transformer. But I ignore that, except if I were designing the transformer and it's core (I am not a transformer designer).
Close rated fuse is T1.6A (the first value above 1.3Aac), or maybe T1.25A and worth trying a T1A after the system is complete.
The maximum current with a soft start resistor in place can be ~ double your chosen fuse, i.e. T1.6A becomes 3.2A
For 240Vac the required resistance will be 240Vac/3.2Aac = 75ohms
That includes the primary winding resistance.
An added resistor around 70r for the soft start that is bypassed after ~ 100ms to 300ms, with a mains rated relay will generally allow a T1.6A fuse to survive many cold restarts.
I suspect that a lower added resistor will not blow that T1.6A fuse.
That would take many fuses and very many restarts to prove.
I use a soft start for all my transformers >80VA and use the same method to determine the close rated T fuse and added resistance.
Here are the values I would adopt:
VA . . Ip . . Tfuse . . Radded
100 . . 0.43 . 500m . . 230r
160 . . 0.7 . . 800m . . 145r
200 . . 0.87 . . 1 . . . . 115r
300 . . 1.3 . . 1.25 . . . 90r
400 . . 1.7 . . 1.6 . . . 70r
600 . . 2.6 . . 2.5 . . . 45r
800 . . 3.15 . . 3.15 . 30r
1kVA . 4.4 . . . 4 . . . 27r
buy a 230:x+x Vac 300VA transformer
The maximum primary current driving the rated resistive load will be 300VA/230Vac = 1.3Aac
Since the rated VA is an output, then the maximum primary current will be slightly higher due to the efficiency of the transformer. But I ignore that, except if I were designing the transformer and it's core (I am not a transformer designer).
Close rated fuse is T1.6A (the first value above 1.3Aac), or maybe T1.25A and worth trying a T1A after the system is complete.
The maximum current with a soft start resistor in place can be ~ double your chosen fuse, i.e. T1.6A becomes 3.2A
For 240Vac the required resistance will be 240Vac/3.2Aac = 75ohms
That includes the primary winding resistance.
An added resistor around 70r for the soft start that is bypassed after ~ 100ms to 300ms, with a mains rated relay will generally allow a T1.6A fuse to survive many cold restarts.
I suspect that a lower added resistor will not blow that T1.6A fuse.
That would take many fuses and very many restarts to prove.
I use a soft start for all my transformers >80VA and use the same method to determine the close rated T fuse and added resistance.
Here are the values I would adopt:
VA . . Ip . . Tfuse . . Radded
100 . . 0.43 . 500m . . 230r
160 . . 0.7 . . 800m . . 145r
200 . . 0.87 . . 1 . . . . 115r
300 . . 1.3 . . 1.25 . . . 90r
400 . . 1.7 . . 1.6 . . . 70r
600 . . 2.6 . . 2.5 . . . 45r
800 . . 3.15 . . 3.15 . 30r
1kVA . 4.4 . . . 4 . . . 27r
Last edited:
Just buy fuses, no harm in it. 4AT?
If the Tiger has a thermal fuse if it were my own I would be happy to use 5AT ( I couldn't possibly advise it to another ).
The risks are these. Short cicuit will blow the fuse. Job done.
Shorted turn giving heat build up. Job done if thermal fuse fitted. 3.15AT is not really protection for that although it helps.
Some purists don't like DC fuses ( me ). This is very bad as the mains fuse is being asked to do total protection. A 1uF 100 V fuse bypass will remove any need to be a purist on the DC side ( not MAINS ). If a speaker fuse do the same. I have even been able to measure that! It was subtle. It was on a Rotel RA931. I fitted auto resetting fuses( nasty ). Once 1uF fitted it was the best sound ever. It told Dan it was too loud. The Rotel is very robust. It seems to take abuse. This was a last ditch attempt to get Dan ( my friends son ) to admit he was the problem. Really nice amp.
On the transistors this would be about right.
=>45V
gain >250
Cob <20 pF
Rbb-<50 R ( low noise,< 2dB typical, this is the bigger deal of all )
Ft> 50 MHz
P> 300 mW
farnell.com/on-semiconductor/bc550cg/transistor-npn-45v-0-1a-to-92/dp/SC13409?ost=bc550c&selectedCategoryId=&categoryName=All+Categories&categoryNameResp=All+Categories
Sorry link failed. CPC BC550CG,
If the Tiger has a thermal fuse if it were my own I would be happy to use 5AT ( I couldn't possibly advise it to another ).
The risks are these. Short cicuit will blow the fuse. Job done.
Shorted turn giving heat build up. Job done if thermal fuse fitted. 3.15AT is not really protection for that although it helps.
Some purists don't like DC fuses ( me ). This is very bad as the mains fuse is being asked to do total protection. A 1uF 100 V fuse bypass will remove any need to be a purist on the DC side ( not MAINS ). If a speaker fuse do the same. I have even been able to measure that! It was subtle. It was on a Rotel RA931. I fitted auto resetting fuses( nasty ). Once 1uF fitted it was the best sound ever. It told Dan it was too loud. The Rotel is very robust. It seems to take abuse. This was a last ditch attempt to get Dan ( my friends son ) to admit he was the problem. Really nice amp.
On the transistors this would be about right.
=>45V
gain >250
Cob <20 pF
Rbb-<50 R ( low noise,< 2dB typical, this is the bigger deal of all )
Ft> 50 MHz
P> 300 mW
farnell.com/on-semiconductor/bc550cg/transistor-npn-45v-0-1a-to-92/dp/SC13409?ost=bc550c&selectedCategoryId=&categoryName=All+Categories&categoryNameResp=All+Categories
Sorry link failed. CPC BC550CG,
A picture is worth 1000 words. The lamp can be outside the amplifer and before the fuse. It could be in the neutral (if by mistake when non polar plug ). A lamp holder would be ideal. I dare say even a 15 watt ES lamp could work. Wait until the filiment becomes stable to throw the switch. House switches can be used. Andrew I am sure can say more. The fuse can be 1.6AT if so. A resistor is less good as it is not designed to work at 230V continuously and gives no indication you forgot the switch. If the lamp is brighter than usual there is a fault. This can save damage if caught in time. It is what engineers do when testing repairs or new ideas. The 1920's lamp is a little bit more impressive than you might think. It changes resistance as it heats. Mostly a resistor doesn't by comparrison.
It might be many lamps need trying. I have a hunch 100 watt will work. In principle banned these days. I see them in many places. There are halogen types that are possible also ( remember heat if so ) . Forget energy saving lamps as I am sure is obvious. If you see a wire coil it should work.
Last edited:
So much useful yet sometimes contradictory advice between posters on here. It's good to have the varying opinions. I like it. Makes you think.
The BC550CG is what I have ordered based on spec, yet an earlier post suggested these aren't suitable. I guess I'll try them and let my ears decide.
Andrew. Have you ever made a transistor Vceo tester? My brother said it was a device with a very high voltage with a very low current. The transistor would be increased in voltage until it failed. When taken from circuit it would be OK as the current was so low. The failiure maked by the transistor attempting to short the PSU.
One could select BC550C.
In am told BD139/140 are selected BD135/136. A machine tests as it goes for this. This possibly means some make 90 V.
One could select BC550C.
In am told BD139/140 are selected BD135/136. A machine tests as it goes for this. This possibly means some make 90 V.
I wonder why. We do get types that are said to be more linear. It means nothing in a feedback amp I feel. BC550C is a dream device. 45V is a bit low.