Yes it was 39.5V, not 395mv. Agree Nigel that the BC239 is much lower on CE breakdown than the BC237. And I was worried too.
However Chris said that Naim used BC239C in its original Naim amp. The part is certainly tested by Naim before getting in the amp.
Any suggested test circuit for the CE breakdown voltage?
I'll put together a small test gig and I'll run it. What do you recommend for a test Gig? Simply reproduce the front LTP with its current source?
However Chris said that Naim used BC239C in its original Naim amp. The part is certainly tested by Naim before getting in the amp.
Any suggested test circuit for the CE breakdown voltage?
I'll put together a small test gig and I'll run it. What do you recommend for a test Gig? Simply reproduce the front LTP with its current source?
Just to quote the relevant part of the email conversation I had with Naim, starting with their first reply, regarding the BC239C in relation to its appearance in the LTP in a NAP200:
Naim: Would you be able to tell us the number on the transistors?
Me: Hi, the transistors are labelled: BC239C.
Naim: ...We have them in stock at a price of £55.56.
Me: OK thanks. Is that the original part for this model? Do you supply the transistors in matched pairs?
Naim: That is the original part. We do have matched pairs. The price stated below is for one transistor. If you were interested in buying from us, you would need to go through a dealer.
To my mind this is confirmation that the BC239C is used. Do others agree? Is it worth seeking a second confirmation?
Chris
Last edited:
Caowei based their clone on an original NAP200 and Naim have been specifying BC239C on all standard power amplifier models since the 1970s. Lift the cover on a NAP140, 250 etc. It's BC239, just like the generic NAP schematic shows. Entry level models Nait1,2 and NAP90 likewise but there could be other types used in later series models that I haven't seen.
What Naim don't specify, is the amount of very costly testing, sorting and subsequent large scale parts rejections, returns or disposals which are legend for the manufacturer and come at a high premium cost to the retail buyer. ₤55 for a 2p transistor surely tells you what crazy policies and lengths Naim adhere to with their traditional products.
In the early years of transistor manufacture, many power transistors in particular, where simply re-graded and marked as other parts by the manufacturer when they didn't meet spec. It seems like a massive waste of time and money to do this as a client after buying large quantities of cheap TO92 parts but Naim and quirky, diehard design philosophies are synonymous in my book.
My inclination is to simply use BC550C, selected for high V(BR)Ceo on test. We've been warned about testing the BC239C pair supplied in the kit, in case a selection test wasn't done but the chances for a higher BR voltage will be better with BC550.
With a test voltage source of at least 50V, simply fit a series resistor that limits C-E current to < 1mA. Ideally, you plot voltage against current with a transistor curve tracer until there is a sharp rise in current from near zero as V(BR) is approached. Check P28 onwards here: http://bitsavers.trailing-edge.com/...0_Semiconductor_Device_Measurements_Apr69.pdf
In practice, you can just measure the C-E voltage of small signal BJTs at a small fixed current of 10-100uA .
Last edited:
Same question as before. Why go to these lengths when there are plenty of more-capable devices out there that could easily be substituted without all the fuss?
Or could they? What is it about the BC239C that compels Naim to go to these efforts? Why don't they just use the BC550C or MPSA18 and save all the time and expense?
I think Martin Clark, Neil Mcbride and experienced PFM forum Naim freaks would be better qualified to comment on Mr Vereker's unique design/manufacturing policies. Suffice it to say that Naim is no ordinary electronics company and always went their own obstinate way - right to the top of UK business success at one time.
I don't think this is just a matter of component capability - more a historical question on a singular policy of product consistency that goes way beyond what other manufacturing designers would have considered economically viable or technically necessary. That is to say that as long as the models were just minor variations on the same basic design, the components deemed critical to sound quality were kept the same across all models. This also makes it much easier to supply and maintain a small controlled inventory of spares for all models, across a global dealer/service agency network.
I'm not familiar with MPSA18 used in audio input stages. Indeed, there are many other similar low noise transistors of the 1970s which could have been used and SE4010, BC109, 149, 549, 184, 414 come to mind. The question of which transistor would be framed in the type characteristics, spec. tightness and Hfe grades and OEM reputation in the 1970s. Motorola was highly regarded in this area of consistency which shows in Naim's general semi choices. Underlying this, the basic issue has to be Mr Vereker's policy of selecting individual parts rather than systematically protecting them with large dropping resistors, voltage regulation or changing to higher voltage types when they later became available. (BC550 was a later Philips product)
Otherwise, in this thread we are on some level of a cloning mission - we're spending a lot more money on this particular NAP200 model, I'd suggest specifically to stay as close as possible to the original because the PCB enables it but where it's necessary to deviate, then we have to make personal choices. If we have heard the original though, at least there would be a reference point. Perhaps we'll like our choices better but IMV, that isn't the point of building a clone or supplying kits designated "clone".
It's much cheaper and easier to fit a couple of H140 boards into a shoebox case if we only want the kit as a pincushion for our own ideas. The looks and clone value don't matter at all unless it's about pics and bragging rights as much as performance.
On a practical note, I tested the 2 supplied BC239 pairs and read 63 - 78V(BR)ceo. Whilst you could infer that 45Vce at 1.5 mA should be safe, I would want proof test results too. That means another big bag of parts, just like Naim. For anyone interested in some general info. on making transistor choices, this entertaining Wiki from Analog Devices is a nice intro
: https://wiki.analog.com/university/courses/electronics/text/choosing-transistors
I don't think this is just a matter of component capability - more a historical question on a singular policy of product consistency that goes way beyond what other manufacturing designers would have considered economically viable or technically necessary. That is to say that as long as the models were just minor variations on the same basic design, the components deemed critical to sound quality were kept the same across all models. This also makes it much easier to supply and maintain a small controlled inventory of spares for all models, across a global dealer/service agency network.
I'm not familiar with MPSA18 used in audio input stages. Indeed, there are many other similar low noise transistors of the 1970s which could have been used and SE4010, BC109, 149, 549, 184, 414 come to mind. The question of which transistor would be framed in the type characteristics, spec. tightness and Hfe grades and OEM reputation in the 1970s. Motorola was highly regarded in this area of consistency which shows in Naim's general semi choices. Underlying this, the basic issue has to be Mr Vereker's policy of selecting individual parts rather than systematically protecting them with large dropping resistors, voltage regulation or changing to higher voltage types when they later became available. (BC550 was a later Philips product)
Otherwise, in this thread we are on some level of a cloning mission - we're spending a lot more money on this particular NAP200 model, I'd suggest specifically to stay as close as possible to the original because the PCB enables it but where it's necessary to deviate, then we have to make personal choices. If we have heard the original though, at least there would be a reference point. Perhaps we'll like our choices better but IMV, that isn't the point of building a clone or supplying kits designated "clone".
It's much cheaper and easier to fit a couple of H140 boards into a shoebox case if we only want the kit as a pincushion for our own ideas. The looks and clone value don't matter at all unless it's about pics and bragging rights as much as performance.
On a practical note, I tested the 2 supplied BC239 pairs and read 63 - 78V(BR)ceo. Whilst you could infer that 45Vce at 1.5 mA should be safe, I would want proof test results too. That means another big bag of parts, just like Naim. For anyone interested in some general info. on making transistor choices, this entertaining Wiki from Analog Devices is a nice intro
: https://wiki.analog.com/university/courses/electronics/text/choosing-transistors
A reassuring hand required + a couple ?s
Before I place my parts order.
The Naim 140 amp board requires 40VDC supply, this is a no load voltage I’m assuming?
I have selected a toroidal transformer, it will operate from a mains voltage of 120v. It is a 330Va primary with a 25V + 25V secondary voltages with a center tap that I am assuming again, center tap refers to a common ground?
For the power supply board it’s self I am still educating myself as to the proper board to use. I am reading about and looking at kits on ebay etc. and also a speaker protection circuit, but as to the AC to DC conversion and the resulting DC voltages? I am lets say not completely sure of my self to say the least. Any recommendation to a proper supply board or specs to complement the above mentioned transformer would be helpful. I have a general idea as to what goes where and why but the specific values of the components etc. I am not far enough along to make such calls.
All of that said if I wanted a supply to use as a test supply lets say for bigger and more powerful amp projects including the Naim 140, could I purchase a transformer with higher voltages and that my power supply circuit depending on it’s components and values would determine the output VDC to various amp builds?
Before I place my parts order.
The Naim 140 amp board requires 40VDC supply, this is a no load voltage I’m assuming?
I have selected a toroidal transformer, it will operate from a mains voltage of 120v. It is a 330Va primary with a 25V + 25V secondary voltages with a center tap that I am assuming again, center tap refers to a common ground?
For the power supply board it’s self I am still educating myself as to the proper board to use. I am reading about and looking at kits on ebay etc. and also a speaker protection circuit, but as to the AC to DC conversion and the resulting DC voltages? I am lets say not completely sure of my self to say the least. Any recommendation to a proper supply board or specs to complement the above mentioned transformer would be helpful. I have a general idea as to what goes where and why but the specific values of the components etc. I am not far enough along to make such calls.
All of that said if I wanted a supply to use as a test supply lets say for bigger and more powerful amp projects including the Naim 140, could I purchase a transformer with higher voltages and that my power supply circuit depending on it’s components and values would determine the output VDC to various amp builds?
Last edited:
The PSU output voltage determines the maximum signal that can be supplied to the load.
eg.
say you have a +-40Vdc supply that droops to +-39.3Vdc when the quiescent amplifier is connected. When the maximum signal voltage is sent to the output, but the load is very high (in the kilo-ohms range), then this maximum voltage will be around 1V to 3V less than the supply rail, i.e. expect ~Vpk out to be around 38.3Vpk to 36.3Vpk
Now change the load to an 8ohms resistive test.
The supply rails will droop, maybe to around +-34Vdc to +-37Vdc. The voltage drop through the amplifier will increase to around 3V to 8V. You could have a high of 34Vpk to a low of 26Vpk into 8r0 test load.
The output power would be in the range of 72W down to 42W from that +-40Vdc PSU.
The maximum output power depends on how well the PSU and the Amplifier are able to deliver current.
Now back to your 25-0-25Vac transformer.
You could end up with +-40Vdc from the PSU when open circuit, i.e. no output current.
This could fall to +-34Vdc, when the load is drawing maximum current.
eg.
say you have a +-40Vdc supply that droops to +-39.3Vdc when the quiescent amplifier is connected. When the maximum signal voltage is sent to the output, but the load is very high (in the kilo-ohms range), then this maximum voltage will be around 1V to 3V less than the supply rail, i.e. expect ~Vpk out to be around 38.3Vpk to 36.3Vpk
Now change the load to an 8ohms resistive test.
The supply rails will droop, maybe to around +-34Vdc to +-37Vdc. The voltage drop through the amplifier will increase to around 3V to 8V. You could have a high of 34Vpk to a low of 26Vpk into 8r0 test load.
The output power would be in the range of 72W down to 42W from that +-40Vdc PSU.
The maximum output power depends on how well the PSU and the Amplifier are able to deliver current.
Now back to your 25-0-25Vac transformer.
You could end up with +-40Vdc from the PSU when open circuit, i.e. no output current.
This could fall to +-34Vdc, when the load is drawing maximum current.
Last edited:
The ebay soft start board has now been converted to use a small isolating transformer, as well as having most other parts replaced.
The before/after images are attached. Hopefully it is easy to tell which is which.
There was no need to secure the transformer in the end as I managed to re-use several of the holes and pads that the dropper capacitors previously occupied.
The fuse has been bypassed as there is a lower-value fuse before the mains touches this board. The transformer is rated at 15V output, but after rectification and smoothing, and considering the regulation value, this is more than enough to switch on the replacement relay.
I've yet to tweak the delay time as it's a little too long (~15s), but that should be a matter of adjusting the values of the components in the RC circuit, I'll probably replace the resistor having already replaced the cap.
Attachments
You must be referring to regulated PSUs. A simple transformer + bridge rectifier + smoothing caps, as in Naim and so many other amplifiers, puts out DC at a virtually fixed voltage. It fluctuates within a range of only a few percent with normal domestic use but at full rated power, this will rise to about 10%, according to how hard the amp. is working at the instantaneous signal level, much as AndrewT described.
If you want a range of voltages to power anything with more than a trickle of current, you'll find it expensive to cover all your likely voltage needs with one transformer, even for domestic audio amps. You could need multi-tapped windings for efficiency and that means using an expensive, special transformer or wasting lots of heat and power in electronic regulator circuits. There is a solution in using a variac to adjust the AC mains supply to the transformer but this adds significantly to the cost and bulk of your test PSU.
In his Audio Power Amplifier book, Randy Slone showed a simple regulator arrangement for a modest voltage range that was based on just one power transitor. Robot Check
I still find this book useful for some of the unique ideas and approaches to DIY and design. Worthwhile at the price.
You might have read here that the Ebay kits and the Avondale NCC200 are not really based on NAP140 since it has a nominal +/- 34VDC supply, which should suit your standard 25V-0-25 transformer better, as you planned. A +/- 40V supply actually suits NAP250,180,200 as mentioned in more recent posts. There is though, a need to adjust the bias by altering resistances in the input stages to suit the lower supply voltage. I don't think anyone has a specific NAP140 schematic, unless it was reverse-engineered from an original amplifier, so this may need to be estimated.
Nice little transformer but why replace the better suited wire-wound surge resistors with MF types?
It's more that I don't trust the originals, given the cost of the board. I replaced them with what I already had in stock.
The unit is working well, the amp is left on all the time so I'd expect the resistors to pass current maybe half a dozen times a year. During testing the power was cycled a dozen times and the resistors didn't feel particularly warm. The delay time is set to 3s.
What would be the advantages of wirewound parts?
Last edited:
I'll test the BC239C, I have a curve tracer
Wow a printed Tektronix white paper from the 69's
Remind me when I was working in a calibration lab at the beginning of my career, we used to have all kind of white papers and app. notes from Tek, HP and Fluke. Great source of information.
Thanks
SB
Wow a printed Tektronix white paper from the 69's
Remind me when I was working in a calibration lab at the beginning of my career, we used to have all kind of white papers and app. notes from Tek, HP and Fluke. Great source of information.
Thanks
SB
I wouldn't be too harsh on Chinese component quality. You may not realise but to be competitive, most traditional US and Euro manufacturers now have their components made somewhere in Asia anyway. Looks and markings can be deceptive.
MF resistors burn when overloaded and are not used in power circuits. MO (metal oxide) types are safer and generally don't cause fires when they fail. Wire wound types have much higher temperature and overload margins and that's why they are ideal for surge limiting. Provided they have welded internal lead connections, they'll be much more reliable than equally rated film types.
Actually, there's lots of commentary, papers, forum advice, wikis and application guides for different resistor types on the web. Unfortunately, most of it is simply description with no engineering basis for making real life decisions where it matters. However, you can look at any soft start kit or application in a commercial amp. and wonder why they all use the more expensive wire-wound parts
