You can use 2N3055 or 2N3773 also.
At headphone levels, the sound pressure in your ears (long term max. 110 dB), can be achieved by a simple TDA series chip amp, costs less than $2.50.
You need at the most 2 watts per channel to drive headphones for comfortable long term hearing, if you are deaf you may need more.
I would simply gut out a scrap pair of speakers, make a box for the amp / supply, put a socket for the headphones, and enjoy the music.
Even an old car radio will work, needs more modification...
And a plate amp, with two 2030 / 2050 /1875 chips is like $2 here, power supply extra...all in all about $ 10 worth of stuff.
Here we have a saying, buying a buffalo if you have grass....you need to do something for the stock of parts you have.
They can be used elsewhere, or sold.
At headphone levels, the sound pressure in your ears (long term max. 110 dB), can be achieved by a simple TDA series chip amp, costs less than $2.50.
You need at the most 2 watts per channel to drive headphones for comfortable long term hearing, if you are deaf you may need more.
I would simply gut out a scrap pair of speakers, make a box for the amp / supply, put a socket for the headphones, and enjoy the music.
Even an old car radio will work, needs more modification...
And a plate amp, with two 2030 / 2050 /1875 chips is like $2 here, power supply extra...all in all about $ 10 worth of stuff.
Here we have a saying, buying a buffalo if you have grass....you need to do something for the stock of parts you have.
They can be used elsewhere, or sold.
The Hitachi MOSFET amp didn't need drivers to achieve really good performance ... and you certainly don't need them for a headphone amp version. So why don't you get rid of them and associated bits that only make the circuit more complex and probably worse.
The only 'improvement' worth having is the CCS in the input LTP. Actually you don't even need that cos your early stages are regulated.
Look up circuits for the original Hitachi HMA 7500 and note the early stages are fed from regulated supplies; ... something overlooked by da myriad of clones. Here's a 'copy' that doesn't show the regulated supplies Hitachi Fet designs You need to see an actual Service Manual for these secrets. Da Paul Kemble performance curves are from an IEEE paper and Hitachi databook which DID secretly use regulated supplies.
Redesign of a SIMPLE good working circuit is easier than redesign of a COMPLEX, possibly dodgy circuit .... which is what you are trying to do.
Last edited:
You can use 2N3055 or 2N3773 also.
At headphone levels, the sound pressure in your ears (long term max. 110 dB), can be achieved by a simple TDA series chip amp, costs less than $2.50.
You need at the most 2 watts per channel to drive headphones for comfortable long term hearing, if you are deaf you may need more.
I would simply gut out a scrap pair of speakers, make a box for the amp / supply, put a socket for the headphones, and enjoy the music.
Even an old car radio will work, needs more modification...
And a plate amp, with two 2030 / 2050 /1875 chips is like $2 here, power supply extra...all in all about $ 10 worth of stuff.
Here we have a saying, buying a buffalo if you have grass....you need to do something for the stock of parts you have.
They can be used elsewhere, or sold.
At headphone levels, the sound pressure in your ears (long term max. 110 dB), can be achieved by a simple TDA series chip amp, costs less than $2.50.
You need at the most 2 watts per channel to drive headphones for comfortable long term hearing, if you are deaf you may need more.
I would simply gut out a scrap pair of speakers, make a box for the amp / supply, put a socket for the headphones, and enjoy the music.
Even an old car radio will work, needs more modification...
And a plate amp, with two 2030 / 2050 /1875 chips is like $2 here, power supply extra...all in all about $ 10 worth of stuff.
Here we have a saying, buying a buffalo if you have grass....you need to do something for the stock of parts you have.
They can be used elsewhere, or sold.
Why argue = waste of time
Lots of people dont want 2 dollar headphone amps or some 2N3055.
Being completely unrelated to the topic.
AT-HA5000 has numerous advantages over a maplin kit
And the Hitachi topology is by far one of my favorites. Not going in to much details for armchair nonsense
Lots of people dont want 2 dollar headphone amps or some 2N3055.
Being completely unrelated to the topic.
AT-HA5000 has numerous advantages over a maplin kit
And the Hitachi topology is by far one of my favorites. Not going in to much details for armchair nonsense
I just like to experiment. The Audio Technica HA-5000 is well regarded, and the clone pcb is close enough to the original to be worth a punt. But I also agree that you don't really need drivers. Here's one I designed and sold earlier, but before anyone asks the lateral mosfet version sounds better - this was designed to be low cost.
@cumbb
I have now done my homework:
Now I have been tricked and the tester has not replaced the components in the amplifier with suitable ones, No!, but simply connected a different loudspeaker box.
What do we learn from this?
HBt.
I have now done my homework:
- Amplifier with the number 1 sounds, according to your description, not suitable for audio as not suitable for audio, i.e. it contains the wrong BJT selection.
Now I have been tricked and the tester has not replaced the components in the amplifier with suitable ones, No!, but simply connected a different loudspeaker box.
- Amplifier with the number 1 now sounds audio-compatible according to your description for audio-compatible, i.e. it now contains the correct BJT selection.
What do we learn from this?
HBt.
This looks good. But why have you taken the gate of 2sk117 Tr1 to the base of Tr5? Surely Tr1 is a constant current source for the I/P LTP.
Got any measurements?
What MOSFETs does Audio Technica HA-5000 use? If they use lateral MOSFETs, (Toshiba had at least one pair of lateral MOSFETs which I've used in a commercial product.) I think your circuit could use these for really good performance.
The gate of the 2sk117 is joined to the collectors of Tr4 and Tr9 as well as the base of Tr5 ( which is part of the current mirror load for the VAS ). The 2sk117 is indeed a current source for the input LTP, but it also forms a kind of dc servo which is "balanced" when the collectors of Tr4 and Tr9 are at 0V. In this case the current through the 2sk117 will be Idss, whch sets the tail current. So you can set the tail current by selecting a jfet with the desired value of Idss. Lindsley Hood was a very clever chap.
I think it's like a burning oil well, feed it oxygen and it will burn almost forever, but deprive it of oxygen and it will slowly die out without making a sound.@cumbb
I have now done my homework:
- Amplifier with the number 1 sounds, according to your description, not suitable for audio as not suitable for audio, i.e. it contains the wrong BJT selection.
Now I have been tricked and the tester has not replaced the components in the amplifier with suitable ones, No!, but simply connected a different loudspeaker box.
A look under the hood clearly shows the use of BC556, BD139, TIP35 ... and their counterparts.
- Amplifier with the number 1 now sounds audio-compatible according to your description for audio-compatible, i.e. it now contains the correct BJT selection.
What do we learn from this?
HBt.
the most typical audio (so called "true complementary") transistors sound different and for this reason I like circlotron and single ended topologies.
If Tr1 is at Idss, the gate must sit at about 0V6. That means Tr5 base is also at that voltage. There is a huge current through the VAS, Tr5, D1 and R12 and very limited voltage swing.
Have you measured the voltage across R12?
If you pick a jfet with and Idss of say 2mA then 1ma flows though each leg of the input pair and the total vas current is about 9mA. I have made this amp, along with lots of others, and it works fine; no massive current through the vas. The 1M resistor and associated capacitor at the jfet base ensures minimal effect on the Tr4, Tr9 collectors.
We talk about audio.
To talk about audio requires audio experience. Before talking about highly complex segments of audio, such as "transistors", audio key experiences are required. If these have not been done, it is not possible to talk about "transistors", suitable for audio, for example.
Aside: if the data sheets of transistors say "audio transistor", it does not mean that this property has actually been determined: neither have the developers sat down and made extensive listening tests in contrast to other transistors, nor do the various manufacturers make extensive listening tests with their different batches in order to sell a transistor that is guaranteed to be suitable for audio.
And: There are no conclusively proven correlations between peep measurements and hearing measurements. The only way to do audio is
to do (it yourself) audio.
The BD139/BD140 are installed in many devices, I have heard many different ones. They are all in the (audio) midfield at best, sounding, in contrast to some others ("audio" suitables) grey, overcast, colorless, lame...: no audio transistor.
By the way: I advise many people to think about the terms objective and subjective, but also about concept and object. And also about the social constitution of perception and categorization of "world". This always happens in contrast, contrasting! For example, most people see black and gray as black and gray (regardless of what feelings black or gray trigger, e.g. biographical, cultural) - Contrastively. No categorization without comparison. And most people can also read the traffic lights in road traffic in an "objective" way! Anyone who claims that contrastive perception - like audio, hearing - is arbitrary, subjective that cannot be categorized, systematized, "objectified" socially, collectively, also has homework to do in this regard;-)
Whether someone likes black or gray more is "subjective"... or the red of the traffic light or the green - or up or down (to the color-blind;-)
The fact that the discourse "audio" repeatedly fails here is not due to the subject matter of "audio", but to the lack of education and training in science and engineering of most of involved. They don't even know about the common constitution of a common "world" view based on common contrasting, categorizing, structuring)-;
I respect your different approach and appreciate the way you defend your point of view. But you will equally understand that we can't build complex (audio) structures without the fundamentals of science.
Many members here do have a high level of education and training in science and engineering, and I don't think it is fair to present them as if they were not.
I would call that mutual respect.
Now, let's get back to topic. 🙂
Many members here do have a high level of education and training in science and engineering, and I don't think it is fair to present them as if they were not.
I would call that mutual respect.
Now, let's get back to topic. 🙂
D1, Tr4, Tr5 and associated resistors form a Wilson current mirror. The voltages across R11 and R12 are 700mV in the latest version I built ( different resistor values giving higher VAS current ).
Thanks for all of the input. Didn't realise that even modern BD139/BD140 are pretty decent devices. Will definitely use them in preference to the MJE243G/MJE253G
I agree with the regulators existing in the consumer-marketed Hitachi products. You're right that little bit of info is usually omitted from discussions about these designs. But I'm curious how you know this tidbit about the databook app notes??
- Status
- Not open for further replies.