JLH 10 Watt class A amplifier

Jump to Latest
#3,062
Bigun said:
+1

However, they do suffer from beta droop and in my view benefit from paralleling.

http://www.diyaudio.com/forums/solid-state/268846-tgm9-my-version-jlh-69-class-amplifier.html
Click to expand...

Doubling up on output devices would split the load and enable 2N3055's to work in a lower current region where the current gain would be higher. I don't see any advantages in having 8 pairs in each output half other than making use of available spare parts. You might include some low value emitter resistors to aid load sharing.

The approach taken by Hood was to substitute a better device in the phase splitter to boost the drive current and increase the overall loop gain.

Transistor datasheets refer to "current gain" where with power devices it can be seen graphically this falls with progressive current increases. At high currents the gain can be very low which has serious implications in high power (Class AB) amplifier design characterized by designers as "beta droop". I think the term should be limited to that context.

With a 10 watt Class A amplifier the term "beta droop" might suggest to some a problem of a mountain rather than a molehill.
 
#3,063
geoff moss design 2005

Hi
I'm trying together geoffs design from 2005. 15 watts.
As I understand it I can skip Q1a and Q2a, these I can use to make the high power circuit?

I've taken steps and trippel checked the transistors to get them in the right place.
I can not remove DC off the speaker out. does not matter how much I turn the pots. 25turns so I can really have a good time of it.
Q2 certainy gets hot, but Q1 keeps hot.
Output transistors is 2sc5200 and I'm using 2sa970 for iq and input, bd139 and 140 for driver.
I've build three of these now and same problem with everyone. Even checked the transistors ... new ones and still the same. checked each solderpoint (tripple and quadruple that).

Anyone throw me a bone here? I've done a few of the penultimate circuit, but these seem to be working just fine with no problems
 
#3,064
Mooly said:
Passive tone controls are OK but lossy which means you either need gain before or after them to make up the loss.
Click to expand...

Passive tone controls seem to be very popular in guitar amps and diy pedals ( stomp boxes ) Not quite sure what the rationale is for this preference as like Mooly said you need to make up the gain lost in the passive network. There seem to be some rather esoteric ideas about what effects certain circuit configurations and component choices have on the reproduced sound of guitar signal chains. I was a bass player though and in general I reckon we're more pragmatic than guitar players! 😎
 
#3,065
gasboss775 said:
Not quite sure what the rationale is for this preference ..... I was a bass player though and in general I reckon we're more pragmatic than guitar players! 😎
Click to expand...

Tradition. Plus it's cheap.

Also "normal" guitar tone stacks are anything but flat when you think they should be. Plus there's usually interaction between the knobs. The Brand-X "sound" is partially associated with their historical tone-stack bumps and wiggles.

Bass players tend to use bandaxall tone stacks. Which actually are flat when you think they should be. Pragmatic.
 
#3,066
Wishborn said:
Hi
I'm trying together geoffs design from 2005. 15 watts.
As I understand it I can skip Q1a and Q2a, these I can use to make the high power circuit?

I've taken steps and trippel checked the transistors to get them in the right place.
I can not remove DC off the speaker out. does not matter how much I turn the pots. 25turns so I can really have a good time of it.
Q2 certainy gets hot, but Q1 keeps hot.
Output transistors is 2sc5200 and I'm using 2sa970 for iq and input, bd139 and 140 for driver.
I've build three of these now and same problem with everyone. Even checked the transistors ... new ones and still the same. checked each solderpoint (tripple and quadruple that).

Anyone throw me a bone here? I've done a few of the penultimate circuit, but these seem to be working just fine with no problems
Click to expand...

The 2sa970 has a different lead configuration than BC560. It is not a direct replacement and you would need to orient the transistor body and transpose the leads to fit a board where the overlay indicates a BC560.

In previous posts I made the point that 2N3055 is good enough for this amplifier. It is because of the limited frequency response fT of 2 MHz which ensures loop gain has reduced to less than one before the phase reaches 180 degrees - a condition that is necessary to ensure stability.

Using 2SC5200 would be like driving a car that can go faster but does not have the brakes to match.
 
#3,067
Hello . i am thinking of making the first version of JLH class A amplifier 1969.It will be my first class A amplifier. What transistor should i choose for TR3 (2N697,2N1613) and for TR1,TR2 (MJ480) ?
For TR1,TR2 i have TIP3055,MJL21194,BD249C.
For TR3 i have MJE340,BD139,2N5551,2N2219,2SD669,TIP31
 

Attachments

  • WW4-69.gif
    WW4-69.gif
    51 KB · Views: 830
#3,071
2SC3902.
Actually, before Philips sold off NXP the last update to their BD139 showed that the minimum gain was now the -16 group (63 min, 100 typ.) All the ST and ON Semi devices I have are ~100 at 100mA although not all are graded ~16. NXP do not appear to sell the BD139 any longer.
John
 
#3,076
Component change?

Many years ago I built 2 x JLH '69 monoblocs. A correspondent on one site suggested (in order to improve HF response) certain component changes as shown in the diagram. You will notice that CI differ in value, this must have been my error when constructing. I invite any criticisms. The PSUs are the original '69 stabilised ones.

These are driving Quad 57s (in a small room) and sound very good.
 

Attachments

  • DSC08081.JPG
    DSC08081.JPG
    197.8 KB · Views: 705
#3,077
Simulation results show a worthwhile fall in distortion from using modern output devices on the original design.

Your changes seem to indicate a further reduction in distortion but also a small but significant peak in the response at extreme HF (approaching 10MHz or so) where there was none before. As ever, you need to fully evaluate the changes, not just as a quick simulation but also when driving real loads... and the ESL is one load that demands a stable and well behaved amp.

My advice... stick to the original concept with the proviso that changing to modern outputs could be worthwhile.
 
You must log in or register to reply here.
Top Bottom