Musical Fidelity A-1 in A/B class?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi

I have a question Does anyone of you have tried to polarize the Musical Fidelity A-1 Class AB for example to 70 - 100mA? I like this amp but it now needs something that much less warms . Is this a good idea?
Listened to it in many configurations and almost always played very well.
I wonder if his secret is a class A or topology?

Looking for a power amplifier schematic about 20 watts to a very small room (about 10 m2 with speaker > 90dB) . I like amplifiers symmetrical design but I can not find anything in bipolar technology
Can you help me?

Cheers

Yarpen
 

Attachments

  • MF A-1 sch.pdf
    20.6 KB · Views: 175
The A1 is a ClassAB amplifier set up with high bias.

700mA bias allows ~1.4A of ClassA output. This is equivalent to 7.8W of ClassA.

100mA of bias allows ~200mA of ClassA output. This is equivalent to 0.16W of ClassA.

If you double the value of the emitter resistors, you will very roughly halve the bias current. This will drop the Maximum ClassA current to ~700mA, equivalent to ~2W of ClassA output.
 
I know that technically there is no problem to change the quiescent current. My question was different - if someone tried this and how does that sound? How much is the original sound and what is changing / worse?

I am thinking of cloning the power amp with other (tube?) pre.
 
I know that technically there is no problem to change the quiescent current. My question was different - if someone tried this and how does that sound? How much is the original sound and what is changing / worse?

I am thinking of cloning the power amp with other (tube?) pre.

Hi,

It what I'd call class AB, as opposed to optimum bias B, I call aB.

Theoretically the sound depends on peak levels, it changes as you
advance the volume, for quiet listening it may hardly ever leave
the class A region, but there will be a transition point, fairly
near maximum unclipped levels where for all peaks (the signal
spends ~ 80% of the time below 20% of peaks levels, i.e.
only in the top ~ 80% for 20%) the class A region is left.

Speaker impedance also matters here, low impedance
speakers will leave class A earlier than high impedance.

Reducing the bias will simply change the level that this occurs.
How warm it runs depends on the size of the heatsinks,
bigger ones will make it run cooler, but don't go overboard.

Decent heat sinks and 0.33R output resistors may be the way ?
0.33R should halve the static heat dissipation for around
3.5 watts class A, which changes the transition point 3dB.

rgds, sreten.
 
Decent heat sinks and 0.33R output resistors may be the way ?
0.33R should halve the static heat dissipation for around
3.5 watts class A, which changes the transition point 3dB.
to halve the quiescent dissipation you have to halve the quiescent current.
If the bias current is reduced by half then the maximum ClassA power is reduced by 6dB (1/4 of 7.8W ~ 2W)

I do agree that a high bias ClassAB amplifier spends much of it's operating time in the ClassA region. I suspect a lot more than 80%, maybe 90% to 95%.

Play average levels at -20dB, i.e. 200mW and peaks from the average of -20dB upto -4dB stay in ClassA (7.8W) and only the peaks in the range -4dB (7.8W) to 0dB (20W) supply their current from ClassAB.
 
to halve the quiescent dissipation you have to halve the quiescent current.
If the bias current is reduced by half then the maximum ClassA power is reduced by 6dB (1/4 of 7.8W ~ 2W)

I do agree that a high bias ClassAB amplifier spends much of it's operating time in the ClassA region. I suspect a lot more than 80%, maybe 90% to 95%.

Play average levels at -20dB, i.e. 200mW and peaks from the average of -20dB upto -4dB stay in ClassA (7.8W) and only the peaks in the range -4dB (7.8W) to 0dB (20W) supply their current from ClassAB.

Hi,

You are of course right about the static dissapation, it will be 2/3s.

80%/20% is true for any signals peak levels, of course if you reduce
volume whilst that ratio remains the same the class A ratio will increase.

As you imply at only - 4dB below maximum volume level (defined by
clipping not being audible) it will essentially stay in class A all the time.

We are on the same page, except for my stupid error.

rgds, sreten.
 
ok then ... i only layed my hands on two of those MF both had the same troubles and that will be a faulty trafo i had to wind new ones that was kinda modified to provide just a bit more current given the available space ...

unfortunatelly at the time i was very young and not expirienced enough ( probably still am ) and didnt make extensive tests ...means that after the replacement of the trafo amp was working make a good profit and let go

( small remark for the good profit at the time the company that was servicing the MF said to me that will not provide anyparts in europe since no one is able to replace them since the electric hazards involved during the repair and i could as well bring it in so they will replace the trafo for about 200 euro ).....
--cost of winding a CE trafo was 18 euro
-- charged the costumer 120 euro and deliver in one and a half day instead of one month (and amp is still working after 12 years )
every body was happy and especially me since i had my @ss kissed by the local MF people ha ha ha ha ha ha ha


beyond jokes the all point is that the cfp are not easy to stabilize and oscilate easy, measures are to be taken in the circuit to avoid that ....those measures usually effect the sound

similar idea ( on the difficult to stabilize side ) is all amplifiers created with darligton pairs tip142-147 none of them ever worked properly and if it did there was not real performers and not for a long time run ...all except one ...The Nelson Pass A40...

now what is common between the MF1 cfp out and the NP A 40??? only one the bias ...A40 runs about 800ma per device 2pairs per board EFP

i was wondering about that and no one come to my resque ...i think that cfp or darligton if biassed high has less margin for oscillation ... please correct me if iam wrong

thank you

kind regards sakis
 
Last edited:
Member
Joined 2010
Paid Member
Thanks Zen Mod, Sakis.
There are some minor differences in the schematics and (presumably) actual voltages.

Now, is the emitter-coupled output stage still identified as CFP? and is the calculation for push-pull class A not twice that already posted? Check Self's tabulations for his Trimodal amp where he uses 1.6 amps to supply 27W into 6 ohms. Only around 1.2 is required for 20W into an 8 ohm load. I think here you can expect ~15W before AB transition.

It really depends on what you rate the amp rather than how much of the available voltage swing will remain in class A.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.