Proper Current Source Adjustment

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
As it turns out I've apparently set the current gain improperly on my 2's although they still do sound excellent. I've searched high and low through these threads for anything on this adjustment but keep comming up with all sorts of other threads except current gain adjustment.

Would someone out there please clear this one up for me and either point me to the correct thread or take a min. and post the formula and or proceedure. My time here has been very limited lately but I want to be sure that my 2's are properly set.

Thanks in advance!!

Mark
 
The one and only
Joined 2001
Paid Member
I'm not sure that the quote addresses the question. I
interpret it as wanting to know how to set the current
gain of the Aleph current source, not the DC bias.

For the former, you generally want to set the gain so that
the AC variation of the current source is about 1/2 the output
current. It can be more or less, but 50% gives the optimal
energy efficiency figure.

You will note in the Aleph schematics that the Base of the NPN
transistor which controls the current source attaches to the
output of the amplifier through an RC network. The value for
C is set arbitrarily high so that it does not interfere with audio
frequencies, and it is the value of R which is the most convenient
spot to adjust the gain of the current source. For the purpose
of this discussion, I'll call it R0.

If you set the amplifier driving a sine wave into a load (let's say
16 Vrms into 8 ohms at 100 Hz), you can measure the current
variation of the gain N channel Mosfets (whose Sources attach
through power resistors to the - supply rail) with a cheap AC
voltmeter placed across one of these Source resistors. With
R0 taken out of the circuit, you will get one AC value across the
Source resistor (say 470 mV, for example). As you put a value
for R0 in the circuit, this will decline, and when it measures 1/2
the value without R0, you have reached 50%. If it measures
1/4 the value, the current gain of the Aleph source is 75%, and
this figure is too high for a standard Aleph. Most listeners like
the Alephs at 50% or lower, so I recommend between 50% and
100% of the AC voltage value compared with no R0.

:cool:
 
Thanks Nelson for making that very easy to understand. I really appreciate that. When I get back home from this road trip I will measure it and see exactly how far off I am. Judging by how good they sound I can't be too far off the 50% mark. They continue to astound vistors that come over to hear them.

Mark
 
Nelson Pass said:

You will note in the Aleph schematics that the Base of the NPN
transistor which controls the current source attaches to the
output of the amplifier through an RC network. The value for
C is set arbitrarily high so that it does not interfere with audio
frequencies, and it is the value of R which is the most convenient
spot to adjust the gain of the current source. For the purpose
of this discussion, I'll call it R0.

If you set the amplifier driving a sine wave into a load (let's say
16 Vrms into 8 ohms at 100 Hz), you can measure the current
variation of the gain N channel Mosfets (whose Sources attach
through power resistors to the - supply rail) with a cheap AC
voltmeter placed across one of these Source resistors. With
R0 taken out of the circuit, you will get one AC value across the
Source resistor (say 470 mV, for example). As you put a value
for R0 in the circuit, this will decline, and when it measures 1/2
the value without R0, you have reached 50%. If it measures
1/4 the value, the current gain of the Aleph source is 75%, and
this figure is too high for a standard Aleph. Most listeners like
the Alephs at 50% or lower, so I recommend between 50% and
100% of the AC voltage value compared with no R0.

:cool:

It seams that setting "R0" to get 25% of the AC value compared to not having "R0" gives 50% share.

Could someone please tell me if I am correct or not before my head explodes?:scratch:
 
Thanks wuffwaff,

With the standard A5 supply voltage +-34vdc, bias at 1.9a, 7.5v into 8ohm load, 50hz I got 99.3mv ac with "R0" (R21 in the A5) removed. I found that putting a value of 453r (as in the original A5) reduced the figure down to 24.9mv ac, and fitting a value of 430r took it down to around 23mv ac.
Is the standard A5 set to 75%? or is there something else i missed? :scratch:
 
Hi,

the standard A5 has to be set to 60-70% to meet the specs (90watts 4 ohms). Although this depends a bit on the bias.

I measured ac-current-gain at a few resistor values for my Aleph5 with BC550C:

390R 66%
430R 61%
470R 55%
510R 52%
560R 47%

I tried this at different bias values (from 1,9 to 2,4A) and loads (4 and 8 Ohms) and frequencies (1k / 10k) and the gain faktor stayed more or less the same. I somehow always thought that ac-current-gain would change with different bias values but this doesn´t seem to be the case......
Between channels there was a difference of around 1%

William
 
If you have a regulated sine wave generator, you can it so the AC voltage over the source resistor is say 100mV which would then represent 100%.
Putting in R0 shows immediately what percentage you get when looking at the voltmeter. If the voltage drops to say 62mV you got 62%. Lowering the resistor value will lower the percentage, as seen in Williams post.

/Hugo
 
wuffwaff said:
Hi,

the standard A5 has to be set to 60-70% to meet the specs (90watts 4 ohms). Although this depends a bit on the bias.

I measured ac-current-gain at a few resistor values for my Aleph5 with BC550C:

390R 66%
430R 61%
470R 55%
510R 52%
560R 47%

Hmmm - what are values of your source resistors and output sense resistors??

Because if calculating AC c.g. based on original numbers in A5 service manual - with combination of 0R47 at the output, 1R as source resistors and 1k for R20 and 453R for R21 - the ratio is 78%!!!
 
Nelson Pass said:
The exact figures depend on a number of things, so it's always
best to measure the actual values.

I've realized that the real life of an Aleph is a bit more experimental than it looks at a first glance ... ;)

I just wonder - After finishing MiniA and A30 the A5 will come (finnaly, I hope) in my budget plans - and before buying pretty $$ 3W resisitors - I want to make some aproximately calculations ...
I'm thinking of using source resistors with around 0R8 instead of 1R, the output ones will be 0R47 - 4 per channel (Mills - allready got 8 of them and they are waiting), so I gues R21 will have to be higher than those 453R - to achieve AC c.g. around 60% ...


wuffwaff said:
Stabist,

source resistors are 1 Ohm, output is 3 x 0,33Ohm = 0,11 Ohm and R20 is 1k.
The calculation is only an aproximation as pointed out a few times by Nelson.

William

Apparantly it's really just aproximation - there's guite a big difference between calculated 78% and around 50% "ideal" ratio as mentioned several times by mr. Pass ...
 
wuffwaff said:
Hi,

the standard A5 has to be set to 60-70% to meet the specs (90watts 4 ohms). Although this depends a bit on the bias.

I measured ac-current-gain at a few resistor values for my Aleph5 with BC550C:

390R 66%
430R 61%
470R 55%
510R 52%
560R 47%

I tried this at different bias values (from 1,9 to 2,4A) and loads (4 and 8 Ohms) and frequencies (1k / 10k) and the gain faktor stayed more or less the same. I somehow always thought that ac-current-gain would change with different bias values but this doesn´t seem to be the case......
Between channels there was a difference of around 1%

William

Thanks again William,

I will have another go this weekend and chart the values as they may help others in the future but for now I have used your values to show the relationship.
 

Attachments

  • ac current source.jpg
    ac current source.jpg
    46 KB · Views: 1,187
wuffwaff said:

Hugo,

measuring 0,62V would be 38% ac-current-gain!

William

Funny, as my AlephX’s are set at 57% at the moment. I believe the values are in the spreadsheet I sent you last week.
These are however measured using the method described in the WIKI.
Nelson’s other way as per post#3 gave me the 62% or 38% if you reverse the numbers. Since 62 and 57 are rather close to each other I concluded that my amp was set as intended and the math in post #9 was correct. It however leaves us with 5% vanished in amp space. And yes, I measured as accurate as possible.

/Hugo
 
wuffwaff said:
Hi,

the standard A5 has to be set to 60-70% to meet the specs (90watts 4 ohms). Although this depends a bit on the bias.

I measured ac-current-gain at a few resistor values for my Aleph5 with BC550C:

390R 66%
430R 61%
470R 55%
510R 52%
560R 47%

I tried this at different bias values (from 1,9 to 2,4A) and loads (4 and 8 Ohms) and frequencies (1k / 10k) and the gain faktor stayed more or less the same. I somehow always thought that ac-current-gain would change with different bias values but this doesn´t seem to be the case......
Between channels there was a difference of around 1%

William

William,

As long as you don't replace the Source resistors to change the DC bias current the AC Current Gain stays the same. You can vary the DC bias by adjusting R19 (A5) using a trimpot, which I know you're aware of. Like you said the A5 needs a highish ACIG setting to meet the specs albeit with more distortion. Without a scope and a function gen the calculated version works well and might be higher and off about 5%.

Mark,
You can max the DC bias by removing R19 and this won't change your ACIG setting. This will increase Source resistors' voltage to about 0.7 to 0.8V. It'll also lower the distortion at lower load impedances. Of course watch out for more heat.

Allan
 
The one and only
Joined 2001
Paid Member
Blues said:
As long as you don't replace the Source resistors to change the DC bias current the AC Current Gain stays the same...

I think this mostly correct, but as the loop involved has limited
gain, I experience slightly different values with different
transistors and values of R19, bias current, and so on. That's
why I always recommend measuring the value if you really
want to know.
 
Nelson Pass said:


I think this mostly correct, but as the loop involved has limited
gain, I experience slightly different values with different
transistors and values of R19, bias current, and so on. That's
why I always recommend measuring the value if you really
want to know.


There's our rule of thumb!...like you said, "What's a few....between friends." The picky will have to invest on test equipment.
 
Blues said:

Mark,
You can max the DC bias by removing R19 and this won't change your ACIG setting. This will increase Source resistors' voltage to about 0.7 to 0.8V. It'll also lower the distortion at lower load impedances. Of course watch out for more heat.

Allan

Thanks Allan, I have 500K in place of R19 at the moment and do not dare to remove it due to the fact that I am running 30deg above room temp if I measure at the heatsink base opposite where the fets are mounted, this currently gives me 0.630v per fet (1.9a) bias. For those that are wondering what heatsinks I used they are the Conrad MF35-151.5 two per mono channel.

I still feel the need to lower my ACIG setting as it currently running 75%, maybe I will like it more at 50% or 60%. I will try a few different settings and leave it well alone once I find a value that I like best, only time will tell. During this experiment I will chart the ACIG over “R0” resistor values if nothing else it will allow me the change this setting without the need to measure in the future by showing me what resistor value yields what ACIG value.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.